Академический Документы
Профессиональный Документы
Культура Документы
nnru;ddc /$zrtc hnr ftwrvr1 nrrv nfl!?r, hr FI ~ r n ~ nwr r!r o n h nlt,rltodr Irr ~ h o r ~ k r
unwrt(i . T/w ~ ~ ~ r ! ~ ~ ~ /,r tor / ,, cr,*ot , r ~ rh~nrt
~ t t ~ r~~ t r~z i i i - r q < z r { t 3 f tI!! t t $ ~ c \~, n i / t ~ d t r ~ g 0 A
~ ! , O ~ ~ ~ I ~,,%U. F
t r < / t n : th,,07ff,q. t . f , i l ~ o t i otlnr<l F III\I,.
~ I ~ I ~ I~ I I- ~o ~ l ~ l ( J 4 . r tnnrti7f f~~rf ~ rql~t t. d Vr
lrrrne l / l p I / ! # ~ T ~ , I ~ i ! ~ , d innd
r f t - . r ~l~~I~r r>~~ v - r r ~H/ ~. r f ~ v ~tJ!r~r t ~ rn~trrrrrl>r~ ~t~tlJ 0
-find tlrrr rrn P I rrr,mr.ly ~rnttmhle.IPPII~ ~ ' r r r f i -borlk.
rt
P R ~ ) F F C , ~A~t Rks RR\HNt. I . N l v F R ~ l T Y <)I L r l l l C 1 t f f O R O i ' I ; H
Y in Social Research
n ~ r ~ n rDd ~?r ; g r~n~ S~~rraI IS onr-stop rritlral guide that d~monstratrs
Rr~~rflrrJr CI
&
the impc~rtanwr,F attend in^ to design issr~csIn sorial r ~ ~ e a r r h .
~ ~ rrollrctinn strat rm and statistical
thlr Fll'ld p r i c > r ~ tclnfa
H David de Vaus
W l ~ ~ r ne tal l~~ t~ooks
r
analysis, de Vaus arjiur4 that, a- s ~ g n l f i r a n tas t h r w Isslirs arv, rnost p p r t l n ~ n nf
t
all ro r ~ ~ r ' n r rtl l~ r h l i l q 15
~ tthe T ~ drsign of thr rewarrh.
~ ~ S I T U C ~ ~ Iand
1'111~bnnk:
* Prvirles t h t~~ a d sit of the irn1)orranrr nf' rcsparrh
~ r11 an ur~rl~r~rarirI~~i~
dr+i am1 it< place in tire researcll pmww
Drscrtbes. thp maln r ) ~ of
s r ~ s ~ a r rrlrstgt15
h 1 1 1 socinl rrwarrh
frE Pulblications
r,onnon J troricatrd Oak$ ~ P I I L ~
T i n t puhlr.;hril ?t?(!l
PWfnct' xvi
SAGI' I'ublicati[>ns Ltd
h nnnhill S t r r ~ t
h(>ndon FCZ.1 .FI1U PART I: WHAT IS RESEARCI-E DESIGN?
1 T h e Context of Design
Description and ekplanation
Uescripfive research
SACX I'ublications l n d i , ~IXvt I td Explanatory resen rch
32, C1-Rlnck Market Thcory tsting and theoq constructinn
Greatc~rIGilfazh - 1 Theory bullding
New Tlelhr 1 1 0 048
Theory testing
Eritirih Library Cataloguing in Publication data What is rrsearch d r z i p ?
Desigm versus method
A catillopir r c ~ o r dfor I ! i ~ sbonk 1% availahlr
from the Brltlql~Library Adopting a sceptical approach to cxplana tions
Summary
TSRN O 7619 5?46 9
ISHN 11 7619 53.17 7 (pbkj 2 Tools for Research Design
Before design
Lihran, of Congress catalog record available Focusing and clarifying the research question
Identifying plausible rival hypothcqes
Tvprrr3t by M;avht=wTvyrscthnq, Rhavades. I'owr,.i
Operationalization
In C;t*j~t Rrrtaln I,\ rltddl~sZ td, GtrLl,ltilrd,Surrr-l
Pr~ntc~tl
Concepts for research design
Internal validity
External validity
Measurement error
Summary
=3 Cawsation and the Logic of Reqearch Design
Inferring causal relrlt;ionshipq
criteria For infprring cause
T,.v~v
,,fF ,,.-I . 7+*,.l-n7
C o r n y a r ~ n gtirnc points Summary
hlaLrn7 m c ; l n i n ~ f rcomparisons
~l Notes
Int~rvcntinnl:a n d indepcntlent variables
Dirnrn.;iclns crf a rrc;c.lrch ~ i c - 1 ~ ~ Analysing Experimental Data
A wngc of rrsearch designs 5 c l c c t i n ~the mcthod of ~ r ~ a l v s i s
ExpcfirnrntaI cle.;ip \Vhat kind of annlvsis is rcqlti rcd: dcscriptinn o r
I,~>nqih!dinaE dec;r~n infcrrnc~?
C r c ~ s - ~l(mal
c t dv\ien W h a t w r t of .;ample do we hL3vc7
C w t stud~rs \'\'hot Icvel of data d o w e ha\.c'
Summary Typc of cornp,lrisons rt7quired'
Vrltr,c Mrhnt type of displav i.: reyt~irtd?
f-lorv m a n y )!,mItFS art? to bc cornpasrd7
P A R T 11: EXPERIMENTAL DESIGNS Arc thc c ~ m p ~ i r i s ogroups
n indrpmdcnt samples?
1s thc dcpendcnt vnr~~lblc nnrmalIy distributtgd?
4 Tvppg of Experimental Design Is thc vnrianct nn thc rtependcnt variable similar
TIw classic c ~ \ p ~ r i m c n t dt*sign
al b ~ t w e e neach comparison grc~up?
3:upcrirnr~tlal ctlntcuts How lnanv indtpenctmt vari,lbles?
Simplct c\prr~rncnt~iE de~igns Summ,lry Froce\s for selecting the right rneastlre
I ' o ~ t - t ~ t only
rt w12hcontrol grnLlp (%her ~ S S L S ~ ~ S
i l m e< > i p ~ r i r nl,lli ~design
I<t~lrusyci-l ~i Changr scorc3.:
\.lnrc, crmlpll-. expt>rrmcnt,~l Jei;iqns (flrwr ~ n i !ci*ilinq cfftctl;)
Trunc,.ilim eff~*cts
hluLt1plc p,st-tt.5t.. Trend analysis
MuItipEc g r o u p Summary
50lorn4m tour-croup d e ~ i c n Nntc
Factnrial ~lcsigns
Summary PART 111: LONGITUDINAL D E S I G N S
Volr\
5 Issues in Experimental Design 7 Types of Longitudinal Design
I ' u r p ~ ~oft lon~itudinnl
~ rlesi~n
Methndological i.;<~t*.:
Dwcrihing patterns of change and stabilitl:
The prnI7lem ot rxplanatorv narrownesq
EstnLlli\hing trmpow8 order
l'mblems with i~itrmmlvalidity
Estatll ishing develr~pmcntal( ~ g ecffccts)
I'n)bIclnq 12-lth r%ulcrnalvalidi?
Est;tbli s h i n ~historical (period ) effcct.:
I'mcticaI i ~ s u c ?
I-low rnz~clld n y i ~ utell participnnts? I , i k course 'career' analysis
I-lori. mnnv past icipantq' Tvp.p.~s crf longitl~dinalr l ~ q i p
l - l c l w 5 h o l t l ~ ipartic~pantqI-r r~cruitvd?
Procptbsti~~ pane! dtwqn.;
Retro.;ppcti\ c designs
i ;aps b r t ~ r e c ntcyst.: and intcrvcntir~ns
Il'llicll nirthc~d(tf Jat,? ccnllcthun7 Qunki-longitt~dinaI designs
S~trnn1.1~1
I'rthlems x\-ith r,~ndornimcl a s q ~ r i r n e n t
knte
Unrveneic~~.; of I ntcrvent ions
T'ht wlf-fulEiIlinr: prnphec~. R Iswes in Longidutinal D e s i p
Ethical i s s i ~ w Mctho~iolog~cal Iqsueq
Vol untarv pdrticip,~tion Is.;ur\ n t ~ntrrnalvalidih/
I n f n m r d ronstlnt 1$4111'< of eutt>rnal vnlrdie
Nn harm t n pnrtlcipant5 Practical Issue.;
,In1f r - ~ ) l ~ f i ( i r ~ ~ ~ f i l l i t ~ *
,+inon\,~~iitx~ <I ,,,, I ,,.,4;.,I;,,., , , I ; + * ,
b>..,n>lm>#7!,
I'n~elattrition Practical is.;uc.;
R~spondc.11 t Zlurden Method clf collecting data
Rcspunrlpn t rr.cal l Samplc si?es
Cnst Sufficient variation in the sample on kc!
Method uf data collection variables
hTumhernf iravcs Infnrmation for statistical controls
Gap betwtcn waves Length
S.irnple error Types of data
Sample s i m Ethical imurs
In~trumenbdesign Summa?
5taifin~ Y otes
I;thzcal issue<
32 Cross-Sectinnal Analysis
Voluntary participation
Informed conscnt
Descriptive nnaly sis
Harm to participants
HOW~ J R V ?
Lei~elnf iictnil
Cnnfidentialih. and anonvmity
Summary
Form of data
Who?
9 Data Analysis in Longitudinal Design Factor structures and scale structures
Misqinq data Haw ~ t . n r r a l STON
l CIOSC~
hwtrccs of missing data Explanatory a n a I y i s
~ ~ ' a1 ~TX&Il'fil'
1 ~ The locic ot stahstical cimtrels
Iiicntifying miwing data bias Multiplc statistical cnntmk
Ikaling with miwing data The elabnration techniquc
M~asuringc h ~ n ~ c Basic approach
A ~ ~ r e g a Icvcl
te versus inctividual level change Interpsrtation
Qual~tativrv c n u s quantitative change Problems with the elaboration model
Measuring change in panel d c s i p s Multivariablr analysis
Dclscrihjng c h a n g ~ Cohort analysis
Tablcs Constructing and reading cohort tablr.:
Graphical Problms with cohort analysis
Summary Summary
Nrlti.
PART V: CASE S m Y DESIGNS
PART IV: CROSS-SECTIONAL DESIGNS
23 Case Study Design
10 Cross-Sectional D e q i p \*at is a caw'
\ ( r time dimcn\ir>n Units of ilna!v~is
Rrliancc on existing diffewnces Moliqtic ar~dembedded units of analvcis
Thl. nature nf ' ~ r n ~ r pin
s ' thc cross-sectional i i ~ s i g n Case studies and theory
Obtaining a timc cl imcnsion: rc*pcnted csnsq-sclct~onal Explanatory case studirh
s h d ips Descriptirtc case 5tud1c.;
Other elrments of case qtudv design<
II Issues En Cmss-Sectional Design Singlc nr multiple case\'
h.lrtlit~dological w u e s Partlllcl nr sequential?
Intrrnal validity Rctrosprctive or prospcctivc?
External valtcli t i ) Types of caw study d e s i p s
CONTENTS
D~scrfptiveresearch
Although some people dismiss descriptive research as 'mere descrip-
tion', good description is fundamental to the research enterpristr and it
has added immeasurably to our knowledg~of the shape and nature of
our society. Descriptive research encompasses much government spun-
sorcd research including thc populahon census, the coElection of a wide
range of social indicators and economic informatinn such a s household
expenditure patterns, time use studies, employment and crime statistics
and the like.
Descriptions can be concrete or abstract. A relatively concrctc. descrip-
tion might describe the ethnic mix of a community, the changing age
profile of a population or the gender mix of a workplace. Alternatively
tFlr dc.cri~~trc,nrnrght a& morr ahshacf qiicsshon\ quih a5 'I.; the 1evt~Io f
inequality incrtx,i+il~g
sr,~.~,il or ~ l ~ c l i n i n ~'Hnw
", sccular i5 societv?' or a] Dtrect causal retationshb
'Hcv\*tnz~clip o t . ~ r h 'i~ there in th!\ cun?mttnih7'
Icc~lratpdtlqcripti~rns thtb l ~ v e lot ~lncnaployrnt~nt o r potet'rty have
hi\tnrically pl,~vcda ktlv rnlr in sr~rialpol~cyreforms (Mar\li, 1487). Ey
rlc*rt~rlmstratinthe er~*tcncenf scrc~alproblems, competent description bj Indirect causal refationsh~p,a musat chain
can c h a f l e n ~ accep'tt~~i
t~ ass~lrnptronsatlout the tvav t h ~ n parc and c,ln
prc,vtrke actit~n.
Good d e w r ~ p t i o n pmvokcs tlic 'wlly' quel;tions of c ~ p l a n a t o r v
rrst.arch. I f we1 deter! p e a t c r S O C I prrlar17ation
~ ~ trvcr the last 20 vmrs
cl A more complex causal r,?od~l
sf drect and rndrrecr causal hnks
(I.(*the rlcti are sett~nt:rtcher and the pnnr arc. grtSinq pnrrnlr) rvr arc
ic~rcrtdto ;ltL '\Yhv i~ th15 liaplx~n~iig?' Rut hcfnre asklng 'why"we muqt
bt* sure about thc fact a n ~ dlnirn.;ions
i 01 the phenomenon of tncreasing
pcll,~ri/ntirln.I t 13 all vrry \t.cll to develop etabnratc tlimrics as to n h \ -
sc~cicbtyrnighl m e n b polari7cd nniv than in the rccent past, but i f the
hisic prcrni.;c is w r o n c ( I c. socltlty is not bccnming nlore p7nl,lrizcd) thcn
atlthnip&tc, criplain a non-eu~stcntphcnclrnenon arc vllv. Part flme or full
Of cnursc i[csmipticm can tIcll;rnerattt to rnindrrqs Fact g'lthering i.rr Ch~ldcare
what C.tV Mills ( l 0 5 U ) callcd 'ab.stractrd p~npiricrsn~' fhrrcl arc plrnty
~ ~ c ~ c l and castn s t u d ~ c sthat rciport trivial
13f twniplt>s o t ~ ~ n f c ~ c t si~l.v<~vs
Empirical Stan
level here
ivornt3rt,urban dwrllers morc than mral drvellers and thc wciaE11~mobile parents will do just a s weIl as those whose par~nt..;are married.
mow than the sot-ialiv stable. Hc argued that the comrntm factor behind rn I'rclprrsifirrn 2: cf;rldren In ~rftmfion.: ( r )~ r r i l( d ) sltn~rl~l
be eqrrtrlllt rworfy
all t h ~ s eobservations was that those groups who wcrp mast suicidal vff That is, children in conflictual couple 6amiEies will do just as
'
wcre also less wcll socially integrated and crpcrienced ~ r c n t c ar r n b i ~ t ~ i t v badly as children in post-divorce familirr where parcnks sustain high
about how to behave and what i q right and wrong,. He tlrrorized that one conflict.
of the c*xplanation.; for suicidal behazrinur w a s a wnw of norrnlesr;ncsq - rn Proposrticrrl 3: rlrrfdrnz if; sift~ntt(lrlC ~ rvill J do ;~vfit'tltnrt t l m v rn s r t t m t i ~ ) n
a d isconncctedncw rrf indi\o~rlt~als frnm tht*ir w c i a l ~vnryd.Of course, (nl That is, those with married pnrcxnts in high crmflict will dcr
thcrt>may Iiavc b c c ~ i ir~therways nf accounting for thcsr observations but worse than those who h ~ v c married . pnrtvts who n r r not in conflict.
a t Ica.;t D ~ ~ r k l i ~ i rrrplanatinn
n'+ was cunsi.;tt*nt with tlic f.lct5. Srr~position4: 1.llrldrrbnirr .srtrrrlhon l d ) nivll i l r ~7rrorstut l t l r r l H F D 171~ s i t ~ r n f t n l i
(b) T h a t i.;, those with divorced p r c n t s in h r ~ hconflict r v i l l d o
MFrrr<cB l a n thore who have divorccd parrtits who .~rt-not in conflict
Pri?osilion 5: rhrlriml 111 sihrntro~(bJ ~ i r r l lrln brffer f l i n ~Hiosr ~ in t t r ~ n l r o n
In contrast, a thcory testing apprnach 1~r;yi~tswith a theorv and u w s (c) That is, children with divorced parunts who art. not in conflict
t h e u n to ~ u i d crvliich o b f ; e r v a ~ ~ nto
n sm a k r i t move< from the fit'ntral will do better thnn those rvith married parents rvhn are in conflict.
tcs t l ~ cp>nrtrcuFar. Tht. nbstanatiun5 ~ h c r ~ ~ lprovidecf a tmt of the worth * ~ c ! t ~ ! ~ i rium < f / t m t l o r( ~r )~ ~(*i!! bP!frr ~TI~III / h ~ ti11
~ ' r o p p s i t i o tti- > ~rf11(8l1t)rt
of thr theow. Usrnx tk?drllrcfir-vrrnsoning to cFcrivc a syt of propoqitinns (iil That is, children t v i t l ~married parcnts who art. not in ronflict
from thtl theory dolb.; t h i s . We nr.rd to dcvrlnp ihcse propositions st1 that will d o betlcr than thmc with divorcrd parent< rvlln are in cnnflict.
THE CTJVTt-X 1 O F DESK;\ r~
J
I Theoy ,1 plan. A ~vcri-kplnn rlctails what ha< tr, be done to cumplctc the project but
tire wnrk plan will flow, E r ~ mthc proiirt'5 rcwarch d r s i ~ nT.T l r p firrtrtrorr of
Inference Deducrion
R S C ~ ~ * R T CrEr*s:,cl~
!? 14 t(7 rf:siIrt7 !ffr?t I ~ R ~~~ii-inrct*
' IFIF? IF?('^^ 1'1111!7!~'.; ?Iff~ Rl15illt'l' / I l l '
r r r r n ~ ~ z b i ~ r w17.~ yws.;~\~/r.
l ~ l r t r n lql~rsfjtwros r~h Obtaining r ~ l c v n n te17idt-ncc
entails syccifying thc type cif ct'idct~ccnwded to a n s w e r the wscarcli
question, to l r s t a thevry, tc) c~nlunte a propammc o r to accurntcly
describe somc phmomenon. fn other words, when designing researcl~
Develop measures.
w e need tn a&: gtvrn this reqtarch q u e ~ t i o n(nr thtclr~),what hpc of
samplp etc. evidence ntlt.rltd to anqlvcr tht. qllc~tion( c r tc5\t nhr theov) r l f 17
crvrr~itrri~;~
rr>rlr/'
Resrarch d~,.;ign 'deals witli A Ioqit-nl problem and no2 a Ioxis!~mE
1 Collect data I problem' (Yin, 1iIH9: 24). Bcf~>rca httilclcr nr arcli~tectcall deveIop a work
plan nr ordcr materials they mtist First establisli thp tvpe of building
Start~ngpoint of
theory burfd~nq required, it.; II+PS and the nrrds u f the occupant.; The* wnrk plan flow.;
From thiq. Sin1il,~rlv, in sucial rtl\tnrch the isques o f qnrnpling, method of
data collcctir\n It..$, qurstitmnnlirtx,nbscrvat~on, dtwl~rnent analvsiq),
p quc.;tit~n-,are all s~rb.;~drarv
d ~ s i of to the rnattcr of ' l i ' h a t evidenre do
1 need t r ~cr~llcrt"
Yo ';in+> ~-rrnposition1~1-nuIt1prtlvide a comprlling test of thv original Too oftcn n~\cnrchcr~ design ( ~ i ~ ~ ~ \ t i i i n n aori rbrl~ill
cs int~rviewingfar
theory. In~lccd, taken crn it.; ozvi~proposition 3, tor cxamplc, would ton early - brfurc thinking Lhrou~h what informn I ion hey r e q ~ ~ i rto r
reve,~lnothing bout t h impart ~ of d i v n r c ~ .F-lowrver, taken a pack- answer t11t.i r ri1c;tlnrch queshclns. \Y ithou 1 attendi nx ti] these research
age, thc ~ r - foi proposihon.; prtlvidcs a stronscr tcst of the theory than anv design rnattcr.; ;it tlic beginning, tlic ctmilusions drawn will norma2ly hc
single prclpnsition. weak and unconvincin~and f;l I l to answer the rficbarch question.
r ' i l t h ~ ~theory
~ ~ h testing ntld theory building arc often yrest.nted A S
c l i sllould be part of one ongoin!: process
alternativr, rnodcs of r ~ s ~ ~ i l rtlic'y
(Figure 1.31. Typicallv, tlieorv I7uilding will prcjduce a p l a u s i h l ~account
or explanntfon uf a set of ohsrrvations. Hnwtqvcr, w c h explanations are
f r ~ q u e n t l yillst one of a number of possible cxplanntions that fit khc datn. Research dcsign is different fr~jnithe method hy which data arc
Whilt plausible they arp not necessarilv compelling. Tkev r c q ~ ~ i r c collected. Mnnv r~searchrnttl-lorlc tcvts confuse rrwa rch designs with
svsternntic ttlr;tln:: 5%-hertdafa arc coIlwtcd kn spcrifically evaluatc tlnw methods. 11 15not nncomrntm to scc research design t r c a t d as a rnoclc of
\\.ell the crplanntion holds wllrn slrbjectrd to a r<lnKcof crucial tfit.;t.;. data cullerti~~n rather than as a Ic7ctcal ~ t r u c t u r eot thc i n q u i v . But therc
1s n o t h ~ nEntrinwc
~ about any rr5tlarch design that rc.qllrres a particular
method of data ctlliection. Al thougIi cross-section~l st1n7eys arc frv-
quently equattd with questioiinai r r s ,111d case ~ t u ic.i d arc often eqirattd
What is research design?
with p;lrtrcilrant olwcrvatinn ( c . ~ W . h yte's Sfrr*r'! Cor'rrr'r Sncicfi/, 1443,
datn for anv dt,.;icn can S r t ct7llrrtc.J with any datn ccrllcction methot!
How ic thtx tcrm 'research rIrn.;iqm' to bc rtqerl i t 1 th~?; book' A n analnqv (Figure I .sl. t l r l \ v the data arc crlllrc.td is irrclel-ant t o the l t l ~ ~nfc thc
rn1cl7t h i p ii-hcn cnn5trircfri1!: a b u ~ l d
in< !Iic~rv1s no pn~ntorilcrunr: JP~F~T
matcrj~lqo r +ctting crit~catdate.; for c o m y l ~ t i u nrli project s t a ~ c l~n!il
s mrt' Failing trl cFl<t~ngulshtret~r-ern r l c s i p and mcrhtx-l leads to prnw
know {what sort of building 1s bring c o n s t n ~ c t r d YIP , first decision i.: e ~ ~ a l u a t i oof
n dt%.;qns. Equating crr~rs-sectionald ~ s i h ~with .; questinn-
whethcr w r nccd a high risr, cltficc building, a factory Ior manufactt~ring mires, vr cnsc s t u d ~ ~with s parlicipant obsc.rv,~trtlr~, mcans that thc
t home or an ,~p;lrttnmtblock. Until this
machintry, 3' school, a r ~ s i d r ni;ll designs arp clftcn ~ ~ ' r l l ~ again<!
~ a t ~ dt h strengths
~ ~ and wi~aknesscsof thtl
1s dolit* rrttl r-annot sketch n plnn, c>l>tain pr,rmits, wurk o r ~ t;1 ~ v c ~ r k methot3 rathcr lhan their ability tc, d r a w r c l a h v c l ~i ~ n , l m b i ~ t l o uconclit-
s
5cheJtilr~cjr clrdcr rnatcri;ilq. won< o r to <t.lcct I-rct\r,een t i v ~ lplnusil-rlc h v ~ ~ t l i ~ ~ t , * .
D@slgn
tvpe 777 Lonq~tvdrnaF Cross-sectronal
a p i n t of c n n t ~ r s ~ .~.m I ~ a sbeen thr t ~ n f o r t u n a l rI l n h i n ~1-rt.t~
study m e t l i o ~and
t k ~ o ~ that
l
embracr t l ~ r s edata
Ousst~unnaire Quest~onna~re Questionnaire Questtonnaire information and explanations that arr 'adequate at thc l t v d o f mean in^'.
Whilc recognizing that rurvey rt.sc;lrch has not ,llwavs hrcn goorl at
1 Mefhod lntphr~ew lntew~ew Interview Interview
1 data tapping the suhicctive dirnt~nsionof brhaviour, *hc argue< that:
(structured or (structured or IstrucTuredor (structured or
COlleCT10n loosely I loosely loosely 1 loosely
structured] structur~d) structured) structured) Making sensr o( social action . . . is . , . hard and sur\rcys have not traditionallv
h c ~ nw r y good a t ~ t TIw . cnrlicst sllrvrbv rfiearchvr.: started n tradition . . . nF
b r i n g i n g thr rneanmg frtrm nutside, ~ i t l ~ bv c r makin): u<e of thv researcher's
Observat~on Observation Obse~atlon Observanon c t t ~ l rnt p l . ~ t ~ ~ i lcxplan.~ttons
rIc . . . or hv htinging ~t trom w b < i d l a n - rn-deptl~
intrnriews <prinkling quts~m. . . Pibernllv nn thr r a w cnrrelahon%drrived frcrnl
the survey, Silrvey res~nrchhrcamc much mirre ixxciling. . , when it h c ~ n n
Analys~s01 Analys~sof Analysis of Analysis of
Clocuments docvm+nts documents
~ n s l u d i n qmr-anincful d~mlw.;ic,ns vn l l ~ cqtr~dvdcki\:ri. [Thi? ha< Ilc,rn donr~1n1
documents
I hvn rvavs, f ~ r < t l vIbyJ ashinl;. the actor cither for kit-r reastin% dirPctlv. or 111
I
<upply ~nfnmatuonabout ~ h tc- ~ n t r a vl a l u r s in ht-r I ~ t t .around ~ t l u i hrye m a r
a w l m e she i s ortenting her life. IThl.;l involvrp collect~n): a .;uffici~ntlv
Unobtrustve Unobtr~slv~ Unobtrusive Unobtrusive t which an actor fiilds h e r w l l Ilint n team ol
c o m p l e t ~ptclure of the c o l ~ t c x in
methods methods methods methods ou biders mnv mad off the mcaninpful dimensinn\. (1 Q82: 223-4)
-
-
The need for rcsearch design stems from a sccpticnl approach ti) research
and a view that scientific knoruled~r.must alrvnvs be prwisional, 7 h c
purpose of rcsrarch dedm is to r~dziccthe ambiguity o f much research
evidcncc.
We can always find sornc cvidenct* consistent with alrnort any theorv.
QUANTTTA'I.EVF AND QuAl.tl.AITVt KPSCARCH
Howeves, we should bc sceptical nf the evidcmcc, and rather than
seeking evidence that ic ron.ci<tnlf with our theory w e should seek
Similarly, d c s i ~ arc ~ s oftcn equated with qualitative and cluantitativta evidence that provides a m r r i p t * I I i ~tcst
~ ~ ~of the tlirory.
rcqcnrch methods. Social rurveyq and pupcrimcnt.; are fr~clucntlyviewcd
There arc two related rtrategie.; tor doing thi.;: elirnina ting rival
a.; piirne cuamplcs nf c l l l a n h t a t i ~ rrst.;lrch
~r and ~ r trvaluatt*d
. a ~ a i n s tht>
t
euplanationc of the er.idcnce and dcliberatrly sccking evidcnce h i t
strmgthq and tt-eaknf-cc of statiqtlcal, q t ~ a n t i t a t ~ \ .reccarch e meth0134
cnu l d rJisymi~.thc theory.
~ n anal\-<I%.
d Case ?turltt9.;,on the ~ ~ t h hand,
cr arc oftcn qrcn a s prime
rauample:: of y i ~ nita l tivc rrsr.nrch - w l i r c h acfopt5 .~ni n t c r p r r - t ~ v approach
e
to data, st~ldres'things' \rf~thintlicir ctjntext and considers thr subjective PLALSIRLE
RIVAL IIYPOTIT~SI:$
mcnnings that pmplc bring to their situatinn. A fundamental . ; t r a t e ~nf social research involve.; cvnlun'tin~'plausibtc
I t is ermncclus to equate a p t r t i c u l a r r e s r a r c h deqign with eithpr rwal h v p o t h r < c ~ ' .We ncrd to cxaminc and cvaluatr alternat~\~e wFa!.q tut
qunntitahvc o r qualitativca methods. Kin /199?), s respected autllority nn cxpl~ininga particular phenomenon. This applic..: rcgardEcsc ot whetlirr
castL slud y drvsign, has shcssed thr irrelevance nf the quantitativt*/ thc data arc- quantitative or qualitntivr; rcgardl~ss of thtl particular
q t l n l i t a t i v e distinction for case studic.; He point< out that. r e s r a r c h d e ~ i q n (experirnmtat, crvz\-wcfional, lono,ituc!inal nr casta
IVI TAT I5 RESFARCI I DFSIGY'
Causal relationsh~p Rut t h w e data art* not cornpclliiig. There arc a t lrnst three o ~ h r ways r
of accoi~ntingfor this correlation rtrithout accrpting thc. caliral link
Academic
ach~evemenf 1 bttn.t.u.n SCIICWI kpr*and achievt*mr.nt (Figure 1.6). There is thc ~r,lt,rtrritu
ex~lanat~rm : more able student< may bc srnt to fee paving private
khe
school< In t h firs1 ~ place. Thew is t h e j l ~ n l f rrsnr~ri-r7s
y explanation: parents
on child's ~ n ~ tab~l~ty
Alternatrve explanallon*s~lect~vity ~al who can afford to send their childrrn to fee pa, i ~ r gprivate schnnIs can
also afford other hclp (c.g. books, private tutoring, quiet studv space,
computers}. It is this hclp rather than the type of school that p r d ~ r c e the s
bcttcr pcrformanc~nF psivatc. schml sh~detlts.Finally, there is lhc family
irnllrc.~ cuplanatinn: parcnts who value ~ducaticmmoqt ate prtlpared to
1 achievement 1 send t h c ~ rchildrcn ti) fee paying private schouls and I E is this family
emphasis on ed ucntion, not t h schools ~ themsclvcs, that produces the
trcttvr academic prrfr~rrnancr,All these explanat ions are equally con- I
Alternative explanation lamily resources
sistent th the obwrv.~tionthat prt va te school .;tudents do bctter than
go\.ctnment schoor ~rudcnts.Wi tl~cjut colIecHng Curther ev~dcncewc
cannot choose betwccn these explanations and therefore must remain
resources ach~evernent open niinded abnut which one ma k t ~ smost empirical sense.
Thrrr- m i ~ h also t bi>rnetlrodologrcal euplanation.; for the Fitiding that
pricatcl ~ c h o n l5hrdt.nt.; pcrfclrrn t,t~~lt.racadcrnical lv. Thew rnrthosfolo-
gicn l 1\+1rt=3might ilntltvnine A n \ a r ~ w m e n tt h ~ ta causal conncrlt~nn
Atternaf!ve explanatron educational values
the rc.;ultq due to q~r~hlirm.~ble
~ ~ 1 5 1 t\ .l r ~ w , ~ y <of rncnsurin): arhlcve-
merit? From what mngc aird numhcr uf schools were the data obtained?
I, Child's I On I ? c w many c a w %,I re the conclusions based? Could the pattern simpIy
be a functirm of chance"ese arc*all possible altcrnati\~eevplnnntions
for thc finding that prii-ate schwl students perfrjrm L~tfer.
Good re5carcl-r dwign will anticipate cornpetin# explana ticlnq Ivfnre
collecting data so that relevant in forma tion for evaluating the rclalive
Figure 1 h Ctrrrsnl ond n t r t ~ - r i ~ l t s at~xplnnnlirrtl.:
f of Ill' lvt7~wn
r~lntin~r~lrip merit4 uZ thew cumprting explanntions is obtained. In this example of
sr!:ou! hflrrnn ~ i j iur7t1rfer?rrrna'lrtt.r'rv)rtv!/ schot,lc and academic nchievement, thinking about alternative plau~ible
h p c l t h w e s beforehand would I ~ a dus to find out about thta parcnts'
financial resource.;, the study rr.;r>urces available in the home, the
study); and regardless crt the method of data collectinn (p.g, observation, parent<' and child's attitude? about ducati ion and the child's acndcmic
questionna irc). Out rn indsct needs to anticipate altcrnativc ways of abilities b ~ f o r cnt~ring
e the schnnl.
in terprcting findings and trr regard anv intcrpreta tion of these findings
as prori.iion~!- sutriect t o further tcst~ng.
Thr fidlirrit of nftirtttr~tv!lttb ior?crq~tt+rrf Althnuqh evid~nccma\* be con-
Ihe i d r ~nf evaluatlnf; plau3ible rrvn l h y p n t h e ~can ~ ~be illu.;kratrii
sistcnt rzmlthan i n i t ~ proposihr~n
~l ~t might he rqunlly consktcnt r\.ith a
using tlic crample of tht. cnrrclation b r l w ~ e nFFC crt >ch~>ol attenci~dnnd
ranginof alternative yrnpositionq. Too oftcn p~uplt' in not even think of
academic acliirucrnent. Many parents a c c ~ p the t causal propos~tionthat
thc allernrttive hhvpcltlicses and simply conclude that since the eviclcncc is
attendance a t f ~ e payin? private scl~nc>lstmprovcs a child's acadtsmic
con~iqtvnt with tlirir ihcon. thcn the thcorv 1;. tnw. Thik form of
pcrfnrrnanc~~ {Flhwrc 1 . f ~ ) .%-hnuls thrmcc.l\-es pmrnc~tt~ thc same nolion
r9lsnning commit5 t l ~ cI(yical fi7!brcl, of iifirrrrrriy t\rra c-(vrs~t71irrrt.This form
h v prclrnincntly a d \ . e ~ i < i n gtheir PI<': rates and comparing them with
nf rc;lqoning has thr follrnving lo~icalsbucturr:
those of otfirr qchools or with natimnl averages. Pv irnphcat~nnthey
prnposc a cat~.;al connrcl ion: 'Send your child to our sclinnl and thcy will
pass (or K F ~grades to gain cntry intc~prt%qti*ous institutions, courses).' a If A i.; true thcn R should follow.
The data t hpv provide artb consistent rvitli their pmpp~lsition that tht=sc a Wc i3bwn.e R.
qchonls producr trcttpr rc.;i11tq 'T'l~rrrioreA i< truc
!\'I1A.S I5 RESEARCH VTS1CY3
If wr apply this logic to ~ I I Pl y e of school and achievement prt~pcwitinn, ilclthing !+ill chnnge the mind n f thc true bclic\'cr. Fxchangp ~llcnrv
tlic logicLil struchlrc of the scl~ool tvpc and achieveme11t arEurncnt (Homans, 1Yh1; f3lnii. 1964) ir hrgt-lv non-falsifinblr. I t assurntas t h d i wc
becnrnt..; clearer. alwavs r n a x i r n i ~o~u r gains a n d a\vid costs. RII! MY can see a l r n r ~ s t
n n y t h i n ~;r* .I gain. Great sacrifict-s to care for a disabled r~latircecan bc
Ini ti,! prnpos~tion: interpreted as a p i n (satisfactinn of helping) rathrr ihan a lvss (incomtl,
time for self rtc.). We need to fr;lmr our prc~~cwitrrrns and define (rur
a Privntc schools pruducp better students thnn do government schools. terms in s ~ r c hs way that thcy art* capable of bcint; diqprnven.
The trst:
THEPRDVTGlONAL NATURE C)F SUPI'ORT FOR TI 1 E O l t l I.\
ICA tjvtt R If privntc wlrnols p r o d i ~ c ebcttcr students ( A ) tl~critheir E v ~ nwhtrr thr. theory is cclrrt>hc+mtcdand ha< sur\,irtd attcrnpt.; ~ L I
~titdcntqshould get bcttcr final mark.; than thoqc from gnviv-nrnen t d i s p r c r ~it,~ thc thcnry remains prt~visiona!:
funded schools (B).
F 1s: Irrrrl Private school qtudcnts dc, achic3ur better final marks than
gcwrrnment school st11dr11ts(observe. H). fnlsific.lhnni.;rn stresses the arnbi1;uity of conhrmatrt~l~ , . . i.orrnhoration ~ i v t n . :
A is t r l t ~ 'l'hc.rcforc private, schools do p r o d ~ i c cbcttcr
'I'lrr'vt~fi~ur only Ihr c~nitortthat the tllerwy has hrrn testcd allrl <iir\ ived the test, t h a t
str~dcnts(A is truc). even altrr t l r t * mo4 irnprcq<~vv ct>rrnbt~ratinn< o f prr~dictrunsit ha'; rlnlv
7 hic . . . 1s t,>r trnm thr <tatus of
nchicvcr! !lrt, ctattz< crf 'not vrt ~Iisct~nhrrncd'.
Rut I hat.? ~lwactyar~lrctl,bhc b e t t ~ rpt~r(cjrrnnnceof privi1fr ~ c l i o a l 'b~inq trrrr, (C~ujknnd Csm(-trr-ll, I L ~ ~?OF~ l -
t htr effect of othcr f.ictors. The problcrn hcrc is
shldcnts might aIso r ~ f l c ~
that ,iny number of cxplanntions may br ccrrrrct and tlir cvidcncr does
not hclp rule out many nf these. For thr sncinl scientist thiq Itwrl of These always may be a n unthcru~lit-ofcxplanatinn, Wi* cannot anticipatr
indctcr'minacy is quite ~tn.;ntisf~~ctory. Tn rCfcact we are only in a position The
or evaluatc r1\?cry p n s ~ i b lc~ j > l . ~ n a t i ~ > n rnorr, dltiwiative explcina-
tn say: ticrns that havr b ~ e nelirninatcd and the more wc h t i v t ~t r i d to disprin~t*
our thwrv, thr. more confidence wtl will have in it, but rue should avr7ic-I
thinking that i t is /7roi7cli.
* If A lor C, or D, or F, or F, o r . . .I then R.
Cl'tcrbsenle B. H o w t ~r r ivt3can dlsprclrlr ;l throrv. The logic of this ic:
v Theriifcrrc A [ox C, or 13, or E, o r F, or . . .] is tnre.
If t h ~ n r vA is true then f l sliould filllow
Although explanation (A) i q still in the running because it i s conststcnt R docs r ~ r ~follow.
l
with thc nbsenlations, we cannot say that i t is the moqt plausible T h e r ~ f o r cA is not true.
explanation. We need to t r ~ to u r prupositinn more thororrghly bv
eva lunting the worth of the ~ l t c r n a t i v epropo<itinn+.
So long ar; D is A valid test of A tlit- ahsence of B should make tls rcicct cw
FALSII:IC'ATION:
LOOKING FOIZ F V l n E N C E TO IIISI3RC)VE THE TllFORY rel'isc ihc Ilicnry In rcality, wt) w i ~ u l dnot reject <I tli~.nr!! simply becau.;tl
As wcll as c v a l u a r i n ~ and ?liminatin~ nltrrnative explanations we a single facl o r observation cFoc.5 nett fit. Befort. rt=jr*ctinga plausihl(b
shot1 lcl rigorc~uslveva luatc o u r vwn theol-iw. Rather %ban asking 'Wh'lt t h e n 6 wc w n i ~ l d requirc rnultiplc disconfirrnnt~~~n.; ~iqingcliff~miit
evrdcncc M ' I I L I I ~ coni;titlitt' Supp~Jrtfor t h tli~~r!'?',
~ ask 'What evidence mcasurr.;, dittcrcnt samples and d~ffcrentmethod.: ot data cc,licction and
W ~ I I L cr>n\.ince
~ m e that thc t h c n y ic ; t v m r ~ I~t iq npt difhcitlt tc, hnd an~lys~\.
evidrncv consistent with a tlieoy. It ir; nli~cl?tougher for a Ihrnry to In surnrn,lry, \vc ~ l i o u l dadopt a ~cepticalapprcrach to euplanat~t~n.;.
suruivc thc f c ~ of t pcnplr trving tci disprove it. We should nnticipabe rival intrrprctatir~nsand c o l l t ~ tdata to ennblt thr
Unforl u n n t ~ l ysome thctirlcs nre closcd svstrmh in which a n y cvidence winnowing t r t t t o f the weakcr 11xplnnntion< and Lhr ~drr~tificatinii of
call bc intwprcted as support fur the fheorv. Such theories a r c <aid to be w hicli altcrna t ivc theories nla kc* 11lnstempirical st*nsc3.Wc also nivd to
non-lnlsifinhle. Many religions or belief ~vstcrnsgan brcarntl c1ust.d ask what data wro~lld challenge tfic explanation and cdlect data tc,
wst~rn.;tv her+ all e\birlc*nci*can be accnrnrncid,~ tcd bv the tlimrl- and evalrlati* thra thrr>rv from this nirlrr d ~ m n n d i n f prr.;pcctiif~.
i
W H A I' IS RESEARCIH DESIGN?
Summary
This chapter has outlined the purpose of rcscarch design in both dcscrip-
live and explanatory rcscarch. In explanatorv research the purpose is
to develop and evaluate causal theories. The probabilistic nature of
causation in socia1 scicnccs, as opposed to dctcrrninistic causation, was TOOLS FOR RESEARCH DESIGN
J i scussrd.
Rcsearcli design is not related tr) any particular method of collecting
data or any particular tvpe of data. Any rcsearch design ran, in principle,
U ~ anyP type of data collection method and can use either quantitative ur
quaIitativc data. Research design refers to the strucflarr of an enquiry: it is
a logical matter- rather tl-rdn a logistical one. Tn achicvc a rcasonabl~research deqign we nerd to attend to a number
It has been argued that She central role of research design is to uf matters before wc arrivc a t Lhc final design. The first section of this
minimize t11c chancc nf drawing incorrect c a u ~ a linferences from data. chapter outlines these preliminary steps that prccede design. IL then
Design is a logical task undertaken to ensure that the ~vfdencccollected expands on the idea of alternative rival hypotheses that was introduced
enables us to answer questions or to test theories as unambiguously as in Chapter 1. Thc second section introduces a number nf concepts that
po~siblr.When designing re.;earch it is essential that we identify the type are f~~ntlamentalto designing good rcscarch - internal validity, external
of evidence required to answcr the research qucslion in a co~lvincing validity and measurement error.
way. This means that we must not simply collect evidence that is con-
ststunt with a tmrticuldr thcory or expianaticln. Research n c d s to be
qtructured in such a way that the evidence also bears on alternative rival Before design
explanations a n d enables us to identify which of the competing explana-
tions is must compelling cmpiricallv. It also means that we must not In the same way that an architect needs to know the purpose of tllp
simply look for cvtdence that supports our favourite theory: we should building before designing it (is it an office building, a factory clr a home?)
also look for e~ridenccthat has the potential to disprove our preferred social researchers must be clear about their research question before
explanations. develop~nga research design.
Fncr~sil?~
at7d clurrfih~~q
the rt"~earchquesti071
The first questinn to ask is, 'Wl~ntquestion am 1 trying to answer?'
Specifying a question is more than identifying a topic. It's not enough to
say, 'I'm interested in gctting some answers about family breakdown.'
What answers to what questions? Do you want to know the extent of
family breakdown? Who is most vulnerable to family brcrlkdown?
Changing rakes of breakdown? Over what period? Where? Or are you
really looking at the causes of brcakduwn? The effects of family
breakdown? AIF tht cffectq or just particular ones?
'1'0 1iarrow thp t o c ~ t sof descriptive rcscarch W P need to specify the scope
of what is to bc described. The following guidelines, using farnilv
breakdown as an example, help narrow down a descriptive ~esearch
topic into a researchable question.
2 What is the tttrrr* f r i r n ~for ~ the descript~nn?Is our 1ntt1re5t in c l i a n ~ c In frnrnin~cxplanntory qnccti(>ns tve need ttr furthcr spcc~fyo u r f~jrus.
nvcr timc or jukt about ro!rh,trrpc~m?flevrl< of famiti7 breakdown? If ~t Explanatory rcrt~.~rch explvri.5 cause< and /or constc-prnces of a php-
3s ~ h n u chanp*,
t over what pcrind? nilmt*iirrli, s o t l i ~ ' researcli q u r s t ~ o nnittst hp clear n b o l ~ tthe stvlc c-rt
3 What is the 8,-cymphiatl location for the descripticm? Is the intercst in explan,ltory rcsrarch and icltvltify which causes nr consrquences ~t will
I
Family brcakdnwn in a particular community, in d i f f t w n t region5 or in\-{=stigate.
tllr suhole n a t ~ o n ? I< i t rrm-tparativc, Zooking at breakdown in Pt4tlrc ilutlln~ncsomr difftlrrnt t y p ~ +o f c\~lan.ltirrv research i t i.;
Jifftarent hpcs ot co~lntrirs(t>.g.l u ~ h l v1ndu5trialirrJrcrws rnpi~Ftl* ~ I C C ~ to
I I intrrwdclunb
~ wrne tprni.;.
industrraltzin~lFrrsusirnp(>vrrished cnunkries)?
4 I low 'yenenrl 1s thc de.;criptirJn to bc? 1jo you want tn be ablc to
Jcscribe pattrrnq for aprr.iFic subgroups (e.g. arncmg those who i t 1 This i.: the vanahltb that is trc,itr.d ns the r*ffi,r.t In
married as tecnngers, amcmg those who arc in i1v,ht-!o relationqhipc, thc* causal rnt~dcl:it is dt-pendent o n tfrc inflticncc r l l somv othcr
\tu-tjnd nrarr1aKes ctc.)? t ,~ctor.The {lcprndent \.nr~a blc is a1c.c~referred t13 a4 the clutcornc
5 What n.qpect o l I he topic a r c vou inter~stcdin? Is thc inicrcst in ra lcs of \rnriablp and ~ncausal d f ~ g m r n ~t s is mnvontionall\>d c s i p a t e d ns the
I~rc,ikdnwn?'Tlrc cxperirncr? Laws? Attitiidcs and hc*lleFs? Y variable.
6 I l o ~ \~dr.;trnt.l
. i\ vcmr intcrll5t' 15 your inlt,rta.;t in f,~rn~Ev 1~roakdc)it'ti or Irr~!cb~rrrrrl~,tll ;nrttdl!r ni\ i* Ihc vCiri;lt~tc th;tt 75 tht, yrta<urncdc ' r 7 / h c L . i t
irr fnrn~EybrcnLclnwn as a rr>flrrhonnf sljmethinq morr abstract (c.g. l < a l w callid tt~clpredictor vdriablc, t l ~ c \ p ~ n m c n t a l\ ~ ~ ~ nor a bthel ~
'it~.ly?l frilgrn~ntat~~ n~, C I ~ .~(lnflrct,
SL I~ r n , role (11 the
~ n d ~ i i d u a l ~ s{hc c,xplar~at~x-y ~ ~ n r r n hand
l c I, { l p s i p ~ ~ l tin
i d cauwl ~ i i a g r a m sas thr> 1
stntc in the private lives of citizens)? v n r i ~ h l e(as In tsducntinn (X') 4 incnrnc level (Y)).
7 What is the ~rnittifa~mlws~s? Thc unit of .~nalysisis Iht~'thing' about I~rh~t-rwzi~r~ r v r k i l d r ~ I w r c vnriablcs come between Il~cindependrnt
which we collcct ~nformation and from which we drnrv cnnclusion~. i.,iriahle anti tlrr dependent variable in a causal cliain. Tliey arc. the
Often this is a person (e.g divorced pcrqcm) bwt ~t may h~ ' t h ~ n ~ . ; ' nln7n.;by n.hic11 carise S prt~ducese f f ~ c l1'. lntrrtominr:\.ariablcs are
such as organi7ations (divorcc courts), a family ax a whole, cvcnts rcbpwsentcd in causal d l a ~ r a r n sby the sl~mhoI%, as illustratcd In
( c . ~d.i v o ~ c ~ s periods
), (divorce in diffcrcnt years), places (cnm- Figure 2.1.
mt~nities,countries). F~fmr~mrr.; rwrr(1llEcs Two vnriables can bc. corrrlatr*d withor1 t bclng
Thtl questic~nsrvc can a n w e e r will dcpcnd on the unit nf analysis. c s u s a l l ~rdtited. The corrc!;ltion r n d p Irt. d u e to tht- two F~ctr~rx bcin):
Wc co~zldcompare divorcrd individuak with nnn-rlri~orced ~nrli- r~utcomesof a third variahlc (see C h a p t ~ rI). This third variable i s
vidlir7t~(indiviii~~al as the unit of analysis). We could studv a serirls of called an exttancous variable and is also symbolirud as Z in C ~ I ~ S ~ I I
~1ivorct.sand ctxarnine whai the process nf b e c o m i n ~divnrced was dingrams, and the form o f this reIatinnship is illustratcd in Figi~rc.
(rvcnt as unit of analysis). Wc might itqr vear as the unit of anajvsiq 2.2.
and track changt.s in d i t - o r c ~rates sincr*. Iq45. Uking cnuntric5 urc
might examine the different divorce rates in differ~rlltvpcs nf mun- , ~ T I F I ~ W or r f f r c f ~ T l l i ~is the Irast focused typt*r-rf cxy7an;ltnrv
Smrr I r r ~ r for
tricl.: w i t h rl vivw to comparing the p a t t r r n s in d i f h w n t typts of r~si*rlrcIi.I t ilivoIv~*< idcnti Fy ing the corc pht~nomcntln(P R. c l i a n ~ cEn~ ~
colrntrics. Allcrnatively familics might br tlic unit5 of analvsis and wc divorce rate dncc World War II) and then searcliinl: for caizhc.; or
may w a n t to Inok a1 the char.~ctrristrcqo f d i v n r c ~ nFantiIleq ~ ( e . ~~. 1 7 1 ~ . constaquc.ncw r l f thiq. Searcliin~for catl.;r.: would invol~.ridentif~jing
fC~milv incomt>, farnil!. ~ p c ,ndhllri* (71 sclatir,n\hrp in tarnill) p c l s ~ i l ~causal
l ~ ~ f,ictors ( c . ~ihanging \'alue5, rlc*clinc~in rcl~gron,
compmed wbtli those of nun-di ~ ~ o r c i nfami g lies. chanpnji population mix, crrsnnmic chang~..;, I c y 1 rt%forms,changr.; In
Thinking bcvclnd individuals a s UII~ t sof anal y s i ~broadens thc welfar~.suppnrt for lonc parc=nts). Wc would then drsign research tn
ranKc nf resuarch question.; we ask ~ n broaden.. d the rangc atld ~vnl~l.itc> which ol tFie~ecau\is< hc*lpq e ~ p l a i nchangrs in divorce mtc.;.
crrllrccs OF d ~ nvailablr. t ~ l'or r.rarnpl~*,ti YeaF wtatrnthe untt of nil<form of r ~ ~ r a r cqhu e ~ t i o ni k ~llustratcdIn F i q i r r 7.7
analvsis we ~%'ould obtain st.ltistlcs f~clmthe rrr[e\.ant nntirrtial collt*r.- rVlrrnafivt.lv normight focn~, nn the rcmqilqucncw rn tlirr than cauwq
tann agencicq ri=l:arcIing d i ~ u r c cfor each yc~lr.We t z r t ) ~Id~ also col lcct of cli,in\:cs in dii.nrcc sate (Fl~ui'c2.4).
f ducatjon Occupat~on * Income
1eve14~) I Changes In l @ v ~ofl
welfare orovisron ( x ) I
I Rise ot I
Workforce
particlpabon
ChalFengo of
IXP
Increase !n
conll~cl~n divorce rate
marrlaqe
Fi+.urc 2 <,*lr, lr~rr=,, 6~ I nlr<r..
Rrlinio~~s
Changes In d~vorcerate
slnce WWll ( X ) - ?(YJ Rejection of
wrongness
absolutes
Infprnal zlnIidity
We need to be confident that the research design can sustain the causal
cunclusions that we claim for it, The capacity 05 a research design to do
this reflect? its internal validity.
TO01,S FOli RESEARCH DESIGN 29
Imagine a rcscarch project that compares the emotional adjustment of our capacity tr, gcncrnlizc m t m widely from a rcscarch study is tl-rc use
children from divorced families and intact families. It finds Hiat children of unrepresentative samples. This, and other threats tcl rxtcrnat valid~ty,
from divorrrd families are less lvcll adjusted than children from intact will be discussed more fully in Chapters 5, 8, 11 and 14.
families. Can we conclude that divorce caused emotional maladjust-
ment? Not on the basis of these rewlts. The design does not enable us to
eliminate alternative explanations. The poorer adjustment of children
with divorced parents might be due to adjustment differences that A further threat to the concIusions that can be drawn from any study is
predated parental divorce. measurcmcnt error. This occurs when we use flawed indicators to tap
A different research shdy may deal with this problem by tracking concepts (see Chapter 1).
children before parents divorce and for some years afterwards. It may
find that these children do show a significant decIine in emotional
adjustment after thcir parents' divorce. Docs the research design s l ~ o w
that divorcc is producing this decline in adjustment? No. The decline in Indicators must meet two fundamental criteria. They must be both valid
adjustment may simply rcflcct a general decline in emotional adjustment and reliable. A valid indicator in this context means that the indicator
as children get older. The same decline may be cvident among children measures the cnncept we snv it does. For example an 1Q test is used to
from intact kamilies. measure intelligence. If it really measures intelligence the test would be
Yet another shldy might try to overcome this problem by tracking valid. If the lQ test measured son~ethingtlsc instcad, such as cducatitm
changes in adjustment c ~ chikdren
f bufure and after thcir parents divorcc level or cultural background, thcn it would be an invalid measure of
and changes among children tram lntact families as well. If children intclligc~~cc.
troni inlact families show less drteriuration in adjustment than children Reliabilitv means that the indicatnr ~'o)rsistcnll~tcomes lip tvlt1-r thc
from divorcing famflies, this must surely demonstrate the effect of same measurement. For examplc, if pctlylc consistently obtain the same
divorce. No We would need to bc sure that the two groups of children [Q score on repeated IQ tcsts, then the test would he reliable. Iftheir
were romparahle in othcr relevant respects ( e . ~age).
. The different rates results fluctuate wildly depending on when they take the test, then it
of change in adjustmerlt could be because those from divorcing families would bt. unreliable.
wcrc younger on average, Maybe younger childrrn show greater
changes in adjustment over a particular period than older children. It V~liclrf!/ The earlier discu~sionof internal validity related to the validity
may be age differences rather than having divorced parents that account of the research d ~ s i p i It
. addressed the question: is t h research
~ design
for the adjustment changes of the two groups of children. delivering the cnnclusions that we claim it delivers? In addition we need
Internal validity is thc cxtent to which the structure of a rcsearch to cxamine the validity of the nzemures uscd in any piece of research.
design enables us to draw unambiguous conclusions from our results. The validity of a measure depends both on the use to which it is put
The way in which the study is set up (e.g. traclung changes over time, and on t h sample
~ for which it is used. For example, the validity of using
ma king comparisons between cornparabIe groups) can eliminate frcgucncy of arguments hetween partners to measure marital happiness
dlternative explanations for our findings. The more the structure of a turns on what we mean by marital happiness. The validity of this
study eliminates these alternative interprets tions, the stronger the measure may vary for different cultural groups and for the same cultural
internal validity of the s h d v . A central task of research design is to group in different histnricaI periods. Measures of children's emotional
s h c h ~ r ethe study so that the ambiguities in the abovf examples are adjustment will vary according to thcir age and thcir cultural group.
minimi7ed. It is impossible to climinate all ambiguities in social rcsearch There arc three basic ways of assessing validity. Criferrt~rlvnlidity is
h u t we can certainly reduce them. best suited to situations where there are well-c.;tahlished measures of a
cnncept that nctd adapting, shortening or updating. It ~nvolvcscom-
paring the way people rate on the new measure with how they rate on
yell-established rncasurcs of the concept. If ratings on the new measure
External validity refers to the extent to whicli results from a study can be match those of an established measure we can be confident of ib validiiy.
generdlized beyond the particular shtdy. h study may I-rave good Criterion validity has two limitations. First, it requires that the
irlteinal validity but its value is limited i f the findings unly apply to the established benchmark is valid. If the benchmark is invalid then i t is of
peoplc in that particular investipltion. The critical question is whether littlc valuc in assessing the new measure. Second, there arc no estab-
I
the results are likely to apply rntrrc widely. The mosE cnmmnn threat to fished nleastlrrs for many social science concepts.
hrrnt,t~tncscrrtcriun Krtwys c i n 1.~2 u.rvcl t o a>-*c\s ct~tr.rirvn <'oli~lit!. ~ J L ' ~ t ~ t t ' ~\ V\ .I l I C r l III,..--
In<tcnd (IF cnmp.lrmg n rnt.a<urt. a ~ a ~ nan c t e\i.;t ~ r 1)runchmark
l ~ mcaqurr, changr c,r just rncasurr~nmt'nt>ise'?
thc. nrw rneaqtlrt can hi1 tri~lfed.T.or eramplc, ;l nrw rncawrc ot manta1 Unrelinh~litvcan stern from manv svtrrccs. Prior question rvt,rrl~ngmnv
happinew could be tnaPIctl nn couples w h n seek rnarit.11 counseTlin~.We cause a r e ~ p o n d e n tto unrlerstand k h ~questicln drtfcrrntlv nn diitervnt
would cxpcct that t h i ~g r t ~ u puE cottplt.~~ , o u I dn n r n ~ a l l vobhin low occasions. IXffcrcnt inttrviewers can clicit ~ l i f f e r ~ nanswers t from a
scores on a I alitl mt-asurt. of marital happiness. I f thc5e couple\ ~ c t u ; l l l ~ ~ person: thc match nf agc, ~ e n d r r class , and rfhnrcitl; of an interv~crr+rr
obtain high %cormon tl>r.mar-tal happiness mensure rue ~l-rrlrldprohabl y a n d interviciver can intlucncc. rctsprmses. A q k i n ~cl~ir~tion.; about which
\ran1 to rj~,eqtion whether thc rnrasllrv w6w rc~FIy tapring tnari~al prople have nt? opinion, 1m.e inrufficicn t ~nforrnatin11 or requirt. too
happincr;s. precise an answer can lcad tn unreliable data. Thc answers to somr
i c l ~ trbnf
r i~lirlit!(cvalua trls h(jw well tht* mrasrircs tap fhc differcxnt q u r ~ t i o n *can be affpcted hv mood and bv tlie particlilar context in which
aspt*cth 01 thc. ctlnccpt a\ sve haw* drfinc-d t t . h Lest of arithnlc~ticskills tliclj nrts askcd.
thEitonly Ierted st~htractinn?kill$ would cli~arlvnot ht- a valnl rncasurr~)f Mca.iurt-s ncvd tn bc both t,alicI and rcli;il-If. . I \ E t l ~ n r ~the50 ~h two
arithmrtic skills. sirnil.arl\~n meakurc o f marital happiness [hat only concepts are related they arc not the qarnc. A mcasurr can bt- rc*lbable
nskcd about t h r frequency of arguments hetwccn partner< would pmb- withotit being v a l ~ d Thnf . is, a measure can be cnns~stt-ntlyrurnng For
ahlv lach cont~nt valrdity unlesq we had deflncd marital h a p p i n ~ ~ q cxarnpic, people cnnsistentlv undcrcstinlate thcir Icvel nl ,~lcnhoIcon-
simp34 as lthc abscnce tlf arg~irnents.Fvleasi~r~s of marital happiness wmpticln in q ~ ~ e s t ~ o n nsuwevs. a ~ r c IIEcnhol consumption rnvasures arp
cottld a l w incltrdr tlic nahlrc o f t l ~ careumcnts, Ieisurt, ;lctivities shared rt.liable but do not accuratclv tell us al-rnut thr~ t r u e Ic.\.tl n! nlcolrnl
by partners, cnmrnunicat~cm,method': cd reqr,l\-ing crmfl~ct,the quality crf ctlnsurned.
thv <c.u~~al rclation~hipetc. Mea.;lirc< rvitl never be p~rfcctlyrclial-rlc anti pcrft~rtl~f valid. ?IIc'~c ilre
C11.thn ~ l i ~ a ~ t ~ ~t tabr3r1i n 'ctmltbt~t' r l t manv snil;l l tcicmctb cnn-
~ m r . the ~ I r~'Ii,ll~ilitv
nut ,ill o r ntrth~liyconcepts ;lnJ thr g-{r.ll iq tcl ~ I ~ I X ~th~b J C and
r-rpts i t C J I I bc. diffirul~to drvclnp rncaqurr15 that havr ncr~w-lvaliditv. valid th.If !hew aspects of measurtnmilnta r c then tltr rcsu Its nf the
r-:vu.n il i r e can agnx*iin the ronrrpi ~ n rnpn<urc. d tt ti sin^ a whn!t- bnttrry s t u ~vi that u ~ e sthrm might plnitsibly 2 7 ;i!trillr~tcr!
~ to poor mra>tlrernent
vf qucstinnq, w t then face t h c prrrtrlcm o f thc ~r'!nt!-r*r ~rtlplrtnnr-co f t h r rathrr than tellinfi u s a n y t h i n ~abortt social real~ty.
vnsitru.; cornyrlncnts of tlw rneaqurc. TJilr ruarnple, s l i t ~ l Mmtxapilws of
thc lrequcncv of argument< bc as important a.: Lhe nature crl t h r arKu-
m m l r , the method of conflict rcsolut~on,the qtvlp nT cornmunicnt~onor
statumcnts about level uf subicctivt. marital satisfnctmn? Error can take different forms and the conspquences of error will vary
Corish.rrct zanlirfify rclirs on seeing how WCII thc rmlilts rvr n h t ~ i nwhen depencllng nn i t s form. These forms of crror art, rnlrdrwr, r-on4nllt a n d
~ ~ q i nthe g rncasurch f t t with theorcticai mpecfatinns. To validatr n ru rrrlnted.
men\urc of manta1 happinpm we might nntic~yatt.,nn the hasi-=of kheo?, R R I I ~ J ~ I I~'rrnr
II i~ that which har no s y t e m a h c form I t meam that in
that happiness will vary i n prcdictnhle wavs accnrJing to sbgc in thc 11fe some case5 a rncawrcrncnt for a variable might be too l o w whiIc in
cvclc. I t the result5 of a studv p r n v i ~ i ctlnhrmatir~n,
t thib coulcl rimflect the others it is too high. The mcasurcmcnt nf somennta's w e i ~ h tmight
validih. of o u r rnpasurc of marital happine~q.HonPevcr,Ihiq approach to display random error. %metimes peoplc~undcresf matr thc,i r w c i ~ h t
a ~ w s s i n g\ralidiv rrfies on the correctness nf n u r cxpectakions. I f o u r while others may overestimate it, but il: thme errors are randnm thew
thcrvy i.: not supportrd this could be for nnc of two reasnns: tlie measure will b~ thc samc number ol ovcr- and undcrestim,~tesand the s i ~ of e tlw
of marital happinew cnuld bc wrong nr thta t h e o v a p i n \ t which thcb s he thc same on avcrngr as the ~indcrwtirn~lte.:.Whcn
o ~ c r c s t i m a t ~will
rnc.isurc i< h r i n ~ btanchrnarked mav bc ivrnng. the avcrAgp (me,ln) ih calculated Fclr the ~ ' h o ql c~~ l l pit wtll b ~ ,~cc~tratt.
.
TIicrta is n o ideal way of asstsss1nK v a l i ~ l i t ~1f. a rncnsurr passi,.; ~ 1 1 bccaure tl~cut-crcstirnntfi , ~ n dthe undcrestirnatrs c,incel each other ottt.
threr, tt3stl; i t f > more Eikclv to bc. \-,hid hut \vtn cannot bv c r r t ~ i n .In thc Furthermnrc, tliesta mis-estimatet arp not cr>rrclatrd i v ~ t hany other
hnal ;~n;llvqr~ ~ t h r r.,~ltditr~f our rnc;lsurc.:
lrc, ~villnrr-rl In ! ~ t . y r rfur characterirtic t e . ~ ~. c n d r r 'l~c)
, but ,isc i r u l ~ random.
. I3rcauw random
emur doe< nnt di+tort mcanq and is ~rncorrelatrd1z71tf1other Factr>rs, i t 1s
R t n l m l l i l i f ~ t A rrl iahle rnca.;lirr 15 nnc that wfit-s the wrnc 'rending' whrn !u.;s sersotis than other fnrms of crror.
u.;ed nn rtlpca tmi occ-astons. Tor cxamplt%, r7<< t ~ r l l i rtl~r.rt-
~ ~ r r r r r c !TO ( 7 ~ - fti01 Cnn.;tnl~trrrnr occurs rvhcrc t h ~ r cis the sdmC error for cvtlry caw. For
t,lmrr,v(-, r~liat.rlt> rnr.1wrc3 of rnarit~3happine.;.: sl>nuld !*ic.lrl tlic 5anw rxamplr, i f Pveryonc ~~ndcrrcprrrtedtheir wcighl bv ;ki l ~ ~ r ~ l l m s
. . . , . -- -*-.-'
111fiermt c x c a r i ~ t w .A thrrmometcbr that rnpas- would have constant errnr. Such error is unccx-n.latcd zvith ibther ~-lisrtic-
I"='c bhp n v x t wrr~ald tcristic.;. Although purely constant rrrrbr will he rarc* thtv'c wiil he
WHAT IS RFSFARCJ I I3ESlGN7 TCK)LS FOI3 RESEARCH DFSIGN 33
variables for which thrre will bpcally be n crwiponenl of constant et.ror Finally, thrw corc concept< that are a t the heart of guod design wcre
(c.g.ovcrstaicmcnls of frequencv id sexual intercourse and undcr5ta te- discilssed . These were the concepts of internal validity, external validity
rnrnts about the amount nf alcohol ct~nsumcd).Because such error is and measurement error. Later chapters will evaluate the various designs
constant it Jws not canccl out but has an effect on sample estimates. using time concepts.
Thus the averagc weight of the sample would be a n underestimate to thc
extent of the constant error.
Corret.lotcd error takcs place when the amount and direction of error
vary systematically according tn other characteristics of respondents. For
example, if women tend to c~vercstirnatctheir weight while men under-
estimate thcirs then this error wouId bc correlated with gender. If the
format or lang~iage in a questionnaire is ditficult then m i s t a k ~ sin
answering rl~iestinnsmay well hc correlated wit11 education. This wou Id
produce results that make it appcar tliai people with diffrrenl levels of
education hehdve or think differently wliilc in fact i t is only their
capacity to undcrstand the question that differs.
A crucial gcml of the design and administration of survcy instruments
is the minirn~zationof the various forms c~frnc.asurempnt error. Achicv-
ing this cn tails paying carcf 111 attcn lion to question wording, indicator
quality, interviewer and observcr training, and to ways of identifying
social dcs~rabilityreqponses ancl ~ t h t ' forms
r of dclibcrate misrrpresenta-
tinn bv respondents. rn many cases it is difficmlt to identify the extent to
which such errors actuclIl!! occur. However, this does nut reduce the need
to d o all that cmc can to minimim their likclihoc>d and to have built-in
checks to identify soinc sources of error. Such checks includc looking for
inconsistenc~csin answers, tising multiple questions rather than single
question? to tap concepts, identifying social desirability response S P ~ S ,
ma king intcr-interviewer chccks and careful field work sapcrvision.
Summary
D r p ~ ~ i d c wrmrinblc
i ~ l ~ u bc. t ~ say that a corrrlation
~ we
s t ctapnhlf of ~ ~ l l o n If Figure 3.2 An indirect calrsnl relnfionship
hehvccn two ~ariablesi s because one is causing the other, we must make
sure that the dependent \ nriable (the effrct) iq cnptrblc o f being changcd. I f
it cannot bc changed then a cauqal accrmnt c ~ tht. f r ~ l tionship
a makes no a) Long causat chain
sense. For example, any causal relationship between sex and income
X--tZ--tW+V--tP--tY
coutd only be in thc. direction of sex affecting income. The opposite
prr)position (income-sex) makes nn sense.
b) Multiple indirect paths
Tl;~orctiinlplnuqihilify Thc causal assertion must make scwso. We should
b r able to tell a story of how X affects Y if we wish to infer a causal
relationship behveen X a n d Y. Even i f we cannot cmpiricallv demonstrate
! ; D T ~X
F nffecls Y we nccd to p r o v ~ d e
d piat~sibleaccount of the connection
(plausible in terms of other research, cu r r ~ n throrv
t rtc.). Fnr cxan~plr,to
support the assertion that sector o f crnploymcnt affects achievement
oricntat~onwe might arguc that the private sector fosters the dcvcl(~p-
ment of an achierement orientatinn by strategi~ssuch as paying per-
formance bonuses, developing a culture of Fligl-rcr expectations, providing
better resources and creating less job security. When backed up by this
t y p ~of rcasrlning, any cnrreIatinn between crnployment sector a n d
achievement oricntatinn can bc plausibly interpreted in causal terms.
TYPESOF RELATIONSHIPS I N A THREF-VARIARLE MODEL
Types qf cnrrsaf pdterns
Anv relationship between trvn variables will consist of two components
~ I R E C TA N D INDlRECT C h U S A I REl.ATlONSHJPS
- a cdusal component and a non-causal {spurious) component, The
causal component can consist of a direct component, an indirect
Cauqal relationships can be cithcr direct or indirect. A direct relationship component or both.
is one wherc we assert that the cause affects the outcnme directly rather I t follovvs then that any relationship between two variables can be
than tFiaother variables. An indirect causal relationship is one whcrc tlie int~rpretedas:
t operating via its lnflucncc on another variable that,
cause has its ~ f i e c by
in turn, produces the cffcct. 'P'l~evariable through which the hvo val-i-
ables arc. rulatcd is called the i~lterrtrrllripvariable: it comes in time a n d in a direct causal relationship
n causal sequence bt.tcuvcn t11c inikini cause and Il~c.rffcct. For examplc,
an indirect causal relationship
n spurious relationship
we might argue that the way the priv,lte sector produces higher achieve-
ment orientation is by making e J t l ~ ! o yfear ~ ~ for their jobs (the intcr- r any combination of these.
vcning variable) (Figure 3.2).
Tndtrect causal relationships may be simple (as in Figure 3.2) nr consist Figure 3.4 il lustratcs the possibiIities where we have three variables
of an extcndcd causal chain or a number of diffyrent causal paths ( F i p r e which, for the purposc of the example, T will call X, Y and Z. The
3.3). I relationship between X and Y could be any of the following:
{,I 1 LJ~nlcl~ r r r r c r r l \ 6r1llr)zv.;
11-1 ti tnt,, I I\ c a p d l ~ l tot
~ Iltl~ngc h . l n ~ c dand
,
i r jk plau~ibtet l ~ a tN c i l ~ ~ tydr t d r ~ c cch;lnqt.s in I In lIip ab.;i*nse t v (
finrttng anv nthrr \.ariablr that is ri5ponsibI~for this rrlation41ip ~ v t .
may cmtinue to arguc that thr vbwrvPd ctjrrclGlt~on IS dircct n i i d is
causal ( F i ~ u r 3e 421.
(%) firrJlrrr! r f i ~ ~ s n lY fcrllo\v<X in time, 1' 1s cLip~blr nf being changetl,
and i t is pla~~sible that X cnuld producr shan~t-sIn Y. I-lclwt.\.cr, In
this case we .Irtl syt*lliiiqout the n ~ e c h n n ~ vbvn which X affcctq 1'. Wc
m;ly think of Z a< a 5inclc mcctianisnr, or a wt~nlclot of i n t e r v m i n ~
varlablcr; (Figure 3.4L-r).
I I ~~ n dY arc not causallv r r l a t ~ dto nrlr annther. Even
(c) G ~ I I I T I ~ I ,\C
t h m ~ ~Y'hmight follc~ir.Y in t i m ~nncl bib sapn'tllc of l l c ~ nc h~~ m y c d ,
both t' and I a r e joint c f f t ~ t sof somp third variable Z.X ;lnJ ) (-0-
/-CLJ I l d i Direct and nblrecl causal r@lat~onrn~p] vnrv purr*ly hecauw Z Iias a simultnneorrq c*ffeck on both S artd S
(Figure 3 4c).
(ct) E,)fI, rf!r,rloitcl i t ; d i r c ~ - t The cffrctof X may bt. partly vi.1 its cffect cm
an i n t e r v r n i n ~variable and pnrtlv d~rcct( F i ~ u r .3.+Ld). c
Eta) Prrt-rt nr;d zlidrvt7c.f ciirlwl nrlri sljrrrrtltr.: TIIV rvlationship betwervn S
and I'coutd cnnsiqt of threc mrnpclncnts:,I dircct 1-at13alp;rrt ( S - . Y ) ,
-111 intl~rcctc-ausal pJrt -1 - Y ) . ~ r l < l >, \purriril< part (Y. Z .: )
I te) Dlrect and lnd~rectcausal
[Frgtrrc 3 . 4 ~ ) .
relahon5h1pand a sp~lrlouscomponent
( f ) L)lrcLt (t11(.<01 ri-hti(v14ti~) ~' {r
t * o r r t l > ! t t f ~:it1111 ~ / ~ ~ lL ~. Oi ~( I v~ ~ ~~(F:iqt~re
[ ] I I ~ ~ / I !
't.4f).
( 1 ) Splrnous (nnn-causal) and r11r~cT { g ) Imlirr,cf crltr,wl ridntw~nshrprorr;brrit+rl'ilr~fiin ~p~irlo~i': c t ~ ~ p ~ i r r ' r (Figure
li
sal relal1on5hlp of X w ~ t hY
3.4~)
-
nred to 5pecify:
1 ~ gS~urtous
) (non-causal)and rndlrect causal relat~onshtpof x W I ! ~v J
1 whethrr selatinnships arp presumed to he causal or spurious
2 whether cauwl ri*laSicmshipsart- expecttd to be direct or indircct
? the mcchanr~ms (inten-rning \-,lr~;rble.;) unrlcrl!.fng anv indirect
ca~rsalrc1ationship.i.
[divorce -
Propositron Dlvorce leads to emot~onalproblems In young children.
emotional problems of children]
Propostiton Dlvorce leads lo emot~onalproblems m young ch~ldren.
Propusrtion Private schools produce high academic performance among Proposrbon Private schools produce h~ghacadern~cperformance
therr students. among their students.
[pnvate schools -. high performance]
Restated proposition Private schools will have hrgher levels of acadern~c
Observal~onThere are many instances of studems from private schools
achlevlng h~ghly. -
performance among their students than wrlfgovernment schools.
[school type level of academ~cperformance]
[unemployment -
Praposrtron Youth unemployment IS responsible for youth suicide
hrgh suic~derate]
Proposdion Youth unemployment is responsible for youth sutcide.
I t will be recallcd that when we say that two variables are related it
research study is providing a compamtivc frame of reference. Dfffcrcnt means that variation or difference on one variable is linked to differences
research designs go about the task of providing compnrisnns in different on the other variable. In this itlstancc we have two variables: (I) parents'
ways. marital status (married versus divorced) and (2) child's emotional
a d j u s t m ~ n t(low vcrsus high). Tf our research design incIuded only
divorced parents we could say nothing about the impact of divorce on
children. Even if 100 per cent of the children from divorced parents had a
low level of emotional adjustment we could draw no conclusions abont
By making comparisons we provide a frame of referencc within which to the impact of parental divorce on childrrm's emotional adjustment. We
try to make sense of particular findings. 'me importance of this can be I would need to compare the adjustment of the children with divorced
seen in khe iIlustratio~~s in Figure 3.5. In these examples the obsesvatinn I parents with the adjustment of children From intact families. The
in each case hardly provides convincing support of the propositii~~. One d$krmcr in adjustment levels in the two groups will provide strong
'reason for this is that there is no frame of reference within which to make evidence regarding the alleged effect of divorce on children's emotional
l
sense of the observations. Thcre are n o groups wit11 which to cornparc adjushnent.
and contrast the observations. It is only hy making rnvrpnrisor!s that 'The propositions in Figure 3.5 only mtlntirln one particular group
o u r observations take on much meaning and we are able to clirninnte (children of divorced parents; private school students; suicidal youth).
al tcrnative explanations. It can be helpfu! to restate thc propositions to highlight the implicit
The propclqition that rii-rlorce leads to e ~ n u f i t ~ r?lnlhl~rns
r~! in ! / n r i 1 7 ccl~il~~rrrr
~ cornparisnns. Figure 3.6 restates these propositions using explicit com-
wcluld encourage us to expcct lo f ~ n d emrjtional problcms among parisons.
children whose parents a r c divorced. But finding such evirlence will not
get us very far. To go further down the path Jt explatlatian we must
MULTIPT~E
COMPARISON GROUP5
make compasisnns. Arr emotional problems greater than, the same as or I
less than those among chilcfren from intact farnilics? Do the obscrvahorls The independent variables above (parcnts' marital status, school type,
among children of divorced parcnk simply rcflcct that ~ ~ l i i cwe h would youth crnyloyment stahls) have all been treated as two-category vari-
find among any group of cliildr~n? 1 I ables (dicIintomies). However, we are not restricted to comparisons of
hnt~r\.~ r h ~ cuth the ~ i i l t c r ~ ~ itr..l\ ~ c t ~rt,~pon+il,!r
< I t v an!, cliltclrct~sc\ In .I+*. Ii'e could ihcn ctmtp,lrc t h t ~ s rgroup.; and sec i f t11c.r~rt3r,rr differ-
,~c,~~lt.rni,- a ~ - l i i t - ~ c t'~ n c ~ ~ i r n i v s in their I ~ v c lof acadrrnfc ncliirv~rncnt.
C'on~pari\cm\ 11f chi Id rt*n frcrnl ci i\-orcc*ii ,lticl 1nt.7~tfiimilrcs ,Ire. T!icre .lrc many prilblcrns with thiq approach The trbviotts clnv is Lhnt
cnmplicnte~lbv tllc lac! t l l ~ tt~picilllvc F i i l ~ t ~ (1I . ~ nd ~\~orccil yarcmtk arc rr7c must have thr r r l e ~ ant information rln nrhich to niatch. 5incte thiq
olcler than chlldrcn from intacl f n r n ~ l ~ eAny s difference< bckwren the will alwavf; btl Itmitrd, t h t groilpq arc. likclv to remain i~nrnatchedon
lcvcl of cmtrtiunal a d l u . ; t m ~ i ~olt thc two s ~ t . ; of c h ~ l d r e ncol~!dbtl diic to important t w t u n k n t i ~ ~taclor.;.
n A n t ~ t l ~prol-rl~m
~r i s that many casrs
.igta diffVrcncc<rathcr thdn t l ~ cm a r ~ t a lstatus uf t h e ~ rpnrrant.;. Crnm both group\ s~rnplvwill not have matches frrrrn thr% other Rroup
rhr s;lmt, prohlrrn can arisc. 1vht.n lor kin^, st tl~rstimc grcwp or and will have tn I*c ~liscarded.1f WL- match nn more than a smnll numbcr
cntrgny of peopIc o\.er time. Icicnl l y rvr th comparisons over hmc thc c~nly of factor\ wc r ~ i loften
l end with only a very small number of people
d i i l ~ r t ~ n cbetst-vcn
~.; thc pre-tc5t 2nd the p o q t - ~ t ~\houlclt b r thc ci7cntwe in thr s t u r l ~slncta the nurntrcr nf cilse.: from both group.; that match c.,in
artb pr~jpurinqa* thr r;iu.;c nt ;InV c11an):c. M o r v t - ~ ~ man!. r, crt~ntsc,in Lit, cluitt. small.'
c i < U r r v l i i t - t ~<.In aiiillint frlr Ihc changc ~ n d tI1c.r ma). conf uqc oilr
Cilnl~TlWtlq2nd ntir m~tirib~ttir)n of 1%hcit lies behind Ihc chnngr i r i lie
rr~~tcrvnt. v;in,~blb.[wc Chnptt-r -1 for Furt hcr di>c~rs<iuri)
Since corn!>ari.;rlnh arc. ccili rl71 t(l g(,nEi r c s ~ a r ~dckicn I1 si0cmuqt cn511re A simpler dnd m o w ~>fft-ctivc wav ~ ) ma f kin^ ~roup:: comparrzblt1 i s to
th,it tfir1v art1 mraningful r l ~ t y rnorr- MT car) rt'rnoIrv u t i ~ n t u n ~ l rand d randornlv nllocntp ppnplc to diftenwt ~ r ~ u at p sthe h e g ~ n n i nof~ n study.
unknown clilfcrtbnctaqtlet~v\.ct*ngroupr, tlits n1tx-t. n f c rcdllcta tho rtqk ot Rv randomly assignjng inditidtrals t r ~each g s ~ i u pany d ~ f f ~ r c n c c s
m i s t , i h i n ~. ; y u r r t ~ urt~lation.;liips
~ for cnus;ll rrlaticliishi~.;- tli,it i?, thc h t * t ~ t ~ uKroup.;
n l d random rather than svstrmatic. 5 3 Ion)= a s
s l ~ n t ~ be
mrjrc wrc3t'linl~nnte~ l t t m , ~ t i pIiiti~iblc \~e cal>lat~ationr;. grotlps ;lrchI.?rgc r t i o u ~ h random , assrgnmvnt s h ~ u f datit~>mntically pro-
r!~c.c' g r { ~ ~ ~ p cr i i hcrvny;lrrlbTt~p+ufilc.c: clil Lvth Irnor\.n mil 11nLnown
l-kit3rt-C I ~ L * tt~urT I I ~ I ~\ I IrLitt3~it*\
for T T I ~ ~ \ I ~ ~ I / t~ hI c> ~K~ ~ m i p ~ i r ~ ~tjlb i l ~ t \
S"""3. t,ritcrn. Fron~a qtn!ist i c ~ ~! ~ ~ r ; p ~ ~ t~3nCjo1n i 1 * t , , a<.;ignment of ycrrplt, hr
qrotl p%12111 111~1kv iht' grt,t1~,1;lilcl~t~cill for rll E intents anti pt~r\'oqt.s .ln~J
yrol idc*s 13,hat Dnvi.i call< '7hc all purpow kpuricwitv in5uranrc of
rCinrlnlni7atic~n' (IQK4: 95). \Wt> cotitrrd tor .In ~ n f i ~ i i nt ~uqm b e r of plsnciblr~
W1~t.nrwruitinq thc xrt)tlp< w v c.ln dcIil-rt.ratvl\. match then1 on rclel'ant ~ P ~ C R ) Iw of t l i i 7 1 1 l nrib(Camphell, 1'W.
rll-al h v p o t h r s r ~rr*ithour S J I P T E S I ~7l*lrai
c h a r a c t e r r ~ t i c ~TIIIIY
. wlicn cc*mp;lsing sludrnt.; a t t ~ n d r n gpj\.ernnlml Thiq 15 thi* npproach takcn in d r i ~ et r i ~ l s .Individual4 are randomlv
qchool.: on lllr onc hand ,ind fvc paying prir.,itc ~ ~ 1 1 1 7 0(in 1 ~ thta ~ ~ t l i r r n s s i ~ ~ cto i of severaI Eroups and cach g r t l r ~ pis thcn R ~ V P R;1 d ~ f f e r -
~ ionc
11rl11ij we s h o u l ~ lenwtre Ilint the two st44 ~ l .;luclPnls f are similar in tl.1.m.; t>n! drug. Sincc the g r o u p shoi~lcih,lvc v l r h ~ n l l yidentical profi1r.s tu
IBF intvlligcnl*c, ; ~ s p + a t i r ~ n sp,lrcntaI , rcquurctbs,t-d ucntinn vnl~res,fatnil v hegin WI th, any difftlrencps in outcnmes bcttvecn the groups should be
Ilistr~ry,gentler and agc. R v comparlnf; like with I ~ k ewr. stiuuld be ablo Jut1 trr the d~ftcrcntt r ~ ~ i h n e n ndrninistcred
ts to rach group. kkI(>~'t'vt'r,
to ~'inlatcthr ~ I f e c tnf tlie tvt* o f ssl-rt,clf on acndc.m~cachirvrmcnt. this approach rs r ~ f t c nnot n p p l i r a b ! ~in social res~nrchhefause practical
7'he ptntrltm in i o m p a r i r ~ glikr with lihtl is tcl e<tahfisl~the idtantit]; of and cthicrrl considerations prr-cludc 11.; a r s i ~ m i n gprr3plc !I>groups and
all thc* t*asiahli** on whE1.11 M'C nvecl b match the grnttps. \Ye can match thcn dc~ingsontetlring tn one p o u p to sec what effect i t ha< (wc Chaptcr
lor fhc charactt.ri>ticc ttmt w r k f ~ o i cnirght ~ rcmtattlinat~~ o u r restillc btlt 5). 4~lrantagc.;and dra1vback5 of thiq appronch tn social rtsearch nre
Ihcrc hc o t h t ~I S C ~ C J T ' ~ that 1t.e have 1101 t h t l u ~ h of.
t d i q c u ~ s e di n C'haptrrs 4 and .'
1 T k r rr11tr~br7r of pn~rrpsi l r tlrr desi,y~r Designs will vary from those with
Method of
no curnparisons ( r . ~single . case d c s i p oi casc study) to those wit11
2
many different cornparison groups.
T/rrb nunrbf>rof 'prr>-tr,5t' r n i ~ a ~ r r ~plrnsrs ~ ? ~ ~ r Designs
)~/ vary from those
1 allocation to
groups
----:
(XI
Random
(one group)
11 'Treatment' Measure on
Outcorns variable
-- I
group makes it difficult to know whethcr the intervention or some other
I "on-random 1 None No 'Ireatmen Mea- on 1
outcome varlable
TIlc basic elements of thc cross-sectional d e s i p arc as lallows: Case study designs might consist of n single cd5e study (e.g. a com-
munity study, a study of an organization) or a series of c a w 5hrdit.s with
1 Instead of interventions the cross-sectic?nal design relies on existing perhaps cach cnsc testing a theory from a difftxnt angle. It is useful to
variations in the independent variable(s) in the sample. think of a case study in a similar way to an experiment. We do not finally
2 At ledst one independent variable with at least two categories is reject or accept a theory on the basis of a single experiment; we try tw
present. replicate an experiment and conduct it under a varicty of conditions.
3 Data are collected at une point of time, Similarly, a case strldy project that entails a single case study is ana-
4 There is n o random allocation to 'groups'. logo~oustn a single experiment. If similar results are found in repented
case studies, or predictable differences in results are found for particular
This design mirrors the post-intervention phase of the classic cxperi- cases in thc study, then we develop greater confidence in the findings of
mental design but without any random allocation to 'groups' being the cases in the same way that we gain confidence in experimental
made. The data for this design are collected at one point of time and are rcsults that are found in repeated experiments.
analysed by examining the extent to which variation in the outcome
variabIe is linked with group differences. That is, to what extent do those
in diffcrcnt categories of the independent variable differ in reIation to
Summary
the outcome variable? Causal relationships are established by utilizing
statistical controls rather than by random aIlncatinn of people to groups.
This design is illustrated in Figure 3.10. Tn this casc thc 'inten~cntion'is This chapter has ~ x a m i n r dways of strircturing rcscnrch designs to help
simply being in a different category of the independent variable. draw convincing causal inferences from the research. Since causes
cannot be 01~t)rxed they must bc infcrred from obw-vations. However,
rncorrtct Inferences can casilv he made. The chapter has considered
ways of structuring research to improve the quality of these infercnccs. A
Case study designs rely less on comparing cases than on exhaustive rangc of ways of interpreting correlations between variables was orrt-
analysis of individual cases and then on comparing cases. A distin- lined and criteria required for inferring that a correIation reflects il causal
guishing characteristic nf case studies is h a t contcxtual information is connection were provided.
cnllected ahout a case so that we have a context withiq which to under- Tl~cchapter emphasized the importance of making meaningful corn-
stand causal processes. pnrisons behveen groups as a core clcrn~ntof drawing causal inferences.
W H A T tS RESEARCH DESTGhr?
Notes
TYPES OF EXPERIMENTAL DESIGN
1 R c r c can be exceptions to thi5 where a suppressor variable may he I
operating to mask a causal relatinnship- See I I o s ~ i ~ b ~(1968:
r g Chaptcr 4).
2 We sl~allsee later that this pmhtem of Ioss uf cases can be overcome with
cprtarn fcorrns of multivarlatc analysis.
1
3 A nalurnIly o c c u ~ring ~nlervcntion1s an event. llint takes place b~twccnpre- I
I
tust and pocl-ic.;t w ~ t l ~ to utl-te invcsiigd tor initialing a n y intervention. An This ch(1ptc.r builds on the elements uf rcscarch dcsign outlined in
example mtght be a study of fnrn~l~es over tilnc In which wrne fam~liescxperi-
Chapter 3 to intrt~duccn variety of types of experimental designs. It.
ence parental divorce. Divorvc is the 'natural intr wcntlon'.
outhncs thc diffwen t environm~ntsin which experiments can be con-
I
ducted and describes both simple and more complex variations of
experimental designs. This chapter does not seek to evaluate experi-
mental designs: that is the task of Chaptcr 5. The discussion also pro-
vidcc; a useful framework within which to undcrstand some of the
strengths and weaknesses of thc research designs discussed later in this
kmok.
Random
atlocallon
(expenmental
I
rre-rest a v ~ r a g e
desrrab0l.f~.SC-P
'Severe
rnrt~atlon' 'Po5T-tr>r ~vareuc
des~raol:,'ys:o?e
Erhr~nqs= €2 - El
, group)
initi~tion'inkc) tlit, grnnp as thc indicator of diflicultv oF join in^. Tlie
cla%%icexp~rimmtal design for such an invttstigation e n t a ~ l s ti111
i r j l low in^ qteps: Random NOnttaton C, Cchalrqe ' -~ 1 1
P ~ t e saVerRgP
I Post-test averaop
1 RandornEy n l l n c . ~ t i npeople
~ to t n w group: a n ctpertmental group (f~~rr; I
and a contri~lgrnup,
2 Ubta~ningrnen<tll.cs of prrccivrJ ijrsirability of gmup rncmbi.r<liip.
3 Subjecting i h rrp~rimenta
~ l groir p to severp inittnlinn reqi~irt*nicnts
and the control group 50 ntl initi.~tic>nrcquircmtwts.
4 Obtaining rntncurl- nf desirabtlitv of group nlcrnbcrqhip from both
?:mtrp aftcr thtb t n i t i a t ~ u n t * r tlic erpcri~ncnbal grouy ha5 I ~ < ~ ~ ~ I I
~ornplcted
C'nEculrltin~the i.lrrr~;~r in dc~irnhilityof mcmbvrsh ip for thc r.~ijr,ri-
r ~ gmtrp bcfort- a n d aftcr tlir initiation and nt the control gmup at
r ~ r t , tnl
Experimental
T I and TZ. inlt~atron'
h Comparing t t ~ Ic1.t.lr of c h a n in~ ~the experimcnta! group with that in I Mean
Z ~ Pcontrol Ernup. I
i
1
-
I
5 The change in thc cxpcrin~entalgroup is greater d ~ a nin the control Labnratorv cxpcriments have been used to great effect in tl-tc physical
group thy 1.5 points). a n d biological sciences. I -lnwever, for a variety of methodological, prac-
6 'l'herefilre scvcritv o f inihation leads on average to a 15-point increasc tical and ethical rcastms their use in social research has been limited (scc
in percei\,ed desirability nf group membership. Chapter 5).
1 Thc groups are the same in all relevant respects before any inter- Because of the artificiality and impracticality of lalwra torv experiments,
vention (TI). social scientists have established experimental designs in real world
2 The groups cxpcricncc the same conditions between TI and Tz except environments. In a field cxpcsiment the investigator creates cxpcrimental
for the specific intervention/heatment. and control grnups through randomized allocatirm of participants. The
experiment also involves a n interventir~n.The intervention is implc-
I f either of these conditions is not met then there are a l f ~ r n a t i vplnusiblp
~ mented in the 'field' where a rcal life environment is trcated as the
h,ypothcses that cnuld account for the different changes between the laboratory. This approach has become especially pnpular among those
experimental and conlrol groups. The different 1~velsof change could he evaluating policy initiatives and pilot p r ~ g r a m m c s . ~
due to initial differences bctwcen the two groups rather than the severity In order to illustrate the nature of ficld cxpcriments 1 will recount t11c
of initiation. If the two groups differ in other experiences between TI and major features of an American shdy of income maintenance among
Tz ( p . g . mixing with diffcrcnt people in thc group) these differences families of Inw socio-cconornic standing. The study is known as the
rather tI1n11 sev~rityof initiatiun co111d account for the different rates of %a ttlc /Denver Tncornr Maintenance Experiment (Hakim, 1986). Almnst
change in thc h ~ groups.
u 5000 low income families partic1patc.d in the experiment in thc 1970s. The
purpow of t l ~ cstudy was to establiqh the effect r ~ thc.
f introd~ictiono i a
Experilr~entnl r-ontex!~ guaranteed minimum income on the work effort and maritaE stability ilf
Experimental designs can be implemented in three different ways: in a participants. Families were randomly alltlca ted to diflcrcn t versions of
laboratorv, in the field a11d by utilizing natural clccurrences. the scheme (no income support, short ttrm support, throt~ghto long
term guaranteed income). Among the many variables measurcd were
LABORATORY
EXPERIMENTS
work effort and marital stability. They were measured at the beginning
of the scheme and then at various points throughout the experiment.
The ldboratory experiment is designed to ensure that experimental and While participating in the experiment these families went about their
control groups are exposed to exactly the same environment except for normal lives and were subject to a11 the varying experiences and
the experimental intervention. This is achieved by standardizing and additional changes (e.g, political dnd social changeq) that took place over
controlling the environment and all the events between TI and T2.Maxi- the period of the study.
mum control over the environment is designed to ensure that the ~ n l y I participated in a field experiment conducted by the Australian
feasible reas011 for group differences at Tz is the differential treatment. Department of Socia! Security (DSS). ?t was designed to see i f the
An example of a laboratory experiment is a study designed to test the participation of welfare recipients {DSS clients) in self-help groups
effect of extremely graphic images of alcohol caused road accidents on increased their living standards. Various self-help groups were esta b-
vicws abnut drink-driving. Participants could be randomly nllncated to lished in 80 locations. Client5 of the BSS living in those locations could
experimental and control groups and asked abnut their views regarding choose whether or not to participate in the groups. Extensive measures of
appropriate penalties for d ~ ~ n kriving.
-d The experirncntal group could be living ytandards were obtained before participation and 12 months later.
shown a graphic film of alcohol causecl accidents. Members of the other In each of the 80 locations a random sample of DSS rEients who did not
g-ruup would nnt be shown thc film. At a specified time aftcr showing the participate in the self-help groups were also qu~stioncdabout their living
fiIm both groups would again bc asked about their views regarding standards. Thcsc people acted as a control or comparison group.
drink-driving penal ties. If the views of the experimental group changed Clearly this was an attempt tc) implement thc l o p c of the experimrntal
more than the views of the control group i t wc~uuId bc reasonable to dcsign in a ficld situation. Ilowever, the design had a serious flaw.
assume that viewing the film caused the cliangc. The laboratory setting Participants were not randomly allocated to groups, so wc d o not know
mabltd the rescnrcher to control everything, so the, nn!y difference whether those who participated in the self-help groups wvre initially
bctween the gi-oups was h e exposure or lack o f exposure to the film. similar to the non-participants. In addition u7e d o not know whether
EXPERlMENTAL DESIGNS TYPES OF EXPERIMEN'TAL DESIGN
participants and tlnn-participants wcre exposed to comparabfe influ- I f WP can antic~patethat an c-r~cntwill occur we can collect information
cnces over thc 12-month period. We observed that self-help group before and after to monitor its impact. To help intcrprct thr rcsults we
participants dici b e t t ~ over
r the 12 months compared to mcrnbcrs of the ct~uldtrv to identify a comparison group that will not be exposed to thc
cnntrol Emup. However, wc could nnt be sure whcther this was due to event.
participation in the self-help groups nr whckher thc sorts of people whu A natural experiment is well suited to s t t ~ d y i nthe ~ impact of
volunteered to participate were more self-help oriented prior to the Ic~alizingpokcr machines on family wellbeing. For such an experiment
cornrnr.ncement of I ~ cxperimen
F t. These initial differences might have we would obtain mcasurcs of family wellbeing before and after the
accounted for h e i r greater improvement in living standards. legalization of poker machines. We w t ~ u l dobtain a comparison group by
While laboratory experiments suffer from problems of artificiality, measuring family wellbeing in a comparable regon, state or country in
field experiments encounter problems horn inadequatc control of events w h ~ c hpoker machines remained illegal. Similarly, a nahlral experiment
between T , and T2.Thurr is always the danger that uncontrrrlled events is well suited to examining tlic impact of the introduction of no-fault
rather than the experimental intervention are producing any obscrr.cd divorce on divorce rates. We would examine the divorce mtes in a
changes. country before and after the introduction of no-fault divorce and
Howevcr, other factors (i.e, uncontrolled events) should nnt be a major compare the rate of change with a cnmparable country without no-fault
prob!em if hofll thp exprrimentnl a ~ llzc
d cntllrol groups ore eq~ially~xpusedEO divorce.
tlrrsc ollrrr influenres. Total change in the experimental group between TI Annther form of natural experiment can be achieved by constructing
and TZ consists of two efements: that produced by the intervention and experimcntal and control groups (Per the 'intervention' has occurred. Let
that produced by other intervening events.' In the control p u p , total us assume, for examplc, that wc want tr) establiqh how women's return
change n d l be due entirety to the othcr intervening events. I f the same to the workforce impacts on the division of housework and child-care
intervening events apply equally to both groups N e can shll jdcnlity between marriagc ppartners. We would collect baseline information about
experimental change by lr~okingat the diffcre?~cc~ in the total change the d ~ ~ r n e s tdivisicm
ic of lnbnur from couples where khe wife is not in the
between the two groups. Ln the carlier examplc (severity of initiation) the paid workforce. We could rehlrn to the same sample sc'17cral years lntcr
total change in the experimental group was 25 points compared to 10 anticipating that some wives will have returned to the workforce. These
points in the control group. The difference in total change between the could be treated as thc exp~rimcntalgroup while tllose couples where
two groups was 15 points - the experimental effect. the wifc remained out of the labour force would constitute the cnntrol
We do run into problems, however, when any of these external factors group. Changes in who did what around the home would be compiired
influenctl one group more than the other. This can occur in at least two in the two groups to assess the impact of the return to the workforce.
ways: The difficulty here is that the groups are self-forming rather than being
created by random allocation. We might be able to match repons,
I One group may be selectively exposed to additional factors more countries or farnilics to some extent but matching has its obvious limits
than the other group. (see Chapter 3). These limits mean that the groups we are comparing
2 The external factors may infmgct with [he intervention. That. is, when a may be dissimilar in important ways. This necessarily compromises om
person is cxposed to both the experimental intervention lrnd the external ability to s a y exactly what is producing any change or any differences
has a unique effect that neither the experimental
factnr this ro~nhi~nztion between the groups. In the gambling example the problem will be
nor the ext~rnalfactur has on i t s own. An example of an interaction finding a truly comparable region in which gambling remains illegal;
effect is evident in the treatrncnt nf depression. Drug treatment on its with divorce the problem will be finding truly comparable countries. The
nwn mav have some limited effect on dcprcssion, and psychotherapy housezvork examplc would encounter the problem that the women
on ~ t own
s may have snme effcct But when the two apprt~achcsare returning to work may bc different from those staying at home.
used together they can have a dramatic impact on depression. The
conibincd impact is much greater than the simple cumulativ~impact of
each approach: it is the mir that has the most effect. Simpler experimental designs
- - -
I
Method of I Tlrnel(i,) Inletventton Time 2 (T,) T ~ m e1 (TI) I ntervenfion T ~ m e2 (T,)
allacat~onto
-
allocation to (XI (X)
groups Pre-test Post-test groups Pre-test Post-test
-- - .-
1
Random None 'Treatment' Retrospect~ve MeasurB on 'Treatment' Measure on
allocatton outcome vanable matching outcome var~able outcome variable
1,
(experimental (experimental
group?
-- -
Random
allocation
(control group)
Figurc 4 4
None No 'treatment' j
-- -
Retrospective
matching
(control group)
-
I
I
I
Measure on
outcome v anabls
. ..
No 'treatment'
I Measr~reon
outcome variable
'
Time 3 (T,i Tine 4 / T , /
nlultiple variation.;. Thcrc may L ~ P qc\cr'~l* ~ l t r r n a t i ~intcrvrntlon.:
t%rallld
tc.
For euarnplc., t\.r may be intrres'led in t h cffcct
~
t.
ot compuI<r)rv votinc
to
1
I:
allocat~on . . >... . , .; 'Treatment' , .. . .,. , ,.. . - , .<. , L : u , r , v r , ~ d non-voting. The control gsorip will tlclicvi, !hat
that will he i r n p t ~ ~for
voting will remain voluntary. After bcing ~ i v e nthis ~nformationprtlpIc
icontro (! in each group will hc nskcd about thcir voting intention. Diffprcnc~sin
, QrouP) thc T? voting intrntions of the five groups s h o ~ ~ reflectld thr c f f t ~ of
t
11
ct~rnpulsionkin thc .;tatid intention to vote. T h i s & s i p 1s d i a p a m m r d in
F1jiure 4.7'.
r -- I -
' !
M~vhodof allocation I T lme 1 ll.) Intervenllon Tlme 2 ( f,l Intervent~on Time 2 (Tzh
lo groups 1X)
Random aflocat~on O.
Pre-test
~nt~ndmnn Compuls~on
wrthout
-- - ";
Post-tesl
on
MPRSU~P
Fe
twperlmentaf group 1 I sanctrons (e~penmenlalgfoupl I Y I
I
I0 VC
I 1 (exoeivnanfal group)
1
(Yi
I
- 1 1
Random atlocatron "'a inter Cornpuls~onwllh
to VC modest frne as ) out~~nin I
{expsrlmental group 3) 1 sanctron (control group) IYI
I
--
- -
Ranrlnm ~llocahon
*-I
".ntclr CornpvTsion wlth
I
",~ n t ~ n & n t l
I , P a n - l o ~n!lr,r7tlon
(con110qmup)
- - --
I
I -
, AS, ,I Nr! 'fr~atrnent' b . l n x l ~ r - 02
Random allocatron
Figure 4.X Solr)~trclrrhrrr-qro~q',lr.;rtv~
(control group)
-- I - I. I
i Method
,o J
of
groups
allocation
l l ; ; ;-; ; tT 1ntey;rlo
Control (XI )
High la\ l o w lbl
Post-fes~
-.
Random allocation # I Mwn-5rl Treatmen
Male (a) Males Mates
(experimental
II I exper~m~ntal qroups
1s 10 IRO mlnuc 7 0 ) Gender
J
I Thrs IS ~nnllut7hlelo (Xzl
I I 1
I lnstrumenr rrnrtlvtty
Randsm allaal~on *Inm , Treafrnenl' Mean=iO 1
(expenmental
!3rou?J) II
Random allocatron Moan-50 No 'treatment' Mean=60 D~fferenceb~!wc.en
control group< IS 10
(control group) (60 mmus 50) J h ~ s
I rs attr~hulahlalo
1
(Figtrrc 4.10). TIIPdcytlndt.nt variable (1')i < jnh snti<facbi*n r n ~ a * l ~ r1311 t~~l
I tnslrument r~ac11vlb a ycale of 0-100, uvith 100 indicahnq vrrIr hrzh sati<f,~clion. 13nc
Random allccation No Treatmen1 I M~an=50 I i n d ~ p e n d m wriabl~k
t 1.; h ~ r e ocf control o~.t?s (me'<~ t w r kv hh>litin t 2't)
rl<~t\rfi~ <imply
d ~ l crintrol Illr* {ither indtbytbn~lr.ntk , l r r -
ac. I u ~ c~ n high
[control group) i nblr i5 gender (?i7).
I A factorial dc*sign entails forming cupcr~niental grcrupq f(*r c-ach
possible conrIli~;~ltio)~ nf thc two indcpr>ntlc.n! variables. ln this casr the
lour R ~ O U Q Sart' ~IIIKC* defined in Figurr -1,lO.Fvrnbolically thc tnro indc-
pcndent v a r i a b l c ~c a n he rcpresentpd a s )ib[contrcll) and X ? (grncicr).
The categories of vacf-r can be repreqcnicd ns 'a' and 'b'. Thus ma 1t.c tvith
For cxarnplc, J lctd~ol.on its own, nffc-crk mood. On their cr\trn, .myhe- high cnntrol can ht> .;\mboli7ed a5 .Y,,Y:, while lonu control I t m a l ~ s
tarnines afsi) affect mnnd. The effect.; arc ninln ~ffcts:Ithey cnch Iiavc an 5vol1ld be symhnli7cci a s xlbS?h.
effect a n tllcrr own lxut the rnixturp (>f ;~lcc>holand arnphctarninc.; can In t h i ~part~culnrcAsr, ~ i n c egender i~ an independent \.arialllt. hlil is
have unique efkcts .ibovc and b e ~ o n dcithi-r alcnhol or ampht~tamines Iinrdly a n 'intervcntinn' we would start with a randomly st~lcctciigroilp
independentty. This will be more than juqt a n additive effect: it will hc a of males and a randomly selected group of fcma les (Figure 4.1 F 1. '1 ( 1 rach
u n i q u ~effect that i.; only produced by tnixiny: tile twn drugs. Sim~larlv, c-rf these gmups wp would tnboduc~tht* 'trpatrnent' of high and l i l ioh ~ ~
giving workers nlortb contrcbl over t h c ~ rwtirk might improve wnrk ati is- control. Half the r n a l ~ swould bc randomlv assigned tn a low cnr~lrnl
laction for cvcrvnnc (main effect) h t it rni~l11have rmlrr. cffcct nn work environment and half to a high crrntrr~l ~ v o r kenviriinmcnt. The
particular type.; of p w p t e such as pmlr.;siclnslq or older pcr~plr{lntcr- same would apply to fcmaleq. 1%'~ w o ~ i l dtlit~shave fmrr g r o u p fvs
action effcc!). tz-hich we would thvn obtain before and aftcr rncawres.
VIE'~ U T L W t ~ tf.lctt~rr;h euperinit*ntal drstgns is simultant~rru\lt. to The number of grotlpc grnrvs i f tit1it.r vLir~al7!c.has motc lli,in !\to
tl\.~rnine t h i~l i r v ~ ~k'ffect5
t trnaln citvcts) ot a nl~rnherut indcptlntlt1nt categories. It ts ptarl1,~psmure diific~tltto v i s l ~ ~ l l w ~ hr ca t \v~)ulcihappan
vnriahles and tcr b t r * how various c.c>mb~n,ition< OF charactcrrrtit:< wclrk with three or mrlrtxvariables. We c a n c l o thi.; ~vrnhnlicnllvwhprt>wi, h n ~ r
togcthcr to product. ,In effect Thc sinlplrst vrbrsiun of thiq design I.; calIt~d three indeprndcnt variahlcs with n v ; l ~ . v ! nnitnlher ~ oi cati*jirlrEcs:
tht. 2 x 2 factor1.1l dzlsi~n,thdt is one b s ~ c on i twn indepcni4tlnt \.arlablcs I
(two categories c,~llcda ,ind b), X 2 (two i a i t ~ ~ o rc~lIect i t ~ ~ a ,Inti b) nnrl N1
cnch with iuqt twrn cat(-xclrics. Thiq st~nplr.? Y 2 d e s l ~can t ~ . i ~ i be l\~ (t1i.n catcgoric-s called a and b). Tkc firqt catcaxor\- (if the first vsri~l,I(ac~1n
cttcnded tn mrirta than t ~ v ovariablcc. cnch (?I which might I l n v v trvr, t,r bc cvmbolr7ed a s S,", and the second cat(-!:(try nq Sib: the h-I s.ittni:t,n. of
marc cate~oriczrir conditions. tiit' C . C ' T C ~ rndey~nrlt*nt
~ variable i s rcprrwlntt~rl.I\ s7,,; ~ n sod on 'f,lblt.
The simyle 2 x 2 d c s i ~ ncan be ilt~r.;tratrd bcqt with an cu,lmplr* in 3.1 i1lur;trates thr n~lrnbcrof groups that cnuld bc proclliccd by c-nch n f
< nne deikndcnt r.;lri;lhlc
whlch ~ v chtir.c htrn i n d ~ ~ e n d c vn at r ~ ~ l , l catill Zlip p o ~ l ; i b rorn2~inaticln~
l~ nf these tl1rt.c- inrll-ycnrlcnt v r l r ~ a h l r ~
Summary
M~~ho 0' d?llncatinn fbme f IT,) 'Interv~nfrons' T ~ m p2 (T?I
70 groups
I Pre-fesi
i ( X , and X,)
Posl-lest Thrq chapter has c1ull1nt.dA v ~ r i e h .of c*\pc.rlm~ntnldesign< af v a v i n g
ci,rnplcrity. Ric 10pc o i drawing c~~i.;nl~ n t t ~ r c n c cfrom
wi tli experimental rlr.;igr~swas out1it ~ ~ d
s data obtalnccf
. various es ycri rntantal
Tllc
Random allocat~on X I , x?, Measure on designs vary in term< clt their use uf prc-ti>.;lr,the number of prc-twt and
i
I
I job sat~sfact~on Males w~th tron p n ~ t - t e s trncn.;t~rcrncntpoints and the nu mbcr of groups cc~mpnred
I 4Y I hroh control
I rvitl~lnthe Jcs~qn. 1\11 cxperirnental r l t ~ s i ~ n wilI s ha1.e ccrmp,lrl<on
I -I Rroupr: and all , ~ t t ~ r n p tot make h c grnrips sq similar as pos~iblr,cvcept
Random allocation 1 ~ e a s XI., X1 h M ~ a s u r eon t r l relation tn t l ~ ct=\p,rimrntal inter\ cntion. r\ltlinugh random ~ l l t r c ~ ~ t i o n
Malcs with lob s;r:rc!actron to groups is fhra t;lvclt~rr*dmethod, tht* rnc,In.; bj. n.hic11 Rrt-lLryq tviFE bc
I ICIW control / (Y) I malie romp-rrahlc w i l l vary dcpcnd in^ t ~ ntht. circumstancr< of tlir
I rt%search(see Cliapirr 3). The riguur of r x p r r i m ~ n t a designs
l will vary
d c p c n d i n ~o n t hc ctmlr1uts within which thc cxpcriment 1s conducted.
Random allocation Measure t
Females w~th job satisfac Lnborntosv setting.: provirlc maximum cnntrol of confound in^ cxti>rnal
I 1' . h~qhcontrol ,Y\ influences whilt* n,ltural crpcrimcnts prtn'idc. the least ci~nkol.
--
s I
Notes
, Ranlnn? ~1loc:ttron yli, X,I,
I
MPRFUVon
job satwtac
1 7111s is <I tncnlifit-,?tion 01 an arpr~nt~nt t ~ a < l t . ( E 111 ,In experiment bv Arttnson
(vl
, ~ n dMi115 (1959) a n ~ lrl~\ruh\c~l 111 Kerlingc*r (E'17')' ' ) I )
2 Untorhindtriv Ihr-sr. p>nFs arr not nlways spi'cifi~da t the hrcinning of thta
~mplrmcntationr l f tht* pilot prngr;lrnmr.; snd t~vaEuations t u d ~ r qarta all too
frt.quently cornmi+ tr>n~*tlrr!tr.r tlitl p~totprngrarnmt. has been de\,f,lc)ptd and
~mplrmcntcd.
7 There i~ also r n t . n ~ i r r 4change that 1s dr~tnto various type5 of rrror (see
C'liapter 1).
4 T l ~ targument h c ~ t ilhaf
~ ns pwplc havt. Oi~ldr~n their ccmc1.m It)r thp
wellbeing nt thtair tarnil\ ii~,~L.rc
them brccmir~morr ronc~rnedahnut sr>rialand
pcll~ticalstability.
f'uptlnmcnb a r r appmpriate whm kfit,rr iq wrnc ar~:trmpnt as 20 thr. Randomization can rps~lltIn the undcrcstimasion of thc tnhd uarrsnl
t.\r-tcnce nf n i a u c , ~ Irtal;i!ir>n5htnb~.t~vt,r.nh' and I tnr.! o t clther canfi.\tt~,~l trr bphvefn thc c\ptrimental i~ r i a b l c~ n tht
rplntln?rrl~r~l d vi~lctlmrvariable.
rt,l.lttlcl h c t o r ~ l tir . l v l ~ r nthe F T P C I L ~ ~nlt4rlct!rcof tht. rl/r .111t1 ij~rectv(\not l f i ~ h the rra l world the rota! c n ~ ~ ctiect
~ n l bctlvcaen hvo \.aria14 t- consists of
~ . t t a = t =I \~ n w d v d I lor(, l O F 1 two rrrm~oncnts:
'I hc. kitrcngth o f cxptlrin~cntsgaincd through randnmizntinn, I~owc*vt,r, m dirwr causal effect
cornrs ,at a cnnsidcr,~blccost. Exprsirncmtnl resvarch 1.; not urell witcad t c l 0 i n d i r t ~ causal
t effect.
~ r o \ * i r l ian ~ r l the
n ~i , . ~ ~ l r ~ r r r ~ l of rl r~<ult.i- to identltv~nr,the mechani\nw
171~ \ r l > ~ i one
h rarr;rt~lt~ ,?ffects anr)tt~er.Yor dcw5 i t allow us to b u t l ~ la Ranclomii.aticrn ensure.; that 111e g r o u p I ~ Vcompare ,Jrcm alikc in aV1
rict~rrilof the crrmplcx ~ r tof factorc t h ~ Ft ~ ~ U C ,given P outcnrnr. relevant rcqprcts. \ T r can only look at thc direct eftcct r l f t h r c F r r l -
E\pvrlrncnts fixu.; ltn the imyart nl' ji~rtone or two fnctrlrs.' rntmtal variable. We cannot lorlk a t the rt3nl world cFfrr1.; crf a v ~ r ~ ainb l ~
v a r ~ a b l envr ~ r ,lnt
, to I?(, \ I I ~ C = t h ~ ilt 1 % d ~ i t t-o tvhat \\.I. t h ~ n ki t is r,ithcr
thin < o r n e t l i ~elw.i ~ ~ 1Vlit.r~the I ~ ~ and I c Gtnlcture (I( a de<ign art. f ~ i ~ t t y
anti f;lil tcr climinnte c-ampcting t'\plnnatiuns nf reqult'i then tkir d c q i ~ n
lack.; ftrkprrinl ?vilrrlit~~.
At1 desiqns !ace thrr,~t<to infvrnal \ratf~i~tv. Carnphcll and StanIcv
( 14t~7)h a w ul~mtifiellthr. thrwt.; disrusscii Iycknw.
g c n e r , ~ l i zexperimental
~ find i n ~ bc.vnnd
s thc spccihc c~xpcrirnrnt. PCRPIC in n*al srxial contcuts. T;ljft,l (11W3) r~mrnii*115 that the ~ . p i c a l
s r r ~ a psycholog!l
l expcrimcnt is hnscd on
Such rncrnory effects can he dealt with in a number of ways. Pilot LIMITS OF RANDOM AI.I.OCATION
twting might cstablrsh the amount of time that has elapsed before
Choosing a small sample can impede achieving comparability of expcri-
parhcipants forget their pre-test responses. Alterna tivcly, if we have
mental and controI group^. For example, randomly assigning a sample of
allncatcd participants randomly and have a sample of sufficient size thc
10 people into two gruups of five people each may easily result in
pre-test can be eliminated altogethct (Chapter 4). A third alternakive is to
dissimilar groups. As the size of the total sample increases the averagc
use the Solomon four-group desipjw. This design overcomes the problem
charactcristics of the experimental and controI groups should converge.
a s it enables us to assess and statistically remove pre-testing effects.
Wl-tere the sample size cannot be increased, the ramdnmized block design
Practical matters will also determine time gaps between stages of the
study. The amount of time availabIe to complete the study and the
can be used to ensure equivalence on specified charactcristics. However,
with large groups there can still he important differences between the
availability of funds at our disposal to keep track of p e o p l ~will influence
groups on variables with low variabiIity despite randomized assignment
the way we finally design the experiment.
(Cook and Campbell, 1979: 342).
The particular research design we adopt does not dictate a particular Random assignment can be difficult to achieve in field scttinp where
method of data collectinn (e.g, observation, interview or questionnaire). practical realities may require negotiating with authorities and other
The data collrct~onmethod we adopt will depend on the nature of the 'gatekeepers' who control access to potential participants. Among the
ubservations wc need to make (Chapter 1). What is critical is that the gatekeepers the researcher is likely to come across are school princ~pals,
method of data collection produces reliable, valid and meaningful data. manngcrs of organizations, and professional welfare workcrs. Gate-
Since the expcrimen tal design relics on detecting differcnccs over time keepers can cause problems by insisting on allocating people to groups
and dlffercnces between groups it is critical that we have measures that ti~cmsclvpsnr by undermining the research by poor implementation.
accurately and reliably reflect change. We must be confident that meas- They might, for examplc, allocate khe 'most deserving' to desirable
ured change is real rather than just an artifact. of poor methods of treatment groups. PdeaFly the researcher should allocate people to groups
informabion collection. but, w h ~ r cthis is not possible, close scrutiny of those doing the alloca-
The need to achieve high reliability does not point us to a single data tion is required.
collection method. The reliability of a method will d e p ~ n don what data
are being cnllcctcd and on the skill of the person collecting them. In some
cases we will achieve much better information by watching people than
by asking them highly structured questions. Sometimes a less structured If people believe that they are in a control group they can fccl badIy done
questionnaire that enables an interviewer to probe and follow up by. Tile resentment that results can lead to a high dropout and produce a
responses might be more effective than a highly stnicturcd one. control group that differs in important ways from the experimental
AS well a s yielding reliable data our data collection method must group (e.g. more negative). These perceptions, rather than the effect of
produce valid information: we must be measuring what we think we are. any interventions, could be responsible for any final outcome differences
There is little use in collecting mliablc data tl-rat are wrong. There is little between the p u p s .
merit in cnllecting data using very reliable instruments on, say, level of One solution is to disguise the fact of being in the control group by
racial prejudice i f thc method of data collcction is getting people to using a 'placebo' - i.e, an irreievant treatment. In dnig research the
express socially desirable views rather than their real views. placebo might be a sugar pill. In social research i t might bc a treatment
that is known not to affect the imtcorne variables and is thus safe to use.
Alternativelv, wc can assure control group participants that they can
have the deqirab3e treatments, sl~ouldthey wish, at she cnnclusinn of thc
expcriment.
Tht. practice of randomly assigning participants is one of the distin-
guishing characteristics of experimental design. For a number of reasons,
however, random assibpnent docs not necessariIy achieve its goal of
ensuring that the exprrimentrtl and control groups ,are initially com- A comparable problem can arise when people allocated to a given
parable. tr~atrnent refuse to participate because the treatment is seen to bc
82 E X P E R I M E V A t DESIGNS ISSUES IN GXPERI MENTAL DESIGN 83
undesirable. Drnpaut a t tl~lspoint would mean that despite initial confident what interxention has actually taken piace and we may thus
mnJom allocation the groups could be different 17efurf the in tervcntion. have difficulties making sense of the data.
This prrjblcm can bc avoided by poqtponing random allocation to
groups (Cook and Campbell, 1979: 358). Having recruited participants
we can ascertain their willingness tn participate in the control ar any one
of a ~iurnber of different experimental trcahnent~. We could then In anv rcscarch there is always the danger that the researchcr's expec-
eliminate from the study participants who were unwilling to participate tations and values will inadverkcntty distort the way he or she collccts
in any particuiar group. Then we could randomly allocate the remaining and interprets information. For example, i f we cxpcct that a partictllar
willing participants to the various groups. teaching method might improve tlie level of children's classroom
cooperation we might kste' more cooperation than if we did not hold this
bellef. A researcher's expectations can affect what they see, how they
~nterpret it and h o i ~l~rrrdthey look for cvidcncc contrary to their
Since experimental designs examine change over time they face the exprctations.
danger of differential dropout from diffcrcnt groUFS ( c . g the less desir- In drug cxperin~entsthis problem is minimized hy using a 'double-
able interventions). This differentidl dropout could be responsible For blind' approach. Nci thcr thc parfirfpan t nor t h ~investigator knows
any post-test group differences (see earlier in this chapter). This call be a which drug or placebo is being administcr~dto which participants. But
serious problem in long term, field based social research. Furthermore, this solution is rarely availabie in social research. Because of the opera-
even where the dropout rate i s much the same across all groups, the tion of thc ethical principle nf informed consent (see later in this chapter)
drc~pc~uts may differ significantly from tl-rost whn remain in the experi- it can be difficult to usc placcbo trcatmtntq. Similarly, it1 many social
mcnt, thuq tlrrcatening the external validity of the studv cxpcri~nentsit iq vrrv difficult to hide i ~ h i c hgroup5 arc suhicct to which
A kev thing to check for is whcther dropouts are dist~nctivcin relation intervention.
to their initial scores on the outcome variable. For example, are those About tlie closest cvr can get to thc 'dr~ublt-bImd'approach in social
with low initial s c r r c s on thc outcnme variable more pranc than others to research 1s to cnsure that the person collecting the data is unaware of
drop out e~therbefore, during or after the experimental intervention? rf the researcl~ hypothesils. But cvcn if this condition is achieved the
they are mnre like1y to withdraw then differential dropout rate between rcsearchcr's values and beliefs can distort what they see and how they
groups could be rcsponsiblc for post-test differences. If participants with report what they sce.
Inw scores drop out this will automatically boost the average score of the We can take two approaches that can reduce the danger of only seeing
group, even i f the intervention itself did not produce the boost in scores. what wc want to see. One is to develop clear operational definitions and
Apart from minimizing dropout the next best thing is to identify the research protocols that reduce distortion of ubscrvations and interpreta-
characteristics of thosc who drop out and then adjust for these charac- t~ons.'The secol~dis to use multiple 'judges' and observers so that the
teristics a t the data analysis stage. We can use the information collected different nbservers can act as a check on one another. By using multiple
at the ptc-test stagc to help identify these dropout characteristics and r~bserverswe can identify the extent to which the patterns are observer
then inctude these variables in a multivariate analysis to see if group dependent.
diffcrcnccs would have emerged if there was no differential dropout.
Ethical issues
In field based expcrimcntal sncial rcscarch both political and practical
considerations reduce our degree of contrid over the intervention. M1hcn t~egarcllcsso f the restarch dcsign, social research shnrrld conform to four
we rely on a variety nf people to implement a n intervention (c.g, a broad ethical principles (Kirnrnel, 1988; Hornan, 1991).
di ffcscni person might be responsible for carrying out intervention in
different orcanizations) we may not be able to ensure that h e inter-
ventinn 1s consiqtently or properly imptemented. When the intervention
is uncvcnly carrlcrd out we dn no! necessarily have a good test of the A well-established princ!pIe of SIK~RI r~scarchis that people should not
impact of the intended intervention. be requircd nr led Lo believe that the!! are requircd tr, partic~patcin a
As far as possible the investigator should maintain Sight control of the study. Furthermore, participanks should know they can withdraw from
wav In which intertren tions arc implcmcnted. Without t h ~ we
s cannot be the study at any p i n t . We shuuld avoid people gaining the impression
ISSUES rN EXPERIMENTAL DESIGN
from the c o n f ~ din which the request is made that participation is 3 the purposc of the study and its basic procedures
required. In situation.; where there is a 'captive audience' and where there 2 thc identity of the researcher and the sponsor
is a power differential hehvcen the researcher and the potential cuperi- 3 the use to which the data might be put.
~nentalparticipants (e-g. first-year psychology class, welfare recipients a t
a welfare agencv) the researcher needs to be especially careful to stress They should also be provided with the following:
that participation is truly voluntary.
The surest way nf ensuring that people understand that paltiripatinn 4 an outline of reasonably foreseeable r i s k , ~mbarrassmentor dis-
is voluntary is to explicitly tell them. For example, at thc beginning of a comfort
study participants can be told 5 a description of the likely benefits of the study
15 a description of how they were selected
',2ltho~~ghyour pdrticipat~onin this study will be greatly valued, you ,>re not 7 an offer to answcr any questions
required to p,trhctpate Ynu can stop a t any point or choose not to answcr any 8 a statement that participation is voluntary, that each participant is
particular questitm. lust let the intcrr~icwerki~ow.' free to withdraw a t any time or to decline to answer any particular
queshon.
Voluntary participatioil can produce a number of problems. It can
While the issue of informed consent seems entirely reasonable and
thrcatcn the external validity crf the experiment. Since certain types of desirable it is not always straightforward. How much infnrma tion should
people (e.g. those with lower levels of education, those from non- we provide before a participant can be considered informed? How firlly
English-speaking backgrounds, oldcr pcople) arc more likely than others informed should participants he? What does it mean to be informed?
to dt.cline ti) participate in studies, vol~lntaryparticipatio~~
can p r n d ~ ~ c e Providing more information need not mean that people wilI bc
biased samplcs. It can also threaten the internal validity if we get differ- better informed. Simply providi~lgdetailed descriptions of the study
eiitia l dropout in the experimental and contrt~lgroups (see eaxl~ersection does not mean that respondents will be a n y mom enlightened as a result.
on problems with randomization). In addition, if voluntary participation indeed, detailed technical information may confuse, distract and
leads to high dropout rates it can lead to too few cases for meaningful overwhelm rather than inform.
analysis. Furhermc>re, providing details about the study - especially detailed
Howcvtr, compulsion is not the answer. Although cornpul.;ion might information about the study d c s i p , about the hypotheses and thcories
reduce sample bias, unwilling participants can undermine the quality of we are testing - can distort the way people answer questions and
data in a study. Since we can statistically adjust for known sample biases undermine the validity of the findings.
it is best to maximize khc quality of responses and encourage voIuntary
One solution is to provide baqic information and to offer to answer
participation.
further questions. If this information could distort the way people
Tt is better to cncuurage rather than require people to participate in
b ~ h a v ewe can tactfully explain that this is a concern and offer to deal
studies. There are two main ways of doing this. One is to appeal to with the questions nftw the experiment is over. h some psychological
people's altruism by pointing to the study's possibIe benefits. PeopIe are
research participants are deliberately deceived since accurate knoudedge
remarkably generous and willing [n participate in studies where they would invalidate the study (Homan, 1991: 9tiff). Where this is warranted
believe it will d o some good. However, in our enthusiasm to recruit
it is critical that participants be fuIly debriefed after the study.
participants we must avoid 'false advertising' which overstates the
'This raises the question of when consent should bc obtained. Most
benefits. Second, we can appeal to pec>ple's self-interest - hnlv parti-
ethics ctmunittccsz require that consent be obtained hPfnre participation.
cipation can benefit them. We might offer rewards fnr participating or at
However, a case can he made that participants do not really l u ~ o wwhat
least compensate for their time and anv costs they rnEght incur by
they are agreeing to until they have participated, and that they can only
pnrlicipating. Somc studies provide token gilts and offer to provide
provide irtfi?~wrrrlconsent affu they have participated.
participants with a siirnrnary uf the main findings.
- w!10 should give consent? Who should give permission in research of
young children, the intellwh~allydisabled and others who may not be in
a pnsitinn to fully understand the implications of participating in a
survey? Participation still ought to be voluntarv but consent may need to
Infnrmcd consent is a close cousin of v d u n tary participation. Typically be obtained from other pcop,plc such as parents and guardians as well a s
participants should be informed about: the particip~nt. In hierarchical organizations gatekeepers may grant
pcrmiq.;i(m tcrr illnior p r t ~ p l eto yarticipatr and t11115,111 w m c srn<c, t h l i g ~
rtlnior p i ~ o p l rkc7 ~ n r t i i i p a t c In
. .;~tch si!uation.; b i . ~~ h r w l t lrln.;ilrrn thal
~ . l i nct,n*ctit frtm both rhc qatck(,eper.: anci thc ;lctu;ll pnrticip,~nt\. :'in o h v r o u ~w n l 117 r\.l~icl~
participanl< onn Ilc- hnrnic.cl i.; h i faitlrrr t,
hunclttr prrlrniwc of C-~niidcntialilvPC-oplrp r t i c i p ~ t i n111 ~ ,r ~ p t ~ r i r n ~ n t s
~ l c u p t ~ t h ~ they
,ire r r ~ t i t l c tti t cannot bth idcntifred a< the ~ r u r c t of . any
yarticu1nr infflrrnation. Frcqucntlv, rm-inl \cicncv tbkl-rurimmts 2,ntalF thp
collecting of personal informatinn which, t f rnLidr rublic, souId be
~ . m b a r r ~ i < sor i nhurnilintin~
~ o r carz<e liarm to p;lst~crpant+In one way o r
In sc>rnc t.xpcrimcntal studie., particirpants arc pcltcntiallv evp~lscd tn another.
Iiarm. !'or k>karnpic, in medical rlipl*rrmcnt.; in which a new d r u ~ic; It 1% r<s~-ntial tlicrtafore t h , ~ infnrmnl~nn
t bc collrcte~lin ruch ,1 w a y that
tria1lr.d. plrticipants at'c pi~tcntlally~ n d ~ l n j i c r r [n d . ~syrhrdirricalruprri- c t ~ r l f ~ ~ l r n t ~coin l i f vg ~ ~ a r a n t ~ . cI!d .i.: ,~l+trtasqcntial l l ~ l tthc pcbol-rlc
~ ~ lw
r n t m t s parlicip,inl+ r n i ~ h !be given stimuli (11. bc iniillcrd 10 L-rc=I~n\re in invt3l r c d ui ctdlrnctin~, an;llvsin~ and rrbportitig tllc find in^.; r w p c t
I \ ~ ; I ~ Gtfial thy? later rt*grr%t ;~ndfind di+trr>ssing.' s u r u p u l ~ u s l vtlic c i ~ n f ~ d t . n t ~ nrcq~~ircmtantq.
l~tv
14'here thvrr- ic ati! d,injict' uf harm trl pnrtic~pantqilic p r i n c i ~ l caC In clrdrr In pn>tcct partlcipat~tc,data muqt be '.tortd in *i!ch 1' way d 4 t13
inklrmt~c!cc,nstxnl rrqtrlrc>s that partiripan15 \-retol~l( r f thzb dancer< hefort, Frcclutlc ;rnv (rnatlthiwi7ed ni+cc.;s. This may rncxan tori in^ data in I r ~ k e d
p,3rticiyatin?; in [hi. etpclrimrnt. 12'htnrtn thc.rt5 is ~@teI?ti~l ior h.rmi , ~ n d c . ~ b ~ t i c(lnd
t s rrvnovin): ,111 ~ d c n t i f v i ninfomatinn. ~ If thc data an., <torell
tl-11. p ~ ~ i - ~ tst111l n agrLbtJ:. t o particiy.itr tll~re is the nccll f(3r c;rrrt~ll elcctrt3nic'dllv thrn ~~a>st\.rrrds and r t ~ s r ~ , p t i canc ~ nhcIp t.nl;ilrr p r i ~ a c ! . I. t
d r b r ~ c h n,lftcr ~ thc c\pt,rimcnt. D ~ l p ~ n r l i non g thr- typc o f pntt.ntlal Ii,irm is h ~ g h dt-siralllt*
l~ that prrsonal ~dt-ntikyingtl~tai3qGinill h ~ rtlspnsr\
thrrc. m.iv hr the n t w l for Irlnl; trVrrn ( C I I I O W - L I ~ t o t*nqtirtA~ h , T~R V t rt*l;lt~ns10 p s r h i i p ~ n l shc kcpt s ~ p a r a t i , l etlforrnat~on
~. about s pt8rs(1n's
p ~ t c ~lallv n t li,~r'n~iul ctfrac-15a r c n~on~trrrcrl. natnc ;lnd .~iliirr~.;\ s h o u l ~ l1-c 3torctl in i r t i c lltt, ,inti 6i'11 Ih(brr ,jtv-wer. la
Lnlikt> o t f ~ r dp.!gns,
r nianv t~~pertmtlntal dc-s~~n.: ~ n ~ . n l \tht~
. c invt>sti- cl~~t'ctir>n> <lioultl I-re .;tot.~tlin n ~ i i i t i ~ n ~tilt,. n t Idcnllficaticrr~nurnb(*r\ ftbr
gGitt>r a c t ~ \ ~ v ldclin):
v w n i t > t h ~ n potcnl~allv
q hnrrnlul to thtl p r t ~cipantq. c r n \ ~ - r t 5 t c r ~ c n cm~uns~t bo ~ t n r e t li l l quch a WJ\J thtir onlv ar~tharizt\d
IZ~li~lr ~t <hrlulrl IT ob\.irru.: t h a ~n1ci n l ~ rriol
t s u b l t ~ pcvplc
t to p ~ ~ f r n t i a l l ~ ~ pcrlplc ran rn,lkc tht* m , ~ f . r h i n ~ .
11.1rrnl~rlinter1 entions. i t <ecrn.; rrsnw~rableto ~ c up t rl\ytr.inicnts ivhcre I'rtn'iilinq .w.;urnncc\ c ~ fcnnfiduntialitv ir irnpc>ttar~ttor mett~ndolcrgi-
thc inlcrvc~itinns~ ' 11 i lwnefi t the participants. F L t~rc~~nlp!t> ~ t ht. g11~r- c-al a5 ~vtlEas ctllical rt-as~~nk. Ii p r t i c i p ' ~ n t s arc. confident that t h r ~ r
antrxetl income c ~ p e r ~ m e n tcli~scril-red r; in Chapter 1 ct>uld bc just i f i d respnnws are* truly corifrdenti,~l (or et0en better i f thcy arc a n o n v m o i ! ~ )
1>ec'd11r~ Ihtx intrr~~cnticm of prrlr-tding a ~ u a r a n t w dincr~rncworllrl rmlv \vt, cnn expect that pcoplc art- m o w likrly tr) p r l i c l p a t c in tlrc + t t ~ d y-
benefit parti(ip,~nt<(Fl.lktrn, Icr8h: 102). ilspcc-ially 11 i t IS al>otrt private matter.. 14't. car1 atso p*pt~ctthal ~f a
t lolvc~rr,thc a r p ~ m e n tt h ~ ,in t inten.cntion w ~ l lbr lrtnpf'icial pre- pt3rstv-i fei*ls that thtir answer.; ;ire trulv confiil~ntia!thrv xvi ll be mtrrc*
*umes that we know thc outcomes of the intenfention. We rn,y trrltrrrr I~hctvto provide k ~ n and k honrst answer?.
that rnconlc maintenancp w ~ l l at , wnrst, bc harm!css, but ra7do not A ~ T L J ~ I J one .;hould not milkc prclrni<lbs nf ct>nfldcntialitv un1i.s~they can bt*
t h i s rrl~t-~rd clt tlwrcn. Even wiih the t,r*st o l intentions thcrc c a n be kvpt. In a stud7r nf 5clioc1l chi2iirc.n v i l r l might Icarn ot thc idpntitv o f a
irn~ntendrrl harmful r.ffi>ct.;. 'Therr it; albo thc isst~eof 'beneficial to drug puslrrr nr a trachcr tvhn i\ 5cw11allv rnolcstiny: s t ~ d m t s In . a qtudv
whom?'. l-lakitn ( 3986) ncltcs thnt thr. income maintenance e r p c r i m c ~ i t s nf pri~cmcrsyc>u might Irarn of plan.; fur ,ln c s c a ~ cnr of thr identitv OF ;I
led to an Incrcaqc i t 1 t 4 i $ r n ~ ~a<e ,SI)MC' wclmrn bt.cam~.financialt!~ morp F ~ S Wwho I ~ murdered an inmatc. In such castt.; vnu fL)c{.thp. pr~>blcm nf
secure. 1-hi+ ) r r n t ~ 1ia1.c I~t>nt.fittdthr wclrncn but thc husbantls (.jnd ~ v c ~ i q h ~Imcturn g prorniw OF confidrntialitv n~;lin<tthc harm olu<ctl to
ti IIL!r ~ n rn;r\.
) 11n1.ehtrld i.cbrl.il iffr~rmtI irrvc c3tl1i~rst-r! r c < p c r t i n ~that From1.t:. 'il;lLr. qurc !ll,it \ - c ~ td<r ~ ntlt rn,lhr
\2%i1n thtare artv perrri\.cad bt-r~cfit\t i ) parhcipatiriq in a n t.uprirnc.nt ~ ~ n r t v l r s l iprttrni~t.+
c of sr~nhrEc.ntiali~
I P . ~ Jn . incr\mrnmaintcnnnce ru~vrimcn!, rnvdical trial\) t v r m;ly FRCC t l ~ t b In promi\ing cnnfirlrntialil\ bc ,1\varr o f the I c p l rami t i c .rticlns r vr>trs jf
pruup M.I'~II ~ C C ' P < Sto th{' tre~trncnttmcr wtnknc~wit h a s brbcn succesqli'r~l: a tlon 11 15 ~mpc>rfant to clLlritvvnur r i ~ h t >to rnaint,iin thc rontidc~ntialitv
(or ;l m o ~ l i f i c a t ~ uofn tht\ trtvlmtnnton the h,~qi.:d wlint wnc: lrarnetl from of d ~ t awhrn wiIrk~ngIn Rnvrrnmctnt agcnclcs or under n ctlntmct
lhr c ~ v c r i r n e r ~ t ) Contracts Frr-q~lcntlvmakc cup!~ c i trtnt13rncntq nticnrt tho owner.;hrp ot
, ~ it 14 irnpnrt~tlt tc, encure tIi,?t tlirasr- prtlr Fsintl- dl3 not corn-
d ~ t anil
promise ~ O I H .~111dvrt;lking'i
r e ~ a t d i r usonfidrntialitv.
~
6
ANALYSTNG EXPERIMENTAL DATA
F.rpctimc.nfat tlrsigns arc valiiablic for isolating the cau5al impact of
qlv*cific x . l r i ~ h l ~1-f~rs
r . 1s rkpcclally thr case n-hcn the!' are cot~~luctcd in
hl~V11vcnntrollcd environinent~.The structure. of the dr.;ign, ~specinlly
tht. t ~ s cof ~ n t d cc~ntrcllgroup--, help< nlaxirni~ethe int(=rnalt.alrditv of
crpuriinr>nt~l rr,.;enrrh. Hnwc\~ci.,the nrttficinlity of s i ~ c hrcsearch and the
il,lrrorvnr~+r l t t f ~ dtlsigrlr
r often rne,ln that, fnr w c i a l w-tencr researil~, T11t.r~.lrr6 I n r ~ e~ l u r n b ~t i tr ~ t a t i q t i c a ltcchn~qurkthat a n 111. used for
~ u p p r i r n ~ ~ ndo51gn\
tal y1ul~donly Timitc.il informatinn th,~tfrequently has a n a l v s i n ~expr~r~rnciital dnh. Thc purpose of this chap1i.r i s to p r o v ~ d c
poor c~tilrnalvalid~tv.T h i s chapter ha.; disrussed come of thc threat< tn some g~liilancrin stalcctinq from amcmg klir.;e mnnv diflertxnt M-a\.< of
~ntcrnalJ ~ ruternal J validiv 111 rcscarch and ha5 rlutlincd thc w a y s in , ~ n , ~ l y s ~expt-rimet~tal
nx dntn.
which c~xperimentaldesims wcrcclrnc t l i c ~ cthreats. In ~ w i scirnce ~l -1-he c-rvre r r f euprlrimcntal arr;llvsi< is thc c i 7 t r r j ) t ; r r q r l r i of grrl;rp<. Dn
scwarch p r n c t ~ ~ a. ~ l n d rlfhical rnattvrq nl-ck mtaan l h , ~ tcuperimcr~tal grtwy.; that rllffcr In tcrm- t,f t h r indrpendi~ntvarl,~hlealso dl t t ~ or1 r the
rlclqi~ns(,Iten ranntbt be used. T l i ~ schnptcr lids crttt1inl.d t h c s ~practical l ~ ~ n t no grcruyk that hnve diffcrcnt ?~sc,-ltrncnth'prrltrrnr
~ l t ~ y c n ~v;lr1dtrlr.7
, ~ n detl~icalis\tres. J ~ i t c r r n t l tnn the rrfitcomil varlnE1!c7 In a rrrll-cltls~cnerlriuprr1rnpnt the
rinly t i c r t . n ~ bc!n,tacn
.~~ Illc g r i n ~ p s~ I I u ~ I ~\lt>
L ~ thtb diFt~'ri'~lt*~~t'~tl7ll'llt5'
4(, rvhzt-h Ihtby a r e t~\yost~tI Tht. morr IVP can be hi1 rc t F i ~I nrtr !;roup ?re
Notes similar cxcrpl in tcbrmq of thc indcpt*ndcllt variilble {tn'tervt*ntion)thc
mnre Lr't, can be crlnfidrnt that thy grtwp d i f f ~ r t - n c e4m ~ thrl nutrtlmc
1 Tlit+c irr.n;Cnt--~c~:,d r a n ~ l r ~ n ~ i t , ~ t.lrtT
r n n rnt~sl evitlrnt with s ~ m p l e variablr are duc tcl the infcrvent~on.
tauperinuhnltnlt l r s l ~ n srn whrr-13 Ihtbrr I\ onlzo nne rndcpendtbnt varlablr 'Tlivv In m.lkrn~comynrii;on~thrcrx thine.; nwO krr hr + - ~ c w c .
can br rr.~luctl~l .;ornPr\.hat In marc cnmplrv twp~rimriitnl dr+im<that 1nt.1ude
\twcral indrpt.nrlcnt variables (sets d~scus~irm nn factrbrrill d p r ~ ~111i sChapfrr 4)
1 Thc comp.~risonfcq-u<~.l;nn ,yrLvrp - nut indi~'idual.;.
mrnt dt.partnwnts) I ~ . ! v c~ r i - t ~ l a l ~ o nctlrics
at rornm~ttrrsM ll~cfimuqt appr0rre 3 M7t*nt)edurayrpf workingri~zthow r n ~ s u l i d i f f t * r ~thercnceds
nc~ tobc
pmject~lrnm a11 etY71r~:per*pvcti~,~~ b ~ f i > rtunding
*> iq rtblcawd for thr str~tty hctwpr-n c~sperimcntal and control grtwps bt*forc i v e attrtbu It, a n v
? n l r \l~lgr,~rnt,xper~rnt.nbcln r~bcdrmcearta c l a ~ ~ rexample< c of thr\. Fee irnprrrtanctl to thc cliff(-rencc* For instance, ~ 2 . rm i ~ h ohsr~rvr
t qrnall
TXn~rrnrrnd(1 Oh3) and Milgram (lW14) bul inlportant dilfcrc%nccf;ur we might rrhscrvc substantial differc~iice<
that might be q~rnplvtFue to sampling and rnr<asrarcrnentrrror.
3 Wr compare grnirps in terms of thc outcome variable.
rt,ttibtic%cstablt~li11 t'tic tind~ncclor n sample can bv cxtrapolaterl t o thc nrdinal arid nurninal (a'tscl called catc.grv+lcalrrr rlunlitatii.~).t 1it.w can he
i r r m n.Ii~chthe s,irnple is drarrn. For c*uample,1r.c might
i t i d i /iol!~ilnf/oclli
~ best expta ~ n r dtzfith thrctl c+nr~rples.
f i n ~ l;Irnt>ngrnernl>cr\ of the samplr that peclplt n . l ~ rarc ~ unenipln\.~d
h n \ ~ nrnrse
i~ hvalth than employed pi*oplc=.How liktnly is it that thr < a m p
link L-lotween t~ncmpluymcntand hralth would hc fnund in thv widcr
o n tllr <.~mpleis meant ti) rcpresent? Inferential statiqtics
~ ~ o p l ~ l a t ithat An interval/ratio \rariablo is one in which thc categories can be ranked
prcn'ide a ivay of t-.;t~mntingwhcthtxr %amplepattcrns will hoSd in tlie tmm IF;(* far Irryl~in some* meaningful npay.In add~tiontcr ordering the
r\.lrfcr population. cntcgorit\ Isom Iaw to I~i~ch it i.; p s i b l c to sprcliy the ot~lotrrrf~irtl~fiert-rrrr.
Idrally, anal?+ s h u ~ ~ lin~.oF\~c
d both descripti\fe and inferential hhzqeenthc cntcgyvirs. Age, when mea4urc.d in ycars, is, an r r a m p l t of
aiinly-;is. We shollld carefully cicscribr thc c x l ~ n to f any diff~rcnrcs an inkerval variable. Wr. cclla rank order the catc~orimof agc ( y ~ a r sfrom )
bratwccn ~ o u and p ~tl~l>nwork out w h ~ t l i e rany di tfcrcnceq arc Ei k r l y to voung~qt(loweat) to nldeqt (highest). Purthrbrmorr, we c.ln specify the
Ioi~ndin the widtbr population. a r n o ~ ~ nzrft d iffcrence brtwccn the catcgoric..;. 'rhe d ifftwncc bchz7cc.n
the ciltcbgory '20 years old' and the cat~gor!. '25 yens< old' 1s 5 years.
i'ariablt.~ \uch a s the wceklv number caf litwrs of paid rvork, I@, height,
weight, Income [in dollam) Arc. all variahlr.: where the c n t q n r i ~ sare
Infrrcntial statistics can imIy be uwc.d to generalize finilings if a ~ ~ o l r n h l i ~ ~ ~ numcric and ~ I i r r thp e ~ntcrvaIs bvhvrcn thc cntegories can bc \pccificd
.WIII/I/~, 2s used. A probability sarnplr is one nf two hl+oad types of saniplr-s. psecis~l!~.
'Tlic ~jthcris a nun-probability sarnplr. A probabrlity sample is nnr in
which t ~ pcrscln h h,lr an cqual or at Icaqt known c l ~ ~ i n (probabi!it\.)
ce (>f
. nbt.11nlngsuch J -,1rnp1e 15 to r ~ ~ i i l t ~ ~ n l ~ ,
I>r.inq \clrxctrd Jhc- % ~ r r c . ittv a ~ ot
~ r a l t dsnmpIc rntmhcrk from a knt)wll and dciincd pcrpl~lation. An ordlnsl \..iriaI,le is c~ni,n.hcrc eve can rank order cateqrrrr~..;irnin low
1Ylit-n w r hnvc probability snrnplt* we can U R Z ~ Ir*gitirnatelyKcncr- tn high. I-lnwcvcs, WP cminnt specify in numcric terms Irtliv ) l r r i l - l t diffcr-
illi7i' I11 the spt~ciCicpopulation I r t m which the sample is d r a w n . A cncc thrrr i'i l~c'kwecnt'nc c,itr);nries. For crcrl~npfc,
when thc variable age
pmhability s a m p l ~o f a country which systematics llv cxcIuded prnplc has catcgorirs quch a s 'child', ',~dolcsccnt','young adult', 'middle aged',
.~gt-clunder 18. thaw withnut telepl~nn~q or thosc living in institutinns and 'elrlt~rlv7tht.variable i.; m~asrrredat thc* 0rcl inal level. Thc catcgclrics
rt,n\uld yietd findrngs that couId onlv bta gvneral17cd t c ~the populofic~nof can be nrdcrt=d from \-ounpa.;t to eldest but n v ccannot sp~cEl\~ prt.ciseI!*
nt>n-tn$tih1tirmali7cdadults rvith a tcycphnne. Tf an cvperiment I < bawd the age ~ a hptween
p proplr in different cate~tjries.If we mva~tlredl e \ ~ I
cln n prohabilip samplc of students in a f i ~ t - y e a rpsyclinlogy class In ,3 of workforctl participation as full hmc, part tirne and not in tlrr labour
particular irniversitv then we can only confidcntyy ge~~eralize to this force this vnriablc would bc a n ordinal variahlc. If we aqkcif I l l r precisf
pop" la tion (unless wc can demonstrn tc that t h ~ s rf~rsi-yearpsychology numbcr of hot~rsworkd tach wcck the varial3le (hours of wnrk) woufrl
studcnts at the particular unir~rsityare r~presentntivcof some ividrr he m e a q ~ ~ r catd the interval level.
pop11Inti~~n).
11 12.t. d o no! hn111- n probability .;ample thew iq little point in w i n g
inlcrrntial statistlcs t r l generalize to snmc wider populntic>n. We <lio~~lcl
limit n ~ ~ r s e l v ctos iicscriptive statistics that sumrnari~c.patterns in that A nominal variable is onv whcrc the different categories hnvc no set rank
sarnpl~.Many rcsrarcliei.~are unablc t o recruit ra~irlornsample%.Npvcr- order. With a nnminal ~arialllei t makes no setwe to sav tIi.lt the cate-
thrlcv- they continirc 10 irse infcrmtial statistic7 inappropriatelv in thrir gorieq cnn t3c ordcrcd tronl tcnv ti1 high 117 .;clmc qenw. I<vligiu~ts affrli-
. I ~ , It ~ t psample=< arr not fea5iblt~thc m w t apyroyrl-
I < i t . IZ'herr prt~'t~al.rl ation is a non~innlv a r i a b k wFit=r~i v e can d i\tin~uish t r t t ~ u v ~cnt~goric.;
*~i
; ~ t vrsrnl 117 test for ~cn~'r~ili7ab111tv is to rcpcat the tl\pt~rimentin diftrrc~nt oi afhlintir~n( c . ~Ic~visli,
. I<ninan Catholic, /Irthrxln\, Prt)t~,.;tn~it, I4arnic.
rcrntibut% a n d wit11 rl'rftcrcnt sarnplrks. n o rrliginn). TI1c.s~catc~trrii.* ha\^ n o nnttjr,~lrank rrr~frr.
_ When sclr\cting a mdhod c>F data analysi.: we must first j~ltbntifythe
level of mcaruwment of c,lch of the variablrns in our analpis 'TO work
out the Icl L'! 05 mcasurtlmpnt rrf an!. variill,lta, ask the c~utstirrri..in Tablr
I t 1 \tatiqE~calanaf!,.;ih thr. l~-i,cl of ? t r r , , r ~ ~ / r r ? ? of IS crltical in
l r ~ lvilri,~!llt'~
I 6.1.
wlc.chng particul;lr .;tatisticar analiw~q.There are thrW main Ic1r.1.; a t In sel~ctirlgapproyriatt>mt*thoclsof a n a l v i < the level of mt*nsurcmfnt
n'h lcll ~ ~ a r i a b l arc
w rncnq~~red: int{.rvnl /ratio (alkn c a f t ~ dcontinuc~u.;), nf IatJr tht. grotipmg v;lria'l.!r ( ~ n d c p c n d ( ~ onrt i n t e n ~ c n t ~ ~ m > ) the
and
92 EXIIEKTMENTAI, DESIGNS
!
(c) Kurtosis Id 1 Spread
For example, imagine that we wanted to examine the effect on uni-
versity course complcsion r;l tes of different methods of delivery of a
course. We examine completion ratcs of those who are taught in face-to-
face situations with those who take their course over the internet. We
might find that in the face-to-face group 75 per cent successfully com-
pleted compared to 60 per cent of the mcmbtrs o f the intemct goup.
These percentages provide a description of the different outcomes in the
two groups (descriptive statistics). We could then undertake an analysis
that would tell us whether these differences behvecn the groups (75 per
cent and 60 per cent) are likely to be due to chance or whether thev are
The i m p a d of an intervention may be seen in its cffect on thc slrnpc of the likely to reflect real differences in the pnpulation from which the sample
scares on thc niltcomc variable. Comparisons of group homogeneity arc was drawn (inferential statistics). In the example just provided, the twu-
one form of comparing shapes ul distributions. Otlicr aspects of the samples test for difference bchveen proportinny u,ould Fulfil this
shape nf a distr~butioncan also form thc basis of gmup comparisons and function (Loether and McTavish, 19741: 189ff).
reveal thc impact of an intervention.
Apart from comparing thc. degree of sameness or difference within a
group we can also compare the distributions using other dinlensions of
the shape of a distribution. These different dimensions arc illustrated in Groups also can be compared usefully by examining thc association or
Figure h.2. i correlation between variables. StatisticaP measures of association test the
42 WHAT IS RESEAIKI-F VFSIGNT CAUSA I ION AND THE LOGIC OF KE5EARCI I IIESIGN 43
Ir~rft.pi,ndrn!varinbles with moviJ than tu7u cate~urics Where independ~nt Young Group 1 : Group 2:
r"; 1 1 1
variables have more khan two categories we can compare more groups
and provide stronger and subtIer tests of theories. For example, in
examining the effect of divorce on children we could simply classify
rnarriagcs into divorced or intact, or we couId classify them as single Group 3: Group 4:
(never married), Intact, separated, widowed, divorced (repartnered) and old males old females
divorced (not repartnered). If we took the latter approach we could make
multiple comparisons. [n doing so we can get a clearer picture of what is
going on. lf u7e Find that it is the children of divorced parents who have
repartnered who havc the most problems we achieve a more focuscd
understanding of the impact of parental mari taI status on children. We
have learned that it is not. divorce per sr but the arrangements that follorv
divorce that are critical. Had we simply compared divorced and intact remeasurement at the end is called t h ~ 'post-test'. Between the pre-test
marriages we might ncver have identified this. and the post-test certain critical events occur - in this case parental
divorce (or non-divorce as the case may be). These events scrvc as th.c
CnwrhEwni effech of clcffeewnf i n d ~ p ~ n d e vnrinbles
nt It is possible thdt it is independent variables in the study - i.e. the presumed causal variable
onIy when people havc a particular combination nf characteristics that an behind the observed change, In some designs thcsc intervening events
effect is produced. For example, it may be the joillt effects of gender and are called the intervention or the heatment (see Chapter 4).
age rather than cach independently that is linked to suictde. Wc may We can collect data about more than fwo time points rather than being
hypothesize that when a person is both male and young, the l j keli hood restricted to simple 'before' and 'after' data collections. We might have
of suicide is at its highest. If, for h e purpose of this example, we think of repeated measurements over an extended period to track 'ups' and
age as a dichotomaus variabPe (15-35 = y o q ; 35+ = oId), then we can 'downs' and to track /rends before and after any critical event. Multiple
think in terms of four groups based on the combination of these two 'pre-tests' and multiple post-tests can help distinguish between short
independent variah'les. These are illustrated in Figure 3.7. term and long term trends. They also help identify the effect of the
'intervention' or independent variable. For example, a simpfe measure of
Comparing tirr?r puinrs emntional adjustment before and after divorce might show a decline in
adjustment. But what if there was strong evidence of the declinc having
A finding that children from divorced families are less well adjusted
commenced well before the divorce? Only multiple measures before
emotionally than those from intact families prestimes that divorce is
divorce would show that a trend had already begun. Similarly, a simple
resyonsibfe for this difference. The problem with a simple comparison
post-test might show that adjustment is lower after divorce but mulliplc
between groups (divorced and intact) is that it does not tell us whethcr
post-tests might show poor adjustment immediately after divorce but a
divorce has actually produced any chnnge in the emotional adlustmcnt of
steady improvement over time (see Chapter 6).
children. Any conrlusions wouId be much more convincing if we could
track the emotional adjustment of children both before and after divorce
to see if thcrc was any change in the emotional adjustment of children
following divorce.
Idcall!! fhe Xrotrps 7uc are conrpnring sr~nulifhp the sarnr in ull rfler~nritrespt7cts
By tracking children over time we could see if thcrc wa5 a change in
except in regard to the indcpendcnt variable. For cxamplc, il wc want to
emotianal adjustment of the children and evaluate whether h i s was
attributable to their parents' divorce.
test the idea that non-government schools produce students that achieve
better academically than government schools we would need to be
confident thal both types of schools contained comparable studcnts. We
MULTIPLE
PRE-TESTS A N D POST-TESTS
would need this so that we could bc confident that the only relevant
Tn the above example the measurement of emotional adjustment a t the difference between the two sets of students is in the type of school
beginning of the study is often referred to as the 'pke-tt.st' and the attended. If the studcnts differed in additional ways, how would we
extent to wli~chtwo v,~rinblr~I-0-vary. Wllct~we say that gcndvr and
income nrc a ~ ~ o c i a t wc ~ r l nlcarl that hcnowinq someone's g ~ n ~ i hclps cr "b changr i n Fixcd ratr C;n>~rp bonu.; Cntl~vidualbonus
prcdzct f l ~ e income
~r bcttcr than i t we did ntit know their ~ c n d r r . ~~LIL~!I~-II\*I~V ~ r o u pIn11l ~ r t l u p(no \ ~:rr\up(no.)
A <in~plrexperimental d ~ w i ~has n hr-o v i l r ~ a b l c-~ the intcr\.vnt~nn
variabltb {independent \.arlabIrl 2nd the outcome variable. If the inter-
rcntilm hnq an effect on thc outcome variablc thcn the two t'sr~nbles-
the intrr\,cntinn and thc otitcornc - will be nsqociatcd statisticnllv. That
is, i f tht~\r, prople in the Rrotrp that recrivcnd rhc intervention (i*xpt*l.r-
mental grollp) scored dif fcrtan tlv on the .oittcornr v a r i a b l ~than thc15r.who
did not rt.st*ive the intcrvcntir>n, the two variables - inten.cntir,n and
outcrrmts - rvtwld be qtatistic~llv associat~d.5i1cli an a s w c ~ a t i n ~can i b~
d i < p l ~ v r dtr.;lng cro.;r;-t;lbulali(~n~, graph.; o r +tahqtical summarit*.: quch
as ~ o r r ~ l a t r ncoefficient.
n
In atl~litlonto rlescrihiny: ~ h extent r to which t h e is an arsocintion
bc-twccn Ihc cxperirnenta l intervention and tl~routcome, wc shou Id use
a n infrrcntiat statistic dcicrmi~wwhether thc as4ociation is likcly to
When deciding 10 display data wi tli a graph we must also decide which
hold in t l ~ c&\rider poy~rlation.Different rnfrrpnt~al tests havc L3ccn b ~ nfe graph lo use. rn Figurt* 6.7 just three of many possible graphical
developrd fnr different sr~rm*lntic~n coeificient~. displays arc pfiw~dcd:histogram?, stcm and I ~ a fpl~lt.; and ho\ . ~ n d
wliisker plnb 4II three grayhq arc based on tht. 5amr d ~ t and a yrocrdr
d~ffcsentl v t ~ y q01 thinking abou I and ~ ~ i s u a l i z ~thr.
n gdistrib~rtinns.
(a) Histograms
Bonus system Statlst~c Std. Error
% productiv~tychange Flxed rate pay Mean -.5455 2 3769
Median .OOOO
Var~ance 186.443
Std Devlat~on 13.6544
Minimum -30.00
Maximum 30.00
Range 60 00
lnterqwartile Range 7 5000
Skewness -.I61 409
(b) Stem and leaf plots Kurtosis 1 078 798
k groductiviry change Stem and L e a f Plot f o r : Group bonus Mean 45.4848 5 9012
GROUP= Fixad rate pay Median 39.0000
Frnqucncy S t e m & Leaf Varlance 1149.195
5 . U O Fxtremes (=<-IT)
I
Std. Deviation 33.8998
1.00 -0 . 9 Min~mum -5.00
-90 -0 . Maxlmurn 130 00
2.00 -0 . 45 Range
3.00 -0 . 223 I 135 00
1. 0 0 -0 . 1 lntwrquartlle Range 53 0000
13.00 3 . 0000030001 Sltewness ,567 409
1. 0 0 I) . 23 KU~~OSIS -.261 ,798
3.00 3 . 445
. iio L' .
lnd~vidualbonus Mean 31.1 51 5 9.071 2
2.00 n . 8s I
Median 14.0000
4. n 0 E x t rPves (>=17) Var~ance 27 1 5.445
GROUP- B r o u g bonus Std. Deviat~on 52.1099
Frequency S t e m & Leaf M~nimum -45.00
1-00 -0 . 0
29.00 0 . 0000111222233333444
Maxlmum 125 00
11.00 S . 55667778899 Range 170.00
.
I
2.30 1 03
GROUP- Individual boaus Skewness
Frequency S t e m & Ledf Kurtos~s
7 . CO -C . 0123334
16.00 0 . DOClUl100111112222
1 3.00 0 . 569 Figure 6.4 Summary staliqtirs: productivity ch~npeIry hnnws scheme
7.00 1 . 0111122
cxrcrimental and control groups should ensurt hidcpcndcncc betawen histogram can give a visual idea of whckhcr Ihc distribution is a bell-
the groups. There should also be independence between cases rvfflrin shaped, unirnodal distribution. A normal curve will also have a skew-
groups. Srnce the sel~ctionof one sample element should be independrnt ness of O (i.c.symmetrical) and a kurtosis of 3 (i.e. mesokurtic) (Loether
of the scIcction of Pverv other dement, a gruup consisting of, say, and McTavish, 1974b: 162). When the sampIc is small (e.g. less than 50)
husband and wife pairs docs not produce an independent samples and is not randomly drawn from the population, and ~ t spopulation
design.' distribution is unknown, the qafest strategy is to use either non-
parametric statistics or a robust parametric statistic.
FAIREDSAMPLES DLSrCNS
In a prlirtld snmp1t.s design groups are made comparable by pairing sample 1s the vnriunue an the d~pe~tdent
varinblr similar betz~wnrack
elernenis rather than sin~plcrandom allocation. Here the selection of c o r n p n ~ i f o~ ~r o~l r y ?
people in one group is not independent of selecting pcoplc for other
gruups. For examplc, wc might have a group of identical twins and Group comparisons can be usefully executed by comparing the typical
assign one twin to the experimental group and the other to thc control response for each group. However, comparisons ot what is typical can bc
group. Alternatively, we might simply administer different experimental invalidated if the groups have very different variances. Comparing the
treatments to the s u m group at different times. By using the same group means nf groups with very different variances make5 it difficult tc>
wc theoretically remove any other differences between the groups establish whether any differences in means are due to differences in
(except of course later treatmcn ts might be contaminated by earlier central tendency or not (Loether and McTavish, 1974a: 176).
treatments to the same gmup). In effect people in one experimental Some statistical techniques require that wc establish that the variance
group arc paired wit11 themselves for the next trcatmcnt. in thc (mu)grtwps is similar. This is particularlv important for t-tests of
Some ir~ferentialstatistic5 need to take into account the way in which differences bctween the means. In the case of /-test5 there are diffcrcnt
peoplt. were allocated to the different experimental condi tiuns. Werc forms of the test d~pendingO n whether there is cqualiv of variances
thev pniwd in somc wav or wcrc they randurnly (independently) selected? between the groups. Most statistical packages include statistical
techniques to test for this condition.
I,q t h d~~ ~ n d e rvarznblp
zf normally dislribzjfed?
Some inferential statistics, caIled parametric statistics, require in tt.riral How many irrdqlrndent vnriablfs?
level ou tcnme variables that are normally distributed in the population.
Unfortunately often either nutcome variables are not. normally distri- The simplest experimental designs have just nnf independent variable.
buted in the population (in the statistical sense) or their population However, many designs include several independent variables. Factorial
!i
distribution is unknown. and Solomon four-group designs examine thc impact of each of the
When the outcome variable is not normally distributed or we cannot experimental variables on the outcome variable. They also identify
reasonably assume a normal population distribution we have three main the impact of interactions between these independent variables on the
options (Loether a n d McTavish, 1Y74a: 2hXff). nutcome variable. In field experiments where random allocation is not
possible we need to estimate and statistically remove the effect of known
1 We can 'normalize' the distribution by transforming i t (Marsh, 1988: I group differences so that we can examinc the 'pure' effect of the cxpcri-
Chaptcr 11) mental variable.
2 M.Tt ran violate t h a~sumption
~ of normality dncl use the paramctr~c When mu1tiple independent variables are used the analysis strategy
statisbc anyway. rt is argued that some of these statistics are robrrst will depend on:
I
and vield very similar results regardless of whether their assump-
tions arc met.
3 Wc can u?e a notr-pnmr~ictrrcstatistic that docs nnt assume a normal
distribution.
I - a the level of measurement of all the variables (both independent and
dependent)
whether we want to examine the rrlrrfilte inlpact of each independent
variable (including interaction effects) or
How can wc tcll whether tl-rc outcome variable iq normqlIy distributed In e whether we want to look a t the effcct ot a s i ~ ~ gindepcndent
le vari-
the pnpuIatinn? Given a sample of sufficient size a line graph or a able with h e effect of other variables controlled.
1D4 EXPERIMENTAL UESIGNS ANRLYSINC: EXPEIIIME!+JTAL DATA
S u n r ~ i r n ~process
!~ for sele~-fil~g
fllc right Inemure
F i p r r s 6.5 through 6.8 are designed tn help you select, an appropriate
form of slatrstical analysis given different types of rxperimcntal situations
ourwme vanable
......................... . . . . . . . . . . .
N u m k r of groups
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.,-
HOWmany '
Indcpmdofll
1
_ (urou~~nq)
variabl* are there7
I'
. . . . . . . . . .
,
,
, Is the nutcoma
. .
, . . .
Graphs 1e.g. rlns grspr.3. stwn and l m f plms for each ', Descdpltve
' L W 8
~1:
'Skewness '
, group, hlstcgrams for Each qroup erc ) hsmlprlon ' Multlp'elpartlat cormlm~on
, l ~ law
' Crws-tabulatons (only w h o ~ o u a r l a bhave ' Gmphs 1p.g llne graphs, sfem and
' Kuflcs~s
' catcqonos) leal p,ms 40, 8 8 m group. hxlugrams ln'e-lbl I
' Ccnlml IOndenv Two-way tiacforlnll
. Knlmogolo u--Sm,rnw . ' Comp~nsonorihe groups rn termsol cerltral tendenv, for ~ a c group
h ptc )
' Crops-tabvlairona ( ~ n l y
where analys~s01 variamje
v a l h i on, shape
' H~sloq'am
I ' Mcasufe of assoclatlon- Pnarson's r, eta I varlablos h a w few categories]
' Stem and leal g o t
'.
,-
~nfererrce
' 07s-way analysis of vaflanoe (oomwnrqmearsl
I -
Commnson ol Iho groups In rsrmf ol
cer>trillfendeqw. varstlofl. shape
. .
-- -. ...... .- .-.
' Kwskal-Wallis K t e s t
.
.
.
..
.
..
. . . . . . . . . . . . .
0- Palmd or ~ndepend@nl
groups 1 Palred groups
and different types ot data. These figurcs cannot coves all pnssibilitics but
provide guidance for a fair range of typical situations. Most of these
statistrcs are easily computed using packages such as S f i S or SAS.
14 -
I
PART PI1
12 -
LONGITUDINAL DESIGNS
--
a
7
-- - TYPES OF LONGITUDINAL DESIGN
+F
-I -G
with thc seIection of the appropriate statistics this chapter h a s outlined 1 h7iIlthe same cases be followed over time? Here the choice is between
the main considerations that need to be taken into account when select- trend studies and panel studies. Trend studies (or repeated cross-
ing the particular method of comparison. Finally, since experirncntal sectional sbdies) entail collecting information from comparable
designs normal1y involve the analysis of change the chapter discussed a samples over time but not from the same people. Annual surveys
number of issues about the measurement of change: the usc of change such as the General Social Survey in the US, the British Social
scores, truncation effects and trend analysis. Further problems witli Attitudes Survey, the Eurobarorneter surveys and the International
measuring change will be considered ill Chapter 9. SociaI Science Program Surveys are examples of trend studies in
which similar questions are asked of a new sample each year. This
design allows the tracking of changes in attitudes and behaviour over
Note time. A panel design involvcs repeated surveys of the samc people.
This design allows the tracking of change at both the aggregate level
1 This ts t r u e if ~ndlvldualsare the units nf analysis in h e ~xperiment.Ef the - and the individual level (see Chapter 9).
cnuple M'3S the unit of analysis then having the couple would not violate Ule 2 Will dafa be collpcted at onr poznt of time? Although panel designs
reqt~isrnrrrits for indcpendencr.
involve collecting data concerning at least two time points, they d o
nut necessarily q u i r e that these data be collected on two different
nccasions. The prospective longitudinal design entaiIs collecting data
on several different occasions. In such a design a group of people will
11y I , , t k I 1 i ~ t v t . ~ l lrr1L
, ht.13
, I I I ~~!L ~ I I ~ I ~~ ~
> Ci \I ~ t b
s ti ! ! L * , I < ~ wi~.t-,il sonit,
Itrh>r!?tinic. 1 !I<, rr7frrl-:'h.il r;?. yrlnt.! ~lc,<i~m =nL6ir-r c.nt,li 1 caT!c.r-ting i1~t.i
un r7nI~r>tlla tlccxiot~. rti~ ~ * changia ~linietw~c>n.:
l i ~ t i and to tlic 5 t u ~ l y
;Irt> rdlt,~ined171,wL~nl:p t ~ t y ! ~to. recall w h ~ tt h ~ n wrm
~ s likt* a t sonic
~vrllttrp0111t "'I h m r ,IF, IVCII <I< af the prficnt. In !hi\ \\'a\. rve can
nl-rtnin cclmc ~ndicabonof thc cvtent iu which the pLarson ha\ c h a n j i d
t n ~ Irime C,tu.;al anntvqit; ~ n t ~ i el sql ; r b l ~ ~ h ~thc n r :trmpor~lorder of rvents T'llis i<
nrt-i-+wr\.\ltbiau>c3 a llasii ttmet ot i , l u s a l rr;lson!nq I < t h t ~a cali+r r n w t
I+ongittirlitla1cttl\i~nswri vnrv OII manv other ~iirnenstonsas wrll. Thcqth prcc(.de ~t.;cltf~ctin time In crtlss-scct~nnal r e s ~ ~ i r cwh1.i-t. h all d a t a art-
~ n c l u ~ it t! ,i ~ lf t b l l t l t t ~ n g * colltcted at cmc.p i n t of time it can be Jiftict~ltto rvctablisli the i~rilcrin
r~-Iiichr t c n t s occur (we C hapirr !(I) f lt~it.rvcr,It>neit~rtlrnaldta.;i~m.
5rtti11,: Thc, logic t h ~ \lonh.itu~li~ial rlt'.;i~n c.ln trr ,~pplii,d in ~ariir.ulnrlv prn.;pt~c~~vt~ ~ltqsignz,i~nablr. trackiilg the rwJcr i t ) which
qc!t~ncs r,~ri::inc from t l h~ 1 5 h l ~controllctl lal,trsntor\. crtztnticm. cvrb~it.;take* place I : Iex,~mplc, ~ \trtdy miqht c ~ t a l l l ~ sa h
(I ~-rtrhs-ztbrt~tmaf
~ f ~ i ~ z ~older.
i i i g Fveqcinc in the ~ ~ ) c i c t yrep>rc?less
, of J ~ P ,m ~ g h tbe
-2
-1
+
5 2
5:
ir rr
7
bccon1111gmore c o n s ~ r vtivv. a ,1 f
M'l~at~t is nccc-wary 50 do tu a n w c r our quc+~tion i s to track diffcrlant ;> fA
cnnwi.vahve crvrr the 70-year per~odand if thev did qo at thc same rate
we roi~Fdconcluclc that thc c h n n g ~tn more conwrvativc posihon is '3
pff~bct.That iq, srorzing older re.;llltq in pci~plcbecnrning
rlr.;~c,l~~rrr?rr~t~tn/ -.
-
-- .d-
mvrc pc~litjcalfvcrlnservat~vi..
In t cxampl~,, l'rtbl~. 7.2 show\ Iliat each cc>liort tl{$corne>morr i L
<: :
histortcal periru-i. Quite <imply, timr, not historical puriod, prod~lce.;
chan~c,
-:.: -
j
:,. G
-.
k l n ~ only
i onc cohort hctln backed (e.g, tliohr Imrn In thc 1950s) we
would h a w observed that as tlwy p e w older they bc-came more LL
, .
in thr* 14hC)s and 19705. To d istinpiqh behveen dcvelopmrntal (ageing) .-.
?
tE~rough1930 and 2000. Differences in the cnnsesvatisrn of thc agc groups T ~ m e1 Intervention Time 2
in, say, Zfl[)O is becairse nf period effects rather than because some people
are older. The nldcr group in 2000 has 'always' becn more consewative. Measure variables ?
1
Measure variables
MULTIPLE
POINT P R U S F E C I ~ I V EP A N E L DESIGN
This simple design (Figure 7.1) requires the collection of data at two This design is similar to the simple two-point d e s i p but simply involves
poinls of time from the same sample. We may make an active inter- more data collection points (Figure 7.2).
vention or we might simply observe the effect of naturally occurring We might h a w spccific interventions between any set nf data collec-
events between time 1 and time 2. tion points or simply collect information about relevant intervening
The simple prospective panel design is very much like the top row of evcnts between each measurement point. Theoretical considerations and
the classic experimental design (Figure 4.2). That is, the design has an previous research would affect what are considered 'relevant' interven-
'cxpcrimcntal group' in which change can be measured. However, the ing events.
design lack5 a randomized control group, which produces the problem Like the simplc p a n d design this multiple point design reqlifses
of not knowing whether any change i s due to any 'intervention' by the sufficient variation in the initial group and wfficicnt diver5ity in
rcscarcher, thc elapse of time or somc other influence. intervening experiences to enable ur to construct comparison groups in
The prnblcrn af not having a conlrol group can be alleviated somewhat the dnalvsis stage.
where a wffiriently large and d~vcrsesample is used in the study. Such a 'nit purpose of multiple data collection points is to:
sample rtllorvs us to drtt~rmincif, over timc, somc pcopIc experience
'interventions' while others do not. Adopting this approach can produce
comparison gro~lpsthat provide a useful, but limited, way of examining examine long and short term effects
thc effvct of specified 'interventions'. For example, we may be interested o track when changes occur (e.g. in 'career' analysis)
in establishing whether or not having children leads to an increase in thc plot the 'shape' of any change
gendered division of labour in thr home. We wnuld begin with a sample tdentify factors that precede any change ((or nun-change)
* Dropouts
I I
Inlt~at I + Dropouts
sample size
I
For cuample, f i t * rnrtv wish 10 (rack the wnv in which marriages change
nvvr time. Wc could lonk rlt these ch~ngesalong any nurnbrr uf
rl in~cncinnqsuch sr level.: o f intimacy, t r h o eucrci.;es ptnvcr, the wav in
r v h i r h ilomeqtic Ial7r~rrr is .;h,lrt.d, happin~*5s,Eer el< n t ctmflict and st,
forth. \Vc cn~ilrlIrask chnr~gt,sIn t l ~ e s casp-ct.; of thr rc.l;ltionship trlrm
cvhrn n ccluple b r ~ l n sto livr t o s t h c r a n d at regular it~trrvaIsthrrralter
- SJV t v e r v two ycar5. WP W O L I I ~ a l find ~ out about otl~ersliai-rgcs
nccu;rinR <ctr<cm o u r tracking \.int5 Ir c. Icvel of ivnrktorce partrci-
1 Final
sample size
1
patlcln, birth of i h ~ l d r ~ nPmV n , pcnple ~ u t -azl ~a parvnl nr p a r c n t - ~ n - l n l v
lolnlrig the h ( ~ ~ l ~ l i oorl dcllnngrd
, econnrnic circurn.;t,incc.;).
II w c wantct-l ti) cxplnrt. !tic effect rrl having n child on marital
l i . ~ p p ~ n\vtn ~ wC O ~ I ~ Lby
~ , hnckin~ couflcs rt8gularly rwcr a long prrtod,
rl~\t~n::rri*hIwttvt-rnn <hurt, rntl~li~irn and Eunq term t ~ l f ~ ~ cIZfe
t \ c ~ l t l r .;PC
i
thth t t i 7 5 , ~ n ddii\rnk that n two-point dr\rt.,n simplv ~ r u r ~ not l d pcrmit TI J2 T? T4 T5
+ Dropouts
--I - + Dropouts
ln~~ral lcDrOpouts
\.l-ht.thcr o r not ;r simpIe trzt)-point (prc-tcqt, pclqt-ttl.tl cjr a rnztltiplr.
Dropouls-
I
p i n t design i q trscx1, a deci<ion will nrcd to be madc about what fo tin
I
+
The main feature t r f thr. multiple cohort ~ L ' S I ~isT t h >prt>,~rl~ rif cuhortq
T l ~ i c cprcad permit5 cornpatiqons hutt\.~cn cohurts and t h u ~hrllp.;
a g ~ i n gand p c r i ~ dcffccts.
d i ~ t i r ~ q ~ t-rct~zw-n
li~h
It i. not nccessi? tn est;lbll~hqeynmte cohorts at thc beg inn in^ of a
v . lotlg as the sample h ~ . ;a suffrclent divrr.;~tvnf aKc grclupq the
s t ~ ~ d 50
colrort~rat1 bc c..;tt~bli~ht~rl sutrst~q~ientl!. at t l ~ edata a n a l y + ~+:aKc
<
the year 2001, thow who were 50 vear old< in 2000 are now q5 years old.
T/ic sample can tell us nothing about the new 50 year old.;: thaw who
h~rn 511 i n 2005.
A Ic5* common but prltc.nti,~ll\-pn\verful drsign i s onc that usta.; nlulsiple The distinctive aspect cjf the cohort ~equentialiief;iq is that twcr limp
cohort4 and ha< a form of panel rcncrval elver hmcn(t~gurr,7 6 ) .One of cohort< drop out of thy studv. This 1s similar in way3 tcr the p a n c l ~in the
th? sllortccrminp I ~ F thc p n r l r l e ~ i pi~ that i f the panet laqh for a n rcitnting panel desim cxcept that it iq coho* that arr rr%placrdrathcr
~.utt>rntlc(fpt.rirlt? tlie 1vIic4t~~ a n t l lajir.: ,ind the pnt.1 tviF1 ntlt ~nclude than whole panels. In tht* etarnple in Figure 7.6 tve drop a rof~orta f c r
\'ounccr pt-ode. Con.;cr(zic~t-ttl~,,ah thc. panel ilxe. nt+thinq I.Ivarncif they rpach the agt. of 75.
nb(wt czmt11rnporL3r~ vcwnc pt~opltb For euarnplc~,r\.e r n l ~ h tcnrld~lct.I ttre oltlcr;t rthurt drops out oT the 5tudv thcv arc rcplncc*d with A
I ( m ~ i t u d r ~ i .ai ll u d ~(71 ;1gt.ini:. I\'ta w a n t tr, knr3~rw h a t i s 11 tikc, t t r Crutv nrr\., vclunc (70 y a r old) cohnrr. This newbcphnrt In t'ffpct \lrllst~tute>f ~ r
rdd H~IW do old*-rp c ~ p l rthink
~ al-rortt ,rqrlnt: and rn~>rt,~lit\.' 13rl pcorlc'k ~ y h at~ tthe f~rqtd n g r w a s the y o ~ ~ n cohort.
~e~t
V I ~ \ V exycriencc,\,
~ , ptrcrptitin* and [car>ch;rnge as the\ gro\\ older' 1%'~. This design rnablrs 11s tn do three things:
11avtl plenty sf time Zn track c h a n j p and decidc tn incIi~rlt= cs>hr>rt<rrmho,
tti thc \,e,$r20fl13, arc (isrd 51). 5 5 . 1d1. h i , 70 and 75. ~ V can P at trrrtl to h , ~ c k 1 It enable< ur, in each tire-war sur\*ey,tu hak'r J. nerv crocs-qectlonal
p . c 7 l > l t l throl~ghto t hc year 207;. 5 n c e n c have- d ~lFt*rent
thing for those who were older except that the panel for that cohort Figure 7.7 Slr~tpler t ~ t r o s p e ~ +prtlel
t i ~ ~drszgrr
~
would run for a shorter time. These panels are rcprcsented by each
row in Figure 7.4. These pancEs allow for the standard analysis of
individual change. The multiple ~ n e l also s prnvide multiple cohorts
to explore thc c f f w t of ageing as distinct from wider societal changes
(see carlier d iqcussion on the advantages of multiple cohort studies).
?I It enables u s to plot thc changing experience of an age group over : Recollect~onj Recollection j Recollection : Recollect~on : Recollect~on Opt~mism
timu. Because wc have rolling cohorts (i.e. we havc information on 5n I of opttrnlsm : of optim~srn : of optlm~sm j of optlmlsrn j of optlmlsm now at 1
year olds in tlie year 2000, 2005, 2010, 2015, 2020 and 2025) we can j at age 50 : at age 55 at age 60 : at age 65 : at age 70 age 75
track how the cxpcrience of being a 50 year old changes over time.
Sirn~larlv,we can track what it is like to be 75 years old over time.
Does the expcritance of this age gmup change over historical time?
The problems tvith this sort of deqign for thesr sorts ( ~ mntturs
f arc
obvit>us.There will be a g r ~ a dcal
t of distortion (deliberate and other-
wise) in those recollections. Not only wiIl pcoplc not bc ablc to rccall
So far all the designs have involved following a sample forward ovcr a accurately how they felt, but we know that recollections of past experi-
period of time. This approach has many advantages as will be shown in ences are interpreted in the light of sr~bscqucntcvcnts and cxpericnces.
Chaptcr 8. But it also I ~ a sdisadvantages to do with matters such as cost, All that these sorts of designs wo11Id tell us is how people construct their
sample attrition and the delay in obtaining useful data abnut changes in past
the sample. The primary advantage of longitudi~~al designs is that they But this problem is not grounds to condemn all retroqpeckive studies.
enable 11s to establish the sequence of events which, in turn, enables us to When dealing with certain events, quite good and reliable information
make strnnger asscrtic~nsabout causal order and causation. can be nbtahcd abnut the scqucnce of past ~ v c n t s- espccialIy if they are
In some studies it is possible to reconstruct data over time by collecting significant events in the person's life. From these accounts nf the
all the information a t one point of time. It is possible to reconstruct the sequence of past rvents wc can r~construct'careers' and in so doing
degree of change and the sequence of ~ v c n t sby asking people to rccnll solve some of the problems of temporal order so necessary to iievclopin~
what happened, and when it happened. causal explanations,
We might, for example, ask people about their current situation or For cxample, we might want to examine the influence of family life
attitudes and then ask them tn provide similar material about an earlier stage on the workforce participation of women. We can ask womcn
point in their lives. We might, for argument's sake, want to know how retrospectively about the timing of particular rvents. We would ask abnut
the birth of a first child affects marital happiness. ,l\ retrospertive design the timing of their marriage, birth of first child, birth of subsequent
would ask a samplr of people who have Jiad their first child about their children, divorce etc. We could also ask about any changes in lr=velsof
current level nf marital happinew and ask them to rccall how happy workforce participation such as when they entered the workfnrce,
they were beforehand. This d w i p is reyresented diasrama tically in whcthcr and \+-hen they lelt thc workforce, any changes in workforce
F~hqire7.7. status (part t i m ~ / f ~ t ltime/casual)
l or the hrning of work (weekend,
Thc same approach could bc applied in the study of the experience of evening etc.). We couId then plot changes in family life stage and changes
growing oldcr described earlier. We might take a group of 75 year nlds in wnrkforcc. participation to help identify the sequencc of cvmts. Given
and ask thcm to recall how optimistic thcv felt about the future when that such tvcnts arc significant ~ ~ e nfor t smost pcoplc wc can reconstruct
they were 50, 55, 60, 65, 70 and how thev currently feel. This design is the temporal order reasonably wcll and build up a picture of work and
r~prescntcddiagramatically in Figure 7.8. family careers for women. The study could also ask these women to
1YI'I-S OF LONGlTUT)IU hl. L31:SIGN
Methodological issi~es
same things differently becnusc thcjl, a s obqervers, have changed in some experimental designs. In other words individuals with extreme scores
way. Thtrc arc famous examples in anthropology where diffcrcnt on wavp I will tend, a t wave 2, to 'regress' to the mean. I-Iigh scores
anthmpnlogists have 'seen' the same community in very different ways. will tend to get lower and low scores will tend to get higher. When
Red field ( 7 930) and Lewis (1451) both conducted ethnographic studies of analvsis of change is conducted at the irrdividual level rather than a t the
a Mexican village 20 years apart and 'saw' the community totally differ- aggrrp~televel wc will mistakenly attribute some individual changes to
ently. Similarly, Mead (1943) and Frteman (1984)both studied Samoans a n intervention or some other factor when it is simply an artifact of
some years apart but nbscrved the culture totaIly differently. Givcn the shtistical regression. At the aggregate level we should not make this
unstructurcd iiature of the anthrupolugist's methods in these cases it is mistake since movement From low to moderate scorcs will balance
difficuSt to be certain to what extent thc different obscn~atinnsreflect movement from high to moderate scores (see discussion of truncation
changes in the communities or simply reflect the different ways the effects in Chapter 6).
observers interpreted their nbscrvations. in experimental designs it is relatively easy to distinguish change due
to statistical regression from change due to otl~erfactors. Since both the
experimental and the control groups will be equally liable to statistical
regression, it will he the drfrrence In thc 1cvc.l of change (rather than the
Since a key reason for conducting longitudinal strrdies is to measure
absnlute amount) that will reflect the influence of other factors. In panel
change it i s crucial that we observe real change rather than simply
studies the absence af strictly comparable control groups makes it more
measurement error. Using unreliable measurement instruments will
difficult to measure the extent of regression.
cause us to detect 'change' whew h e r e really is no change. For example,
we might measure a person's level of prejudice at TI and again at TI. If,
l~owevrr,the measure of prejudice is unskabIe, how can we tell whether
any change is due to the unreliable instrument or real change? Equally, i f
thc measure .of prejudice is insensitive it may fail to dctect change evcn The concept of mortality - thc differential drc3ppout in the experimental
h u g h there has been changc. and control groups - was discuqsed in Chapter 5. Different outcomes For
Measurement error can lead to serious overestimates of the amount of the two groups can be due to diffcrenkial dropout at some stage betwee11
change at the individual level (i.e. identifying individuals that change). pre-test and post-test rather than the intmn,ention.
Using unreliable instruments cause many individuals to appear to The same problem arises in panel studies since dropout will not be
change at least marginally. In descriptive analysis the problem will not consistent for all categories of people. The importance of the prnblem
be as pronounced a t the aggregate level. IS we look at the mean scores of depends on the particular analysis (comparisons) undertaken. Identify-
a panel ur of subgroups of the panel then the effect of mndnwr measure- ing the biases introduced by the dropout and then statistically removing
m e n t error will be muted. Those who appear to change in one direction the effcct of such biases can minimize the impact of differential dropout.
will cancel out error caused by penple who appear to change in the other
direction. For example, in a panel of 1000 people the average score on
n measure of prejudice that ranges from O to PO might be 5 . If, due to
,measurement error, 200 people show u p as becoming less prejudiccd and
200 show up as becoming more prejudiced (and the degree of change in PANELATTRITION
one direction is the same as the other) then the average prejudice score
will remain unaffected. A major potential threat to external validity in panel studies is the loss of
BI! g~aphingthe change scores one can see if the ubserved change i s cases over time (attrition). Thc longer a panel study exists and the more
due to random measurcmcnt error. Tf thc observed d ~ a n g cis due to waves it has, the greater the danger of attrition. Attrition is a threat to
random measurement error the change scores should form a norma1 external validity because it is not random. Some types of people are more
distribution each side of no change. likely to drop out tltan arp others. As dropout is compounded across
waves the sample becomes increasingly unrepresentative.
There are several ways of responding to attrition and its threat to
external validity. One way is to replace dropouts with people with
The concept of statistical regression of extreme scores to lcss extreme similar characteristics so that the known biases are redressed. This is
scores over several measurement waves was discussed in Chapter 5. more easily said than done. We may not know what biases are intr*
Kcgression is no less a problem with lnngitudinaI dt!signs than with duced by dropout. Also it may not be possible to recruit the appropriate
ncw <nrnpItx n.tmll,rr~. I f wta 115crcal~l~icc-mcnts thcn the data for thew tht- .;tl~dy7ni):ht nnlt bc* cornparabl~tn Ihosc nrnorlji ~.Fliti1 rcn n t tnarri,lgt*
nr*w rtacruils r \ . ~ l l bc ~ncnrnylt~tc,
thu\ Ilrnit~nl:~ i n . ~ ~ to
v szg~rt'gdte
r~ ~vI'I brca kdown more gcntbr;ilIy
anal! ~ 1 . i ( C ~ I Chaptcr
- 4). t-lcnrr\-tr, tlrere I* no teaqtln rzqhr rrinditiorlin~shnulrf aFfcct the r(*ltl-
I'his probltm m n he a d d r c s s ~ db v w t - ~ g h t i nthc ~ samplt* to take t ;r,rr~nl*lt*.;.
tic?n.;lfrl~b~,l.tir't+c.~ M'r m ~ f i h w
t , ~ n ttn .;tu>if h i ~ hcon t l i d tl~vorci.:.
a c c t ~ t ~ noft .rttrittm. Wt.i,shting mtailc ~tati~ticstlly boo.;ting tht. propor- produce pcwrrr adimtrnent nnionc chi tdrm Illan Itir\, conflict el ivorcc.;.
tion ot ntndtxr-reprr~riitc~fSrrvlpF In thta sarnplt* so that t h e ~ rprt~pnrtion WhiZr prticiyatinn in the \ t ~ l d y rnw rnoilrrntc tt>c IrvcF nf c r m f l ~ c t
reflt>ctq thpir pmpnrtirm in tlrr p n p u i ~ i l ~ o(C'hnpter n 'j). I f tlilc srlmplr i.; k h u c m di\*orcrdparrnts and may r 7 f f i ~tlie t ~~flustmcn ofchilrlrrn
t thcrt*
\vt.i<htpd ~ r c ~ i r a t r lthcn v i t i n n still prn~~iiit*
a uscfuI bask tor gentlr- is nil rc.;lrm why y a r t ~ c i p a t ~ o nill ,iltcr tlri, l l r t b b c t w c ~ nconllict aird
aIi~,?tion. adl~~strntmt. F r m thr>~ighthr absnltrtc levclf; of conflict and ndlustrncnt
Thc probl(*nis cauwd bv snrnpl~n t tritiun n l w can bc addrcsqed hy may not r~~fll>ct thi> Icavcl.; in the ~ r i i i e rpupulatlrm of d i v t r r c ~ sthc
acknt31vledgin~thr I71awq crist and nctqepting that tlici* will aftect s~t;- relnt~onsliiptlt*tw~cn1-onflict and adlustrnmt +fioul~lremain t~~iaftcctcil
n i t ~ c , i n t lu~u t i~paClt!' tn genrralizc ~t rhc ~ l ~ ~ cca I~C Ti C~~I. l~o w h ~rv e rn. i Cot~.;cqi~entl\, fIt15 n,l,rttcvrd~lrjcnn br ct-ncralii.cvl ttl the, ~ d e pt)p~tlatinn.
r
hhc I * X ~ ] R I I R ~ O ~Irl7rl
V b ~ n sr n ~ ~ not h t rnattcr. 11, for cxamplr, a panrl tn tt~osec i r c u m s t ~ t ~ cwlicrr
cs panc.1 cnndjtioning 15likely to I7ecomr a
studv is bin.;cd bec(1i ~ wof ttii. diqpr~jportionatedrnpout of voungcar serlou< probltm for thtx rxtrrnnl vafirlitv uF a .;luJy i t m,Iv be appropriate
peoplc (the! m o r e ;lrr)i~ncFmarc nn~tart. thcrdorc harctcr to ktytbp track t r f tc) E)IW diffctcnt dr.;i~m.A rc*pt.attrf cmqs-stnct~cmal d i , ~ l ~mav w nfkr a
nIocrtinic), thi5 age bznq only mnttcrs i f R E P is linked ~ G the J variables t h a t suit,~lllca3tcrnntivc. rotating pnnrml cle5i~n ~ l s op r r r ~ ~ d eas way n i
are Ilcing tnu;lrninrd. If' X affracts Y, rtynrillc.;s rlt agp, tlwn the age bins estimating thcb cvtent to which ccrndltit~ri~ng might bc sffrctins r c s u l t ~I t
L ~ U P I(> attritlnn doe\ not mLittt3r. 1 % "can ~ cclnildentlr ekpcct t h a t thtl dm=c this by comparing r c ~ ~ l l in t . ncn. panel.; w r h tt~o.;~, of 'c~pt-rlenccd'
rcyal Ionship br~trvemX nnd Y izrill PX i.;! in thta w11lc.rpopul,~iinn,clespitr. pat~t~I.;
2 1 ciiflc~r~ncc,~.
~ in the* ~ i g cpro'filc~of t h t , ramplc ,~nclp c l p ~ i l ~ ~ l i l ~ t ~ .
In thwe circrrmstan<cq in { v h ~ c ha \,irnpIe Ilia< ictr~~-ti is proriuced trv
attrition ma\ attect t h t h v a t t r r r ~ sWP &jrl>pxarnlnlng, \z,tA nccd tu rcmc.rr.c
strttiql.ic.~llythe effcct of thiq b i x and wta if llic rclatinnqhip btatween X Panclq can br.cclmc un rrprcwntative i ( thr population From which tllc
and \ rchrnalns.Iree Ch,?ptefi Lp and 12). I t the ri*lation4wptwtwcrn 4: a~lci pancl i.; clr,~wnchances. Ccrt'lin h-pth$of pt-uplc nj,>\ Ieavc the arcv
1' rcm;lltls aftt+r removing thc cffect nt any variable cln w h i c t ~there I;. wlicrc the r;tucly i.s k i n g cnncluctcd. As a n*<uttthc pancl mnv orFcr-
snmplr hia.;, wc can w k l y grnerali7r to tht. wider pnpulatinn despi tr reprcscn t thrw leaves.. (outmigmnts). Simi larlv, n r w tvprs of prt~plemnv
sample bia.;. movta rn to tlic area and tht. pnnel wilE con~c*i[~rcntlv r~ndcr-rtbpremt
thew inmipatlts.
Whrrc. pcyjtilation changes arc c r e a t i n ~a substantial problcnl for tlir
prcjtbctthe srdtrtion lc to 113t*r j ~ n a m i ccarnplt-. Dvnarnir sarnp1t.r can bc
The csrlier discc~s<ioncin the i.ff~cts rd test in^ ~ n it< d irnpIir;ltinn~for a c h r c u ~ din ~ r ~ w - ruavs.al Yhcy are: rotating ppant.1 c J ~ ~ r ~ decign<
ns; thnt
internal validitv ( a t t r i h u t i n ~chnngc to a n inlcrrpentitin or othrr facttjr aIlrrw for r~placemenlcrr sarnplc si~pplcmcntLit ion; ancl repca It'd crow-
whpn I I is ~ ~ a l drte l v to s i m i l v kin^ qtlrdiedl nl~rrha.: implicatinns irlr sectional ~ ~ r n - c ~ ~ ~ s .
ekterr~al\,af~dlty.J i j~nrticr;mttr.rt in a l o n g i t u ~ l m ~ l
study p r i > d u c ~change.
* I f w r use dvnarnic panels wc* need to hu clrar about the scmrcc of nnv
(because i t alrarts participants to ntattcr%the!, w o ~ r l dnthrrwise nnt think changc in crtrr depenili1nt vnriahles. Changes nn the a g ~ r c p i t elcvel in ;l
about) i t is diffici~ltto <;1rthat c l i a n ~ t wc?b~er\,cdin thth .;tudv trill take C I I R J I ~ ~ C pant1 mat- rrfltyt rinr of ~ \ Y tI l ~ i n pc-lianc(, In trrdrivri~rnls,tor
p l a c r In the xvldcr pnpl~lation. c h a n ~ c qto the, c.clrnpo~iti.nnof thr panct. l-tjr t~x;lmplr~,~ v cr n ~ g h thi-
H o w c \ r r , th~.; is mnrc ol a problrm ,it th~.descriptive thin nt thta t r a c h ~ nlevcnl.;
~ of tife hnilrfacticin in a p a r t i c ~ ~ l local a r rcgion nncl ohscrvr
P \ ~ ! N L I ~ L )Icvcl.
T~ TF ~ ; l r t i i i p n t i t ~inr : ;I pnnsl *tudv afftact.: the brahavicr~ls that, 01 rLr tirntg,thy i l ~ ~ t ~ dIVVC~I q p of 11 tv +,I tast,~~-titln
dc*cliiltw. m h ~ cilcd inc
< ~ n ;~ttjt~rdc-
d r l t p a r t ~ c i ~ n n ~t t s i \ r i l l n1,rI.e it d,fi~cnlltto j:clncrall/r. from ma!' br Jut, t o i n d i ~ i t l t ~ , i l <bt#comin,gIr-\ satiqtlt=tlrlr tv ~lt'uplc\v110 h~c1
thr. p i n ~ tc,
~ l thta pcjpu [ ~ F arm.
I H r ? ~ c \ ~ ccrn, n d i t i n n i n ~sh{bil not affect thr high I t v e l 5 c1I life sati\tlcticln Ii%.i\'ing[lit> ;lrrn;l( ~ n dt h ~ ~ r c ' 1t I~l v~ h panrll
tvav lb~ctclrs ,>rt>linked to nnc .~nnther.For euamyle, in c1rdt.t to rxplvrc At thtl same time ptt>ple lvitli low I t ~ 1 7 r 3 5 r r f < a t i ~ f ~ ~ z t irn~l!, o n rno1.c
the rnlpact ot thr 1~rcaLclnu.nid nlarriaj:r irn thc wrllbclng ~ i c h i l r l r c nn o t * into ~ h area c (and thu. ;Ire rikcruited into o u r d v r ~ a r n ~y cn n ~ l ) Ij~.clinint:
.
miglir t r x k fnmilies w h r ~are t1\pcrienclng marriagv Lbr~akijown.li being aggrcnKate Ic\.c,lq nf li fr .;atisftiction ~ ( 3 1Id1 bc 11111' entnribl v to changcs tn
pnrt elf the sturtv makr.: parcntk and ~-1i1ldr~n morp nlsrt to mnttcrs r l t thc crmrnyo~tirt>nof thc panrl , ~ n dnirt tu c h a n ~ t ~i n\ 1I1c lettbl.: r ~ Iilr i
adiuqtrnc~ntat711 ~ ' ~ l l t l c , ~t hn tq~ nthe lc>v~I< o f a t i l r r s h t ~ n !we c)bwnreIn s a t i ~ l a c t l ~c nx p ~ r i ~ n c!.I,~ ~icnlr l r ~ ~ t d ~ t a i ~
15SUES IN LONGIDU I'I N AI. I3ESIGN 139
Practical issues. thrldren o r wc~rkplnctw that it thc rt-qpondcnt mores rvr can ctwtact
In both the clrsicn and rutuuhon stage4 of R longitudinal study thcrc arc
I tlicw clthcr backlip cuntacts to find the* rcbcpnndent. Public rcctlrd.; can
also be uwd T ~ l c p h ~ directvrie~,
~ne e l t ~ t o r a rolls
l and ~ u b l i c l va\ratlabl~
m a n s practical rnattcrs to scrl~re.Somc of these arr to d o with ma\irnizin~ records from IIK.~II cr~uncils can help IOC'R~C r ~ s p o n d ~ n\vhn t ~ Ii;1ve
internal and cxt<*rnalvalidity. Othrr.; h a v r more to do with tlir cost OF moved. Of course, i t is easier if panrl participants tell 11s t11c.v hnvc
implerncntin~~ i i f f ~ r c idesigns
it and thr practicalities of gettiill: tlic siudy rnovcd. A t each intrrview and any othcr cnntnct i t is helpful to provide
completed. change of address notification cards to participants to mail back tcl the
.;tudi manasrr.; i f thcy move. Keeping u p c o n t ~ c hetwtvn t suntcv wJves
is anither helpfrll strategy. A useful tracking device is to send n birthda!,
card to participants. I f tl-tis card i> 'rcturncd to sender' we arc atcrtcd to
I
S ~ n c echange i< thi. core focus of Inn~itudinalstudies rve n w d trr be sure the need to st,lrt tracking them down ~ n ~ m r d t a t e l Thc y . card dl.;(l pro-
that obscncrl c h a n p ~reflects real change rstlrcr than chanqc- in tht* \.ides a n opprlrt~~ntty 'to scnd a c h a n ~ tof address card and ;tct.; n s nn
data arc c n k l ~ c t ~ Idt .iq eqsential thcrcfnrc to ~tnndardizcthc way data are acknou,ledgrmmt of appreciation fnr part~cipntingm the httrdy.
cchlccted for thi. ~ ~ i r i o uwave5 s of thc study. I t is necessary to stan- To srducc n ttrilticln i t is essential to maintain the respondent.;' motiva-
dardiw tlic rncthod hv which data a r cnllccted, ~ the way qitcstic~nsare tion to continuc to participate in vrdcr to rcduce attriticln. !'mvid~ng
asked, the order in wiiich thcy are asktad etc. Wc need to clcvclop clear incentives such CISrnnney for the t i m ~pcople qpend on the s t i d y, .;malt
and cnns~stcntmlc.; n h u t what to d o in particular s i b a t i o n s and rcli- tcrkens of thank< or summaries of key wrvcy rcsdts can hclp achieve
giously f~llorvthcm. The time nf vear at which data are crjllcctcd may t h i ~goal. WhiEr rnnt~rlnl incentivw can hrlp maintain participation,
nccc! tn tv ct)n.;r~h=nt fc3r each wave. Bus\- times, glmmv t1nrt.s n i the Frobdb!v thy rnrrqt citcchrre ~ncentivc I < for pcuple to f w l t h ~ ttheir
\'ear, tirncc (?I .;!rtxsr { r . ~ Christmac,
; ~ n rdj t financ~alyear), \.ac,ltilrr~timp ~nrtisiya!ion r< both important and cnn<bttclivc. The morc pvoplt. !PC!
Inay all atfect thc k ~ n d sof rcf;pon<r<wr PI. lhev are participant.; ~ n pda r h e r s in rc*sc;lrcli ra tlier than 'sublcct.;' the
Jt is not nlwavs pos5ible to a c h i r v t ~ t ~ i n d ~ l r d i z a t iThe
o n . Ivss .;tructurcci better. T ~ nvcrridillg
E ~ ( 1 should
~ 1 bc ta nmkc parhc~pationin t l ~ t study , a
the data collcciioii method the more difficult i t is to stan~iardizrpro- positive exp~ritlncc~ rt>t1-trrthan an unwrlcnme burden and intrusion
cedures. Whl*rt* more rluaIitahve and subjective techntqucr; S L I C I ~ ds (Dillman, 1YTH).
ohscn~aticmand ~rnstntcturedintervlcws are called for, thc ~ r c n t c rthe Attrition is a t - i i ~ g ~problem
r in somt- kinds of st~tdiesthan othtars.
~ossibilitvthat changcs may be dup to thc FPOFIC collect in^ thr. data and Sh~dicsof peoplc w-hcr are highlv mobile, marginal or membcr.: o f
the rnrtlmdq thcm~r~11.e~. 'extrerne"roup5 can cxpcricnce high attrition m tes. For e~arnplc.stutlieq
C o m p l ~ t eqtandardi7ahon is not alwavs dmirablc. For examylc, it can tlf parenk aaftcr clivcirccc inrariablv crp;lte the probPern of m a ~ n h ~ n i n g
interfere with our r p ~ t a r c hgoals in lifc crrurse studie5 in tzthich pt~ople contact rvith non-rcqiclent parent< - especially non-resideni f ~ t h c r < .
arc tracker) as Ihry grow older. In longit~ldinalstudies peuplc's clrcum- Longitudinal s t l ~ d i c sof potip': such as criminals, drug USPrF arid some
stances chanjir 417 we may necd to aciilpt our indacato~sto reflt=ctsuch S O ~ ~ ofSuncrnptcrycd people havc very high attrition rates induccd by
changes. In thtb same way that tlic indicatnrs of i n t e l l i g e t i ~nrc ~ age high levels of niubility, illness and dcath. Panel studics of prnple
specific, sc) our ind~catorsof change may also need to be l i f r stage e~periencinglifr transitions can also havc highcr khan average attrition
specific. Mcaqi~rc.;elf family w e f l b e i n ~and cnhesion mav well nccd to he rates, Researchc*rs investigating young, pcnple becoming ad (11 t, ynung
adapted according to the srape in ithp farnilv life cycle. pec~pleentering Ihe labour market, per~plcafter divorce, or Itho<e p r o p l ~
In o n ~ n i n qv n v 1 studies there rq nftt3n a cnre oi standard rncazllrtSs who leal-P the IsElnur force all face thr prrrblrm that people expcrtcnctng
cr~llectedin t ~ n c rz7,;x.c
l~ ot the 5hdv. In ,~clditittnmodules ~vhichd l ~ n g e theqe transition.; a r c hiqhly likely tn c h n n ~ nddrt-5.
c As a r c ~ u l tihcv s;ln
frcorn time t r ~timr rErts.IISO f+equrntlv i n c l i ~ d ~ d . be lost to tlic panel studv.
Sample attritioi-r prod uccs problems with both sample size and rcprcwn- Panel surveys can impose a considerable h u r d ~ non participants. Thi.: is
tatir,cness and tf~tlrt*fnremust be rninirni f e d . especially so i f rach wnve uf the shrdy i s closta togcthpr or i t inttkrvitaws/
To rnin~rnirrattrihcm we muqt kecp track of panel r n e m b ~ r sTrackrng qir~5ticmnatreqarc- Icnqthv or mtnis~vt*.I-1~a1.yrespondent Ptlrdrn rs n
dernnnds considcrablc resources. A rangr of t c r h i q u e s is ttwd tn a s ~ i q t n a Ims of data quality
problem b~caustli t can prtduce hirh a r t r ~ t ~ oand
w t h track in^ prclp!c. Additional infnmmntion is collect$d a b u t parents, as the burden increases.
140 LONGITUDINAL DESIGNS ISSUES IN LONGIDUTINAL DESIGN 141
Therc are a number nf ways of reducing rcspondent burden. Some information needs only to be collected once (in thc first wave} thus
questions d o not need to be asked at every wave: every second ur third enabling us eithcr to collect extra informatinn in suhsequcnt waves or to
wave may L7c sufficient for some questiims. Alt~matively,a set of qucs- reduce the length (and cost) of subseqtrent data collections.
%ionsmight be asked of only half the panel and a different set of UltimateIv the cost efficiency of the panel design will depend on the
questions put to the other half. Although we have f ~ w e people
r answcr- difficulty of maintaining the panel - how much time is needed for
ing the questions we map stiIl havc sufficient nrtrnbers for meaningful tracking and the cffnrt rcquircd for 'call conversions' (convincing panel
analysis. membcrs who initially refuse to continue).
The cost will also depend on the method of collecting the data. Of the
three main methods of collecting data in panel studies - tace-to-face
interviews, telephone interviews and self-administered postal ques-
tionnaircs - face-t(>-faceinterviews are the most expensive. However,
Rchuspective panel design? rely on respondent recall. This has two
they tend to yield the best response rates. National panel surveys
problems. The first i s that pcople will recimskruct past evcnts in the light
typically rely on some form of cluster sampling (Moser and Kaltun, 1971;
of subsequent cvcnts. This can result in distortion and reinterpretation
de Vaus, 19%) where scts of interviews are concentrated in a number of
by which recalled events bear little resemblance to actual events. For
geographical areas. By this mean5 it is financially feasible to conduct a
example, a study that asks people who 11ave divorced to describe aspccts
number of face-to-face interviews in the one area at the same time as
of their marriage will be affected by the divorce - by the nature of the
divorce, who initiated it ctc. maximizing response rates and minimizing attrition.
Chrcr the course of a panel study face-to-face interview5 m a y became
The second problem with studies requiring respondent recall is tele-
less financiall~~feasible. Despitc cIustcring inter~iewsinitially, in parti-
scoping and reverse telescoping (Menard, 1992). People can rnisremem-
cular areas high rdes of geographical mobility can lead to 'dccIustcri11g'
bcr and think that events l-tappenrd morc sccen tly than thcy actually did
or scattering of the sample, thus increasing interviewing costs con-
(telescoping) or longer ago than they really did (rcvcrse telescoping). A
siderabIy. Fortunately, deveIopmcnls in tclcpllonc based interviewing
number of techniques can be used to reducc this problem. Getting people
(see next section) can alleviate decIurtering problems (Buck ct al., 1995).
to Focus on certain rn~mclra'tllecvcnts and constructing a calendar of
However, changing from tacc-to-face to telephone interviews can intro-
thcse key evcnts can providc anchor points to help remember the
duce instrumentation problems as d i s c ~ ~ s s cearlier.
d
sequence o f events more accurately. Fur txarnple, we might be interested
~ between family life stage and the work patterns of mothers. To
in t h link
provide anchor points we cor~ldfirst ask mothers to provide key dates
such as the year of their first job, the date of birth of each child and so
There is nothing intrinsic in any research design that dictates that the
forth. Since thcse are likely to be dates that can be accurately remem-
researcher use a particular method of data collection. Tn principle, data
bcred we can then, piecc by piece, reconstruct how long after these
can be collected using structured qucstionnaircs, observation, structured
memorable events other, perhaps leqq memorable, events took place.
or unstructured interviews or any other methcd.
However, when it is proposed to utilize a panel design there are real
advantages in using computer assisted interviewing techniques. Here
thew are two a1ternatives available: computer assisted telephone i n t r -
Panel studieq are typically rnt~chmore expensive than a single cross- viewing (CATI} and computcr assisted personal interviewing (CAPI).
sectional study. 'Thr expense of panel surveys sterns from the use of Therc are tWn advantages to using computer assisted methods in
multiplc data collections and the cost of tracking and maintaining the panel designs. First, these designs can involve extremely complex sets of
panel. questions, with many qucstions being relevant to only small subsets
However, if the choice is between say a five-wavc panel study and f i v ~ of respondents (depending on their charactcrisbcs or the patterns of
repeated cross-sccrinnal surveys thc cost disadvantag~of the panel change we have observed). For exampl~,wc may require a special
design is less obvious. Tncre is some evidence (Buck et aS., 1995) that the module of questions for families that had experienced divorce between
cost of panel maintenance may he no more than the cost nf recruiting waves and require diffcrcnt questions for r e s i d ~ n t and nor?-resident
new samples. Furthermore, the cost of collecting the same background parents. With cornputcr assistcd in t ~ riewing
t m e t l i ~ d sall the possible
data in each cross-sectional surrey will limit the amount of other qucstions arc programmed inlo the interviewing package. On the basis
I
information that can bc collerted. Ln a panel design most background of answcrs to earlier questions ihc cornputcr selects which questions
1.UNGITUDINAL DESIGNS ISSUES IN LONGlDUTINAL DESIGN
the interv!c\vt.r necds to ask. This simplifies the task of administering In establishing the gap between waves, respondent burdcn and panel
con~pIcvquestionnaires and reduces recording and interviewing errors. must bc considered. Short gaps can exacerbate these
A further advantage of computer assisted inknriewing i5 that in the problems. On the other hand long gaps can create tracking problems and
secmd and subsequent waves the answers from previous waves for that increase attrihon.
respondent can be fed into the intenricw. These can serve as excellent
prompts that provide respondents with reminders as to how they have
previously rcsyonded. For example the CAT[ or CAPT interviewer can
say, 'Last time you said . . .' or 'Last time you indicated that vou were Sample error is, in part, a function of sarnpIe size. The smaller the sample
planning to retire. Have you actually retired yet?' Not only does this the greater the chance of sample error. In prospective panel surveys
provide respondents with a reminder of how h e y were feeling East time, (without replacemcnt) sampIe error will occur when the initial sample is
but it personalizes the interview and makes the respondent feel that their selected. Subsequent attrition may produce bias and will increaw sample
previous answrrs ~ R V Cbccn taken seriously. error as a rcsult of reduccd samplc size. If the sample is replenished
sample error due to overall sample si7e will be controlled. I-tow~ver,
Nu nr bur nJ wavcs ~eplacemcntwill be based on new mini samples and these replacement
The number nf waves utilized will partly dcpend on the purpose of the mini samples will be subject to sampling error and can therefore aifecr
study. [f we are interested simply in 'before' and 'after' a particular cvent, the precision with which we can confidently gmeralize our result
two waves should be sufficient. If we are trying to track a process of beyond the panel.
change, multiple waves may be needed. For example, if we simply
wanted to knnw how well people have adapted to retirement, say two
years a Ftcr retirement, h ~ waves
o may be all that is required. If wc wan tetl
to track stages of adaptation to retirement we might require multiple Sample sizc creates a problem with any type of desip. 5mall samples
waves to detcrt the ups and downs of the process (see Chapter 7). encounter problems with sampling error and with insufficient cases in
The number of waves will also depend on funding since each wave special subgroups t c ~allow meaningful analysis.
adds comideralsly to the cost of the study. Respondent burden must also In studies of change and transitions, for which lon@tudinal shdics arp
be c~msideredwhen planning t h number
~ of waves. cspwially useful, the actual number of people in a panel who experience
Many large scalc prospective panel surveys adopt an open ended a particular change or transition can be quite small. For example, we
appro ad^. Funding is available for an initial set of waves and decisions might want to use a panel design to examine the effect of divorce on
about how Ic>ngto continue with the pane! will depend on the capacity to children. If we used a general population panel of households the
sccure hrther funding for subsequent waves. number of households that actually cxperiencc divorce in a given period
will be yuite small. For example, on the basis of the Australian annual
divorce rate of 1 2 per 1000 couples (i.e. in any given year 12 out of cvery
1000 married couples will divorce) we would require a very large panel
Setting the gap hetween waves will bc affected by 'political' considera- indeed to have enough couples divorcing in a year to allow meaningfu1
tions such as the pressure to come up with some results. The nature and analysis. Even if we had a panel of 5000 couples nnly hO couples would
content of the study will also be crucral in setting the gap between waves. divorce in a year. I f we followed the panel over a five-year period this
Studies of high frequency events or processes of reIatively rapid change wouId increase tn something approaching 300 couples. But even this
may requit-c only small gaps between waves. The theoretical model number may ccmstltute too small a sample if we wanted to look at
being tested will also dictate the gaps. For example, in a study of adap- divorcing couples with young children. The numbers would be even
tation to retirement that anticipated different phases in the retirement smaller if we wanted to look at fathers of such children whu had primary
transition le.g honeymoon in first 3 months; distress by 6 months; reqponsibili~for caring for the children.
rebuilding new- identity by 72 months; stabilization by 18 months) thc -Where uur interest is in small groups we may need to refine the initial
model would dictate the gaps between waves. panel more (e.g. just couples with young children rather than a random
Where changes are expected to be gradual there i s IittIe point in s a n ~ p l eof hnuseholds). Alternatively it mav be possible to oversample
f'reqiient revisits. The famous '7-UP' scries, which tracked a group of certain tyres of people (e.g. couples with young children) to maximize
people every seven years from the age of seven, is an example of a long our chances of obtaining sufficient numbers rrf pcoplc who subsequentll~
term study with large gaps between waves. I
experience the changes in which .rut arc interestcd.
Thc qi/c r,f thc iniiiol G J ~ Fir~ill
I P ,11.;t> ilct,d to taktb t ~ r c o n nof
t tht. likclv implcmi~ntcd.Sincr th1.s~ principlthq wcrc discu.;scrl in Chapttsr 5 this
attrition %in(-rattrition tv11I Jepcnd tn p , ~ r ton the dumtion of thtn .;tudy section simply hiyhlrglits matter< th.~t arise particulntlv with Iunqitu-
and tF~cnt~mhtlrand irt,<ltrvniv at IV.~\I..;. the i n ~ t ~ pa l3 ~ r slzel uv~llntrd dinal dt,.;igns.
to takv ,iccciunt of cqflRT,TI(YIiittriti~nr.lttT.;.
Missing data
Summary
Missing data arF a problem for any s t ~ r rof analysis but are particularly
Wlulc experimental designs are ErequentIy not Feasible in social science problematic in panel analysis. There are two types of missing data: item
research, Iongitudinal designs provide an important way of tracking and non-response and unit non-response.
understanding change a t the individual and societal level. Whilc longi-
tudinal designs can encounter some problems with internal validity Sourcijs of missinx data
because of the absence of randomized control groups, there are ways of
minimizing these effects. These have been discussed briefly and will l-re
deveEoped more fully in Chapter 9, Other potential problems with inter- I T ~ M NON-RESPONSE
nal validity have been discussed and ways of structuring longitudinal
Ttem non-response occurs where individuals do not respond 50 a parti-
designs to overcome thesc problems have been outlined in this chapter.
cular question. In somc cases this is because they are not required to
i, As il general rule, longitudinal designs tend to have bettcr external
respond to thc. item (e.g. questions about a spouse for people who are not
validity than experimental designs but the threats to externaI validity
married) while in otl~ercases some people may refuse to respond to a
remain and must be efirninated. The chapter outlined these threats and
question (e.g, income).
indicated strategies for minimizing them.
I,ougiludinal designs can be quite compIex to implement and involve
a widc variety of practical deciqions. The practical issues and ethical
issues that arc particularly pertinent to panel designs wcrc outlined. This is a problem that is peculiar to panel studis.' Since panel designs
jnvolve collecting data from the same individuals on several occasions
=mdmeasuring change over that time, we encounter difficulties if wc fail
to collect data frvm the case on the second or any subsequent occasion.
There are turn korms of unit nnn-response. One is where people drop
out of the study altogether. The other occurs in multiwave panel studies
where some people miss some waves but are not permanently lost to the
panel. This is called wavr non-rpstxl~~se
(Rose et al., 1991).
LONGITUDl l\lA L DESIGNS DATA ANALYSIS TN LONGlTUDlNAL DESIGK
Missing data can also introduce bias into the sarnplc. Where iten1 non- Essentially imputation involves constructing responses for the questions
response is systematic rather than random (i.c. certain sorts of pcople are which people do not answer. There are a number of forms of imputation,
more likely than others to refuse to answer questions) unrepresentative some of which are discussed by H e r t ~ l(1976) and by KaEton (1983a;
samples can bc. the outcome. For example, i f higher income earners are 1986). These include the following.
more likely than Iow'r income earners to rcfusc to reveal their income
we a r c left with inaccurate income eqtimates. ConsequenMy, any analysis Samplr nzwi? flpjlroach Tf we do not know the value of a person on any
in which inctlme is a variable will be affected by [he loss of cases due to pivcn variable, a reasonable guess for that person is the average respclnse
item nvrn-rcspnnsc. by other sample members for that variable. With interval levcl variables
Unit non-response will introduce other sampIe biases. Those peopIe we could replace missing values with the mean for the variablc. We
who drop out of a panel are unlikely to be the same as those who remain. could use the median for ordinal variables and the mode for nominal
Loss of panel members may be due to death, gcograpliical mobility, \,ariables. Although we will make errors in our guesses this method
changed family status, or some other type of dtanqe in a person's produces fewer errors than guessing any other value. Where the sample
circumstances. is relatively homogeneous bhis approach may result in few bad 'guesses'.
7 hv prtjhlcm t ~ 1 1 1 1~ r l c ~ p t t ltlliq
~ c apprrwch, hrliccavrr, i < t h , ~~t~redl~iths
sample \ .~ria't-rTtt1ern !hta \,+r~ablt. rtlr tvhrch t h t ~mlc.;ing ~ l a t aart. I-+t*ltig
tlrtirn.~h+rl. Thl.;, ~n turn, rr~illccr;tht* ctlrr'lntion t%ctwccnfhfs v a r ~ n b l r
and r ) t l ~ \,ar~alllta.;.
r This i s a rnori* cnrnpl~~x
I < i - , q r t ~ ~ ~rrrrrl~,.;is:
~~ou metllnd rtf impirtntir~nthat
intnlvcs the 115enf rc!:rr-ion t o yrdicb thc i alue of tht, F e r w n (In thc
I I I : C h 7 ~ivav of cnt-rcnrning fhi.; probll-rn IS tc, u w nliw,lnc 3 al~lcqkarial7lth. C7n thc l,n.;i~ ni a \\.hole set r l t charactt~rr~tirs we
group mcLinsmth1.r than t-ht*overall <ample mcan, In du !hi+ we would c,tn cshrnatr how, on ;IirrraRc, a person with those c't~nrarteri~tic.; nrc~ulcP
l o S ~ I I T I ~ ~ I ~into
~ l i v ~ ~1171- . croup.; on thr basis 11f a harkgr~jilnd varinlllr* a r l r n r e t t h r question nli rvhicli tlicv have rnisqing data. In this rrqpccl i t ia
Ic E. ~;I.~Y. aze ant1 cthnic~tt.)that corrclntes rvrll rvitl-t thc nli<l;ins c a l u t c o r n p x ~ b l tt>c thc
~ grrwp means approach. Thc prnhlt*m with thc mcthod
t.ar1.2bTv \\e wu~tlilthen obtain the mtvn tor tlir- missinq d , ~ t a\~asi;rfilr i\ that 5111ccr e q , ~ ~ < sI <~ u t ~ o n a ctim~lationnlatri\ in rvl~lrhthp
bnsctf
for c.i<.h u t thp.;~,sclhgn)llp+. l-nr c x n m p l ~ ,if nrv ~ v a n t c dto cactinratt, the p.i~rwisrarnibtlind of I i ~ n d l i n gr n w i n g data (st117 below) I.; ~1st.d.thl.ri. v\,1I1
inctr~il~b Inr p~%rjl,lt'w h o cit>clrnrd to ntiswcr t l ~ ( r jncornta qllp~tion i\*l> nt)rmnllv I-rtl home lo%% of caqcs.
micht rilvidt. tl11. <ample itlttl vcrup\ according tn their chtlt~cationIrrcl
and t h l . ~ ,within taach cduca hon !r~,t*l, Furthcr 14 ivide t h t ~5;lrnplc lntn t t i i i For <om? stnblr*charaztcr~sticsrvca can 'carry
mallas ,?nd fen~;tlt*s.For cach cclucat~cinletpelatid ~ e n d c r.;trhfir0~1pw v ci~~cr\nbut%s trnm ~ l nrarlier w.ivr (Kmc rt dl., 1YQl) or check [lack from
cnul~i1ll7t;lintht. rnt-an incomr Ievel for Illnqc ruho did dnwvC'r 1 1 7 ~income ' k~11,~t.qucnt w.~vcs.'Hti~i~cvc.r, flrr variable\ that a r c Irkcl y to t l c n subject
q t u ~ ~ t i r wW. c r n i ~ l i tfind thaf thy Incnme Irvcl I i l r tc.rtjnrt edt~salr\iI t o c h a n c ~ ,lnd
. for nunrrric \ . ~ r ~ ~ ~ in [ d p;rrt!ck~lat
c=< t t can be incnrrccl to
fcrnti It-. Eq VO,O131) F a. rvhilr trrtiarv c.ti~tcatedrn,~lesenrncd. on avcw>:t-. 'c;trrb nrcgr' rerponwdc from ;I p r e r - i o ~ ~w\ a r e . Ovt-r time i t may be
<75,0111);lnd 30 tc?rtl~.U.;ing thy ~ 0 ~ 1 1~ 7i ~ a approach n s w r n'111lld slrnplv pnwlble ttl fill in sonlrl gaps CAI~SI.YI by M ~ ~ V non-rilqp{rnse
I* l-rv Inrjking at
~ ~ ~ b \ l i thew*
( ~ l t 111~~111
t ~ ~.11111-.for rni.;.;ili~value#\ Ior pct,pl{. ,ir-cordln< i ~ ~ a n ~ t!he q ~ r h r - q u c nwt a l m eand r r n y i r ~ ~ nrn~tlpoinl
1111, r ~ r c v i r rrvavca ~
the Crrrrrp inhi tvhich t l i t * ~tcT1. r,,ltae fnr the- rnisqinq uvar-c.
Rich glr-ad\~;lnt,~yt~ of th[\ approach 1% that it t3\.E~gerntr.. t i ~ Jcgrt>r
~ a r,t
hnrnop~t~rlity wlfhin grt>up-- and t l i ~ ~ ccan ; ~ 1 v t ~ t ~ ' ~ t i rthc
n ; l vananre
t~~ r r : f~J.ss.inl:~ i t l t i l Thi5 is not a n imputation L ~ p v r o ~hut
I g ~ r o l ~ i ~/lrru c h pro-
b~t13~rbr~n g r o u p and tor ihr *arnplc nvt-rail. Thi.; c5rror,in turtr, can 1nHntt1 ~ ~ d ac W.IV s nf rnatmgin): with miwing data when ct>n.tt~rctinc scales. I t
the cc~rnal,ltion.;ir*hcn using tfic r-ariablr t ~ which r t h e mi- in^ data 11.ir.t- rntailh c.alculatin): q c n l ~ScorcG bv calculatin< ,I perwn'5 mean <illre h r
bccn cstimatrd tjtd- ;-l~ril~/l? h cr- wItiL.11 !lft*t/ I i ~ i ~ /?~,r,i > : v d r i ftn17Ti7i-r.q.5 ~ 1 p p 0\vta
~ i v t w con-
qtrurting a scale from 10 qucsticms cacll h,iving pn\<lble scorc.: ranging
I f I i t This approach i s rimilar 11) the f i n j i ~ p trl>~ii0 to 1. : Ip e r w n who c-rbt~inrda ~c4,rt. nf 2 oil all I0 itcrnq tvoul J
rncarl.; npproaclr 111 that i t rc-lit*~ on ~livldingtht, +arnpl~ i r ~ t t rL T O U ~cln a c h r e ~t* a total qcnrc of 7n ovcs ,711 10 itpnl.; and a mran SCOT^^ of 2. ,I
the I,J\I< n t o t l ~ c characti.r~\trcs
~r that a r c Itlel! tr, bc corrtll~tcrlwit11 t t ~ c ptJrson who n b t a i n d a scar? t r f 2 on cucli nf c ~ ~ ttcms h t l ~ did t not
r n i s s i n ~data variable. But ~t d ~ f f r r sin that i t does not involve sutl.;ti- ~ i n r w e rthy rtvnaining two items would ohfain a Scltnl wore of l h but a
h ~ t i n gI E W g r ~ l ~
nlcan
p for any rniqsirlfi clata. In<tr*nd,wF~l.nr\,e Incattb a nlcan of 2 Calculatinr: a scnFc <core in t h i ~w a y avoids the prot~lern(11
casc iz.rth m i w v n ~data a n ,I particul,w vasiablc we ~vntaldrook .lt thc rniqsing d a t ~whcn ctrn5huctinc .;caleq.
vnlur on the qamc variablc of the noarcBhtprpr-ding casc In [hat gat7up
and give that snmr ~ ~ I L to I C thy caqe w ~ f hthe m i w n g valur, l'his nlcan.:
that thr* missint: lire5 ;Irt7 rcplacrrl w t h n t.3rit-k of d ifh-mnt ~ r , ? l t r t a <
rantlonllv chrv.*r~from tt.lt7~1nt h ~ t+irtl~rl'r\ipt > t c ~ s cl ~Firc ayprc>acl~ \\'eiqhtrnq I < ;I \\.a\ <,i ,~Rluqtint:a sarnytt- to aHo\ro for yo5sit.lltb Iliac d u c
doc..; 1 1 c ~ t atrcvt ,,~rnplror gri)uT ~ . ~ i r ~ a t , ~ift ~hast v . n o t n f f ~ r ton thc t o ttrut ncv\-ri7';pcln\ta.W c i g h t ~ n,I ~+amplt, chcruld m,ihi. I! rnclrta rl-prcs~n-
stret~gtli of rol.rt~lativns ,~ncl avoid.; nnv Inss of c a s w I h p i t c btxing l . ~ t i \ eof 11it. populatlc~n it is rlcsignpd to reprt*srtlt st3 that rclialllr
~crrnt~it'l~at morenrr,mylcx to taut-cutr, t h ~ >1s a h ~ q h l vd r ~ l r ; h ~rnctlrt>rl
lt~ r>t vktlrnatt~(,in be m,idt* from l l i ~ a<,~rnp!t'to the pc>yi~laticlnFor I-uamplc,
hantllltlg rni-lni; d a t a r l n , l ~ i n c;I wftlatliin trl i\*hicll per cell1 or the p o p i ~ l a t ~ oarta n male and
50 per c t ~ i t.Ire icrnalt~.Hnwt*~.t,r, when ,I <amplo ot 1111spoytllallon
nt qpnrr A relatrd approach rrqi~iresthat wc cnlculnte the tnlbnn
Arjr7r?~,q13 ~ l r a w ni t 1% tnund tli,if, cithcr I i r ~ c n ~ ~ofs tnun-rcqpimw
~ or bccause of
vr thr r n c d ~ a n(;1.: approprt,lt~~) ot IFI . case< tr-~tliina gi\i*11<pan t r t flit t i r n p u t , (dl yrr cent of camplc ~nt~rnbcrq art, tcmale , ~ n t llil pcr ct-nt mnlc
c;lhc \vltli mlwlnl: 14dtd d n ~ lwhstittltt, this v;llur for thr rn1wnq d ~ t . 3 %nee yt-nt11~ 17itl rr,lnted t c 1 m,lnv otlm.r vnr1ab1t.c fhis b i o ~ir.111 affrr-t
I
valtlt-. Thiq ayprtv;rrh is r t ~ l . ~ t ~ v cslrnplc
lv and ,~rclidstlct7,iting varlarlre r.\tirnrrttbs rrthrr varl.~t.rleqtsrm the s a m p l ~tn thc p c ~ y r i l ~ t i r ~ t ~
D A T A AKhl,YSIS IN t~CFY(;ITUI7IYAI.DESIGN
To adirt.;! tor tl~rI)i,14 IYC ~x,nllldnrrtl to rvclght each of thc trrnnlcs to Measuring change
~ ca hl ~would bp ~ ~ ~ ' i g tc)
count for lpsc 111,111 0 1 1 ~ ~ P T S C I HI % ~ ~ I I C #t ~ ~ m lit~~l
count for rnnrl>t h a n oiir person. In thw ~ ~ their a v pryrrrt~on.;in the
<ample are adjustc~lto equatc with ihcir proportions in the prjyulation.
Tu achievc this rand wc weight each casc hy a spccific weigh^. A wcight In studics that txxtcnd over time we can think uf change at two Icvcls:
is achieved hy dividing the population pcrrcntage for a category by the aggregate and individual.
<amplep~rccnt"~r In this uaqe the formi~lamtirlFor males thrc~ulclbc FiO/ Change at the agErrrgate level is al?io scfr+rrcd to as net c h n n ~ cor
40 which gi~~r.; .I nrt>ight of 7.25. For ftmalcs the ratio M,OLIIJ br 50/h(l niacro c l i a n ~ etvh1lr5, change a t the initividual Ir\,el is also rt.fcrrr\d tc-r a<
whish gi\.rl;a tr~eichtc ~ 0.53 f (de Vaus, 2olrl). g r r r s ~change oor rnicrr~chance. The meaning id the two ways of t l i i n k l n ~
In wave 1 of a pllnel study we ~ ~ ~ on ~w dl ltod rzpeightthe samplr on khe ,3t>outcElange can bc moqt enwly described W Ith a n example. In WJI r 1 of a
bnqis of tvhat w t a know about the ~ h a r ~ ~ c t c . r i s t01~ cthe s pnp~~l,liir>n and .;n~dvabout ppoplr in thc paid wutkfirtcc~\ v t rnlght identih~thc prrctXnt-
l l i ~chilracI~ri<tic~ o f tfinqe \%.hi) rcspnnrf ;II the first yh;l\c (r E. sex, age of peoplc worhlng full time and thc pi>rcmtcigrorki king p.lrt t t n i ~At ~
rcgion, ethnlclty). Tliis rcquires that wtt I~ilvrreliable knuwlrdgt. 1)I the wave 2 we c o l l ~ r tthr same informarion a n d find that exactly thr w m r
populatiim cliaractcristics. pt?rccntages c ~ propl(- f work full time and part time as in the F~rstwiivc.
In sitbsequent waxFpswe will need to niakr further weighting ad just- While it would b~ correct to conclude t t ~ att the aggregate Ic\lrl lllrm
ments to takr ncctwn! of biases that ma), crccp in hecausr of .r ttsiticln. has been no changc hctwccn wave 1 and w n v r 2, it does not fnl low thaf
Given tIip amount of' ~nfc~rrnaticln available from the lirqt 1var.c rtrt* \.\,~lj ntr individuals lia\,c changed their level of rvorkfurce parhci~.$tir~n. If 70
!lave a grc.dt drat of infnrrnation that can tlr lrscd to calclilatr t c c i ~ ! ~ t In s. pcr cent of full t i m w~nrkrrs ~ had t h a n x ~ dto part time and prr CL-nt~f
these cases thr rvciglit IVL>II~LI be a c h i ~ w c dbv dividing thr t\..l\.r 1 part timers h;ld hrct>mc*titlP t~merswc 1 \ ~ 7 t l l C ( have a large number OF
ycrccntagr 11!. i l i i~z , I \ Y 2 (or 7, 4 ) ycrccril.lgt. Thi5 wilI enLlblt,lht, IV.I\ r.; t ~ ~ t l ~ ; ~ ~ rMTJ~CI
f r r n ill~l ~ ' ! t c ~had
f cli.~ngecl 111t)irI c ~ c luf workic~rcrp r l i c i -
after wavc 1 to rrnl.rin rrpresentativr o l wnvc 1. So Iring an w n v ~I is p k i o n , but at an ; ~ ~ ~ T c ' Rlcvel J ~ c 11 wnuld appear that t h ~ r ~~ . ~ 110 1 %
represel~tativcc l t I hi# populntlon then tl~twtbwrightings cnn hill p rnch clrangc. At the aggrtal;.alr l ~ v e lchangw t y En~li\.ldualsin onc tf~ri.clinn
wave remain rrprrsc~ntnti\~e of the pryml"ion. can be cancelled rl~ltby changes in thc opposrtr direction.
Analysis can bfcomc quite complcx w11t.n mtllhple waves arc analyseif Distinguishing Lbctw~cnaggregate and Individual change is irnpnrtant
a t once ~ i n c reach waiee mav nerd to bc wcigf~ted difft'rcn tlv. Thcsc both at a theoretical Icvcl ;lnd at the rtwarch dcqign level. it 1% easy to
strategies for wrighting ~ v i t h i npanel <unrrvG are discussed In dttail In miqinterpret IOI\? Icvrls of aggregate c h a n p , I f tve interpret I Q ~ + ~ re-
Kalton (19S3a: IqSh), 1,~'pkowski(lQS9),I.vlin et a!. (1994) and Ruck et a1. gate change as thc ,lb.ience ot individu,?l changc rvc will arriipcat ~ l ~ r l c r -
(1 495). c m c l u s i o n ~ahnu t the phenomenon. Fur rtample, the ~ercr.ntagt. of
people who rccimivc welfare and the charactcrivtics of such p ~ o p l carc
Fairly stahlc ovcr tiinc.. This has Icd snmc prople to believe that wrl (arc.
r~ciplcntsare a stnhlc and unchanging group. In reality, there is a large
h~rnoverof thosc (In w ~ l f a r ewith rela hvcl y Icw continuouslv d c p ~ n d m t
on welfare over timr (FIakirn, 1986). Stabilitv at the aggregate lrvcl will
indicate snrnethin~:about the nature 0 1 tlir macro system (i.c. thcrr 1s
something nbnut tF-tc str~~cturc of the s t ~ r n land wclnornlc s y t r m that
Vot only d n c ~bias ~ t i our ~ atr~lity
~ t tt, n ~ ~ rcliahlp
k c estimatt,.: frrrm a produces a given 1ct.d of pcrrple on ~\.i*li,~rtb) Fln~vp\,pr,it d n ~ nots t r ~l l ur
%ampleto the pcrp~~lat~rln but it ran .11+0 d~qttrrtthe patterns of rt*lation- .ihout the indrviduals and the naturc of changes thev experlcncr.
ships ilnd I-rrtrz~trnvariables ~ 1 1 1 i 1 1 101 sample. Tt 15 not alv,.al;s The type of rr>t7;lrcll design we adopt urill dcpcnd nn whctlir~rour
pm5ibIe to rt-wcil:ht samples reliahlv. tlowl.vrr, we can en5urc th,it the Sncu~is on the aggrr*gn!c or the individual It%vrlof changt.. h srrir* nf
pattern of rr.l;ltion.ihips wtl find betrvcrll varinblc5 is not cllw tn thc rffcct rcpea ted rros<-sccr rclnal ~ r v c v is s qil ite appropriate for backi tly: ~~~~~c-
of the variablt* It~rwhtch we have bia<rd information. gat' change. Sincv qilch n s u n e y mea\urta.; c . h , ~ na t~a~grclup Ik=rr,l tvc
To do this we cuntrtd qtatisticallv tcsr var1.1bIc4where t\.r rithrr know need ccrrnpa~abl~ groups river time. Hc~wcvcr,11 we are intcrc~trd in
o r believe thertl m a r be a bias. Bv z t s i n ~<tnti<tlca! controls [ c k ~ . ~ p l c17)
~r change a t the ~ndivrduallevel wc have to u\c n pane! d e s i q bt~;tt~.;c lllf
tt-c can Inuk a t r<=latinnshipswith the r t t t ~ ut' t other rariabfr. rcnm\?rd. Game rndividual~arc inrcrlvcd in cach %\'.~I.L*.
Rv r c r n ~ \ ~ i nthc t ~ cfftlct of a
g cficcts ot other .r.ariablt,\ wtac;ln e l ~ m i n q tkc D a t a will he analvsrd diffcrentfy dt.pr,nd~nc on ~VhrthrrWP a r c t - ~ l m -
~.ariab!ewJit3r~sarnplt~hias is evident ining aggregate o r individual level c h a n ~ rI4'twn draling with agcrth}:ate
I ~ , \ , t h l change MY' ;trc reqtricttd t r j nggregatc rntLasurc.;. I-.or e ~ a m p l t tviicn
,
c ~ , ~ n ~ ~rtqys~jiate
ninz rhangt. i v c ~ 7 i l FI-umpnrc yrrltrp nlcans nvcr timc,
rhancc.; i n Rrrllry vnrlancc, nvrrall percvnt,rces i n particular cntc:~nrli*ia t
rl i l ((*rent timc points ( e . pewcntagc
~ unemployed, ptlrcrntagc on wcl-
(,in. txtc.). Whcn adoptinn a n aggregat~change stratcgv we can compare trmr scri#% .inalvs~c;latcnt
whgrnupr; to ~ c icf the ertcnt and direct~rmof ag~rcr:atrc h a n ~ cdiffer ~ a r i ~ b4ru1-tural
lr rn(dt.1-
bchvwn sul,eri~~!ps. For example, we rnifiIit track c h a n c t r , at an agere- htl\turc t>f nominal and A Z ~ C ~ r* I AVCI )\ 4,
' Jltlr
ink-rva! (r#+nhnuotrs.~nt! rvpcwrln I\ llh Jummr
1 ; ' 1 1 ~ Icvel, in tlw 1ci.cl of rlfrtirnism of b i ~ + i n ~otvnrxr<
ss and empln\.ccr. categtrn .I11 vnr~ablc+
bcfor~and aftcs the election n f a labour oriented go\)cmment. Ry corn- Wnrnlnal (L-alegorical) A N O V A , nun-parantetric
r,?ring the difftmrences in ~ g ~ r t . g a tch.ingc e for thc rlifftvent group< we AKCIi74,rlrirnrnv varialdt.
can brgin to ~dcntit!. qnmc r j f tlic fiic~orsthat affiact c~ptimtsm 1 1 : Ilrt+
il~vK,'~l?tl' IL+:~,>~, Diwriniin~~i~l L1n21y<i5:I ~ ~ ~t ~~ r, I !
probit .~n.ilvsis, Eoh~stic
f-lowcver, nonca irf thesc htrrltegies ill lows US fn identify the cximt r e p s c i ~ u i ;h n r a r d / s u n ~ l v : i l /
and multidirectional n a h ~ r cof changc n t khe individual Icvcl. Wc event hictr lry ,~nalysis
cannnt identifv thc d e g c c tn which ind~viduals changc, the numbt*r wIln V i x t ~ r r r -1 1 1 nominal dncl LLV-lincw .inalv~is;k o ~ i ~ t i t ,
cl~nncc in nnc rl~rcction or annther. or the mwlt~riirectionaS nature ~ntewslit-vntini~uus.~nrl reFrclw>n. lnzard /.i~~wlr.aI I
of change (infliws a n d ryutflorvr;). Whcn we can ~cltantih i~idrrvrlrrni< cateE(.clrlcvI) rl-en! 111~lorv
annlvcis
Non! tnal (catcmrical) Log-lint'.~~.analysis; mulli.;laltn~
whtl cliange thcn our analyq~rc~lnfw~rso n identifying t h charactcriqtics
~ haznrcl/
life t,lhtr n~o~lcls;
who thaiip' to small and large eutr-ntk, and tlio.;t*who
01 rlinngerq: t h o ~ c sun:ivnl /#>writ histon' an.hv.;ir
~-li,irr~cin oppo.;ifr dircctiijil, To uondisvt dnalvqic nt t h i s 1t.t t.! lye
rl*qlltrt3pane1 c3altl For Indi\,~iI~rnls.
'Table 9.4 Cllarlge.; irr z-scorcs of rndrcidrtals Tablc '4.6 ~Mrnnzscnrr filr 1 1 1 ~ 1 1~ r t r l fT L W ~ I ~ P I T ,
7 9?0-2OllU
Case Raw score 2-wore Raw score ;-senre Change in
IVY0 (income) 1YYU 2000 (income) ZWO :-score 1W ZOO0 Change
1 (female)
2 {male)
3 (female)
4 (male)
5 (fernalp)
b (female) Table 9.7 Fmnle rnrnrnp as a pmprf~ilnof male e#r?iings
7 (male)
8 (malc) Year Colrrinn 1 Col~rrnil2 CuIumn 3 Cnlumn 4
4 (female) Averape FT Rv~ragcFT Female as prnportion r i f Abwlutc
10 (malc) male $ pa female $ pa mnlc (crdumn Zlcolumn I) $ gap
people who gained ernplrrymcnt Getrvccn wavcs, thnse who becamp othrr axis will represent tllr particular {ray in which ~nrlividualchangt. is
t~nernploved, those who remained cmploved, a n d thosc wlrn have being cmccptunli7ed. A ~ a i na, \vide rangy of poq.;il-le ~v;lt.'; of rcprcsent-
rcmninctl uncrnplovcd W e might find thnt thosc who eainrd ~rnplov- in^ ~iidividualclian~c*i<pc5rsilrlc. %me commtm approachez arc ill~is-
rnent showed a real irnprcl~rrnrntin sclf-rsfrrrn, and those who hlst thcir tratrd In Table Cb.10.
jvh shnrv~c1a fo.;~in self-cstet.m, while thr stabfc emylaycd and ilnem-
yloyrd showed no sclt-cst~cmc21nngt~.It tvou Id br tempting to ctmclude
that it i q tlic diffprcnceq in emplnvnient that account tor the d~fterc~nt Graph5 a r c a g00d way to rtlprcscn t chnngc. The most ccmjmnn h-pe crf
s self-eqtcvm of the p+c~ups.
c h , ~ n ~ ein p a p h for $ispl;ly~ngchan~e;ciq Ihc trenii graph, This stytion \rill p r o ~ i d t
Thc problem, howevcr, is that whe du not know that ctiangrs in a br~cfintrcldttction tcl trend graphs A filllcr dircusrion of thc.;t* ~raphli
emplrrymennt s t a h ~ sare the only changes t h e people a r c e ~ p c ~ i c n c i n g . and other w ~ y of s d i s p i a v j n ~chnn~tngraplilcal'lv can bc found it1 Hrnrv
Nor d ~ we 1 know that the four p o u p r are comparable En otlicr wavs ( r . ~ . ( 3 99,s).
gpnder, age , ~ n deducation). In experimental deqijins this prclbleln is The sirnplc trend b ~ ~ i p1%-ill h h a \ e mo ayes: tlir X-nxi~ar-ros< thtb
l~anillrd 1.y random nllocstion to RrtuFs. Tvplwlly In panel ~1csim;nsthis Iyclttcln~ant1 the Y-a.;i> on the ~ i d eThv . S-axi:, ~ v i l lrrprcsent time rvliilr
I.; not \x)w~trlc. thc Y-ari.; will rt.flwt ~iq ~ ~ ~ t i ton i t vthe vari,~bltthat iq b ~ i n ) ex,~rnined
:
n l c maln i\rav of r m p n n d i n ~tn the problem of compamhilih i s to i15c for cl-ranct*.Figure 0 1 ~lluslratesa srmplc twnd gmpli.
st;ltistic-al controls to cr;~ithli<hcnrnparnbil~t\' among the );snup%. Thic T ~ L)'-axil;
' c~3nbc r n ~ a r u r e c u.;ing
j a w ~ d ranee
e of mraxurcrnenl t l r ~ i h
gi?,ll cat1 be achievtbd bv u.;gng one form or anilthcr of rnr~lt~vnri;ltc am- ! Fig~tri.u.1 it 15 m r a s u r t d as a ratc - t h t ~nitnlbcr crf r i i v n r c ~I,clr
n ~ ~ 111013
Iy'is This s t r a t e p ~~inablcs
trs trl rcmclve thc. effwtq u l kntl%zbn d iftertnnct,? pn~~i~Entirm aged 15 or nver. Flnwri.cr, the Y-axi5 cnuld Iiave denoted the*
h~twt.r.11the ~ T L ~ I&CP F ~ Sit. i~ estaitrlish~dthat t l i ~p u p ' ; are qirn11~1rIn trri~~tllrrof Ji\ nrcc.5, err snmc other rvay t r t lndrcatin~tlir prc\.alenrtb t r f
the> .;pcc~ht-tl rc..;pccts i t is thpn pa.;siblc to cstabllsh hvw much the d i vnrce.
~ o i t p qCl~itcrtn regard to self-estwm ch.~nges.Thr log^ and tccf~niques Figurta 9 1 prllvicics trend data f-r A -
of thjs approach will bc dizcii.;<ed more full\. in Chayt~.r 12. @j-,31>17~ ""- '
DATA ANALYSIS IN LONGITUDINAL DESIGN
164 LONGITUDINAL DESKNS
Year
1998-9 201113-1 2002-3 '
Mean 5 10 15
Varianw 80 8tl HU
9G change ( ~ mean)
n 1on./, 5o0O
horn previous per~od
rat^ 12 pcr lOOO 12 8 pcr 1000 13.6 per IORO
Gaps (c.g bctrveen 510,000 $7,5110 ?%5,5,0Cf
male and ferndie
avcrdgc income)
% in~pmved[from
+Males --t Females All 1 ,
. .- - - - -
prcvious period) 5% 15% 2% go .
"dj
"<)
change
cliang~ngby X amrn~nt
"'%1
10%~
2wyb
5Ys
22';'"
7.5%
.
70 -
Vean improvement -u
t
* +
of improvers 35 prr~ntq 40 p o ~ n b K
Mean dect~nenf decliners
~5 ppinnh
10 points 10 points 10 points
-
:so --
- -
-k5 50- - - - --
(a)
$00 -
1 - ~ a l
-
e s m -
---.
s 2 4 year old malo sutrae rate a$ tat1001 male age standnra.rpd so~carlefa!-
-24 war old male u n e r n p l m l ralp ns rutlo 01 uvprall unemploymrrll raw
- -
Bk soh-
p- 6, I~
1
L-
w
E d 0 1 .
g 20'
;
I
- . - --
1-7
2 ' -- - -
-
0.5
0 , - 7 - - 1 , I-T--l I , I I I , ,-7 1 I
."
h9 '6
24 ,GO ,$ $
$ ,' \d6 4
,' 8 ,,# &
,' @
,' $
0' @
,'
Year
'u
z n t ratios, 20-24 yenr d d s ,
Figure 9.4 Malc s u i r i d ~duntlrs nrrd n r ~ m p l o ! / m ~ mfr
% 60- - - - -- -
A~tsfralio,7966-90 (DHSH, 1995)
E
d
.. 5 0 ,
a)
U
a"
axis has been contracted and the Y-axis stretched which has the effect of
, exaggerating the trend rate of change.
I Trends for multiple variables can be plotted on h e same graph a s
illusrrated in Figure 9.4. The figure plots two variables: suicide rate and
40 - unemployment rate of young people. In this figure the trend line for
suicide of 20-24 year olds has been plotted alongside the unemployment
rates for the same age groups. Plotting trend llnes from different vari-
ahles on the same graph can highlight the way in which trends on
--
different variables co-vary. However, one must take care not to assume
thrt t this co-variation of trends demonstrates a causal relationship.
In F~gure9.4 each of the variables is measured on the same scale, so
each variabIe can be plotted against the same Y-axis. Sometimes wc may
20
1975 1980 1985 7990 19952000 2005 2010 wish to comparc trends on two different variables, each of which is
Year measured in different units of measurement. Tn such a case we can use
two Y-axes - each with a diffcr~ntscale of measurement. In Figure 9.5
Figure Y 3 Etrlplqrlment o f p r o ~ l unged 55-59 lry ,yeyellcIL.r,Aw5trnlia, 1975- the left hand Y-axis (primary Y-axis) is the seric>us crime rate per 500,000
2010 trrr'tttal n ~ r dpredicted) fIattmed scale, {b) e . ~ r r d dscale (de Vaus and
[i~) populahon, while the right hand Y-axis (secondary Y-axis) represents the
Wolcntt, 1997) percentage of ex-nuptial births (birtlls to unmarried mothers). Again,
168 LONGITUDINAL DESIGNS DATA ANALYSTS I N LONGITUDINAI. DESIGN 165
3.5
+
2
3
Xm
m
c 2.5
g 8,000. .-wa
m
-
9 2 2
0 7,000 0
51
$ g s,ow X 1.5
rn = r 7
j 5,000 r
n al
ur
g
U
4,000-
3,000
w
2
.-
0.5
0
-5- I.-
0
5 1970 1995
1,000- Year
< - 0
Flgure 9.1, Crude drirorce rntr, Austndin, 1970 niid 1495 (de Vaus and
Wdcott, 1997)
Year
Summary
CROSS-SECTIONAL DESIGNS 1 1
I
1 Level d marital No children Level of marital
CROSS-SECTIONAL DESIGN
2 1I LWBI 01 manta!
happiness 1
1 Haw Orst child
wrthln last 12
months
Lewl of marrtal
1
ween 12 and 60 happhesq
rnths previously , (Y)
I
, gramme (35 countrics in 2000). Rcpeatcd croqq-sectionat survcvh provide
an rxcclIent snapshot of values at different points in a pcriml id rapid
change. '
Dia~ramrnaticnllvthe repedtcd cross-scrtional design can be repre-
w n t d as 3 ?;rricu;cri independent <arnplc.s IFipre 10.7). Each timr point
d r a w cm a nC*w samplc. Tvpically, car11 <ample will provide a wide
cross-scctinn of rastas so that useful crtlss-wcticonal analvqiq r a n Irc con-
. can track changes by compi~ringpattcrt~s
ducted witllin rnch s a n ~ p l eWF
in each sarnplc with those in previcn~sor subsequent sarnpTps ( c . ~rather ,
than just <)TIC aEtl group khev include a widr range of ages 5 0 that at any
gilVenpoint n t ttmp the age differences c a n be examined).
M'hile this itt-sikm does not prmit thc trackins ot individual change
o%.erhme i t enable-< l o n ~ i t u d i n a lanal!'siq at the aggregate Ic~~ck Ise
Chapter 0).Fwr!hrrmorc, these r~pca tcd cn)ss-srctinnal q h r d i ~ scar1 bc
Minr, ever, mn*t 01 t1ic.w thscat? Ihictt>r\, m , tcrr;ltron, ~ instrulni3nf d v c n ~ ,
rV<rLv.\rtln,
+tL~ti<tlc~1 inort;llilv ant1 tc*tini: ctloct.;) a r r w bi,ca~l<t, of
thc n ~ r r - t ~ ntvtt*rnc*nt
~~b ttt thcsc dt><ig~~+. 7 I ~ L >il\ re not pr(~L~1~-111\ wvtli
srtlr5-scctioiial ilc.;i~n.;. 'l'hr main bhrt.,~r\ t i ) ~ h i r~ t c r n avalidif\ l of cross-
~ r c t ~ o tdwigns
~al .ilcani frr>it~h v o ~ o u r c c q :~ ~ r o b I ~in ~ restabli3liinji
n< c.luse
ISSUES IN CROSS-SECTIONAL w i tlic~uta tirnc. Jirn~~tision; and problem\ a t thc I c t ~olf meaning (Mnrqh,
1482). Problen~s.lt thc Ievcf of cal~sc.arca, i l t cclune, an ~ s s ~ t~c\ ~ 1 1 any
li
DESIGN ci~-.i~m b u t cnrc+st~c.~lcru;rld ~ s i v .ire s p r ! i i ~ r l . i r l v prone to problrnis ~t
thw level. !'roblrn,s a t tlic Icr.cl of mi,sntnq a r p also common t o ill!
dc*i~msand ~ r i c t -\ v ~ t fequal ~ furct. for t.\prrlni~-ntat. I o n g ~ t i ~ J ~ ,n~anl t l
~ r ~ ~ r ~ - s c cdt--i):ti\
~~onnl
h.l;lr~h(1 L)F;2) ]\.I\ 1s.l-~tfcnclearl! anrl fort-t*f~rl I\ about thcsc trt.t~13rrlll-
1i.m.; Shc takcs I.;\ilr> r v i t l ~those who t i in i.;.; cross-s~ctit~nalc i r~v r y
Cross-scction~ld r 4 ~ n sarc probatrlv thc most widely urrd d r \ i g n r i n
J ~ r i g n sh e c a ~ l wof thcii. ,~llcgcdprobl(~rnswill1 causality and r n t w n i n ~ .
wcial researvh. C ~ I I P r ~ a s o nfor this prlpularlty is that the!, cnablt- the
Shc argucq that Iliic d t ~ q i j i nis not nt.arly a + flawcd as i t i s fr~~c1~scwtlv
rc~enrclierro r\llt,iiii r c ~ z l t s ~ l a t i v c ~quickly.
ly 5ince data arc ct~ll~,rtr-J ;~t
p u r t r a y ~ dto bt, Thc f r ~ l l ~ b w i ncliscuwion
g Jr,rt%.son Mdrsh'\ ~ r p ~ r n t ~ n t .
one point nb titilt1 t h t w is no need In .cv,.,littrlr various tollow-up stngcs or
intcrcenti~m+hr.tt)rr* a n n i v s ~ n qthe d.it;t. I t iq aJqci h i e t h ~ t tjt11t.r , t hlngs
being r c ~ u a (l c K. 5,implc s i x , p n y ~ ~ l , ~ t r osna,n i ~ l ctype), crclss-<r.itronaF
(lc-ignq a r ~ . nlrwr- r o \ t rtfcctivr th,ln c t > n ~ y , ~ ~ a tcl l\ty. c r ~ m ~ w t . land l
I n n ~ ~ t u d i nral rll + i ~ n +'Tl-rls
. 1~ becauhcn crti.;s-scctir>n;il des~q~i.; d o not r n t n i l 'fl~crci s nt?den\ i r l th,~!
~ irn5s-.;ection.11 51rltlic. I , ~ c cprnbIcm\ ~rl~aritilvin,:
t l i t ~cusi 17i r1~[11~~ift~cl iltlt;l collections, (rar-A~ngrcspnndenb r>r t>f c r p r r i - cn~~.;nl >~ivI m i ~ l i1t ~ crrrrlh-
vari~l,l(-+.F\,t.n I l i o ~ ~ gh lvio t ~ r n t ~ - c ~ l . d t\r,i~~iai)lrs .
liicntal intervcntio~is. I,itcd cir gruz~psnligl11~ l i f f t lon r an o u tcnnitb v~ri;lbIcwc cnnnc~tIjr L;.I rc that
C:ross-scctit~n~lltlcsign~can bc irlra l for d c d p t i v e ar~alv.;i.: I f we thcsc dil!crencc\ art- r l ~ l ct c ~a mri,si~llilrk bi,lwc.c,n ~ h vari~blt.5 r (RlCilurrk,
sirnyIy want tcl ~ l t ~ s c r i l lthe t . c h ~ r a c t ~ r i ~i )tt t ac ~p0puldtf~1n. thtbir a t t i - 19h-l).In CI.r,~ptt~r 1 4 ) 1 ~ n t r o d ~ ~ can
c ccl r a r n p l r r ) f the effect af parc~nthr~oit
h~dcs,t h r i r v c ~ t i n grntcntlon or thtbrr bu! inK Fatterns thtxn Il~rcnlss- .tnt~is (have nc-,children; plslpst child ~ 1 n t 1trnp t ~ v e x old and eliIcst chiltl
.;cctional stir\ PI 15 n nlnGt satisfactcln i v ~ ovf obta~nln:: 11734 c f r ~ ~ r ~ p t i v c bct.c\wn 7 and 7 r.r>r.: tdd) and marital h,>ppint*s<.Fcen i f tht*thrrr*gmup.;
nntc~rnlation.nut crt)<~-sechonaldesign* .Irr not restricted tcr di-.;cr~ptf\~e 11 i ttered in tcrrnh t ~ nzarltal
t happinex\ rvv i ~ o u l dharrc difi~cultvi r ~ c > r k ~ n g
nnalyqis :I<~ 1 1 t~1 ;lrgued below, yrt3pt.r analysis that uses qtatl.;tical out whether t h i ~ dil trrtsncc M'RS CIUC' t v t l ~ t i cliift3r~nt
r parental +iaf11.;. 'l'hc
controls c n ~ b l c *crtwq-s~cticmaldnl,i 10 p r r r v ~ d cr.aliz;lblc itifc~i+rn;l tion 1I1i-PPp o u p ~ w ' r i ,likoly tn differ i n w,iy< tvlhcl. 117i117 just thc~rpart~nthnod
J bout causal prt)crsscs and for testiiig cn t ~ s a lmadrls. h t a t ~ ~-s 1 1 0 ~
I ~ ~ i ti hg y \lave been tc?gcthrlr, their age profllt., Irbvcl uf
'There arc., I i o ~ ' ~ va~number r, oi mrt h n d o l o ~ c a land p r q ~ ical
c t iqsues wtrrkfnrce partrciyntiiin and so forth. 71rf,<t, ~iifferenccs,rathcr than
ot w h i c h wp need tri hr aware when ~ i s i n xcmss-secticln~~l designs. An parcr~thoodshtli\, c r l i ~ l daccount for diffc>rt.na*hin marital htippintlsa.
;EztBarencss of thtvsr* 1s5uc.sshould help m i n i m i ~ ethe ~ h o r t c o m i n pof this
clr5ien.
Thi, prnbtrm r r f c r v i f r > ~ ~ n r l i n~g. a r i ~ > t lir; In c m ~ . ; - ~ ~ ~ c t ~ r r n n l
l ~ t,lcklcd
\
dc*kigns a t thy rlrrliro~rr~lir.;~.; r ~ ~ tli,~n at tlit. data c v l l r c t ~ o nrt,Ij:tb
~ t i rathrar
as it i s in cwpc~rinic~r~l,il r l ~ ~ s i 'Thc
~ ~ s.inlirllt>n
. 15 nchieved lv m a h i n ~
t l ~ rgroups a\ \itni!ar 4% pmstble by ~ t , ~ f r ~ t r r - , 1.crno~ing ~iE~/ ~liffrrt.iicr~s
s rfirln IIII;'~'bt,rvl cmthl,'r'h'r/W hcn jvc cmnpart- ):rtvu ?I;.
hctwccn j i r o ~ ~ priflr4r
Wc encnr1ntt.r p r r ~ t ~ l t ~ with ms intcrn,~l \>,ll~dih.rvhcn llit, I O ~ I L - nncl MT need t n hnou, Iiirw l t i l l ( . I ~~lifferrncc41 II~J h r to ~ Y t l f l / r ~I I I I ~ I CI*> i ' ~ / / y
;friiitlrrt' OF t111. ~irv5ti:n docs n o t enatllv 11s tc? clitjc~scunarnbigunu4v unp rrr!rurl. 5inct. rrthtsr I h r n ~ ,trial. nut bc cqual n t v d to remnvc n1.Tnv nt'
c'rpl.inatlon oi r w r rcsitlts (31 er anctt h r a r t.\planation. C'arnphr~TI and tf~rditfer~nccta \ pr>\-r27ly ~ n cnrnp;rrt7 d !]kt- rv1tF1 like.
C;t,lnlcr- tlLhr7l) h.n tx ~dcntlfiedA n u m b v r c v t t;lctors that threatr-n intcsna! I n outllnc c ~ fhrviv I.; dnnc \ r - t l l 1 ~ .tli+ru<wd in Chapter 11.
~ t ! - thr<ta Im\.c b w n dt>cusst~lI~Idt%!;liE in Chaytcrs 7 and S in
~ , . ~ l ~ r l dnd I lt~iz,cvcr,tlic h;lkic IticEc can bc 1ntrnJ11~t.tEnow For the c l i i l d r t ~ n.~nd
relntron tcr I o n c r t u l l ~ t ~and
~ i l experinrcnt~ilrli-.ignq. r n a r ~ t nli;lppirat.;s
l t ~ ~ r o u l dcr~n~l,.irr~
r u a r n l ~ l \uc tllr marital h.1ppinr.t.; of
tlirtbt> t ; r o u p . 1 1 i c 1 ~ 1\vitliot~t
~ *In{ ~ - t ~ i I d r ~t l -i t~~ ci t, -\Y!I~-vtbldtv\tc11iTd [ \ Chrlrl's acaOPmlc
Model 1 ach~euem~nt
11ntlt'r tlnp !par ~ t .lpl t n n ~ Ilio\tb
l r\,htr.;v t<lclt.\l c h ~ l d1.. bcluvcrn 1 a n d 5 school
\.i..trs (111.1. %rrnvl\nc r n r ~ h %a\. t th,\t ,>n\ rliibt.rt~nc~b~ rn ~n;lrlfdlIia~j~jnc+.
3r(' nt\t ~ L I L .to c l i ~ l t l r cIwt ~ ~ r i t i r to t h c d ~ i f c r ~ ngc n t proiilt..; 43f tlit. thrr-t*
gn'~ip\ pm't~ll'.. I h r ~.lrgllr , h ~ i flrt~sc
t r t ' l ~ uha\ r trtal'n 1larcnt.s Iongar
Parental value
art,, r?n arcr;lgc, oldcar than !hr~.;c w ~ t h u u tch11clrt.n Thcby ~ i r q u t h~, ~ tit I \ placed on
tht, aytb ot partners, not thc prcqrnce of c h i l d r ~ ~ n~,v l i i r ~ infl~rt.nutl< h ~ducat~on
r n , ~r t t l~happintacs. ach~eveinent
Il'e idn It,<t thr+ pos.;il~il~l\hv r c m c > v i n ~the, actv J It trrtncc..: from ottr
cornptiri5t>ns. Wc c.ln make a nurnhcr rlt nbsfrictc~lcninparlconc. Firhi wcm
171 1<11tt t v b l . c \ t p(-t>p11+ L ~ ~ l>t>t\vet*n
c ~ ~ l 20 *ln{T2 1 , i r i k I \ t y S \ v i ~ ~l1far , t ;(sft!~~)r !hi>
p a r l ~ c u l ~ii:~.
~ ~ r grrb11l-rlhtb mnrital h.ipprntn\% ( > + Ihc thrt*t, Errwry, d i f f t . r ~ ~ l Parenlal
l ' h a j t IS, !vt3 rc,slrlct o u r cumparlsclri arruqc the t h r t ~ t ~- ~ I \ p3 L I to f l ~ t ? \Z ~ ~h ,o Model 3
rP.sources
art, !nrit-11 Ihv 5.1rnc J ~ L ' .1 4 ' ~ 'C(TIII J r r \ p ~ ~ dtliic. t ch3rnparrcon +cparattl\ Ch~ld'sacadem~c
within caich ag<*h+rcli~p(1, E. 25-2'). 311-11, 74-W). 5tlppc)sc w e l ~ n dt h a ~ , achievement
;rvll!~?r t*.~c.h Srl-rttpinK, thtyrP.lrc strll dilft'rc~ncr\In rnarilal 1 ~ a p p ~ n c r ; s
hctkvcrti Ihr- SrtrupG ( I e bt,tzvt~t.ntlic clitldlt-w, I he nt,zt. parl>ntsGindtIic>st,
with (llilcl Eigctl 1 10 I; ~cdr.;)dr+pitt, ~lirnrrc o ~ n p a r a l ~agt3 l c ~ prt,irlu.;. M'r
(-an thcn *it It.;l~t$ b ~ i t thiq rtal;lt~~>nchir
thn Ix~t\vt~t*r~ tnar~t,ilh , ~ ~ p t ~;lnCj css
r d r ~ ' r ~ t h o t )+!a1115
d i- Itri! tlllr' t v the ~f t l ttbrcnt JCI. prtrt ~ l r + c l t l h t ~thrrt, ,~r!:t~v<,'I~nrling,I c c ~ ~ ~ r c l , ~ b! irot ni v c ~ rkrntlk~nq
~ ,inrE lung c.ancc>r I11 nil
~ r ( ~ k 1Tlit, p . IOZI( nt fhiq i \ ~ i n i i l , ~tor fht> it~i:i~- o t i ? r i ~ t , - ! t ~!:~L>LII?, ~tc tllal I < w-,ay FrI%vPsthat smoking cause3 tung sanccr, 11~1i f ~ : r l r . ; mran !hat thc
~ o n l t s t i ~~l\c'rf
v t ~ ~in c-\p*rrrntan tal d r s j ~ n s . l i v ~ n t h t camnot
- ~ ~ ~ btl rul1.d n u t ' ( l W 1 721
I-hrbnlclrcb~,.rrialllc. Ihat art1 ~.c>t~lrr>llt-~l for <t.~li~tir.all\, lhc rnrlr.taw t bt-an
I,e conirelcnf thnlf t h r * fiii,>l rr,lati~>nslirpw r l ~ r ~ cI,ctr-,.~.cn l N arid li i h
r ~ fwtinql a t h , 7 I r~*i~jh(m,h 111 b t l t n . ~Y ~c ~~ ~n\'. ~~ l Crt~v.-<rt-tio~r;II cis til can h ~ l p~b~f,lbTi<I~ t t t t~C.FLIS~I r e l , t~ron5hip.: clo v n t
TIM-prohlr~ln will1 till\, htlwtwtnr, ir that wit c.ln rinlr~ ct)~itrcll for cuiqt. L%'hrle i! is truv that "I c ~ ) r r ~ l ~hctrvci~n ~ t ~ o i tlvn
~ ~~;lriahlc< doe.: not
r a r i ~ b l c <!hat 1 ~ 1 I~a\,r. % thottp,lit oi , ~ n d,~l,o~it 1\,111ihnvt.li,~\.r ~nforrnat~un. ~~.;tal~lisli caurntion it is a prc-rr>qlrisit<, for r < t a b l i ~ l , ~,Incat1\a3 ~ rt*la!ton-
Rnntlrrm aFl(>r,7tionin ,In c~\pt~rirnt~rltal de.;tgn r4ft.c titmclvconkrr3lq Ft~rall s h i p If thcrv i s nn rrrrrt4,ltitln rva* cart br, cont~dc+ntt h ~ tthcrt. i.: ~ r r r t ;I
1 nr1,117Tc\ - Lnnivn ancl ~ ~ n k n n w\l;ht>n n wt. c t ~ l i t r t ~fob.! vari,~L>l<>.; st,it~~- cxrqa1 rcl,lti~>nshi(> I I I r r ~ l r r ~ r i t r r rv.lrinb!c+
y a.; cauqcs can ht,clf jiist ar much
ticallr at tht, dst,i an,1!?.<1\ qt.1gp t r y t a n ncvrr br, .;tlr~%t l l j t \tmc11~1\-p + c i c n t ~ f ~importance
c as I(~callngC L I U S C l y t r r l ~ , ~ c cmanu-
~ t~%~r.t . ~ ~ nthi* c
c-cmtrtlllvrl f ( v all rclc\.;lnt 1~ari.ll7lcs 'rhcrc nq alwav.; th(* p(>wibilit\r thal t,lctrircrs ut+ru~fdconqidcr n fintlinr: of ) r r l ctrrtrl~tit\nbrt\\.tr~n.;lnokin<
,In> rtb!;ltitlndirp Lrrc h ~ v fnunll r brh\,rt1nS nnll k' cc>liI~1 be d ~ ~ cL sr7inc I
and lung cdnccr lo Lli>an imporla11 f i ~ r d i n ~
zrncnnkr,! lud t ariablt*.
But t l r i ~is ilot to ctmtlenln crn~s-+-t~ctic~t~aI i i c > ~ c n$1.;s bt~tngof n o u+t. in h'f0111 IS: A PKIOItf R L A s O M l V G A K I ? 4.0 110(' R E A + V V l N G
cb\~lc~rinq c L ~ ~ ~rrala~lt~nd1ipr.
\aF In~lc~d .;inc.t*
, ira sosi;ll +cier~sc.rrse.in-h
t l i ~ c.1pat.1
r ty ttb ~ r ~ l t ~ n .t3iIl1ert v c p r ~ c i ~ ~ ~t>r i lt~t111c~ITv
lv < i ~ i%~v~i ~ i a n i p ~ ! I ~ ~ t e Thtl moqt ~ ~ s c f tuvl , i ~ti3 $(I al,nrrt tht- cau.;.~l nn,ilv.;i< of crtws-3r,ctirmnl
wI~ti111on~ ,I< r q ~ i ~ r t ' lby i 1~xpt~t.imtvlS111 t l ~ < ~ g n~cc ..11111ltt.1i Ilit~rei h c~Xlc.n d.1tcl i s tal d w t v ninelt.ls, o r tlo~r-rhart, c>r -path dicigrLin~i. of tlw tvpe
t t - ~ vii, C T T T ~, a l ~ c r ~ ~ , i ttt3 i ~ <.rm--\tb~
t*\ lit>~i,i! r{t>\iqri+ elluhtr.~t~,tl in I rqrrta I 1 1 & ~ l r ir%l,ctt
l hcrtl In tt71.; Elr~uk 'rhc t a 4 o! !I>cd8?ta
RJ t l i ~ rI han rejty-[in\: crtl--vv~:l lon,i I tltb<iqti+nc heqn!: t i - t - T c b k + for ~-aut,ll .in<ilvct it thrn lo cvaluattb l i n ~ \ n,cll ' t h a w t~iodr~lq Cif tl-11. Jnlti.
an,iF\.<r+ 1%c n i ~ ~yrvctvd sl -\ qtc.~n,>trc,illt, . ~ t i ~~-arcfuII\~
i !\ 1111 tlw , i n . ~<lI ~ \, Tht. Itc.rn1 4 tfl~.clrir+ ,ind nlod~bls2 % a ):uidc. t c ~,inalt.~i\i + ,it1 ~ m l ~ r , r t a n t
, ~ n dsoilti-!>! ctati\ticnll~ ttjr t,>~ttirkt?i.it mrchf yl.lu.;~lrl\- t , \ ~ l , i r l l the t+>,rl tt>r t 4 i t g Ftar\oii at1.11y5it1t: crt>c+.;e<l~otr,lId a t ~ '.I t ~s tiit, rntwlr.1 !halt
A l t h ~ ~ i ~Findin):
gli th,l 1 t ~ v,i w riahlt?q ,I rc ~ r l r r t d ~ ~ tdut>+ r r l ntyt ez!,ihl i\h rht,rc I.; a dif t ~ r ~ ~be!rvcr.n r ~ c ~ .n J V rorr reahonrilg a n d rrrl Irut r c a + n n i r ~ ~ .
carlbt, I! t1tw.; nltun that ,i i . l t l s ~ 1 ~ ~ I ~ ~ ~ \ R i\J ~~I~Ir w > R . ~ E ~\ bl c fTtlrkV~ ,4 r)rprt rt*n%onrnc in\.nlrcs yroptr.;im:, r ~ n the twsi.: nf thc.c~r~tik.al
ISSUES 1?J CI<CXSSECTIOVAI IIFSI(;N 181
cunsirlcration~ and prcvious rcscarch, that X will cause Y and will Evcn in thcrsc c~rcum~trlncrs in wliicl~we cannot develop a new study
thcrcforc be cnrselatccf. This reasoning is ~ndependcntof the data we to establish causal direction wu.c can shcd stmw I ~ g h lon the rlircctic~~~
may have. Ari hnc reasoning, nn the other hand, is the process whereby, by mllcchng rrtrnspectivu infc>rniation and 23v asking people dirrctly
nn the basis of an observed correlahvn in nur data, urc make up a causal about any effect childrcn may have had on their marital l~appiness.This
story as to why the correlation exists. approach does not yield unambiguous data hut, bit by bit, i f builds a
Thc psoblcm wilh nrl Iloc redsoning is that we can always make LIP a picture regarding which causal direction is the more likely.
story r e p r d less of the correlation that we find. 1t iq not hard to t h ~ n kup Establishing causal dirrctinrl, however, is not always pmblcmatic. 5)mc
a qtory that is consiste~itwilh a correlr~tioll - it is quit^' a different matter independent variablcs arc fixed (r.g. ge~ider,race), others are not r ~ a d i I v
to anticipate, on h e basis of rcasnning and theory, that a correlation subject to our rnan~pulation( e . age, ~ religinus denomination) and for
should xis st and then find that it does. others the time nrdcrings are straightforward (e.g parent's education
I n silmmary, cross-sectinnal data can be quite effective in dcmnnstrat- prrccdcs rcspimdent's educn tion, cduca tinu l r v ~ lubually prececlrs the
irlg where a cai~salrelaticlnsliip probably does not exist. F~lrthermosc,it tvpe of job we gct).
can hclp evaluatc which, of various competing theoretically derived u
priuri models fit the data best. Finally, wc can use cross-sedional data to
evaluatc and modify a given R prrori model.
By establishing causat explanations on thc basis of cilrrelations, cross-
sectional designs a r c nftcn accused elf treating human action as dcicr-
Prc)perlv d ~ s i p c d,~nticiyaling
, rivdl causal explanations a s far as powiblc and
building in wavs to test t h ~ m [crc>rs-sect~onal]
, surxbt.ys can provide evidence
mined by cxtcrnal forces and of neglecting thc rule crf huinan agency.
, Marsh puts this vicw most cIedrly:
for arid against different causal ~nodels.The t.\.rdence i s not prouf, huwever; it
I S only as goud <I\ L 11c mnclr~lis. ( t I c ~ r s 1982:
l ~ , 117)
The aim of c x p l n n n t i o ~i s~not just lo show high cr,rrelatir)ns between v ~ r i a b l ~ s ;
sociological cxplnm tirjns IYLIII~L~ i1Imo5t i ~ l t v a yturn
~ (wt to be r~ flop R I I V W 011
~ ~
thrse grounds. It must also show how the actions of the peuplr involved were
the actirms of ctlnsciuus hurndt~hcings, rcwting Lo a n environment, trying tr,
Another prnblern with cros<-sectional data is that Even i f we estabIish makc sensc c-rf it and pursuing vasiuus goals anrl actions with morc or lcss
that two variablcs are correlated this does not establish their causal success. Only cxplanatinns thal take cognizance of the meaningful aspect of
(Iircctio?~.Tn order to illustrate this point let us use the parental status and social action will satisfy us as human observers. (IqH2: Y 8 j
marital happiness example again. Suppose (hat the cross-sectional study
showecl thd t the chiIdless couplcs arc less liapyy than those with babies Herbert Blumer (19Fih) c o i n ~ dthe term 'variable analysis' to describe the
and that corrples with the oldfr cIiildren are t l ~ chappiest. The problem is type of analysis and understanding providcd by cross-sectic~nalsurvey
that evcn i f we assume that this 14eflectsa cai15aI relationship between the designs {hewould also apply ii to longititdinal and experimental designs
two varial-4es we know nothing about the direction of the relationship. . arglles that ai~alysisusing discrete,
of the type already c o n s i d r r ~ d )Me
What comcs first - childrcn or marital happiness? Do children increase quantifiable variables is inadcquntc for at least thrct rcascms. First,
marital happiness nr are happjlv marricd people more likely than the
unhappily married to have childrcn? It l e a w s oul the actual compluxiher of acllvity , ~ n dthe actual prorrsscs 05
A panel survey could resoIve this questiun. We would know the level internchon In which h u m a n 11fehas ilq bring. (195b: 101)
of marital happiness before children and again afterwards and scc if thcre
was any clerrig~.We could also check whether thc initial level nf happi- !%cond, i t ignores the actor's ii1tcrprct;ltion of khr r ~ b u tion
a and b ~ h n v -
ness was linked wrth the likeliliond of subsequentiy having children. lour and tlic way in which thcse intcrprctatinns impact on their actions:
However, t h c r ~is nu simple way of doing [hi? when all data a i e collected
n t and aliorlt the one point (if time
The i~-tdepc.ndcnt vnriablr ih put , ~ tt l ~ rhrginni118 of the prucew uf
Onr way of dppronching this dilemma i s tc3 dcvelnp a priuri rnodcls ta ' tntcbrprctdhun anti the dependent varlahlc. at the tcwn~lialpart of thr process.
tcst. WhiIc support for a n 17 priuri mociel docs not prove causal direction Tilt, inten~eningprocess is ignored . . . as sumethii~g that need not b r
it a t least prmidcs a kheorcticnl basis for arguing a case and provides considcrcd. (lq56. 47)
empirical data that a t least arc consiqtent with this. Hairing found data
consistent with thc thesis o f rr partic~tlarcausal directipn the next step By ignoring the i n t e r \ , m i n g process one cannnt ~ u ~ d c r s t a ntuhy
d the
would hc to adopt a dcsign that can establrsh for causal direction. independent and dependent variables are or arc not related. Where
I H2 CROSS-SBCTIONAL DESIGNS ISSUES EN CROSS-SECTIONAL DESIGN 183
ht~rnaii hehaviour is intentional and based on l~ndcrstand~njis of a thetr behavio~ir.Experiniental ~ v t d e n c cshows h a t in certain euperi-
situation thcn thesc. intentions 2nd uiider~kandingsqt~ouldbe part of mental situatiims we can drmc~nstrateen~piricallywhat causal factor led
the explanation of thc beliaviour, L'lumer argues that variable analysis to given behnvlour but thc indi\liduals themselves are unaware of what
usually fails to do ihis. this factor is. This objection has same merit for somc behaviour in some
Third, Blurner argues that variable analysis w wnches people and thcir situations but it does not mean that individuals, in many situat~ons,
b~haviourouk cjf their contexts. By trying to work with discrete, quanti- cannot supply at lenqt some reasons w h y they have behaved as they did.
fiable variables onc loscs touch with the 'whole'. By dealing nnlv with To say that they do not have access to the full range of factors that
artjh'clalr); isolated hits of behaviour the full meaning ot the behaviour affected their actions is not to say that they have 110 understanding rrf
will be misscd. By looking a1 particular aspects of proplc's bcliefs and their bchaviour. If we were to ask a mother with young children why she
actions without looking a t the context in which thpy take place i t is wanted to work part time she woulci no doubt be able to identify a
pmhLiblethat we will niisundcrstnnd lhc meaning of the RcI~aviour. number of thc rp;lson3 behind her decision. That she might not be aware
In the cnd the qucslion should b e n o t whether I l ~ cmeaningful and of or he n b l ~tn artrculatc all the intluenccs is not to invalidate the
intentional dimension of behaviour should constittlte part c ~ fsncial irnpnrtancc of those she has identified. A third objection is that indi-
explanations, but whcthcr this aspect of bel~aviourcan L7c captured using viduals may indeed bc aware of the reasons frjr their acticms but, because
the data coll~ctinn methods usually associa tcd w it11 cross-scctir>nnl the reasons may not he all that nc~hle,social desirability factors wilT
designs. prevent them fmm reporting these reasons.
Thcrc Ilavc been a variety of ways in which thosc uskng cross-sectional Whilc these objection5 havc some merit, it can still be useful to ask
survcys I ~ a v gone
e about {'roviding 'lncnnit~gful*
rxplnnations. As Marsh people for the masons for their behaviour. Marsh argues that 'actors . . .
observes, thcse do have privileged access Lo their own experience but not to all the
drtcrminanks of their o w n brhavinur' (1982 107). Hcrwever, M a r s h says,
a p p r n ~ r l l c sdiffer in whnm t h r y bclirvr thc n>mtjlrig ~ h u u l dbe s~rpplredby w e shnuld nr)t cxagscratc what we will discover and we must b e a t the
ant! how rni~chand what tvpc. of e\,idencc is required before nnr can r~alidly explanations provided by people thcn-rselves with circumspection.
nttribuir mcaning to action (1982: 1111) A final approach to includi~lga meaningful dirnmsiun to explanatinns
IS for khr rcsearcher to supply the meanings (1982: 111). Thcrc will be
She identifies a number of approaches. The first, and least satisfactory, is many situations where we simply cannot reliably ask individuals for
that which she calls 'facesheet sociology'. Tliis consists of correlating a set their reasons. Often we can only identify reasom by Icmkinp, at a number
of sncio-demographic characteristics (c.g. sex, class, ngc, rcligion and of peopEe in similar stti~ationsand reading off commonalties b e t w ~ e nthe
ethnicitv) with the phenorncnnn one wants to explain. Tlie more one can cases. The indtvidual actor will normally not havc acccss to the addi-
dctect cori.elations the more we have 'explained' variancc in that phe- tional insights available to thc researcher who can ohserve from a range
nomenon. However-, this approach doe? not provide sociologicalIy of cases.
meaningful explanations. 'To find that gender is corruIated with Burkheim's (1970) study of suicide provides such an cxamplt.. R v
rcliginusness is not to u n d e r s t a ~ ~why
d gender is linked to religiousness. studying thc statistical regularities he identified the social groups that
We netad to ~znderstalidwhat it is about gpnder and religion that explains were most 'suigenic' - the groups with the Iiighcst suicide rates. By
why women are more religious than men. Typically, as Marsh points out, asking what these suigcnic groups had in common he sought to provide
this approach to supplying mean~ng1s brought from outside the data the reasnns for suicide. He did so by proposing that in thew groups
whcrc, in an nrl lint way, t h ~ rcscnrcher
' si~ppIiesthe explanation horn people were most likely to experience anomie - the feeling of normless-
thcir stoc-k of plausible sociolngical idpas. While these explanations ncss and a sense of not belonging. He argues that thrs subjective feeling
ma!; be consistent wiih the data they arc rarelv ct~mpellingsince many is distresqing to the extent that penplc commit suicide. Durkheim sup-
'cxplanrltiorls' that arc equally consistcnt with the data can bc dc.vt.loped. pIics this rncanlng mthcr than i t corning from the data. rt is Durkheim
A second a p p r o a c h i5 to ask indi\,idu;lls tn s u p p l ~the
, meaning. We rvho rirylrrs r a t h e r than (~tnylirir.rlll!td ~ m a ~ l s t ~ nthat
f ~ , qthc suigenic groups
c-a11 ask them w h y they acted a r they did. Tliis approach is criticized suffer more anomie. However, he ncvcrtheluss attempt5 to lnakc sense of
h!, thnse who argu13 that uftcn people d o not know why they acted a s the empirical correlatio~~sb! supplying a sublectively meaninghi1
they did. Thc response to thiq objection 1s that if people d o not know why tf t I ~ cl~nksbthvcen rnernhership nf certain social categories and
a c c n ~ me
thcy acted a s thry did then these 1s not a pmblern sincc wr cdnnot s a v s~~icide.
that that bchavtour is in any sense mcaningiul A s e y n d objtction 1.0 Marsh (1982) discusses a t lungth a study of the social basis of
asking people tn give their own explanations is that t h ~ ymisunderstand depression by Bmwn and Harris ( 1 978) where i k was bclicved that those
ISSUES IN C'KCX-SECTltJYrZ1. IlESIGN 165
ctlltcrlng From ~lcyrt~c..iunr \ r l [ not bc full\ ,Ir\,artAo! all tht- tnctors d a l th,it
~ .Irka rcflccti\ i3nf thi. p o p t ~ l a t i ~rm
t i t h r rvt3rc.c f c ~ ~ to ~ rcprewnt.
n ~ ~ !
i l > n t r , b u t ! nto
~ lhr-rr rlt:prrsqic>r~. I;rn\tTn a ~ i ~I4;lrr1% ! adopt an ,ipproa<.h For thi\ rcmnn c r c w ~ - ~ c r l ~\tnrv~ys
r ~ ~ l ~ lim'1,
~ l ! ? C * C C ) ~ Ct l i t x d i v p ot
~ ch~~it*
nlrf ~znllkcthat uttl~;.c.d by Uurkhr*ini in his shtdv (Ti suicide - but w ~ t h For h t l ~ d i ~zvli~re
"i acctlr,llcn ricscripticln is r c q ~ ~ i r r(r d K. pol^ tical pnllq,
ant* diffcrcncc. I'licv cuIlwtvci ~nft~rrndhon a b o ~ ~what
t tlrr!! thought househnlil taxpend~turc . i ~ ~ t v r y xWliilv). csir\r-srctio~ral rlra51~msIiavc
W O L I I L be~ ~ h ~ c n c i a lp r e ~ i p i t ~ ~ t ifactors
nq behind dt-pression ( l i k tvcnts potential wt*aknesws in unambigunu4 y irlc.ritiiying cari.;tl+ (a matter of
rvhlch acted 3.; p m ~ o k i n g aplnlc such it< famtlv problems; poor internal \'alirli?) the!. src* qtrilng at descr~ptiun.
) ~ t > ~ t s l nepr n; p l ~ i ~ mtc strcswrq;
n .;cvcrih; of qtrr*ssom; access to quppnrt
t ~ ~ t w o r k cThcv;) tnfrrrrna tion about I he\(. tac tors becarl5r t liry
tlio~#ghtrt would Fwlp make qcnsr, a t a sublectivc* level, of why some
piwplr were drprcqccd and why othrrs uterc not. I3,v planning ahc.ad and Practical issues
,~r?ticipatingthr pr>s.;ihle rcawnc that might lic brhind depreqsinn thrv
trvrc able to dt?\clup a n accn~rnlof drpre3sion t l i ~ tried t SO h u ~ l din the Many c ~ thtl
f practical i.;\u~-\ Ji.;cusscd in thi.; srctinn a r r no1 prculiar trr
t,upt>rit.nces uf !ti[, .ictcws. It is not n f~3olproofw a v c ~ ftapping thr .;t~b- cro5.;-sc~-lirjnal dcr;igns I w t a r e impnrt.~nt is5u.c~ for ; 1 2 ! research
~rcti\.c.l!+ m c a n i n ~ f ual y x c t of brhaiuiour but I! dr*n~un~tr;ltes a n attrmpt rcgard lccq of the d e l ~ n .
111 n cross-s~ctron~il surLbevdcqign tcl Include hi:: form of explanation.
I r i summarv, tlliarc i s nnthil~):vnlit-rcnt in cross-st-ctitlnal designs that
prc=cludes t r y ~ n gto a c c ~ s sthe r n r ~ n ~ i i f idirnenh~or~s ~f~~l of human brliav-
lour and b u ~ l d ~ nthesr g into t ~ r p l a n a l ~ o nIt
s . rnaktx:, qmsc to emplov 3 Thcrc ic nc~tliingabout tlit* It~grcof cm~+srartic~nal desicn.: that r q u i r e s J
r,irigc i l f stratqqcv5 .;trice no stnclt. approach i s w ~ ~ h c n~i tt qsliortcotnrnq~. partir~~F,lr method of coll~.r-!in): the d , ~ t a I'hc c.;scnt~al thin,, in crcvsk-'
Case
I
I I Mate
I
Mldqle S4SK , 44 , Agree No
I openion I
Consewalrve
I
Iestt.ii, ant1 thc c ~ m ~ p l e x 01
61- I 4-
Female Workrng1 - 52% . -
56
I
-c
D~saqrne D~sagr~r
- A -
Uon- 1 facr-tn-facc hv kai ncd ~ntcrvlewerr, (2) over the ti*l(&phr>ne
inttbr\.ier\.~~r~, (7).;tllf-adn1ini5trrcclr by tht. rr<pun~lcmtivllerta the q u e -
with tr,iincd
I '
I I I I
Conservnrtve
I ti(v111air1,i\ nr~rmallvrt~ci.~\le~I
Fach rncthud 13f ,>drnini5terrng qucqt~r~nnairpz
'qnd returned IIirt~~r):hthr>mail.
l i , ~ *i t \ a d v w a t ~ * anii
<
dt'ii*;lct~rsan(+ thi, mt.tliod\ havtb Ihcir <trcncths a n d wcakr~cssc.:
(Dillman, 79%; Ro~crs,107h; Crorta., l't70; C;rovv.; a n ~ fKniin, 7'4:ct). In
s c l r ~ . i i na~p a ~ - t ~ c u mclhod
l,~r of nrltrlinistmticw vrltl n c c ~ lto hv aware of
th~-+i.rt3lklti~ia ~trcn):thsanrt wt-akntkssrq and huw t h ~ vrntght applv to thc
pasticul.il- slir\.r,v in which vrw arc1 t ~ n g + ~ ~'rl~c.;t% c c I . strt-ncthr and weak-
rtLw-p< c,ln btb p-~>upcrlinto f ~ v cb t o ~ dc..~teg~rriec. T a b l ~i I. 1 prtn.irltr5 ,I
! i (I>INV
W I - , I ~ A V \ * tl\ a ~ l ~ ~ a of v ~ P J C ~(if flit, tlirc,c*r n p ~ l i t I~I l~~~d r i ~ t n i k t r ~ ~ L i ( ~ i ~
r;1tta4cm ear11 nt thc critlaria. \t fttllt>r~frccuwcini s ; l ~ ~ ~ l a binl t 13illmnn
-
(ItVS) , I T > Liltz ~ V c ~ (~l ~~~\~ J 1 1 ) .
Tlitl sample <we dcpenil< 017 fund.;. timi*,acitb.;s ttl pottant ~ a parhclyants,
l
pl.innt~cl method.; i l t aniilyii\, and Iht- clugrtv' nt p r e c i . ; ~ ~arid
~ n acr.urncy
rrquirt*rl. 111 ~ c n p r a l ,rllc Earger thc qamylc the bPttt*r, Iluf br\.rmd 3
ctarttlin point ~ n c r c , ~ s i nthc
c s;lrnplr silt, ha.; .;mnllcr and rnrjrc marginal
bt3nrfit~.
The Inr):rr thr, sanlplr v r r thc anore it can I-(.i;uhdEviilcd so thn! mcnn-
i n ~ f u lwbgrnuy cnrnparistir~scan br mad(,. If, 111 a gcner,~lsrirvev OF
h r ? ~ ~ w h ( ~ we
l d s ,uPantt>rlto comparta thv wellbeing nf ~hlldscwin couplc
h t ~ i ~ s ~ h n~l d1 s1 t hthat (1C chilrlrcn in 5olc p-trcnt Iiuuw*hold~,I\.rV wnuld
1lt.t.d t c ~cniurc Ii1;lt thr urnplr 1%tn~lrll ~ c lai-!:c cntluqh !(t \,~~alrl \ ~ l Cistont
f
nun~bc>l.+ IT( suit, p n w t hriu.;e't~t)ltls. I f wi, wantrd 1 0 ronlparta Iatht~r
ht-,~drdm i l rnut1w.r h ~ . , ~ d c5171r. ~ ! prt'nt ht\r~\rh.rllkl<I\.+: \vtrt~ldt ~ l r dtn
<c.crurt a vcrv largt3 tniti,il \amplc. simplv t ~ ttnd t c n t n t ~ hI,~tlirr1 i ~ a ~ l r d
5ole parpnf hc*u<chold>.
Adopting thy .;tratcgv of ~ncri,,isin):the ~rvernllsample s i x tn ol7tairl
.;~rfirc~cnt hard to t ~ n dKrrluyr; c,in b r ineif~cirr~t1%-c <-an t - i d up \v"h
more of thc ' t _ v p ~ ~rr~spoilrlent*
l' t h a ~ iwe a c h t a l ! ~I I ~ V ~,An
. alltarnalivc
appronch i~ t(t cwrrsatnpl~.flrc hard t r l h n ~ lpoupq I t 1 en<urp \t~tflc~c'r~t
Sat!sl;uc.ton
['or-r
53tu4.l!..hTv
I.rxd
Coatq i l
t rrl
C,lr,d
find in the sarnptc. IC 50 per cent of the sample supported a particular forms including iucr~nvc.iiirnce,invasion nf privacy and time. How long
polit~cianthen wc cur~ldbc a l m u ~certdin
t that the real level of ?upport in should inlcrvicws las! and how long can a qi~estinnnairebe?
the population would bc somcwhcrc bctwccn 40 and 60 per cent. If the rt is impossible to he definitive. The optimal lengtli of an interview or
sarnpIe was increased to 400 then the hand of seasonable certainty would qucstionnairc will vary widely among different types of people. Some
bc 45-55 per cent (50 r 5 per cent).' people \till be busier and unable to spend time. Others will encounter
language difficulties and find the interviewlquestionnairc more
demanding. It will depend on how easy the questionnaire is tn follow
and how skillcd thc interviewer is. Even more important is how interest-
Since the analysis of cruss-sectional data requires that we anaEysc ing people tind the topic. Undoubtedly, the length of a market rcsearch
variation in one variabEe in terms of variation in annther variable, our questionnaire about something of little interest to people will need to be
measurements must bc sufficiently scnsit~vcto tap the vaiiatioi~that shorter than one dealing with rnattcrs of direct concern and about which
exists. For example, i f we were trying to explain variations in income and they have first-hand experience.
sirnpIy collcctcd incomc datn using just two broad categories, SO to
$7 n0,UOO and $1 00,000+, wc. would mask most of d-te variation that exists
and have litllc variation in 'explain'. By including mast pcople within the Types of dntn
one category (under $100,000) we would be treating them a s though they There is natl~ingabout thc logic of cross-sectional designs thal rcquires
are the same. Conscqucntly, we would bc unable to explain the variation the use of quantitative or qi~alitativcdnia. Thc kcv thing i s that data are
nlifhi~;this category. Similarly, if we warrted to measure marital satis- systematic - that we have information on thp same matters for each case.
faction and this was done in such a way that everyone dppeared highly Whether these data are in thc form of quantitative, numerical or coded
sahsfrcd we w t d d bc ~rnablcla 'explain' variations In marital sa tic;- data, or uhrther they are quote< frnm intcrvicws, cxtlacts froin diaries or
faction. Ef our it~rtrun~ent is too blunt to pick u p variations then we c>bbscrvations,is not irnpurtant FLIT the logic of t h e design.
cannot explain variation since we have not measured it i l l the first place. Analysis is simpler (but not ncccswrily better) if data are collected
The same applies with the key exp1,anator~variables: we must ensure using predefined variables and categors~s.This is wliy q ~ e s t i o n n a i rRTC'
~'~
variation. For cxarnple, if wc thought that variations in incomc might so widely used in cross-sectional designs. But even in structured ques-
he due tn gender it is extremely tlnlikely that, in a sample consisting tionnaires we must make choices between open ended and forced choice
mainly of males, much of the variation in income will be due to gender questions (see Fodd y , 1993: 126-52 for a ful l ~iiscussionof the purpnrtcd
differences. I f we thought thal incomc diffvrtmces are partly due to advantages and disadvantages of vpen and closed questions).
age differences but then tested this using a sampIc that was fairlv age Structured questionnaires are one of the simplest ways of obtaining a
homogeneous, wc would find that age was not very important in structured data set. Howcver, we might use lcss structured data collec-
explaining income variations in thiq samplc. tion techniques and frnm this construct a structured datn sct. Using in-
depth interviews, for example, we migl~ t identify from the interviews a
for statistical cu~rfmls
I~~futurr~aiion set of variables of interest to us. or each of thcsc variables we would then
Because of the nat~rrcnf crnss-scctio~~al designs we must br ahlt to construct a set of categories and then classify each case. For exampfc, I
introduce statistical controls into our analpsis (Chapter 7 2). This is conducted a set of in-depth interviews with adults that examined the
essential if we arc tu get any clarity regarding causal prcresses using nature oi relatianships with their wider family. As part of the interview
cross-scciinnal data. I f we are to intrnducc statistical controls a t the we talked about relationsliips with parents. After completing the
s i s these must be anticipated brfi~rrcollecting d a t a sincr we
a ~ ~ a l ~stage interviews I constructed a set of variables, a number of which were about
cat1 nnly conhilf fur factors about wh~cliwe liave information. This can adult-parent relationships. Prom my interviews I had idmtified what r
only br achieved by first reading the rclcvnnt li tcrnture and i.hinking thought were a number of different type5 of rclationsl7ips. I u s ~ da range
thsclu~htthc thcorctical mc>dpls h r are plarnning to evaluate ~ n antici-
d of pit.ct~of ~nformatiniito identify the type of relationship I thought
pa ting what variables will requlrt. coi~.trollingfor statistically. applied in a particular case. 1 (hen dassificd each case r)n the basis of a
whole set of clues scattered thrnugh Lhc intcrvicw. I rcpeated the same
process for many other aspects of the interview ( e . ~ relationship with
siblings, nature of marriage, degree of involvement in local community,
Whcn data arc collected directly from respondents wedneed to be mind- ct.iih;?liiy of work in life). In other words, 1 war: able m rlsc a range of
ful of the burden placed nn respondents. This hurdcn can take many q~~alitativc information t~.?produce variables and to classify individual
I ISSUES IN CI<ClSS-SECTIONA I. UFSIGN 193
192 CROSSSEUTIC7NAL DESIGNS
I
1
I
cases. I was then able to systematically analyse pattcrns in the data (de Bccause cross-sectional designs rely on existing variahons rather than
Vaus, 1994). intrr>ducing il~terventionsthey avoid manv of the potential ethica! con-
cerns about harm to participants occurring as a result of the inten~cntinn.
I t is also an advantage that they avoid randcrm allocation tn experimental
Ethical issues and control groups. Randurn allocation can give rise to serious ethical
problems (see Chaptcr 6).
As with a n y rcscarch design the researcher using a cross-sectional de5ign
will need to attend to matters of confidentiality, privacy, avoidance of
harm In participants, and infnrmed consent. Summary
Tn some respects cross-sectional designs can minimize some of the
ethical problem5 that can arise with experimcnta1 and longitudinal Sincc cross-sectional designs lack a time dimension and randomized
designs. Since no tracking of participants is required it is much easier to control groups thev encounter potential shortcomings for causal ana-
ensure anonymity. Thcrc is no nwd to kccp l~stsr ~ names f or to match lysis. 'These shortcomings represent threats to the internal validity of
names a t all to responses. Furthermore, since the data collection is a 'one- cross-sectional studies. This chapter has outlined thtse shnrtcomings and
off', cross-sectional studies ran be less intrusive and, in this respect, can identified some ways of minimizing their impact. While cross-sectional
be less of an imposition on a person's privacy. designs present some peculiar weaknesses in relation to internal validity
In some survfys cenfidcntiality issues can arise even when information it has been pointed out that they also have some particular strengths.
is provided anonyrnrrusIy. In a survey of an organization or a specific Because they do not have a time dimension, they do not suffer from
region or in a national census it is conceivable that individuals could be many of the other threats to internal validity t h a t we have encountercd
iden tifird by examining a scl of charactt.ristics. For i n d i l iduals with with cxpcrimental and Ioiigih~dinaldesigns. Cross-sectional designs are
rcIatively unusual characteristics tt may be possible to idcn tlfv individual a simple and cost effective design and consequently are often able to
people. For example, In a national census data will be collected for each employ exccllcnl s n m p l ~ sand avoid thc problems caused by sample
person that lists their geographic locatio~~ (down to relahvel y small attrition. As such cross-sectionnI designs can have strong external
geographical arcss called collector's districts), their ngc, gender, occu- validity.
pation, income, education, household characteristics, marital status and The absence of randomized control groups means that the analysis of
so forth. With detailed information of this type it would be possible for data from cross-sectional designs relies on considerable sample variance
someone with access to tht. data to identify some individuals such as the and on the creation of post hoc comparison groups. This has implications
doctor in a particular small town, the school teacher, the clergyman ctc. for sample sizc, sample type m ~ dthe practical strategies of implementing
Sincc some survey data and census data are made publicly available cross-sectional studies. The chapter has outlined these practical and
to researchers for Further analysis it is important that no individual is cthical considerations.
identifiabre by means of cross-classifying detailed information in indi-
vidual cases.
, To guard against this problem, data from census callcctions and Notes
surveys wilP be 'anmymized ' before they arc released to other pcc,ple for
further analysis. This is achieved in various ways. One way i5 to omit 1 n i s needs to be q~ialificdto the extent that snppmsor variables may mask a
certain key inlormatinn, such as any geographic identifiers, from any
causal relationship. Strictlv speaking we would n w d to be able to eliminate the
influence nf all suppressor variables before we could bc ccrtaln that the absence
public release 06 the data. Another way is to reduce the precisic~nof the
of a correlation b ~ l w c c nhvo vnriablt-j really rrflccts the absence RE a C ~ U S ~ I
data by collapsing categories. For example, instead of providing data relatiunship. For a fuller cfiscussion of suppressor var~ahlessee Rosvnbcrg (1 968:
u ~ t specific
l~ ages o r birth datcs, only age g o u p s (e.g. in five-year age Chaptcr 4).
gruups) will bc reyeased. Another stratcgv is to omit any linking infor- 2 These estimatrs assume thdt we are dealing with variabfes where there 1% a
ma tion between cases. For example where informatit,n is collected from W/30splil. 'Thc cstimatt'd sample sizes will be different In more hnmogcncous
each member nF a household the publicly released data may makc i t pnpula t~ons.
impossible tn identifv which indiviclual cases belong t c ~the same house-
hold. By controlling the deta~land amount of puhlicly released infoma-
tion in these ways it is possible to prevent the iclentificatjon d particular
inrlividuats from an anonymnus data set.
levels in particular locationr: and , i t a pnrticular \tag{, nl Ihr Fanj~lvlifil
cycle.
Since thy.;(. sirrvets are nnt primarily designcd ti) rlc\,rlnp cxyl;lnn-
tinns c,f bcha\,loi~r,problcrnc (11 causation, of ~a~1*;12 o r d ~ or
r the inflw-
ence of unciintrr>llcd \-ariabl- do not matter. TIlc ki~vthing is to rn.;rlvca
CROSS-SECTIONAL ANALYSIS that sarnplr%,irr Inrgc enorrgl~5cr that rve can p a t sufficient ntrmbcr.. En
the varicius si~hi:roups to prrrvidtb rt-liable descriptions. The 0tht.r kcv
thing is ti) ohtaiii qarnplcs In .;llcli a way that wr3i a n l?;t~ncraI~ze bryclnrl
the samplr ttr thc population frtrrn which thc. sa mplr w a s drawn.
?hc an,ilvsis i ~ w then
c bvcnrnrs a matter of t171* npprnpriate wnv': tct
~lc.scriPethr data It 1;. beyilnJ th~b+ctrpr of this I l i ~ > tt,
k go through nll thi*
Crtr5h-.;tu.trrrnaI Jeqigns arc r ~ d.suited
l t r b 14t~~cr1ptive anaivis - at Itust a be i i r ~ r r i l ~ c r l .
warTsIn tv111t h d ~ i can
for dcscriptionq uF thing< ; t ~ thev are at n glvcn point of time. Flowrvcr,
crii~q-st-ct I ~ a v cbrrn cri t i c i l ~ t t tor their weakncss n t the
~ n n a ld~s~g115
level nf t*rplanatory, causal analysis. Somc rt*scnrc-lier? argue that, for
scvcral rr*n.;ons, cross-sectional designs are ~ ~ t ~ r t l i lat c sthe
s c.xplnn;rtory C n u n t i n ~i.; 3 Ilnsir aspect of dcsrrjptive analyqi.;. Tlic aim i s to eqtnblisli
q t n of~ ~'1 ~ t u d v .To support t h ~ i rl u d ~ e m e n tIhry cite thc diflicirlties liow m a n y pcople have pnrticul,lr cliaractrri~ticr, hnve n particular
c m ~ s - w cI (tma1 data presrv~ilor t ~ t ; t a b l i s l ~ ithe*
n ~ c a u ~ a 1order ot c t . m ts, opinion, f r c l ~ s ~inc a g i ~ e ni \ r a v ;lnJ $0 forth. h,l;rrk(,t x\lrvPyG wilt find
for cun troll l n for ~ all thta variablrs that miell t 1-c inflz~cncin~ corrcln ticwq, out hnn, man!, pcnple like a particrlldr product, a pr)litical poll rvifl c011~lt
nncl tcrr r%\tnbli<hing,~ z n , i r n % ~ r ~ u ~cntv.,lli u ~ l y , clrder. .-\-\ltIiou~h t l ~ c t ~ o w111.111\. pi,opIr ,"tend to 7 r ) t ~tor n spccifir p;rrtv clr can~ttd;ltca, ,I
j ~ ~ d g ~ mIt~vst~ eon c l the- I-r~tic~smsis too hnrqli I1it.v arc ~-alnab!t* hc~ca~~~c census w ~ l lctrunt the nt1mhr3r of pcople in tl~cn l,c*ptll;ltiun, tvhilst .I
they C ~ ~ J M ~ l*l c tion
n tc1 inl~tlrcn
t weakncssrs uf tV~i% cross-sectional dcsign hcluse!mld i*xp~.nd i ture SiIrvLay ~ v il l dcscriht. h o ~ vmany pcople 1iai.r-
that murt 271' addressccl. parhcular patflmrnsof e x p e n d i t ~ ~ r v
As a result wc rnunt how many pcoplv gi\re p a r i ~ c ~ ~answer5 lar brcausc
we are intcri>stcd in the t i r . ; l r r l l l r l r r l r i nf cases. Mrhat is tlic h.pical rt.spon<t.
Descriptive analysis
widt.1 y ~ i s r dfor these ptlryoscq. Thcir wideaprtud use in markct rcqmrcli clustered lowards one end ot n ~ i ~ s t r ~ b u f( op n. ~ low . inrnnle a d ) .
and pol~ticsl polling b ~ a r this out. 'l'hcsc .irtl contexts in which In othcl. wt)rrls in addition tu nimplc counting wtTwill look at thc sirnpr,
rt.scnrchrr< want simply to dcscribe such tI11ng5 as: who u s w o r likcs of any distribution and try to summari7e in rnmningful wavs what tllr
particular products; aspcct'i af consumer brlinvinur; voting intcntinn; distributinn Irx+kslike. Thew ~ n ; l l smay be acliirvrd with varini~s
level of rtittlrt.>t in politics; trt-lings toward< plrtictrlar politicians; and m a y st~tisticq.'To stlmmarire wlir?f i~ ~ I I J J ~ C RIn! a ~iistr!b~tiiun w e liunuld
attih11k.s tcwards a rnngc of iqsues. Typicallv, this FTe elf rnqcarch use a n appropriate measurc. of ctmtral tendency (mi*nn,rned~ano r rntrc"l1).
fc~itsc.;on question? of 'Irr>w m,lny' and 'which tvprs" of pctlpZc do or 'To sum marirc thc spread and i r n t t ~ ~ l t i l iin l ! ~a d ibtrili~~tion we would I I ~
t h ~ n krn pirticular ways. an npprt~priatcmcasurc of d ~ s p ~ r s i o(ne . ranpa, ~ intcrquartile rangc,
'Fhc ciln.;IiG i h anothcr 1-xnnlplc of a desvriptivtn crclss-sectional tfc~ign. variancc and rlt~ndnrddevtntltrn) Other aspect5 of Ihr shape of n clislrl-
I t s pfinlarv f~mction is it1 prt~vidc <In accur.llc description c,f fhc
bution call IV .;llinmarizetl cvitli qlatiqtic~sucli a.: skrwncsr; and kurtrwi.;.
pnp~l!ntic~n of a cotlntr~f: hurr mJny pcnplc lhttn~arc> in tllr p~rpulatiun, G r a p h ~ c drrcpr~sentations rli.;tribi~tions can offtlr~c o i i v e ~mrlrc- ~ and
d n ~hmvf man\ people p o c \ c - ~ parhcular chLlmct~ * r ~ c f i(e.9.
t > svnrfrr, arc,
be mmo i~nliritl~clt. u n d ~ r s h n ~ i , ~than l ~ l cf;iimn?ar\, .tatr.;ttcal mc;lwrL*>.
e d ~ r c ~ l t ~ rIrl.ing
m, in p , l r t ~ ~ = u Itration.;
t.~r chtc 1, lloucehold c x p c ~ ~ ~ l i t u r c
Tic c h ~ r t k ,Ivr chart5, 11nc gr;lyIi\, h ~ ~ t n q a r n lieu s , and tvhiqkcr ylllt\,
survcv* provide a filrthrr t*uarnptcof c m r ~ - w c t i c ~ nclata a ! bcinz II~C'LJtnr stern and I c ~ fplot.. are ius! somc nf the graphtcal mr;rn<c>f reprecrntinfi
dcwript i v r purposes. In thrsr. \11~vcys ~ o ~ ~ ~ r n snei~m ktot mnp
~ thp
distributinnh. Mo.;t cvrnputer b ~ s c d<t*~tisticnl p;ick;ly;.rs and sprcndsht~c~tc
cxpcnd iti~rv patterns of indrviduals snit I ~ u u ~ r I ~ r > lSuch d s . .;urvrys producc n rangc3of such praplis. AltcanlativeIy, wv may choose to pn3hrntit
ennbli~.(int>-graineddc.;crip~ir~nsof the expcnd i t ~ i r cpatterns of cl i fft-rent
the data in tabular form. Frcqiiimcy distribuhon. arta lllc normti! upavof
tvp~':of Iamilicb and Iwia~c~h~drls and for pivvjlll~1231th different it~comc
doing this.
CROSS-SECTICINAI, IIESTGNS CROSSSECTIONAL AN AI.YSIS
i i rm>dincy~iro<kril,y
Table 12.1 ,411~ l l r t ~ t r f l t mof rr rrlntto~tsl~iy
Recoded version
M't~endescribing distrtbutions we have to decide how much detail to Male Fcmnle
provide. Continuous variablcs such as age (in years), incnrnc (in dollars)
and hours worked each week can be analysed as cnntinuous variables, or Shonglv agree 5W" 60% M1%
Agree I0 45
collapsed into groups. Altcrnrltivcly, age might bc collapsed into cate-
g o r i e ~S U C as
~ under 20, 20-29, 30-39, 40-49, 50-59, 60-69, 70-79 and Disagree
Skrmgly disagrw
30
113 35
' 1- Dirnpree 40 40
go+; income might be collapscd into income groups (Iess than $5000,
$50[10-10,000, $10,001 -1 5,000, $15,001-20,000) and analysed in thcst.
grouping5. Manv ordinal variables might be collapsed into fewer cate-
g~jries.For example, respondents in a survey are asked liow strongly they
agree or disavcc with a particular attitude statement (strnngly agree,
agree, can't decide, d~sagreeor strclngly disagree). We would need to
make decisions about whether to collapse down to thc cat~gories'agr~c', Using this approach we focus on the distribution of a variable as a basis
'can't decide' and 'disagree' or, alternatively, if we are only interested in for recoding and let the distribut~onrather than the substantive meaning
agreement wc could simply use the categories of 'agree' and 'other'. of the categories dctcrminc how we combine categories. For example,
Nominal variables, such as country of birth, may havc a largc number of rather than collapsing income in the way described above we might
categ-nriics - many of which can be combined for the purpose of simple simply have two categories: people below the median incornc and those
and cTear presentation. Wc might group countries into regirms (e.g. above the median it~come.If t h median
~ income was 535,000 then the
Western Furope, Cpntral arid Eastern Eumpc, North Arncrica, Other hvo incoinc categories would be helow $35,0011 and abovc $35,000. Or we
America, Sor~thAsia, South East Asia, Oceania) or according to levels of might divide thc sample into deciles: the bottom 10 per cent of income
development, or whether or not thcy arc English-spea king (Chapter 6). earliers, the second bottom 10 per cent ~iincome carriers, through to the
The d e p e of detail we retain will depend on factors such as sample r o f income earners. The specific dividing Iines would be
top 10 p ~ cent
size, the number of people in particular categories, and the way in which determined by looking at the actual distribution. Using this approach,
data are to he analysed and presented, and on the main points that need definitions of low and high income are not imposed arbitrarily by the
to be highlighted. researcher but are defined mlafiz~e[yon the basis of the distribution.
The way In which variables arc collapsed can have important impli- Table 12.2 illustrates liow the same variable might be collapsed differ-
cations for patterns of results a n d care must be taken to not to simply ently depending on whether substantive or distributional methods of
rerode to cithcr crcatc or mask a pattern. For example, Tahle 12.1 recoding were used.
ilIustra tes t h way~ in which coding can mask a relationship.
There am two broad ways in which coliapsing of categories of a Form of datn
variahle will be done.
As well as deciding on how to present and summarize distributions and
how to group values, there may be a need to attend to theform in which
the data are analysed. Sometimes it is desirabfe to change the way in
Using this approach, categories are combined because nf thc similarity of which our measurements are expressed.
the categories. Thus combining the categories 'agree' and 'strongly agree' We may need to convert variables so that thcy are a11 measured on a
bvould be done because thcv botli reflcctcd asreemen t. A1terndtively, uniform base. This can be necessary when analysing data in repeated
depending on the purpose of thr analysis we might corntlint. 'strongly crosssectiona1 surveys where the value of, say, lncomc needs to be
agree' and 'stron~lvdfsagree' as t h y both indicate people who feel adjusted fnr inflation. In cross-na tional studies some variables, such a 4
strongly about t l ~ cissuc. k i t h a variable such as incorn?, substantive income, may need to be converted into a commun currency (say US
recodlng would combinc categories c>ri the basis of the actual income dollars). This was Jiscuss~din Chapter 9.
values. Tl~uqwe rnigI~tcollapse income into categories such as below Data may also be represented by exprwsing individual scores, not in
5113,000, 51 0,000-19,994, S20,000-29,999, $30,0130-39.999, $40,000-49,999, tern~sof the units of the original variable ( e . ycars
~ for age or dollars for
950,000-59,999, $60,000-69,999. 970,nOO-79,994, $8U,Ol)ll-fi9,9'dY, $90,000- income), but in terms of how the score of an individual (or group)
99,949, $100,000 and over. compares with the vpical vaIue for that variable. This might be done by
do1F.l r+, Il ~ . i ~ t < l - h r n . ~vtln
r k ~ .and fr;inil;. Instca~lrrvt, I\ Iruld Iw tit-al rtiq ~n
1 hv cclmmrm 't+tsrrtrr~c!' nt dc\.iatinn< Frrlt~ltht, miv,ln
A v ~ r i a b l c15 <tL~ndtirdi/l+d \lrnpl.r: 1v. s i l l ~ t r a i t i n gthc typrc,il \.,iIuc
Irnran or mw'rnnl frclnl c;tch pr>rsonrsacrual ccl3rt. on Ihdt x . ~ r ~ ~ b lTilt' tb,
rtb.;~rltinfi vn1ur for cnch ycr+cm I* thr-n d i i . i d t d bv n rntlasurc c ~ \.nrintinii l
(5tanriar~1dcvintlrrn 1 1 n?t5;rn rvas itsccl a n ~ lintcrqliartilr, range - nr
~\.linlc*rcr- if [he rncdi;ln trrn< 1 1 4 )
Tle~(-rihtn c l~~ + t r ~ l l i ~ t ~nocir-trw < ;l ~vli.olc~ .;tirrfr; I+ t\~prc,~fl\ t)nlv thra hr,t
part I ~ Fa Jrst+ripti\.c. n n a l <i+ ~ florr, clftrn th.111 ii(+l\\.I., \tm.int stat. bur\*
r l ~ s ~ r i h t ~ t riit icm frar~ anlong lit C~lrcnt.;ub.;rls i j t tli<bpctrplc rmnl3rj<ing tht*
<t~i<F\..Fttr r \ < ~ r n p l ccf41 , tllr incurnr pattt~rnkt ~ rnvn f ; ~ n drvtmicn ~ l l f t t > r '
By hnw nluch? I n rvhal dircction7 Art- Ihtarc rlifftlrt-nctl.; ,lmclnr: tilt*
yclurig, r n ~ r l c l,1l1:~td ~ n rthr. l rldcrl!7 thtav dibtcr asrordrng to rr>gion7
C~vnp,~sini; ~ ~ ~ l ~ on q ra oy t ~ m t~wvar,,rblc
p ~ linctrtnc, c t ~ ~ q c r ~ ~ l t i 5 n 3 ,
rclic~nu\nps\,l i a r ~ p i n r ~ <i\s f";\r~ntlnll,v l n d ~ \ c r i p l l v ecbuerc15ethat rntail.;
r<t,~l,li~li~ rl-hirli
n ~ tvpcq ptv,pTtl ha! c p,?rtiuul~irrIi;lrLictcr~~hcs, att i -
rc~-c.rprc~siiiq 1*,1ch pcrkrrrl's inctrmc If~lrt . \ , ~ ~ n ~ ~it1 l t "tr7rtii<
l t r l tie\\ m~rcli tuilr... cir I)ch.~~~crtir.: It \trrp+ ~ l l r ) ~rrft ,iftcrtjptin!: t r y c,\~l,rrn r v h ~snnlt9
I o n - r r c>r 111):hc.r i t i c th,In thta avtLrngr \r;llt~c.O n t ~rz-av I n \\.h~rhthlc is >:nrtrp ,ire cliFfr>rcbnt Irvm r~tlirars(,il t l u r ~ ~ gthat h IVCIIIIC~ IJC'tht' o l 1 v i ( ~ 1 \
dtlnt. 14 b~ {,\yrr,\+rnq a pcr\on'+ ind lvidunl w r r r Iir tcrn-rs ot t h ~iumtirr ~ next +try) 12'11arta .;irnlllv ~l~,st-rltlin:: F;Flrrbrn%, \\ Fll~h n~,t\,in tk~rn.![-ad r ~ b
(I( <tilli~IilrCId c \ ~ i i l t i ( ~~t t ~lit'<
s ~ ~ O Vor P~ P I C I W t h t ~IT~CYII a u c o ~ ~ l i(31 t * why p-nul,c ~Clffer.I'Eir, I , Y P T C ~ ~of drascrilrin~I+, lur~v.yt-ttv+, tl~c
lZrc can go one ~ t r pt i 1 rthr-r anif stnrtrft~rrit:t>~ . i l r i , i l d p ~ Standnr~l
. i~atton flr.;t rnd ~ ~ ~ t ~ n5t.l-p c aITrltlr
l ~ l h~1 t!tb\~clup~ng
~ p ~ l ~ l ~ ~ n a t ?Tor o n c~.x a t n p l v ,
11.1.; thv effrct of cxprcssing c~,ichyrrscm's ccnrc rt,lt~liiirto that c3f other.: in brfvrc . ; I T ~ I e ~ \ yIl ~a ~ n~ ~ n t l rdifft*r(.ii~cc
l. i t 1 it~cornr> M ~ L mu<t
~ flrkt
the dl.;trihutinn I t has t l i cftrct ~ t ~ cllrninatlng
t difltkren~-rsrn thc 11n1t.; of ~ ~ t ~ i f ~tliv l i st*xtrllit
li ; ~ t l dnaturr 05 tlic LI i l tpc.r<.nct...
m c a s ~ ~ r c nt ~ acrosk
r ~ n cniln hlc*\, nc-ms+trrnt. anti air(?\.; d l ffert-nt iv.~riLillles. l ? c ~ r - r i p h t r at-rrrs\
n~ * u l > g r o u p cat^ bi*cwrn l n v ~t~krrig l a ~\.idt* ~ a r i e rrf
t~
I t mah'lcs LIK tu cornpart, ;~pylcsw i t h oranges. U y s t n n t l ~ r d i ~ i nincome, g data analvs~\I c c l i i i i ~ l i ~ bu ~ ts~ l i ( n rtnhcarr-her w i l l a l w a v ~rmplu!. F-riv~riatc
tor curnplr, wt. cc>illd then t>xarnlnihchance+ in !lie incornr* po?\lhon c d or n ~ ~ l l t ~ \ * ater !mrthuils
,l o t ontl w r t n r nt-lother. Tlic tt~cliniclirr*s\ \ r i l l varv
rncn and wornrn (3vt.r firm- i n diffCrt*nt c t n i ~ l l r i ~If+ w e c'(lt~ld~ 7 ~ p r t ' s s hrlnr siniplc crt~<+=knbulatlrrnq and s w n l ~ a r i ~ o cltnrnc.rnt; to rncjrc. c n r n ~ l t l \
c,~chman'\ anrl each w c m a t i ' ~i n stniidar~lizcdtr~rrn,t1ri.n for any ycnr or n ~ k itivaria
l t r ~ r l i n i q u c ssuch as niuJt ~ p l crcgrt~ssici;lnlld d ~<crrrnin;~n t
country rr.I3 ctr~llrfcalsulatc thc a\.tbrage : r t ~ n i l a r J.;crrrc t ~ tmrn arlrl of anal\-si<. Thr analr~l.,m i ~ hrtr l v t>n wrnm.lry .;tatisfirs ( r - g. rrrnly;lr1.;rm
wonlcn. We c t ~ ~d~o lthi< ~ l c3vt.r a largca ntirntlcr of year.; and nnt b r of rncvtl.;, corn>l,?tinncntbfficirnk, annlv5i.; of \~nriance),uatl ~ l a p h i c a l
w n r r i t ~ dahout thc nvcr,lll tnfftuf of inflalron. rrprfit>ntatitlns ( c . ~ wattrr~mrns, . co1nplt.x I ~ n cgwph.;, sirtstcrt*d I w r
M'r MFI>LIJII stlt' thlft, 011 nvcrajic, hvrntat? ~ ~ o u l ibc 4 rclci.i% infi incnrnvs r l i n rtql nr rrly tin t.~bular prt.w~itaticm ( t ~ t - t ~ ~ vrra<c-tnbtrlatirm ay to
t.cblow thc ;IvtaraKtl and m,~le.; nbovc lhc ,jveraKe. I ~ v htnndardiring m u ltiwf,i\r crc,s~-t.lbulatiotix)
c s i n n rr.,~dilvstm(,
t t ~ r ~ a l l l 1l.e r\.hctht~rthta cap bt"tt\~rcnrncr a ~ i dxvtrrnvn Tht. I-htlir-t* l ~ t . t r t
c9t.11 t h c ~ rV J rtrjkab .~prnml~.I~c\ I\-111 <1ef~-nCIrtn thc
,
I n a r r n ~ u i t rerr
~ \ i * i d r r ~ i ~ i )(>v(*r
: tin~t.. I n ttrc un+t;liitla rtj i ~ t x i l f~lrtv 11 c l ~ ~ > ~ ~ n c t r , r i thr- t - ~ It...:. Il~~t.1mc~a<~irrtnt-nt)
~ t ~dnl<i ,init 11ie tvpi. of
L\ IW Id be rnclrt- diitlcu It tri r%ht,lblish$In\ Insrtqki<th o r drcllnt, bt-ca~rwbtht, atttl~cnt-ct ~ ir,li1t'F1 ) thl* an,~lysl. 1s d i r e c t ~ dChapter . rl p r r ~ v i ~ f ian
rk o u t l ~ n ~
' \ aIur,' of rncc)mc.: c t l n n ~ c 5'Iht* avtxr,iqc E~CIII~IT ~T~>~I\.CC'II IIICinctlrnlx.: nl a mric,.p i l f alr,>Tvcrc trnrhnrtluv< ;lntt t l i t , d a t , ~~on-.~d~.r,~tfo~i~ Ih;lt tviI1
c l b r n r b r l ,rnd wurntan m,iv h ~ v ~ricrt*,i~ecI c I n i1t~c)luIc~ l o l l tei.ms ~ ~ r ' t l ~ In~t aill-ct clnc's c h o ~ c vI ~ t ~ t w t ~ t lc i~L~Itl~rn.r
~n t j x (15.
penplc resptmd to sets of questions. I t is an approach tr) analysis that is of inferential statistics are cc>mrnonlyused in descriptive analysis: inter-
frequently used in s t ~ ~ d i cofs ntiitl~dcsand valtles and personality val cstimatcs and tests c ~ significance.
f
characteristics. Essentially it involves looking for pattcrns in sets of Interval estimates allow ws to estimate within what range of the
attitudes people hold. It seeks to drscover whcther there is an underlying sample figures the population figures are likely to lie. For example, in
structure to tlie pattern of responses to questions. For cxarnple. we might a sarnpIe we might find that 60 per cent say they intend to vote for a
ask people to tell u s what they ronsidcs to be the more important attri-
butes for children to possess (e.g. good manners, obedience, neatness,
! particular political party. Wc can use interval estimates to estimate
within what range of the 60 per cent the population figure is likcly to lie.
imagination, independence and self-controI). By observing the pattern of Using a statistic called the standard crror we would be able to say, f0.r
answers, wc might find that stme people emphasize good manners, example, that we can be 95 per cent confident that the real population
obedience and self-controf and place littlc emphasis on imagination and figure will be within a certain rangp of the sampIe cstimate of 150per cent
independence or vice versa. In other wolds ccrtain variables cluster ~ per cent or within the range of 58 per cent to 62 per cent).
( e . *2
together. %me people will select charactcristics that reflect the under- Another way of using inferential statistics is to use tests of significance.
lying factor of uorifo~mit!/whilc others select charactcristics h a t reflects Tl~cscarc frequently used in conjunction with correlation coefficients.
an ralltunnrny dimension. They provide an eshmate of the likelihood that a correlation at least as
Similarly, we might hc intercstcd in the types of things pcoplc look fur strong as that i~bscrvcdin the sample will also bc found in the popu-
in a job. We might ask about the importance uf the following things in a lation. For example, a correlation in a sample could reflect a real correla-
job: tion in the population or it could simply bc thc result of sampling crror.
Tests of significance provide an estimate of the likelihood that the sample
correlation could be due simply to sampling error. If there were a
1 good pay
rcasonalrlt chancc that it could bc due to sampling error W E wnr~ldnot
2 opportunities to use initiative
feel confident in predicting that the sample correlation would be found
3 having rvspc~nsfbility
in the wider population.
4 the feering of achieving sornerhing
5 absence of too rnuc1-1pressure
6 gcncrous holidays
7 g o d hours. Explanatory analysis
We might analysc answers to see if there is a structure or a pattern in the When trying tr, make statements about causal relationships, cmss-
way individuals respond to each of these job characteristics. We might sectional designs must rely on static comparisons between groups or on
find that there is a structure: sump people tend to list the intrinsic job correlations hehvcen variables where measurements are made at the
characteristics (npportuniti~sto use initiative, responsibility and feeIings same point of time. The problems with this have already been discussed
of achieving something) while others list t t ~ cextrinsic characteristics at length (Chapters 3 and 11). The basic strategy by which these prob-
(pay, holidays, hours and no pressure). lems are handlcd is by introducing statistical controls into the analysis.
By looking for a pattern in the way people respond to sets of questions
we are able to develop a more sophisticated understanding of the TIre logic af statistical controls
structure of attitudes and behavioilr.
The logic of statistical controls is best understood within the context of
the logic 01 experimmtal dcsigns. The experimental design relies on
comparing change over time in two or more groups, which were initially
As well as describing patterns in n sample we will frequently want to idenhcal in all relevant respects, Different groups are exposed to cxpesi-
generalize to the wider population From which the sample is drawn. We mental treatments and a c o n t r ~ Igroup is E X P Q S E ~ to no treatment.
1viI1 want tn know if the pitterns we have described in the sa~npleare Following the treatment, any differences between the groups on the
likely to hnId in the wider population and, if so, how close the popu- outcome variable is taken to be d t ~ eto the effect of the treatment. How-
lation figures are likely to be to thosp we havc foi~ndin the sample. ever, ii ! h ~p n u p s were different to begin with we would not know
When we use a probabiIity sample (Chapter 6 ) we cap use inferential whetf~erany post-treatment difference5 between groups were due tn the
stntishcal tcchnjqt~esto answer these sorts of questions. Two main types treatment or tto the initial differences.
I n rror.-w.c.rt i,?tla\ rft.ciqns tht, l o ~ l r1, to I-tvnparc crclllps \Y hicli art*
~ ~ ~ ~ (1t1r bt/,l.r bwis
l ul e\~.;tinq ~iiflrrcnces(the indi.p~nrit'nt variatll~:)
\t'p t ) ~ ~ Cc>mpJTta
r i OLIt<t>m~** ( t lie C I I ' P C ~ L ~ C I I I \)arrt~blc)In t hcsi* grr ii~pq. Thc s t r t m ~ t . an.ilv<i\
~t rxcusc whrn ~t 1%. f ~ ~ l ' r r L >fo * .7 tch on a tvhvlr 5t.t t ~ i
l ~m
\ l ~ r .difft,rcncr< rnav be interprettd .+ b c ~ n qduv t t l thc influence of variablrls ~ r ) ? r l r l ! r t r ~ ~ - ( l r r 1~ ht*
f ~ t . tirnrtnvnri,~blt,.;M.V C C I I troJ
~ a t once, ~ I T c ' more
dlftcrmt grm111 rnt~rnbt*r3k1p. Thus diff~rrncc.;in lhc incornc (drprnden! s ~ m i I a rthe grcrul,q wt5vtrrnparc \+,ill his. I-nr rx,~n~pl(b, wtnniiglit think !hat
\.asinl~le)tlf nralc< and femaler [indc~pcnclcnti ~ n r ~ ~ brni>:ht l t * ) I3c due t~, incomc J ~ l t ~ r c n s klct\\.t.cn cs men .lnd \\,nnj<.n .Ire ttue tn thta curnula t ~ e\
the cftect of ~ c n d c rorr incnmp H ~ M ~ ~ the V I male-fi.rnalc
.~, difference.; efft3ctq of gi-nclcr drfCprcnncc.; rn hcrurs o f rvt~rk,tvyt*. t > f job<, atnoun1 of
might trc J u t , not to ~ c n d r p7t.rr Gr but ZII oihcr dilfert>nucs bc.lv\rcrn thr wrrrkforct. ~ ~ u ~ r r i c n and c c * 19 li~ca~ion,ilF,rckgtnund. I\'(, .~rnulrlh!. to
g r o u p . T11c nialr p o u p rn~gfithi* (11d1,r. br bcttcr educated, includi- rcmtrvc3all tliehtb fartors frcm uur cornpari.;on.; of rncn and women. We
nlorc t'ul l ttmr* workers, bc rlre1\vn from d iftt-rent c r c ~ ~ ptiom ; l than thr m i ~ l i dt i v i ~ l co u r qnmplil into m a n y grnupG so ~ h a in n r Kroup rn1g1-1'be
tr-n~alegrrltlF. Tn i\wrk t n ~ t ~.hekht*r m.ilc-fcn~alc rltffYrr.ncr.< In xvorner? ~\.nrkrn);h1)1 tinlrl in yrofr.s51rm~Ii c h c \viit> ! l ~ \ ' ctrrtiary ~ quali-
irlcornc ~ l r cdi~tatv ~cnt1c.rwe r ~ t w !to ct)mp,ire curnp.~ra~~lr* nl.iIe5 ,111d fications ant! 11or.c a l Icnst 1-1 irrar.; worktorcc rb\perrclnrt%.Th~.;t> rcrulll be
tt~malt~s. compnrcci w i t h ,I c o n i p w h l c group o f men. ttnr~thtl.FaTr I I F KriIup n l l ~ h t
?litb c.xyc.rimt.n tall cit-.to,n norrnallv achievcb.; ctrrnynrat,~l ~ t yhehzeen bc < i r n ~ f , ~ rthct o ahrn,c. in 211. r r ~ p e ~ t s . e \ c c pthat I t l i ~nwn nnd w n m m h,lve
CrnuFG by lantium a q s ~ g i ~ m t lr ~t igro ~ n a l stink to
~t i ~ p sC r o r ~ - ~ c c l i ~rlt~.;ign.: trntlc jobs. M . I I ~r~tticrpair5 01 grtwps c-clulJ hc comparctl, R y ct>ntrcill~n~
d c h i ~ v ccrlrnprlrt7b1l1 h- h v I ~ I ( I ~ C I I F (Chapter
I~,~ 3) t w t , slrlctl Ihrt' r t . 1 ~nn for or rc>mt>vinginal~a-fcrnali. dit1t.i-encus we [.,in ,in.;ivcnr tht* q i ~ c s t l r ~ n .
ctiqting diffen-ncr<and all rlatn arc* ctdlccted . ~ at single time pcr~nl,tliv 'M'lrrrl r7llrt.r Ilrlny. nrtut-,~rlt~l, d(x&cgcndrr*mahr*a n v di Fft,rt.rir.t*to i n c o t i ~ rlf, ~'
mntch~ngis dtlne at tht, dala analv\is s~;lgc - ~ l t c rrlata CITC collcctcd ~%.hi,na11 irthcr t h i n ~ sark- ctlir;ll, rncn ant! wonrcn *till havta d ~ t f c r ~ * r i t
s , t~her than bt-lr r e . ~nc~lnic*r, !hr*t~ rvt. h a t t~evt~lenc-i~t>f gcntlt*r~li*crirnin,it~or~ Shr\ nlea,n$ that
Th~krnatrhing will alzva];? he s I~niittlJ, ~ t ~.I dIrs, . i d ( ~ q ~ ~mt*,ltls a t r ~ Illkin gt311~Ec.r iirtlf il: rt>spc~t~rlhlta ftrr tllr ~rh.o~iit,111lfcrr-nccs If,~ r ~ h ccz-c i r r-tlnlnd
r,indc>inalh~cat~cm o f achic\.jng grtwp corny;lr,~l)il~tv IZtbcnn on11 maitch !r>rGI sci c)t \,anahlcs, gcntlcr d lfertanctb- 111 rnctrmc 11 ~+alywar, 1k.e \rcwlrl
in r~l;ltionto .iriaI~lcs~ b n i ~\.l,ich ~t ~trtc>rnrntit,nhn-- brtan i-t>flect~d.7'his bax t47i1( I h c > i rri 1i,11 i t ~ i . o n ~~ i~i,( t ~ > r ~ ,tt*L\rta
~ ~ ~ tt!l{d4, $ (!I(, v'i r~~iL-l~~<\VL> h ~ v t -
nican5 wc n~listanticipate rc,lt.vont r;lriabl~l.;on w'lli~ll+t:rrnips bl10~11C1 ht* conlrnl lcif. I'lit 11 ~ffcrr-nilv,wtb wc,ulcl s a y tha b t hc. c{lnll.n'flctl iF rrabltbh
;l
n~,~tchcd. Our ,111ticipatirln r \ , l l l aluva~=,be I t ~ sthan pcrti3ct. crpl(ir?~ t h t x irliti a I c 1 I' ttt,ren~-~s.
-
I f M'I* wt1ri3e ~ a m i i i j n g!he i r n p ~ c nf
t gender nn incorntb wc would ncc.4 C I I c o i ~ r s rrt, is not pcxqil7le to control for c.1 erv posh~blev;lriaihlc, so I h i s
to con1p;lri. mcrl and tvonicn ~ ' I 1 ( workcd 1 ,I sirnildr number of h r ~ f per i yrr-ibili~!~ atwal,s rc,main* that any gcandcr r ~ l ~ i t cincc~~nr- d di ffrrcnct.~
~ v r t = kIt wnrilcl be inappropriate tn I-ornpre a grrxlp crf W I ~ C I Iw l i ~ r e cor~ldhtn dut. to thc<t* rrncuntrt1lletl vari~bltn+.
m a n y worked part timc ivitll a R ~ I ~ L I F )rnt-n w h ~ 3Wcre mainly full time Ilirria a r r .I vari~t\.of lvily\ in w h ~ c h.;tal~sticalcontrols A r c marJc a n d
wnskcr<. It maLcs m n ~ wnsc t to comparc. mvii ~ v h otvnrkrd f u l l titl~e fhprc. art- a I ~ r i c . l \OF wav\ of interprei-tng tht* ~ . a r i o t ~ pattern\
< of rrsullq
wi tfl u.cirncn who workrcl (~113 timtl and crb~t~n~t, incomc dif lrrcncrs il~dtc i l ~ i1~ t~tltaincd.Ejnv cmmrnon nlctliod o f ct>ri~rrrll in!: for v;lriablr\ ih
brtwccn !Itrsc ITVCI FroilFs. We ro11IiI thrn c(>mp,lre men and r%,orncsn the ilrabomti~tntcchniqlrc.
w h o wnrkcrf say 2q-75 hour< a zveek, and then th(3sc that work 75-21
hours, and so forth. Senqiblt. a n a l y t s rill seek to C ~ ) H I J I R IIikr P z1)!/1t
The elahoratinn technique
I f income d ifit-rt.nct=spcrsi-t hctwcfn rnvn ant1 warnen ichrn wr match
for (control) hot~rqof ~ ' ~ tlit'n r k wc h r i t 7 ~tllaf~ flle initial inctlmr, difftbr-
cnccs an3 1 r r 1 l dllr~to dkffrwnf t~o~rr.; r>f wvrk trf mptl anti ivlrrncn, T h ~ r tih, insufficitwtspace licrc to cxplorr all the Jcatall* nf t h i ~trachniqrlr.
Ilthrla~ghthrrt, m<iv Ilr' nthcr %actr,r, ttwt ma\. ,iftrot ink-r>mrrlifttxr- An ruccllr~ntImt~kt h ~ fullv t dt.\cril.rs thv mrthc~tli\ !hat Atr<~~~t~er~.
ent-tl~l i t ~ t w ~ t 'nlcn
n ,ind wotncn, rIur ,1n,rly5i< will a t Ir>,lstt i . l ~ . ( ~ elilni- Us. 1 tl,yrr I$ S I ~ ? ~ A~ J~ I+ /TI J! IFL!~~S).
II~~
natthrl hours of \x,nrL. a 5 t h r t-\pf,qt~ntit>t~ t<rr the incc-tme d 1fitwnrt3s \Ye
micht iv;lnl t t r cl>rnp.lrr tncmnitl drt it.rtwc.rs .Irnlmq rncv and wclrncn ~ T I ~ R I
sirn~lar~rcc~lpatltm.;. Wc r4t)ulcJco~itrollor (rnntc.11) t)c-cupation aritl cnm-
part- profe-~c~naE mr1n \v ~ t hprcrtcs~~nnalwclrnen, clerical men with
clertcal women, and 50 on. If men nt~tlwiinien in tliu same tvpr t ~ fjob
ha\ i> di l f ~ r r n rncr)mtas
t thcn \vc r%mcrul~l h a v r tcr say that inrclmr J if!t*r- 1 C'rvrilrrt t tl~r' I~rr~rrrrnlt'nijr~l!/.;r~ This will cx;lrnlnin tht. rcla tir>nship
cnrr>+hchs-xn m r n and \rfomrtl art. r r t ~cluv tn the rliffr~t'nttvpes (if jotw bt9trvt.r.n tht. rnrlrptmdtv~tand dt.pcn~lent x.ar~;lfilr.( c . ~ ~ t w d c s rand
helrl by rncn and wnrnen. iilcnrn~,).'l tiis . ~ n!aI s i s might be i r l t h ~ fnrm
, nl corrtalat ~r,n+,t;llll(a.;or
204 CIIOSS-SECTIONAL DESIGNS
3
(see below).
Drdw a r~rodrlof the rclatinnship you are proposing between the three
1 fi) Cnnd~tionalhblcs rt.lat~onr;hirof g ~ n d r rand Income cot~trollingfor occupition
There are four main ways of thinking nbout and interpreting the results
of elaboration analysis, as follows.
RII Gender initial relationship was due to the hour5 of work variable as represented
Male Female
in the figure.
This is also illustrated in the cross-tabulations in Table 12.5.In this case
Low tlie zero-order relationship shows that a greater percentage of men than
Inconir women are high income earners (55 pcr cent versus 40 per cent). The
Iliglr
cross-tabulation for part timers shows that virtually the same percentage
of men as women are high income earners (31 per cent versus 33 per
(h) Conditional rablcs: rrlatinnship uf ~ n d e snrl
r income contmlling fur lt~c~ticm cent). Similarly, among the frdl timers, the same percentage :ef men and
uf women are high income earners (65 per cent versus 65 per cent). In
0t11~rwords, in this hypnihctical example, when men and women have
comparable hor~rsof work the income differences between men and
women disappear. T1-2.czero-order relationship is due to thc relationsh~p
betwecn gci-rder and hours of work and t h e relationship behnre~nhours
lncnmc of work and income, The fact that there were far more part time w o m m
High than part time men in thc sample generated the correlation between
gender and income. Had the sample had the same number of past time
men and women then the initial relationship between gunder and
Gender
income would not have appedred.
Male Femalc
I..ow
Income
High A spurious relationship is one whcrc the correlation b e t w ~ c ntwo
variables is not because of any causal relationship between them hut
because both the variables are related to, or outcomes of, some third
variable (Chapter 3). For example, let us suppose that people who attend
religious scl~oolsare more religious than those who attend secular
TND~RECT CAUSAL RELATIONSHIPS schools. If this reiationship bccause religious schools make people more
religious or is it because both factors are produced by some third factor?
Elaboration analysis can also be used to clarify the nnturc. of the xero- Ln tliis exarnplc the modcl is proposing that the reason for the rcla-
order relatinnqhip. In this case we might be asserting that gender tionship between type of school and religiousness is trot because one
casually affects income. But we might want to know by what mech- cailscs thc othvr but becausc religious parents arc more likely than non-
anisms it has this effect. Dofs gender affect income because gender religious parents to send their duldren to reli*ous schools (Figure 12.2).
affects hours of work, which, in turn, affects income (Figure 12.1)? Herc Religious parents are also more likely to raise religious cl~ildren.Thus
we are controlling for hours of work in order to see if the relationship the reason whv young pecrple in religious schools are more religini~sthan
between huurs of work and gender is due tu the relationship hrh~rcn those in secular schools is because religious parents sent their children tu
gender and hours OF work alld the relationship between liours of work such schools nnd raise religious children.
and income. If the relationship between gendcr andfincome disappears Resulk such as those in Table 12.6 wauld be cnnsistcnt with this
anlnng those with cnrnpar~lbl~ hours of work [hen we would say that the interpretation. The initia! relationship disappears when .controlling for
zoe CROSS-SECTIONAL DESIGNS CROSS-SECTIONAI. ANALYSIS 209
Low Child's
Income religionsness High
High
(a) cnndltional tdhIcs: reIat~an.;hipof school type and religinulncs.; ctmtrolling for parental
(b)Condltlonal tables relalionshlp of gcndcr and lncome contrulllnp, for liours of work rcliginusnms
I .ow Law
Incotnc Cliild's
High rcligic)usnesc High
N N
-
Controlling for parental religiousness parental religiousness. Among those with religious parents we see the
qnme level of reli~inusnessamong chi4dren regardless of thc type of
School type school attended: 65 per cent of those children in religious schnols were
(religious or highIv religious compared to 63 per cent of tllose in secular schools. The
Parental
same patteyn 1s repeated among those with non-relipjous prlrcnts: 15 prr
cent of children attending religious schnols were rcligieus compared to
18 pcr ccnt of thosc in secular schools.
religiousness
SLIMMARYOF INTERI'KE'CATION OF DIFFERENT PATTERNS
r r l n l ~ o n ~ l ; ktalren
F i p r e 12.2 MorJr,I~irlircnfir(q~l~drrect tp sdrod /ypc and The strategv in elabnrdtion analysis is to compare the pattern in the wro-
stridr~rtrc.lrgror~rnt~~p order relationship with that in the conditional rela tionships. 711e
Multivarjahle analysis
Thr ~ l ~ b n r a t i nmodel
n ha%a nurnbrr elf prrs2llcm%that can limit its u w - Thcrt* is in.;ufEicir*nt qpace hpre to descriht* tht. rnultiv,iriatc
fulness a4 a tz7av of contsnlling ft,r varinbli~qin crosh-wctional analv5iq. fcshniqtrt><.XLin]: simpltl yct sr~phi5ticatcdintroduction< are available
h rnajnt prol~lcin is that this sort uf malcliing rcquirrs very Inrgr ( K e r l i n ~ c and
s Pedfta7izr. I(T3; T~lrachnickand Fidell, 1Q89, Grimnl and
samplt>s. Bccause it involvt*~subdividing the snrnplc ~ n t osrparntr Yamold, lLW5).
groups for each catrgnry of tlie test varrabIc rve can quit-kly t n d lrp with
quite small ~ ( w for p an.?ly.;is. If thba te%t \*arrat.l~has, <a? fivr
caleqoric., caili subcroup c;En t-e ~m~11. I f wc rzant to match fnr h v o
\.ariallle\ a t once thi.; probFem is compcwnde~d.7 he iliificullics crc,ated bv Cohort analysis
runnrng nut of c a w s ;IrcL so wriot!q thal this form t r f nnalvsrs iq lim~trri
Frt~qucntZvto r n a t c h i n ~cln just onc v a r l ~ h l ent a time. Cross-~c.iticlnal rlcsii:n5 art, Ijmitcd in thr extent t 1 5 rvhiih they can
\ < rvtall .I< rttnnrrlg ntrt oi cnq= thc rr<trlts o b t ~ i n e dfrom t*labr)r.itir>n pruv~dr~nfnrmationabout c h r ~ 1 g trx e r t ~ m t I~lm\,evcr,
. reptattad ~-ro--
~n;llyst'h can tlrcnme vrrjmcomplex and cfifficult to ~ntcqlret.\Then 1r.r <cctinnnl < t l ~ d i e scan yrm-idc valuahlt. in<iellt.~a b i l ~ t tchanqe Cohurt
Iia\-rn man! diflcrrilt cr~nditicmal tatlle.. thc logic invnlvcd in ~nterprt*tinc annlvsi'i can I I ~ CJ ~ ~ ~ i ofi ' rrpeated
< cros+\ecticinal .;tutlic.; to drscrrhe
tliv vt~ryin):pa t t ~ r n s(,in I,ert-lmr \*crycornp2vu. It unn heccsmi. ex treniel v o,ygctrp~rfrc h a n ~ colfer trmr and to Fdrntif\p the cxlmt tu which this
dhfticillt to taupla~nand prtvqtlnt tlie resirltq of anvthing but the quitc change is attribtttablc to pt'ri~d cttrcts, agclnj: ,~nd cohort c*ffrcts
I
q~rnplctlYnhrrrat~onanaly\e< [C'hnptvr 7).
212 CROSS-SECTIONAL DESIGNS CROSS-SECTIONAL ANALYSIS 213
Clear period effects are evident if, within any particuIar year, there are
no differences between the age groups (i.e. within a column) but there
Ageing effects or developmental effects are thosc where change is are differences between each column. In Table 12.9 we see that in 1Y70 all
attributable to people growing older. These can be detected by reading agc groups had the same proportion of r e ~ J a church
r atkenders. Age
thr standard cohort table diagonally-from top left to bottom right to identify groups were indistinguishable. This irrelevance of age persists in every
lntr~coho~t trends. Reading down the diagonals of the table enables US to decade. Period effeck are evident by looking a t the trend across cnlumns.
see how a particular age cohort changes as it ages (Table 2.8). Iiow have In cach decade the proportion of regular atlenders changes, and it
people who were in their 20s in 1970 changed as an age group as t h y changes in an identical way for each age group.
havc moved into their 30s in 1980, into the 40s by 1390, and into their 505 A n examination of these p e r i d effects maables us to see how people
by the turn of the century? within the same age group differ at different periods. Thus in Table 12.9
Table 12.8 shows perfect ageing cffccts. Tcn per cent uf the cohort of 20 we can see that people in their 20s in 1970 are quite different from those
year olds in 1970 was attending church a t least monthly. Ten years later, in their 20s a t the turn of the century. The age group is constant: the
when this c o h c ~ would
t have been 30 or so years of age, the church penod in which this age group is living is different.
attendance of this cohort increased to 15 per cent. Another 10 years later
(1990) this cohort - now in i t s 40s - displays a further increase in church
attendance. The same ageing pattern can be seen in any left to right
dnu~nwarddiagonal.
The concluston one would draw is that as people get older they Cohort effects arc cvident in Table 12.10. That is, wit hi^ a l l y grzlun vpnr tlie
increase their church attendance. This is consistent regardless of the a ~ groups
e have different proportions of regular attendcrs: the age
1
years one examines, or the age group examined. group makes a clifftrencc to the Ievcl of frequent attcnders.
214 CKOSS-SECTIONAL DESIGNS CROSS-SECTIONAL ANALYSIS 215
nl I ~ m trwu?i!lrly by ngc
Tablc 12.9 Colrcrrl table ofpercrntlrfi, flttc~rdirrxrl~~rrcl~
~ d 1 L?711-2110!1: p e r i ~ ~E ~~JtC Z S
culiort a ~ Ilcnr, I Cohort effects are evident where there is change within a column and
within the rows, and the direction of change in CI column (c.g. from low
I to high per cent) is opposite to the direction OF change w i t h ~ nthe rows
(e.g. from high to low per cent).
l'hlq ch,~l.trr ha.; providcd guic~elmras fnr thc dcscripti\'e and prplanatory PART V
, ~ n a ~ y ~ i i-f, datGi gcncralrd by croqs-?t>clinnal Arsigns. It has not
sctnsirlcrcd the spclclfic tcshniqur;, but manv of thc .;tahstica! $;.l~~dcliiics
pr~ivirlcdIn Ch,~ptes6 apply to croqc-sectioiial analvs~s. CASE, STUDY DESIGNS
The ctzapter consid~rcd the particular problems of using crom-
.t.ctiunnl data for causal nn;llysi.; - pmblems- creattd bv the absence nf J
tirnc dimension to the d e * j p and the lack of rnndomi7ed control groups
The genera! wav nf dealing with thcsc prohlcms is a t the analysis s t a ~ c
r a h r ttian at thr r c s x ~ r c l rlrsip
i .;tag?. Thc chapter otltlined thc. loqic uf
ilsing ~ t ~ l t ~ s t icontrols
cal and rnultiv;lr~ateana1ysis r;o that cm.;~-wct~clnnl
data c.in bc uwCul for ci-tuqal analysis. tt tllrn described on(-particulnr
form rrF rnultit~arintean,llvsis - elahoratinn analyqis - to prilvidc an CASE STUDY DESIGN
itnderct,lndin~ctt the \ r a w in which the pmblrms of cauqal annlvqi~cnn
1~ tacklta~l.Finally, the chapter d ~wussedn rnetlioil oi a n a l y i n g d a t a
f rmn rrgptba tcd crow-section~ls t l ~ ips. d In p.irticular r t fc~uscilon cohrwt
nnalvsr> - how to conqkr~ictand intcrprct ct~hclrttahl~.<and tn be awnre
c ~ Ihc
f 111nihctl t h i ~typc of analys~z.
For rn~inv~ ~ ~ 1 1t . h5 c~a w *iudy h a s been Ihc rlglv ducklling of rt.sr>arrh
d e s i ~ nMust rtwurch rncthudu t r r t 5 eithtbr ignore c6i*csturli~~+ or cotifuse
[hen1 \ \ , Ith ntlit-r tvpcs o t 5 ~ ~ ; i rt-*tvrcI~.
al \I Iwn cawa *tud! JL,+~LW~ 11~1t.e
been disrussed thry havr generally been scrn - trom a r n r t h u d o l o ~ ~ c a l
p i n t nl view - as 'soft'options %me curnmrntatca< believe* cnse 5 t u ~icq l
+auld bc w e d rmly for cxplora tcsv research: ke genera tc Iir.pothesrs tnr
futurc tcqting ~ 'th i rnLJrv r1gor011rrcsearch designs.
The rnrthodoloby of c a w study design has not bccn neglcctrd because
c a w stud ics h,>\rt, bccn ~ ~ n i m p o r t nin r ~~t n c i a scicnccq.
l Indcrd they have
been f~uldarnt*nt,llto thy <tlbstantirfc2nd mcthodolo~icaldct'clnprnent of
the st~cial scicr~ct~s. In hcrial anlhropology studies of trihrs linve bfcn
c a w studies prrr t,.rrr/lrrlcin.Community .itudies sirch as the Lvnds'
h4iddlr,to:l*~1/lu?ql and small GrOUF s t ~ ~ d i ~like . ; M'hyte'r; 5trr7t.t Corrrrr
S o o r f v (1943) have rnacir major crmtributions to lllc development of
ii(7ci~~Iogy as R d iscipIiiir. P~yc'l~onnrlTysi~: ha4 thrivcd on a case shtdy
approach (fr e d , l Q 5 ) Educativnal r e c a r r h , cvalnahon rcwarch and
organi~ntionalrtksearch haire all made ~ r t m s i v e11qr of caw studim to
instcr f h ~ i de~cli>prnen
r t.
The I ~ c ki3f a - \ sbernatic- rll~cu.;.;ifmof w.t. studr. d r > i v q 11.1q hequn to
be a d ~ l r t w e dIn rtacent trrnr>\. I'in ( 1 'J89; 7Qq3)h a < provided ,I pnrticuIarl\-
useful tl-catrn~ntof casc \ ~ L I C I V Jrasigns ;lnd .I porvi,rful dcfrwce of their
value.
This rhaptcr rramincc n range of w ~ v sof dcvel~ryinc:case ~ ~ I I J V
d e s i ~ q Thi5
. i< a difftnrrmt task to that ~xndcrtakcrl for cuprrimental,
loneihlclinal nncl crc>ss-\r~ction~l ~lrqignsI~iyn~r.;e,ns Yin points nut,
'unirkc 4,tht.r sr.;parch strr7tept,s, the prjtcntial "c,rtalog" r r f rewnrch
desi~n.;has n d !,t*t been dtnc.elript>d'(1Q.q'): 27).
CASE '4 UDY 171.51GNS CASE q1'UUY DIISIGN
f'or ruarnplc~,tht- c a w ?hzcly mrglit begin ueith tht. talloxc~ngqutb>t~on: RegLtr~ITt-+of the c r , r i ~ l i l t v *of~ ~t l w t l ~ c t ~ r vthis , ; ~ ~ ~ r oto~ icase
cl~
is the cjfr3it ~ l de\-crlvcrl,
'\\'?l;lt i c i h r ~ base2
~l icmtrnl of staffing on tlic studit- bt,crri.; with il C . C ~1+k ~ - \ ~ ) ~ r t a t ri l~r~. ri~i \*, ~Fr~lrn
d prr.\ trru5 rwcarch
qt~alityof e d u c ; ~ t ~ nin
n a scIio~17'Dt'\~oll,rd or ~cI11>idhasrd staffing and /nr thc~rricq.lye approach t h e c a w y t t ~ d vwith thc purptr.;c nf test in^
cc~ntsulinvolvch 13~ich<chon1 appili nting, dism k i n g ,tnd promoling stalf OUT tl~txor! in a reaj l i l t > <ituation. Clc,?rlv the selection t ~ fthr real life
a s well as scttlng pay levels ant1 Icnchtng conditiitns. Devolvtd qtaffing situation (tlw cast) wnuld nr.cd to match thc c a n d i t ~ o n sundtxr which the
svstcms can br contrasted wit11 cmtraFi7ed systcrnq where public service theory priipcnes particular r.rutcomus (Chapter 74).
burt=aucrats managc these staff in^ rnJttcrs at head ofiice l e v ~ l Wc . rn~ght
I7ccln rvith thc prrrpr~Gtiont h t ~Jr\'nl~.edcontrol will yir.Id much Iwttpr THEDRY
Iil'll n f Y G CASE V r t r n I F 5
r d ilea tional nutcumcs than c e n t r a l v d contm!. T l ~ crcnqoning \vnitld be
i lint Iclcal conlrol makes peoplc h r more accorrntablc, enables a schnnl to
us in^ ;I thc=tjrvbuild in^ appr<x~chtn c a w rturiics we seltxct cnscs to help
Jevclop and refine tllc prr~p.,m~tir>ns and d ~ v c l o pa tllcor!. that fits the
build a staff prr)Filc appropriate to its necds, and makes it much ~ n s i c rto
case5 wiL\ttrtly. For exaniy-rl~.,cvc migh l .;larl with thc silnplt. pi.oposition
rrward qtaff wht, nrc a c h i e v i n ~the- oiitcornes dcsirrd bv the schord. The
that this intmductirln drvnlved svstc\nlr wi I t irnprrwc. cd iicationnl
~ n ~ p l i cthconp
it hcrc is that edizcatinnal clualih iq fundn&entallr. a rcsult
nutcom~.;. LVc might .;c*Ft~-t a schmd rvlwrc 5uch a .;vstcm has Lwcn
of t ~ n c h e quality
r Ins opposed to thc nature of studcn ts, school rP\cllr re[.<,
intrcducrJ and find that t*rl~~caticmal outcrvncq did, in f;lcf, improve. We
nr~,lnitationals t n ~ c t u r eparcntal
, ini.olvemmt) ,-inti !hat teachcr q u a l i b
iz a function of tmchcr accoz~ntnhilityand teacher 'fit' with t h schtjol. ~ rn~glitthm find a scl~rx)lrn which [fir inttndrrctivn (]I fht. de\~ol\'ed
staffing \vstcm did t ~ o f~>rt>duce better educat ic~naln ~ ~ l c n m rIl'hy r . did i t
The selectinn oC a case to test this theory would 11.i111 u s kr, find a sclioul
tlrnt has intrtld~icccl'I cievolvcd staffing system. 1x1, thnroughlv invt~qti- producc grind otltcomr7.i i l l wmt. cascs and not others?
Our an,~lvsi.;of cnch c,l\r would aim tt, highli~litclitfrrt~n~+t~s bt*hvt.cn
g d ~ n f the i school (including thc 'cmhrdded elcrnrlit?' such a s tpnc'hcrq,
cases 5%-hcrci t did and Jiii nnt work. I h r analvqiq wntild , ~ l < oi d ~ n t i f v
man~tycmcnt,partlnt.;, past and prthwntstlrdrnf5) \rfc 1vnu2d buqF(1 u p ?,
cornrnt~nal~ti~ls among r,twh+ \\ here rlcz.r>l\.t-dqt,+ttingna* ~ u c z c w f u Tand
p i c t l i r ~of t11~rlu;lltty irf edur.-ltioii In the schcrol (houueverdcfintrll ; ~ n d
commn~r.~l itit,.; arnnng c,l.tn\ whcrc ~t 1x.n.; u n ~ ~ r c c c ~ s f( tt, t. l~ . t hc wav in
n i ~ out l ~ ~any l111L.s [hat exist bc*twt-ci~the cducatinnal q u a l ~ l y, ~ n %hc
~l
which tlic -y.;tem was irnplrmtmt~d).Tlirsca case shidivs coi~ldthen bc
staffing systpm 11 pi~qsibIew~ wtn~lrlcollect histnric,ll data re la tin^ to the
used tcl dcvttlop a sct o f 1>ro1~o~i tion5 r ~ l > to lthy
~ conditions and con tcxt
timtl before tlic dt*vnlved s y ~ t c n lwa5 introductld lo see if tlicrc w , ~ s
rmder wliicl~dcvolvt.rl .rtakfing system.;, Ir,i~lt o impmvrmr>nt~ (Chapter
c\r~rlcnccof improvement after it\ introductic~n Wtn ivcluld also qcck to
~dcntifyw h a t r.l.;r% happen in^ In the schnol that might hay[* Icd to 15).
Rie Jiffcrcnce b~t\vrtln thc t h ~ o r y trsting and thc~rrv building
irnprt~r.erncn!s In thr yualihf of education err factom that r n i ~ l i thave
a p p r c ~ a c h cis~ that in thr fr~rrnerw e Irr4~r)l with a set c)f qt~itpspecif~c
prc\,cntcd imprnvements dcqpitr the introduction of the rlevolibcd
propc~qitic>n.;and thm qrc if thcsc wnrk in rrlal world .;it~tntions.Ln the
systpm.
theory building model wr begin with t ~ n l vn question and perhaps a
Tlie point nf tlic care s h ~ d vr n ) i ~ l dbe to qrc i f thc thenry actti;lllv
basic propor;itinn, lnnk a t rt=al cases and rwri trp with '1 m ~ r wspecific
worked in a r ~ a lil fc. situation. I f i t did work then thc theory i s supprwtcd
theon rrr s r t of propnsition.; it< n rcsrrlt r)! rtarnining ; r c t ~ a cascq. l
(nvt proven) 1 f i t did not work thrn we wouId <ct,k tn understand, f r t m
il c < ~ r c f anafvv.;
~tl (if the case, w ! ~ ythc predicted outcome f i e t t ~ rcifrtca-
tic~n)did not evcntuatc. Is it trccalise the theory 1s completely wrnng?
IJclrs the thct,ry r r q t ~ i r esome rc?finernt.nt? Is lhc I hmry applicable only Whilc tlic t w o above c a w stud y approachrs ,Ire rlrrory ~*c,ffirr*ri in that thc
under specific c~rc~tmstsnces? goal is to i ~ s cthe case to trst, rchnc and ~+cvclopthenr~ti'al gcn~rali7a-
In.;tead of thi+ wrnptc prnpnqftinn ive mie;ht dcvclclp a rnnre complcu tions, fh~lclrnical ca5r >tudv use:: theoricc: vrrv differentlt.. Clinical c a s t
modrl that spco flrlt what wc+ultl h ~ p p c nrr-ill1 t h t . intrcldrlctmn of n s t u d i ~ ;lrcx
s r i ~ ~ it e. ~ ~ t r ~Ttl !~ t USP
' ~ ~ther~r~tah to understand a c.;lst.
dt~\ol\-cdstakting stern. The rtrnd4.I rnlght a n t i c ~ p n tpo5itil.e ~ effect< in Thc w.lrr in xvhich a r l ~ n i c ~ ndeals n wlt'fi .I iiicnt illu5h.1ti.c t l i i q s ~ l of e
p,irticnlar ar;pcch of educatior~alnutcomes and ncg.11ir.c e f f e cin ~ r,tht=r% cast. sttidy. 4 child mar V I ~ I aI p ~ v c h o l o ~ ~because .;t Irc is prrf(mning
It inight specify pnritive effvcts c~nlyafter a givrn ptriod of lirnt*, and poorly at srlinol and h,i+ ht,~.nmed ~ s r u p l ~ The v ~ ~ tnqk
. nf thc cl~nical
only when thc d c v u l v t d systcm ;1. i r n p l ~ r n e n t ~ind n particular wnv (r..g, psycl1c71uci<t is to work olrt w h a t is going c~nin this casc ,~ndw h y it is
! r ~ t hbu!l teachrr and cnrnrnunitv cor~sultahon,wrth appropriate. cliccks h a p p ~ n i n < .tZ goor3 clit~ici~ln ~ 1 1 sfart
1 ~ ' t t 1 1~vmptams.wlrat is happen-
anrj balance?). In t)tller weord<,thz= thtw? xve spvcitv might nnttciy.~tc ing a t <chr?()!' In {\.hat nril,i.; of schm~liq tht~clitldperfr3rniin.: ptrv-lyTThc.
I h n ~when cond~tinn. ;i + H + C + D are met thvn thc devolved ?v.;tt=rn clinician rr.il! undt~trbttdlvti.~vt.a rancp ot pi)<~ib!c e ~ ~ I i l n a t l t ,tor
i i ~Forv
1 ~ 1 1 1~ r h i c v c t s aspcctq f and l' aftpr a p'crincf ot V ZivrArq.
I n ~ p r r l r ~ r n e nin schcml pt%rfrrrmanceanrl d ~ ~ r u p t i v~ eE I I ; F I . I On !I Ischncll.
~ 5hr1 \\,ill gather
CASE 5-I'Ci131'DF.SlC;U
ut chararlteri.stics we e ~ p e c !trill
t grr together tn a c a w . Thcv represent a example. In other sltuatron? we mnv rely on a single cr-rtrcnl case. This can
p i r r or theoretical tvpp t h ~ i~t ilcot necessarily found in that purr fc~rrnin be appropriate when rve have a clear tlieorv with well-Filrmulated
empirical reality. For example, while Wcbcr identified the charactenqtics propositions and wc haire a single cdse that rnects all the requirements of
of a ' p w ' form oi bureaucracy ( G ~ r t h and Mill$, 1946: 1Yh-204) we may thc theory. Such a case can provide a rnodcrately convincing test elf a
find no cxan~plcsof aciuaI bureaucracies that function in exactly the way complex theory (Yin, 1989).
Wehcr postulates. Multiple cases, strategically sel~ctcd,c;ln providc a rnrrch tougher tcst
Deductive typologies and ideal types can providc a useful way of of a theory and can help specify the different conditions under which a
analysing casc studies. The ideal tvpe can serve as a yardstick a~ainst. thcorv may nr may not hold. Fusthcrmorc, lnultiplc cascs a r c essential if
which tn compare actual cascs (Chapter 15). The template pmvided by the case studies are heing r~scdfor inductive FuToses (see Chapter 15 O n
the ideal type can guide the way in which we investigate the casc; i t can andlytir induction). Given sufficient resot~rces and access to cases,
guide what we Pook for. multiple casc designs will norrnnllv bc more powerful and cnn~incing
and providc rnorc i n s i ~ h t sthan single case rlesi~ws.
I Iowever, when using a multiple case dcsikqi tvc should endeavnur to
An inductivelv derived typologv is one in which we start with a cluesticln treat each case as a single cnsc so that we are able to establish a full
and then examine cases in the light uf thc question. A comparison of account of that case before engaging in cross-case ctmparisons. The
cases can then highlight rlusfrrs nf similar cascs. For example, 1 com- unity of thc singIc casc shauld bc respected (Stake, 1994; Yin, 1989: 56-77,
menced a set of casc sttldies with thy question, 'How d o aduIts pet on
with their cldcrly parents?' I s h ~ d i c da mnge of cases and found that
cascs diffcred along a rangc of dlme~isions(u.g.lcvcl of conflict, tension, h paral1c.l dcsign IS one where all the case st~rdies arr clone at once (e.g.
drpendencc, balance, r n j o y m c i ~t, inh~nacyand intrusit cness). I then ddtcrent in17rstigators eC3chdoing a cast.). On1y ( ~ thc
t ct)nrplct~r>nuf thc
noticed that cases tended to cuhibi t diqtinct cl t~stersof ~Iiaracteristics.On investigation of tach r a w s h ~ d varc cornparisnnq madc bctwren the
thc basis of thesc conimnn elements I d-l~rlract~rized particlllar cascs as cases. Thls strategy can be appropriate when the project adopts a simple
heing of a particular type. Mrurking from a question and actual cases 1 theory testing apprc~ach.
then deveioped a fourfold case based typology of adul t-parent relahcm- A sequential d e ~ i ~i.; nonc where case studies follow one another.
ships that T called parent centred, child c e n t r ~ d ,remote, ~nterdcpendent Using this approach one investigator could, in princ~ple,conduct cach
(de Vaus, 1494). case study. An advantage of the sc.qucntial approach is that the selectinn
The particular characteristics of each type nted not concern us here. of each casc and some of the issues examined can be infc~rrncdby
The point is that typologies can bc developed in different ways - tither putzlcs identified in earlier cases. One case can ihrow up idcaq t h a t can
deductivtly nr inductivclv. R~girdlessof the way in which they are influence the selection of subscqucnt casts and that can be followed up
developed they can provide a helpful template for conducting, arralysing in these later cases (Chapters 14 and 15). When s d n p t i n ~a more induc-
and reporting (Chapters 14 and 15) descriptive case studics. tivc, theory building approach a scquen tfal d e s i p is mnre appropriate
than a parallcl approach.
WIlcn designing case studies tlierc, are n numhes of elements, apart from Most case shrdics, a n d all explanatory caw studies, will incorporate a
!how discusscd abnve, that can he built intn thr design. time dimension. Without tliis dimcnqion anv adcquatc causal cxplana-
tinn is not possible. Casc studics can provide a p w d wav of carrfulfy
mapping t11c scqlrcnce of events, which is the basis of c n u s d explana-
tions. It has to be decided whether this t i m ~dimension is obtained
A case study desigr ran bc based on slnglc or multiple cascs. A single rctrospectivel y nr pro5pectively.
casc design will norrn~llybc less cornpcll~ngthan muEhple casc d e s i p s . A rctraqprctirlc design involves collcciing, on thc one occasion, infor-
Using the logic of rep!ication (Chapke~14) a single caw represents only mation relating to an extended period. This r ~ q u i r e sthe reconstruction
one rcplicatfnn and does not necessarily provide a tough test of a thcorv. of the history of the case This rnlght be done ttirough the use nf archival
However, W E mav have little choice. Limited access to cases or the records and drvcurnents, or interv~ewswith peuplc whn participated in or
extreme nature of the c a w may mean that w e ran nnlf study a single chserved past events.
T'lriq J t , \ i ~ n)la\ lht*clbvit~usprut7lcnls as\r~crali,d with 11)~s01 c ~ ~ r l c t l c ~ ,
rec~rnstruchr,nr l t thc pact rn t l i ~Ihgllt of t h v prra?en!, a n d m i + t a k l n ~he
%pqupnnn in rr,hnt=heiprnt+rvc~~rreii. l-lt)~vc\~cr, in manv qitualions tf~rrc,is - --- - -- ------ ----- - -
l ~ t t l ccl~oiccll~rtt o dr,lw on pcopltb'.; ahiliiv to rccall thr p + t With <.ilse I E rplannlny"
- - -- --
I
.-
Ll~rrnct~vn
- - _-- --_I -- --- rlp'Annlnry
- ,I
+tuifrc<tlrc ilsc ut rnliltiplc sourccc of rvicfmcc can t-educe t h t ~ r n b l c r n s
that qcl 1vit11 t h i ~approach.
c - - --
T~mp
-1-
/ Ravosp~fw~
-- - -
h s m w e
-- .-- J - - -1
Relm4~~-1w.-
-
P~FIIPE~IV~
- -I
A ~ ~ r n ~ ~d uc ~y n s ~involvcs
~ t ~ r t s ~ c k i n gc h a n ~ e sforward over t i r n ~ .t t 1- - C a r crde,
I I I 1 1 I
- 7:-1
-
I I I
ha.; thc obx tolls atl\'nntayr of enatlFin~,thc inur.;h-q,ltcw to I(wk at e\.rnts ?
) ; ; I ?
thev tnccur rather lhan relying rm partial and rcron.;tructcd nccoui~ts.
I3t*pc*ndlngon the i<srrc.; uncfcr consirleratic~nand thtl t y p s 17f c,rses, I--,-
1 3 1 pwatta I wJwb8ql.-
*-
phml,om , s w , w n a
- -1- -
I w l m a 8I
parfilb
I- t --i - I I -1-1- I
pri)<pcrtt\.rcast* \tud1e~might la+t for l;r.lrs. 'The u b ~ i t , u s~Fisailvan~agc
r a n bc I l ~ c t i lnc and rc<ourct's rryuirrd. 'I hew c u n ~ t m i n t scan <cvt>relv
l i n ~ the~ t nrtmber . ~ n drangeof ca.;~.q that can bv5h1~lied.Thi\, in turn, can
rcqult In \'rl.v I1rnrtc.d teql. ol a tlicor\e.
lipgarrl lcrs nf w l i ~ t h ~ar r e t ~ t ~ s p v c t ior v ~ pr(xpt=ctivc approash i~
d . gi>:clis trl build 11p J d e a r ant1 reaqnnnbly ~letailcd picbrirc, r > f
~ ~ d o p t rthr.
thc qetlilcncr In whfcll cvr-nt.; took place a n d nl thr conlcxt In which thcv
i ~ r ~ t ~ r rTu ~ i ?b.( ~ i l dmeaningful c a ~ ~ + t1xplanationq
aI we must grt tlic
serlucncc and thc ct~ntextr ~ ~ h/It .con\-incing c a u ~ a r\planatrt~n l ivjll bt,
~ b l t .to frack t l ~ i b 'ql(vy' bv which Ijnc (or a sct t 1 0 c'vcn t(3) rb17tl1nrtlip
producing .I particular rjutcomr.
Jr5criptilfe o r rxpfanatnr).
a tlirory testing or thi.nnl building
r singlc case or rnultipie casr
= hrd~sticor cmhcddt-d units of analysis
pr~11'1 or ~ e q u e n t i ~case
il <t~~dicl<
rtltrthycct~veo r prorprctivr.
B t . f n r ~cnding this chapter i t is nwrth saying what case study designs are groups Post-test ,
1111k.
'Experemental' Group rs untfom >re on
In terms of the variable
Not thr onr-shut casc 5trd!1
proposed causal /)
vanable
The taxonomy of r e s ~ n r c hd c s i p s outlined by Campbell and Stanley
(lqh3) and hv Cook and Campbell (1979) has led to misunderstandings
of what a case study design can be. Using the template of the classic
experimental design they identify a design which they refer to as the onr-
slrtrt msi3sf~r~-2r!.This is rcprespntrd cliagramat~callyin F~gurc13.3.
As a simple example of this type of d e s i k ~we could consider a group
of children whnsc parents divorced w~thinLhc last year and for whom
wc obtained a measure nf emotional adjurtment. In effect this design
consists nf members o f an 'experimental' group only and relies only nn
'post-in tervention' rnfnrmation.
The earlier discussions of the logic of txperlmcntal, longih~dinaland
cross-sectional designs indicate how this deqign tells us nothing of value
ahout thc impact i ~ fparental divt~rccon the eniotiot~~l adjustmrnt of thc log~crvI7~reby~ n i t i allypotlles~s
l can be s~rbju~tcd
clr research qur<t~nn\ Lo
children. We cannot tcll from this design whcthcr children with divorced empiric~ltcsting. Dccidin~betwccrl single and multiplc casr sturlit.s [st,?
partanthare in any way ~ l ~ s t ~ n r t frum
i u p othrr ct~ildrcn,and whct1lt.r hle1.e latrrl, srlecting s ~ r c r f cnscq
~ c 10 he .;tu~lied,dcvclopin~a case study prutcjml
and dehl-ring r~alcvantdata collection stratrgres 1e.g. the period to b~ covered -
is any tl~njtgr!in emotional adjustment follt~wingdivorce.
rrtrospech've, prospectirc] are all part o f case study d e ~ i v (1993
. 33)
Representing this type of design as an ekamyle of case study design is
unfortunate and provides a rnislcading picture of thc ~ o t e tial n of a well-
Any method of data collection can be used wikhin a case study design so
cot~ceivedcase study design. lndccd Cook and Campbcll (1979) haxfe
long as ~t is practical and cthical. Tndeed, nnp of the distinguishing
acknoruledged this and have renamed thc design as a L I I I E - ~ ~ U I Ip~1st-Ir5t
P
features of thc case study method is that multiple methuds of data
only design, and now stress that it should not bc confused wit11 true case
colltrtion will often be employed (see Chapter 14). Case studies seek to
study design.
build up a ELITI picture of a casc, its subunits (see earlicr discussion on
units of analysis and embedded designs in this chapter) and ib contcxt.
Casc studies have frequently been cqua ted with qualita t ~ v emethods,
e<peclally participant nbscrva tion and unstructured, i n - d ~ th p intcr- Summary
views. This identification is no doubt partly because some of the early,
influential case stt~diesused participant obseriratinn rncthcjds (e.g. Sfnxct Case study deqigns constitute a major deqign for social r w ~ a r c h Thcy
.
G ~ r ) l rSorir.t!l
r and Middlefuua),Sirnilarly, most ethnograpllic casc studies offer a flrxiblr approach, which can result in an extensive variety nf
used in sncial ,~nlt~rupologp employ participant obser~.atior~,and ethno- particular designs. They make usc of a wide variety of data collccticm
graphieq and case stiidies have sornctimrs been lakcn to be the same methods and they are particularlv suited to using a di\*erw range of
tliing.' I n other contexts case studies are cqrlated wlth unstructured in- units of analysis. C a w studies In social science rcscarch should bc
dppth intervlrws that enable the rcscarcher tn bii~ldup a much fuller fundernen tally theoretical.
picture of a rase than is possible with more quantitdtive based methods Case study designs are particularly suited to situations involving n
of data collection. In general there k a tendency to equate 'thick dcscrip- small number of cases with a large number of variables. Thc apprc>ach is
ticln' (Ccertz, 1973) with case studies. appropriate for the invcshgatlon of case5 when i t is necessary to under-
I f WL' equate case s t u d ~ e with
s a particular data cnllcction method we stand parts of FI case withln the conkpxt of thr whole. Casc studies are
misunderstand case study d e s i p . Yin argtles that casr s'tudy dp5z4yn iq dc.~ip,wcRudto qtudy vvholcs rather than parts. They are ;lFst~partic~rlarly
232 CASE STUDY DESIGNS
Note
Methodological issues
Case sti~dicscan be utllizcd for both types of explanation. A case qtudv They stress thc importance of looking at parts w i t h ~ nthe contcxt of
adopts an idiographic approach wIicn a full and contextualized the whole.' Thp casc also must bc seen within the contcxt in which it
understanding of a case is arhicvecl. But case study designs can adopt exists. By examining this context fully the rcscarcher can gain a fuller
R nomothetic approach. Thiq nccurs whcn particular cases are used to
and more rounded picture of the causal processes surrounding a parti-
achicve a more generalized understanding of broader thcwretical cular phenomenon. Yln argues that
propositions (see drscussion of case studies and theory in Chapter 13).
By dcvcloping a full, well-rounded causal account, case studies can
a major rational? (or using [case studies] is whcn your investigation must
achieve high internal validity. By seeing particular causal factors in cover both a particular phmontpnoi; and the roiltcxf within which the phenom-
combination with other causal factors we call assess bath the relative enon is occurring eithcr bccausc (a) the context is hypothesized to cmtajn
importance of particular causes and the way in which various cazlseq important rxplanatory ~nformatiunabout the phenomenon or (b) khe boru~~d-
interrelate. As such, case studies can achievc a sophisticated and aries hehveen phtnomennn nnd contcxt are n n t cle~rlyevident. (1993: 31 )
b a l a n c ~ daccount of causal processes and, in so doing, avoid mistaking
cause with cvrrcIation. The study of context is important because behaviour takes place within a
context and its meaning stems largely from that context. The same
WIIOLES,NUT IUST PARTS behavrour can mean very different things depending on its coi~text.
Furthermore, actions have meanings to pcnple performing those actions
The designs discussed in earlier parts of this book Fwus on 7torinblw and this must form part of our understanding of the causcs and meaning
rather than cusps. They examine how different haits nf cases arc associ- of any behaviour. To simply look a1 bckaviour and give it a meaning
ated with one another (Mitchc.11, 1983: 192). Tvpically, howcver, thesc rather than take the meaning of the actors is to miss o u l on an important
trails are not considcrcd within the contexl of the case cd rvhicll thcy are source of ~tnderstandingof human bcllaviour.
part. By wrenching traits out of the context in which they occur we strip An example can illsistrate this point. 'rl~lc research li tera turc on
them of much of their meaning and ccrnsequcntly risk misreading therr extended families assumes [hat cxchange of help and carc between the
meaning and sipificance and thus misunderstanding their causes. ge~cncrations(e.g, adults and older parcnts) reflects the quality and
Dlumer (1956)uses the term 'variabIe analysis' to describe analysis that strength nf intcrgcncrationrtl ties. However, if we examine the meaning of
f0cusc.s on variables rather than cases. HE argues that although such intergenera tional help within its context we might interpret its meaning
analysis can establish statistical relationships between variabIes they do differently. Within a particular Family thcrc may be a great deal of
not providc much insight into causal processes. Hc argues that: intergenerational help: t h adults
~ in the family are attentive to the elder1y
parents, visiting often and helping nut where they can. Rut is this a sign
Thc i~ldrpendentvariablr i s put a t the beginn~ng part of the proccqs of of family solidarity, closeness and caring, or is it a sign of control,
interpretation and the dependent variable IS put at the terminal part of the dependence, manipulation and blackmail (adults help out of guiIt, con-
prccss. The intervening prncess i s ignored . . . as sometliing that need not be ccrn about hhcritaiice, worry that siblings will be favoured unless they
considered. (IY56: 97) help)? b the help reciprocal? What wcrc the patterns of help in the past?
What is the history of parent-child relatinnships in this family? What
Although this is a somewhat simplistic version of the statistical analysis particular forms of help arc given? What type of help is withheld? Why
uf relationships between variables, it contains some truth. Whilc many of do people give the help? How happiIy is it accepted? What hnppcns if
the deslg~iscovered earlier in this bonk can isolate variables that produce help is not offered? What are the norms regarding family loyalty and help
part~cularoutco~nes,they are not so gi>i>da t telling LIS 71jh!/ they produce within thc community, class and ethnic gmup in which this family 1s
these outcomes. situated? What are the rrrles nf inheritance in this society? What are the
Case sti1ciie3,on the crther hand, en~phasiyean understanding nf the
legal obligations, if any, of children tn carc for elderly parentq? W e would
whole cnse and swing t h r case within its widcr context. Goudc and Hatt need to address thcse and other questinns b e f ~ r cwc could build u p a
(1 952) describe cnse studies as picture of the rneantng uf intergenerational help and the yossihle causcs
of help in a particular casc. To isolate the behaviour from this broader
context and to strip it of the meaning p v e n to it by actiws is to invitc
a w ~ uf v organinng soc~alrlatn - . to preserve the rrnrtnry cknmctrr n f t h r socml misunderstanding, and thus threaten the internal v a l i d ~ qOF the study.
.
olyn-t IY'IJTX S ~ I E ~ I R. !. il is a11 approach whrch vlews a n y socyl unit as n whole
To take a further example: our research goal may bc to gain an undcr-
(1952. 331, their rtalics)
standing of drug addiction. To make sense of adclickion in a particular
236 CASE STUDY DESIGNS ISSUES IN CASE STUDY DESIGN 237
case we would need to understand its social and institutional context drpend, in part, on the type of case, the data collection methods, and the
Was addiction the outcome of medical treatment to contrul pain (addic- nature of the outcome variables.
tion from trcatrnent)? Did it originate wrth emotional distress following A life history of a dead person, which relies on documentary evidence,
relationship breakdown (addiction as escape)? Was drug use a taken for is unlikeIy to suffer From reactive effects. Retrospective designs will be
granted part of the immediate social context in which the person lived less prone to reactive effects than prospective studies. But a study of a
{addiction as conformity)? Did the person comc from a very anti-drug tightly knit group in which the researcher uses participant observation
background (addiction as rebellion)? The context is all-important for and interrrifws will inevitably affect the way in which the group
understanding the phenomenon (addiction) and presumably is crucial in operates.
shaping appropriate ways of managing or treating the addiction, We can try to reduce the threats of reactivity by using unobtrusive
In summary, explanations bastd on case studies involve much more data collection methods (Kellehear, 1993). For some types of cases this
than explaining variation in one variable in terms of variation in another can work well, bl others it may not be possible to use unobtrusive
variabfe. Case sh~dydesigns are devised ti> yield a sensible, plausible methods for c i t h ~ rpractical or ethical rcasons. If data can only be
account of events and in this way achieve internal validity. They achieve ohtained by interviews or 0bsEr~ationit is very difficult to avoid the
explanations by building a full picture of the sequence of events, the reactivity of being a 'foreign object' (Denzin, 1978: 200).
context in which they occur, and the meaning of actions and events a s
interpreted by participants and their meaning as given by a context. In
the end an adequate causal explanation is onc that makes sense. It
invnlv~stelling a plausible, convincing, and logically acceptable story of
how events unfold and how thcy are linked to onc another. While caw studies may achieve excellent internal validity by providing a
profound understanding of a casc, they havc been widely criticized as
lacking external validity. A profound understanding of a casc, it is
HISTORY
AND MATURATrON nrgucd, provides no basis For generalizing in a wider population beyond
The way in which 'l~istory'and maturation can threaten the validity of that case. A case is just that - a case - and cannot be representative of a
causal explanations has been discussed in previous chapters. Exptri- larger univers~of cases.
mental designs seek to deal with the problems of hiqtory and maturation
by using randomized control p u p s and focusing on d$ferenc~s in
change between these groups rather than on absolute chang~(Chapters 4
and 5). If is correct to say that case study designs cannot provide a basis for
Case study designs try to deal with the 'problem of history' and making statistically valid gmcralizatinns beyond that particular case (see
mattrration by looking duscly at the wider context and exploring the Chapter 15 on statistical analysis). A case study may consist of just one
extent to which these sorts of concurrent cvcnts contribute to observed case (one community, one organization, one pcrsnn) and we cannot be
outcomes. That is, rather than diminatinx historical /cnnt~xtual and confident, in any statistical sense, that the case represents a wider class of
maturational factors from the analysis (by contrnlling them out) the case cases. Even if we conduct multiple case studies, say 10, our n remains too
study designs incllide them in order to enhance our understanding. In small for credible statistical gencralizatinn.
this way we can arrive at a fuller and richer understanding while, at the Ijowever, case studies do not strive for this type of external vaIidity.
same time, avniding the threats t h a t history and maturation present There are two types of generalization: statistical and theoretical.
to the internal validity of our case study conclusions. The quality of the Sfufisfic111 genmnliznti~~zis achieved by itsing represents tive random
case stucly will largely rest on how well i t identifies these liisto~ical/ sampleq. Oh the basis of statistical probability we generaIizc our findings
contextual and maturational factors and includes them in any to a wider population that our sample i s designed to represent.
explanation. Theoretical generalization involves generalizing from a study to a
t h ~ ~Rather
y. than asking what a shtdy tells u s about the wider pupu-
lation (statistical gcneralizntion) we ask, 'What does this case tell us
about a specific theory (or theoretical prc>position)?' In Chapter 13 1
Simply doing a case qtudy can produce changes in the case and we can argued that casc study d e s i p s are fundamentally theoretical. They are
confuse the effects of doing the study with the effccts of pther variables. des~gnedto heip develop, refine and test theories. They d o this rising the
Whether or not this will be a problem in a particular case study will logic of repfication.
r"'
238 CASF STUDY DESIGNS ISSUES IN CASE STUDY IIFSIGh 239
de%cribe tlie normal methud o f selezt~ngc a w s in case sttrdv dtbs~gns. knn~ctvnrkwith n~hicl.rw r Art, ~ v n r k i n For ~ . example, i f qtwial cia+\ $\.a<
Flakim 11 9x6) uses thr tcwn frrrlrsrd .ca~rrydili,pto dcscribc thc snme prooeG3. ;Irr in>pr,rtnnt concept in tllr shrrlv ~t r v w l d hc neccqsarv to ~ n c l t ~ ct~sta.; de
T h c ~ cterms r ~ f c rtc) tlic strattyv uf scltacting caw? that wi31 prciviilc frnm vach cyf the drffrwnt s o c ~ a lclasses. Tlie number of cnscXs would
illuminating examples of a tr-pe of caw (x In de~rriphvrcasc qtllciy dfpi-nd tvi hntv .;wial clnsc wras concephiali~ecfand (HI ]low many cln\<
d r s r ~ n s n) r that will provide nppsnpriate t c ~ t snf ,t t h r o y . cat~goricsthi:, concrp~tralizntion included ( c . ~ .jlrst rniddlc~L - I J . ; ~ and
working clasc, or a more r ~ f i n o dclaw c l , ~ s ~ i f ~ c n tthat i ( ~ ntnclude< upper
c l a w and \.arton.; cla.;~iSicntinnq rvitlirn thc middlc ant1 working cla<ws).
Wr co~ililgo on fortwt-r and chrck t h ~ n g smat under anrr nlrmbcr ot
T h e i > no ctrrrcct rrl~rnbcrnf caw< to i n i l ~ r d ein a c.l\cl .;t~td\-di*'.iqi. rlifi~rentcontliticln* in the end we mu<! m a k e ~ l ~ d ~ c r nat)ouf t ~ n tI~~ k r l v
C,IGL~ ktudt rlt..;isn< can conqi~lof a sinelc E ~ P v. ~r mt11t1yIr~
~ c w t - (Uh;lytcr
~ \ a r i a t i c ~ n;rnd
~ hallc good rt.a<rm\ for t.atx,cttnq [lint thrsca difft.rt~n~
121. I4'ith r n ~ t l h y lcaw ~ ~ *tlldic< tl~cnlimbcr nf is a nlntt1.r r ~ f ci~ndrhc~ns h t the p ~ t t c r n s W
r n ~ ~ *3ffect . r nlstl nimt I7e gi~rdedb) n halt ih
judgcm~nt()'in, I t r R q ~.X). IZqisnlfrcant factor in d r t c r m ~ n ~ nthc*
) : number prnctlcnl w ~ ~ t ht ~hnc-rinrtraintq
~ of time, mnncv and .rccr.;s tin relevant
r p t co'itn.;~ ~ hc l h-lel r i ~ o r l rwith whrch t h e prtlpositions a s r t r l 1-c Icqttld. C J ~ L ~ ' ; . l'hc critic-al thin: ic t t ) 5trlrct the mo<t ~ t r a t e ~ c-;lscs
ic trr twt 4 ~ 1 r
Ll<inr: the Ior;ic o f repllcation n singlc rcp2icatrun tctl5 r l s ~c-rrnttIi!t~g but proprrhrhon\ rathcs than aiming ftir a l a r ~ cnumbtar of r a w < .
n ~ p m t t dr c p 1 i ~ d t ~ ogi\ n ~c' marc corltidrlncc* in trnd ingq l i w f r seck
tlict~rctic~il replication5 wrbr v i l l npetl to cnndl~ctaddf tinnrtl caqe s t u r l i r ~
Whcrtl lvr fii~dcase* tIi:lt not fit with nur tbxprct.ltions thi\ rn,I!.
p~crri\~cr thra r~rri!to rr~nductaddit~cln~ll c;lsta sltt~riib'; to enable c r q trl
r,,tl..... 1..... .,L-- . . I . ...I 1 S T -. . c
govt.min1: sd~rrcrT\I l . 1 ~bc'ttcr ~ t.ci~icntlnrlal ou tcnm~.;w e nerd to knnw
<chlcl-rschur>l<h a v e ~\*hicllsytcrn. IYtb r n i ~ h t2150 need to kn(~r%d, for
~nfc~rrnnticm. I.
G a i n i n ~rclia1,lc informat~rm a p r r r c q ~ ~ i + i tfnr
rrsc;lrch. Thi* mc;ln5 that tlic infurn~ntlonnjuct nvt s i r n ~ l vbe .In artifact
e ~ood
In mv view this is not thc task of thc social scientist. rt might be thc the face o f cnnsidcrable social incyuality, does the communily maintain a
role of a biographer, a novclist or an activist. The role of a social scientist strong sense of cohesion?' These questions guided the study a n d pro-
is tc> develc~pand evaluate theoretical generalizations that enable us to vided the framework around which the study was reported.
understand whole dasses of cases - nut, in the final analysis, individual Regardless of which approach is adopted it is highly desirable that
cases. the logic of the research design remains clear. The reader shcnlld bc
We do not have to have a welL-fo~mulatcdtheory to tcst. Our questions clear about the prnposition(s), and why a particular case was selected
may be such that there a r e no obvious theories to test. l3ut this dws not to trst the proposition. Tht reader should be able to see that there is a
release u s horn the nrcd to be well read and well prepxed theoretically structure and a logic to the selection of cases - that they have bcen
beforc going into the field. Gasteur observed that, 'Where observation is sdected for a purpose rather than simply because they happen to be
concerned chance only favours the preparcd mind' (quoted in Mitchell, available.
1983: 203) If wc are unaware of rclcvani theories, concepts, dcbates ancl Any report that merely provides a number r)f stories is incomplete. In
thc Iikc, we will probably miss the significance of much of what we reporting ctlse studies it is critical [hat wr relate tl~ecases to propositii>ns
might come across. Without having some idea of what we are looking for and that we seek to draw comparisons between cascs and arrive at
we will not know what we have found. generalizations or more refincd yropoqitions as a requit of the caw
studies. W c need to tell the story of the set of cases - the generalizations,
thc propositions and the questions they answer and the propositit>ns we
end up supporting.
Case s t ~ l d i ecan
~ involve the collection of a vast amount of information.
This ~nfrsrmatic~n must be carefully prcrccs5ed and distilled bcfnre it can Ethical issues
he yrescntcd. W e cannot simply describe the casc and somehow let the
facts speak for d~ernsclvcs.Any attempt ti) prcsent all the facts will result 'Thr s a m e ethical principles apply Lo case study designs as to the designs
in an indigestible mess that is unreadable and unread. discusscd in previous chapters. The particular ethical iqsues confronted
The facts do not speak for themselves and we should not pretend will vary, depending 011 the typc of case study design adopted and t l ~ r
othenvise. Describing a case always involves selection and ordering and particular form of data collection used for that design.
at least implied construction of causal sequences and interpretations. For cxarnple, the ethical issues w ~ l lbc cliffereit if the casc study
Therc are many different ways in which case studies might be pre- involves an active intervention or a passive intervention, or requires that
sented. We might describe each case as a whole s n that the reader builds W E obtain a retrospective account of an event, intervention ur change.
up a clear picture nf the casc. Alternatively, we might describe good The ethical issues will also differ according ta whether data are cnllccted
examplcs of particular types of cases and then cornpatc thcsc different by interviewing individuals, using informants, analysing official records,
types of cdses and draw general conclusions regarding our theories. passive nb3cnration, conducting a survey of casc elements, or using
Hochschild's The Second Shiff (1989) prnvides a good example of this type participant observation. Since case studies frequently employ a range of
of reporting. She describes a series of marriages and from each of her diffcrcnt data collection techniqucs fur the one study it is likely that a
cases she extracts ~ C S S O ~about
S the domestic division of labour and the greater range of ethical issues will arise when using a case study design
procrsses by wluch marital inequaIity in this sphere is maintained. than with other designs
Alternatively, we might repnrt the results of a series of case studies by TIicre is no point In repeating the ~tlncalmatters identified earlier in
extracting ilrern~ffrom the case studies and reporting the findings of the this book. You sliould read thesc earlier sections carefully and you
case studies thema ticall y. For example, Vnughn (IY84) cnnd~~cted rrtru- should see how they apply to the specifics of the part~crrlarcase study
spectivc case studies of coupIes w hoqe Intimate relat~onshipshad cnded. Jesim you adopt. H o w e v ~ rsince, case study designs frequent1y involve
From thcse casc studies she Identified a set of underlyng patterns and ~k--trliningmore in-dcpth information and Crcqllcntly employ participant
stages beneath cach disinfegrahng relationship. Another approach is to nbscrvation techniques, some ethical issues arise that are less of an ~ s s u e
focus on a series of qlrfstionl; and rcport the cases in terms of how they with other designs.
shed light nn thesc questions. For example, Dempsey (1990) conducted a I argued earIier that reactivity can be a significant thrcat to the intcmal
single case shldy of a small country town The guiding questions in the validity nf casc s t ~ l d ydesip;ns: thc pwsence uf a r~searchcrcan alter thc
case study were questions such as, 'I-low unequal arq people In this dynamics of the cascs being nbserved. This can be partlcuIarl y so where
comrn~tnity?','What are the key bases of social clifferentiation?', 'How, in participant observation techniques arc used. One way in which some
I
I'
246 CASF STU13Y DFSlGNS ISSUES IU CASF STUUY I3ES1GN 247
il
I
1;
schnnl and you learn of the identity of a person sell~ngdangerous drugs
~ d e n t ~ tays rcscnrchers to avoid their r r s ~ a r c l ~ cidentity
r intruding. For to students? What do you d o if you discover, during your rcscarch, that a
example, an rnvestigator may want to examine sexual harassment in the group participant is about to commit a csirnc? Is tlre betrayal nf trust of
workplace. If he went into a companv revealing his idcntity and purpose thnse around you more unethical than one's responsibility to those in
i
!i
it is likely that people in the nrganization would behave differently while danger?
1; the investigator is present and thus undermine the study. As an alterna- What is the responsibility of the researcher to othcr rcscnrchcrs? If one
I tive the investigator might undertake the research by obtaining a job
with the company and covertly observing what goes on.
rcsearchcr betrays the trust of a community, even if for ethical reasons,
does this thm undermine the capacity of other researchers to study in
I The ethical issues of deception, and failure to obtain informed conscnt, that community? Once a group has learned tcr distrust social researchers
1 i
I
obviously a r i ~ in e this situation. The dilemma it higl~lightsis the tradeoff
that occurs between ethics and internal validity: keeping to cthical
guidelines can threaten the internaI valldity of the study.
There can be little doubt that revealing one's identity and gaining
informed consent can undcrmine a great deal of what Punch (1994) calls
it can be very difficult to conrfuct any further research with that group,
and this may havc further ethical consequences. The behaviour of some
researchers w ~ t hindigenous groups has led to profound mistrust by
some indigcnaus co&unitic.s of outside researchers.
How do you resolve the nbjcctivt. of feminist research to promote
II
'strcct style' ethnography. rn some cases revealing one's identity can identification, trust, empathy and non-exploitative rcla tionships between
I ! iI
I
ticipation take priority over tlie value of research exposing I~ighly
unethical behaviour?
Even if a researcher revealed his/her identity and gained permission
This chapter has emphdsized that thc intcrnaI validity of case studies
relies on a Ingic that sets this design apart from the other designs
'I to undertake the study, the question arises of how to get informed discussed in this book. Case studies focus on a large number of variables
I
II consent, and From whom to obtain that consent. If we were conducting a and thc way in which they intcrrclate. In contrast, the other designs rely
study of sexual harassment in a workplace, would we obtain pcmisslc~n on cnntroIling out the influence of other variables and on focusing on the
I
from the gatekeepers in the organization, or from all people whom we influence uf a very limited number of variables. One of the strcngkhs of
I might observe in the coursc n i thc study? case studics is that they atternplt to understand the significance of
Do people need to he informed that their public behaviour is being particular factors within the context of the whole case rather than by
observed for the purpose of research? Clearly m many cases this is screening out this context. As such case studies havc the pc~tcntialfor
simply impractical (u.8. observing the behaviour of people at a party, in a good internal validity based on a more thorough understanding of the
dernonstratron, shopping at sales). In my view abservtng public behav- meaning of partic~~lar behaviour and events than the other designs
I I
iour raises few cthical problems. So long a s individuals arc not identi-
fiable there should be n o danger of harm to participants. After all, if the
prclvidc.
The external validity of case studirs is based on the logic of replication
:I
li
I
I
bchaviour is public there is l-rardlv anv invasion of privacy.
Anothcr ethical problcm that is more likelv to arise whcn using nbser-
vat~clnin a case sh~rlyis the problem of what to d o when we observe
illegal or i2arrnful bchavirrur. Should the cthical isstres of informed
rather than on sampling logic. Case qtudics are used to generalize to a
theory rather than to a populahon. Cases in casc study designs are
selected strategically rather than statisticaIly and a r c selected as critical
tests of theoretical propositions. Iswes in selecting cases, the seTectinn of
1 consent and voluntary participation take priority over exposing highly research methods to collect information about the case, and condtlcting
I
I
I ,
unethical h ~ h a v i n u r ?For cxample, what wuuld be your ethical respon- case studies were discussed. Finally, particrilar ethical questions that can
sibility if your case studies of families revealed cascs of dpmestic v~olence arise in casc studies, particubsly Lhosr employing participant obscsva-
or c h ~ l dabuse? What should y c d~o when conducting a case study of a tion and ethnogapliic methods, wcrc raised.
-- -- A
t Thr. crlriccpt nf n wholt- IS, .is C;c?c>rlr dlld tlntt (1Y51) point oul, n ctmqtrflct
1tqt.lt It 1.; r l q I! Iin <It-hntbqi\har the '~r,lu>le I.' 7'11eill-lincat~rmI,~*twt>rn t h ia5t1
~
It h ~wlx7Te)
. and I!< context 1%. intlwti arl~~tr,ir>.
2 'Tvyical' casr.: qhr~illdnot bt, ct~nlu.;rclwith selt*cting '~rlcalhpical' c a w < - CASE STUDY ANALYSTS
C.-X(-: th.it mn! n,prr<ent ail idral tvpe (Chapter 13)
Statistical analvsis
Independent variable
Independent variable
I
Pattern 1 Pattern 2 Pattern 3
Dependent variable
Dependent vanable Y, Pattem 4 Pattern 5 Pattern 6
Yh
SinipEi, pntfer~ls At its simplest level pattern matching involves one Thtvry A !,oral control will lead to higher commitment to work
independent variable with huo values (e.g. male and female) and one bccausc effort and 'fit' is swn and rewarded (and lack of effort and
dependent variable with two possible values (bchnvus in one of two not fitting in with school needs is punished).
ways in a particular situation). In this case there are four possible difier- a Theory R 1,ocal co~itrolmakes people feel more demoralized, and
cnt patterns (Figure 15.1). For a givcn case with a given d~aracteristicwe vr~lncrahlcto local politics and prejudices, and dues not recognize
could predict nnt of hvo patterns. Which pattern we predicted would wider professional development etc. This leads to a lack of commit-
depend on theoretical considcrations. ment and a lack of professinnalism and to plaving politics to win
The pattern wc might predict for a particular case that we know has Favour rather than fostering performance.
value X, on the indepcnd~ntvariable is that the case will have value Yb
on [he dependent (outcome) variable. The prediction takes the form: if X, A slightly more complex version of ~ssentiallythe same situation would
then Yb. arise if either of the variables had more categories fe.g, medium leveIs of
For example, our theory of work perfomancc in schools might lead us performance and a hybrid of local and centralized staffing modeIs). The
to expect that when staffing appointments, dismissals and promotions more values either variable or construct possessed, the greater the
* are managed at the local school level rather than by a highly bureau-
number of possible patterns we could predict. For example, if both
cratic, centraliled and remotr system (X variable), staff teaching com- variabIcs had three possible values there are nine possible patterns we
mitment ( Y ) will be high. If devolved staffing systcrns are symbolized as could find in any particular case (Figure 15.2). Tn this situation we would
X, and high level of teacher commitment is symbolized as Yb, nur still predict a particular pattern for a givcn cast.: if they were X,we might
prediction is: predict Y.,
Fnr example, we might have a theory about the impact of parenting
When X , (local based staffing system) exi4ts then Yh [high levels of tenchcr styles on child behaviour. Each of these two factc~rsmight havc thrcc
cornm~tmcnt)w1l1 finllllow. c a t e p i c s a s illushatcd in Figure 15.3. Using this example we might
predict that authoritarian families will produce children who are
We would also cxpcct that when Xb (centralized system) exists then Y, rebellious (pattern 1) A case study of an authoritarian family will help
(low levels of teacher commitment) will follow. address whether this proposition receives support. Repeated case studies
Our test, even in this very simple examply of pattern matching. would of autl~oritarianfamihes wilI provide a firmer test. A further proposition
be more interesting and more powerful if we were tTting alternative map be that permissive families will produce children who constantly
theories that predicted different patterns. One theory might predict engage in limit testing brhaviour (pattern 9).
Parenfal style of aulhoritv
---
Panem 3
Low 1
I
Hfqh , 101 1 Hlqh 1
c~wesltld ius.
19 70 7 , 24
~ r fpattrrn.. can hc prcrlictc~c!if h.rmoindcpcndent rari~blcqa r p i~scd.a
diftcren t outcom t1 is prcd~cted for each a ~ ~ ~ ~ b r ~ r of ~ r ~indgendcnt
rirn
v,~riablrs(Fiprc 15.4).
Here wc might make different prediction<, depending on the corn-
bination of parcntat authc~rityspir combined with the child's leve! (if
anxiety. We might predict a diffr~rentoutcomc pattcrn of pcrmisqivc
parenting d c p c n d i n ~on whether we a r e looking at n farnil11 where thr
child IS ,inxinu< rrr s~?ct~rr/cunfident. We rniglit predict that t h anxiouq
~
child w ~ t hpermissive parents will exhibit the ' t ~ s t sparental l i r n ~ t r ' is not present we will nnt get n spcciflr outcome (see I.itt*r for ft~Ilcr
responw {patt.cn 24) while the confident /r;ccure chiFil wi tll pcrmissivc discuwion).
parents may set her own limits (pattern 17). Another approach to cnllecting and anaIvsing case studv data i i s to
We cnutd extpnd this example tu include more indryendent variabEcs lonk for cases with particular outcomes (c.g. Itiglllv corny l lank children)
and thus anticipate even more pl~sriblepnttcrns. and predict !hat thesp children wilI be anxinus and hnvc authoriiarian
Figurr 15.4 ill Listrates ihtt logit-nl possibi litie- - thr different patterns parent<. In term.; of the prediction5 in F i ~ l ~ 1r5.4 c ruc have predictctl this,
one cclttld conce~vabIyfind. Of course, our thmretkat rcasnniny: will I ~ v d but t f we find case<of f i ~ ~ hcompriant ly childr~nwho are n r ~ t h e anxious
r
u s to prtdict pnrtrcrrlar pattwns i n particul,ar circurn~tnnces,!hat is, for nor h ~ v au e thnri tarian parents thcn o ~rt t htwry 11rt.d~ refining-
particul.7r h-pcs of casw ;I< illustralrrl in tl~cs h a d ~ dcclls in Figi~rc15.4. 0 1 1 ~ .way of tc?tin~, a theoretical propns~tionis. t i i demnn\tmte. t h r o u ~ h
Harjnt: cstabli<hcd tht* pmsitrlt*pattern* ,lnJ the. prediitl-t! pattern+ c a w %tudic<,t f r d t ~vhcncertain crmditlclns apply (e.g. at~tl~ori!;lr~an
nTcurould then wck 011 t C ~ S that ~ S rnahl~ u s to test the pmprw~tions.For pmrcnting and child anxiety) tllvrc will J I bc~ ;l sprcific c~~rtcr~rnr (h~t;hly
taliarnplt., we wrluld conrtz~cta c.,isc stud\, cm a familv in wliiclt authcwi- ctlmplinnt child). The prohleni with this w r t of r n , ~ t c h i ni q~that i l only
trrrian parentint: and child anuiec occur. O u r p r e d ~ ~ + t ~I\-ot~ld
r , n be that ~ 1111 cwtcorne. It dtw\ not It-I! U G ~ v l l t ~ f h ~ ~
sptu-ifies sttfhc lr'tlf c o r l ~ f r l t ofor
the ch~li4~roulclhc v c y ctmlpl~ant.if khis proved t o be h ~ c w . e would the<c ;ire the ~)!tl~/conditinns under which the outcome occllr5. As hitch i t
hnve achieved a Iltrrnl rr,j~l~cntinrr.We m i ~ l i tthen Iwk for a casr of does ilot rcprta.;cnta rlrmandln); t ~ q ol t n propositfnn. A nmorc. dpnl<~t1rl-
authnritasian parrmntln:: ~ ' l t ha cliilil I \ not an~iou5.Our predrcfitm ing tc+t tvnulrl involve Innkinq Inr caws !rehere tlii* particular outcrvnt3 is;
here mij:lit be that the chitd twirl not bc c n m p l i a n t ~This would hc a prrwnt and tlien a5king whrthrr thiq cltitcomc rlirl!, mcilrs unrft*r the
tlrc*orr'irti~lrrgplirr~lrnrr.TIl.lt is, whcrv tl-re prticular combinatinn of factnss khcorc*tically prcdictcd ci~nditions.
IndependPnT vnnables Independen! vanahles
--
---r 11
I
Condlt~on1 Condition 2
Work
l qual~ty
Staff
firrr EVP prcd~cta pattrrn of outcomtXsin c a w < that meet condition 1 0rg~ni;tatirlnz u-ith high dc~rr*csof supewl+lr,n and direction arc
ancl n cliffwent pnttcrn tnr cases that mcrt condition 2. Thc pnttcm is a p ~ e ICICCJ
d to bchave differrnttv. fhcy are prcdic?r=dto have a drffcrrnt < t a t
p~rticitlarrornI~i?rnfifl)lr b l ~>l~l.comes. In t l i i q c a w wc prcdict t h ~ tfor cases of outcolnr chnrrlct~ristics:Pow morale + low qualrty of work +high staft
rnr.rat~tigcondition 1 (on the independrnt vrrriabld) we will f ~ n d the turnover + moderate Ievelc nC cfficitncy + low Fcvcnls of ~nnovatiun.
1561) CASE STUl)\r 1 )I!SIC;N';
7rt.rtrl nrmlrr~rs 7'rrnR analysis i< a n cuaniin~tinn of thc directinn nf Cl;rtvmlii~yicnl nrtnli/rl< Chrnnolo~ical ,analv\i< involves prt=d~t-tinl: n
c!i,lnpc in partirufar 1-ariable or w t nf \.,lrlablc.;. Mre addrr- I ~ I rllic-q-
?
L * of e ~ ~ c n tI qor rl-en a sequencc nt trt'ndq) in\.ol\ins a n?irn1.rzr
.;r7r!~trvi~t-~7
trrvn ol rvhctlwr tlic Irrsnrl i< i~pr\,ard(htt,cp tIr gr,-1ilu.71),shot\'* r l r ) ch,tncc. (if dlitcrcnt evtant.; or i,,iriL~blcs That is, \\T w c ~ l r lprcdict ~ r . h , ~t \ tr~i!lrf
variable ( u p ant? clo\+,n) or drrwnwar~l(htc~bptrr graduall. clinngc, nt what t-vr-litc \teould take plactl, in w h ~ turclttr, 111~1
Prcd icted trc*nrl.;c;lii rangy f r u n ~tlic srrrrplt, t c ~the higlily t - o t ! r / l f r . ~ .TFir .;txqllcancenf evcwt* might bc a corrst7nrrrl rbfl'fTc.r srqucnct. nr it rnrght Ilc ,I
simplest fnrnlr p n d i c l a trend in tlnc ili*pc~nJrritvari;lblt~.Wta ~ n ~ i : thrn l~l J c s c r i p t i r ~S~P L ~ ~ I I V ~ C -tl1,lt
L' proposes partic~rlnv<tr7"yrS~in a prorr%+
prcd ict a d it'ferviit par lcrn of change t ) i i t311s\,a la tllc in dificrcnl crlnttlrtq. l'xnrnplcs of n s t . ~ ~ vcvioii r~l of chrr)nt~loj:lrill;111,11~si<
might 111, rm~t1~,lc
Ior e ~ a m p l c ,w r nrkli-anticipate that cnrt~lmrr)t< in a partluttlar t v p , c j f llrat prcrpmc predirtnhlc Gtazeq in b t ~ o n i ~ r l ap , marijuana w r r t l < t - ~ - L c ~ r ,
tli111-erqih.cotlr+LaIthta c a w ) ~ v i l l irwrcn.;c
. In ~lirchnrF tern hut d+-c-l~rlc in ICkhhk.~ t d q e =~n t l l t a rlrvntpgsatinn vf int !mato rc~latiniwhiy.: (\'auqhn.
f h c Imger t t = m ~rlnc,tht*r course m ~ g h t ~ r ~ l d ~ cto t umarntaln
d ~tabtc I W h ) , stages in thv procrw ot adiustmcnt to rt-tlrcrn~nt(Xtchltav. TLr7rlh,cbr
en rdrnent !cvt4c ~ r ~ l ~,~tir~thcr
ilc tteprc ) C r turr+r i c prcdictkd t o txtyrar!tlrlcc chnngt+ in the rcl=ltton.;hip~ beh\.een ,~rlultsavJ thcir pnrcntk .I\ pEirtant+
'J-l.irp Incrcaqeq in [.ti rt~/nientIc\.cls. ilKe l Massdcn ant1 1 2 t ~ a m <1947). ,
2hZ CAST: STUDY IIPSIGIGS
R r g ~ r dlcb~s ot whether thC analvsi.; involvec a causc and effcct nnt find case% ~vht'teMT 1?,1\.~'thc c f t t ~ t swifhclt~tour prt>\t~rned
aeqi~wicvof evenis, or a set c l t predictcd stages, the analytic q u ~ s t i o ni.; cause."If we ,foil tn find any huch caws we havc achieved lurther
wr!l~fh~.r t l ~ cpredictcd p n l l ~ ~ l 'rlfclra71,iy
h occurs in the casts w h ~ r etheory thcnret~cslrepl~cntic~n.
rvt3uld rxpcct it to occur. Or does t h c pattern nccur rvht~nw e would
predict it .;hould ttrrr c ~ m r ?
T l i e Corms of the* predicted patterns can vary. Yin (1989) indicates four Analysis for theory building: analytic induction
tvpt's of w;iy5 in wli~c-lrevent< might br predicted ttl change in r ~ l a t i o nin
~ a c hntl~rr.We can predict chmnolo~ie.; in which: At ib heart, analytic induction ic 'a s t r a k c ~ vof analvsis that dircrts the
~ apply to all instances ot
inrtrs tigator to iorrnulatc g e n ~ r a l i x a t i o nthat
.ir>mc must always occur ht-fore othrr cv~nts,with tlic ntvrr<cb tl1c1 prckblc~n' (Denzin, 1978: 191). It is a rncthod that can bc 11st.d to
+,rf~/r,r~~t' bt.~ncimpmsibl~.. rirliicve d c ~ c r i p h v rgmcrali7ation.; o r tn arrl1.e at ca~rcal explanatinn5. I t
wmtx event? nitlst ahvnvx h t followrd bv other r\ ents, on n m n t i r r ~ r . ~ ~ ~ t / ;.E n q t r a t c p that move< from i n d ~ v ~ d ucaal w s and ~ c h tos identifv what
t,n51.,
tlic cases havc in common. Thc common talerncnt prclvidcs tht- basis ot
v somc events can only follow other evpntp after n spccifled pt.;qn,~t+nf timt*;nr
tli~oretica2~eneraliz~ition (see Iyiguru 1.2).
rrrtain f i n i l * / r 7 r l m lIn a cnsP study mav be marked I-y clawe.; t > f events that
diftrr ~ u b = , t a n t ~ a l from
lr t h w e of u t h r r time prnani.. [stages]. ( 1 999: 17'3)
Analytic inductinn Y G a stTafPg~that s r ~ k . :In arrive a t genrral~zationq
that appZv tn d l ca<i3s,bn t h i ~rr.+pect it tI4ffers from the s t r a t r g i e esf
nri.~lysisdisc~~ssed In previous wctions of this hook. In Chapter 3 1
dislinfiii<htd hetwecn deterministic and prnbahilistic notion5 of causn-
I n Chaptt*r 13 1 n r g u ~ dthar gi>neraIizin): from caw ~tudie*;rrdit.s on t h r t~rm.rJw I T I I > ~ V Gof ilrldl\-if: ~ I ~ S C L in S ~ ' ~ wctions a w bawd on
I Cprcviouq
repiication logic 01 cxperimrnts rather than the .;tatistical 317gic of sur- p n ~ b a b i l ~ ~causat~c,n:
tii r4.e cst~rnatc1~1irtht.rone group is mclrc Zikely
veys. M'r gain mntidcncr in e ~ p ~ r l r n v n t arcsu
l It? not ji~st from thc tll.in 0 t h ~~T ~O Y P S to I ~ t h d v cin a part~culnrway. Analytic ind~~ction,
elegai~cilof the cxpcriment but from our capacity to predictably replica tc hnwcuer, st.ckq to achirvc univprsal gencr;lli;lations.
! replicate results (i.c. mr anticipate that
r ~ s u l band to pwtlictably f ~ i to Denzin ( l Q i R : I i ) 3 sunimnri7cs c i kc!~ ~tcpqin thr process of analytic
the intcmcnhon wll ha1.e its effect ttnder spcci fic conditirms hut not indirction:
under other conditions).
Similarly we gain confidpncc in case study findings wlicn we can 1 Specify what it is you are s c r k i n ~to explain (the d ~ p e n d c n variable).
t
accuratclv predict which k p c s of caw.; rvill d i ~ p l a vparticular patterns 2 Formulatr an initial and prc~\,isionalpc>r;sible c\planatinn of thc
and ~vhictrcasee will ~ i o fd~\plays ~ ~ f fpatternq.
ic phentln~rnnnvou are seekin< to explain (your thcory).
Where. we have a predfcterl .wt of outctxnes and a singIe causal factor a 3 Conduct a shtdy of a caw scnl~lctedto tcst your thi~ory.
similar logic holds. We would ask the following <or% of qu~stions: 4 Review (and revisv i f necessary) your pro~isionilltheory in the Iight
of the caw o r exclude thc casc as tnap~mpriatr.
1 DOCSthe full set of outcomc charactrristics occur when thr prcsurneJ 5 Conduct further case studiw 10 teqt the (seviwd) propositinn and
c a u w 1 factor is present? IT so we have confirmation of our theory. reformulate the proposition as requirrcl,
7 Mlc wc~uldthen find another case where the prr.;urned causal factor i~ h Continuc with caw studies ( i n c l u d i n ~l m k i n ~for cases thal might
PPPSIVI~ and stc whether the full wt of outcomtSs is also prt-sent in that disprnvc the proposition) nnd revl\t. the prnposition until you
cart1. Zi so rvc h a v e a litcral replic~tionof thc previntrs case a n d ~chievcaa causal proposition that accounts for all the casrs.
furtht-r confirma tion uf o u r theory.
3 D{JwrJ get caw< whew the presumcrf causal factor is prcscnt but nnlv 'nie prncr,w might bc illuatr,ltt.*d uqinc the hvp.~"th~tical eiampIe in
w r r r . of the pr~bdictedcrutcime characteristic+ are preq~nt7t f we F~nd Cftr7ptt'r 13 about tlita impact (rn the qtralin' of c d ~ ~ c a t i nr)!n d ~ \ ~ c > l v c d
surli I-ases t h r n rc-e hn1.c Ialled to rraplicate thr throry and we would \.rr.;ur; ccntralizcd sclionl st,^ tfin): sy5tcrn5.
c11hcr rqect or modify tlic theory. I F wr could find no c ~ s c swhere thc In this c*xample tlic quality ot ~ d u c n t i o nis the pl~mnrncnonwe arc
c a u $*.a<~ prcwnt and the full qet of nutcornrs was rrrrt present tlicn wrking tcl csplai~i(.;try 1). O~rrprnvisinnal (partial) cxpfanatinn iq t h a t
I\.(, hazre a t hccrrchcal rt>rlication. ~Fr~,nI\-cd .;tnffing s\..;tems 1 ~ 1 1 1~7rnducvgreater i r n p r o w ~ ~inn thr ~~
4 1Z'r rvould thvn seek to dtnd a caw in which the presumed cai~\al ~ U J I iry o f t.rIlrcaZ-ton in ,I schoo[ Ihan will ~~t~r-rt~~lizcrl
.;v~tern$(%trap 2 ) . 1%'~
factor ES !lo! prtxscnt. Wi* would rxpcct that thc fullI set of outcc~mrs would thprcforc anticipate that cases [st-hnols) w i tli devolvtd +ystcms
~ i l u l dnot 01-stir whet1 the, catlsu M3,r.. not prc-rcnt T h a t i.;, we shn~rld ~vouFclhn1.1. bcttrr ~ ~ u ; leducaticmli~ tllnn thore W Ith ccnSrali7rif svstcm.:
( ( ~ t h c rt h ~ n ~k si n g e q ~ l a l ) ./\ltcrn,~hvt~Fv,we might anticipntc that of thc ~ i l i c m e thc process bv which i t \\lac tn c > p r r a t ~M'c
~ wo~tlrithcn
schools th;lt intrcduced a dc\,ol\,cd st,~ffrngsystr-rn wcluld exhibit an Ilr in a position (step 6 ) tcr mollify o u r initinl, pro\,~sionalprt>powtion ti)
lmprcwemcnt In thr qlralrty c ~ feducatitm and thnt this improvcmcnr read ~ o m c t h i n xI ~ k r :'Whtn ~mplcrnenfcdin a clirnatil of con.;ul tatit~n
M*LIUICI b~ greater than in schools that did not intmducr such a ~ s t e m . and ctmsrnsus, de\.nlr,ed staffing svqtcmc wrll prtxlirce imprcrr.ernr*ntqin
WP w o i ~Jl thtln sclcct CI caw to tcst our prol.713~1tinn(~tt7p3), W r might cducntion;ll quality.'
find a schcml that ha< rect~ntl!. ~ntrodlsceda Im-alfy hilsed staffing .;vstenl. Mnr\+evr,r, our ncrt caw CII~~SCS problems. Ills. have Iwated ~ n o t h c r
Having d c v c l o p ~ da detinition of what c o n s t i t u t ~ seducational quality casc ~ ~ h c tht. r c l n c ~ l l ybaked staffing s y t e r n also h a s brcn a d~s;~.;tcr.In
and worked ntEt horv tn measure thiq, we could then conduct a caqr thi5 case, t-etlwcver, thr 5yskern w a s irnplrmentcd only after cascf 111
rtud v to src if the introd i~ctionof the ncw staffing systrm had led to tl~r const~ltaiiunnrld w ~ t hthr full agrrt8rncnt of ~ t a f f,lnd ollicr.;. mi.; is not
prcdtcted Irnprtm.crnentk. Let tts auppi).~ that thc prrndictclri improvc- \chat ~ v expcitcd.
c ll'hy tines !I115 caw not fit' rtlrther r u m Ination ot the
nients had rn fact tnkcn place. This lends support to our pwpositicm hut case rt.r.c*,ilsthat t l ~ t~chr,nl
* hw had a nrbwprincipal appointed sincc l lie
the support is t~ardlvovc*nuhclming and probably wonld no1 convince a .;ystrarn IVJS ~ntrmiuced.This principal i< not t n ~ s t e da n ~ il< bcl~cvetito
.;ceptic (stcp 4) Thc sceptic ct~uldsav that one case hardlv proIVrsthe plav faunuritr.; H r is a l w bel~errcdto victimize particular popular staff.
point, that there rndy have h r ~ n a gencra! Irnprtlvclm~.ntin c*ducationrrl Mavbe thcse aw the rcnsons whv the systrrn d t ~ lnot work. Chrcking
qualitv in all --chonl?: over thc sarnt. prriod, and that qchonls that had back nn t*arlirr cascs, the succew stories all had princip.lt.; rvhn wrbrc
kept the old system m i ~ h thave also exhitlitcd an improvement in widcly liked and trnstc~i.
c d u c a t i ~ n qzlalih*.
~~l At tl~c*vcrv Pen\t thcy m r ~ h .;,iv t that thrv c~>uld hnd ntr thp two failurcs h a v e anvthing in c n r n r n ~ n ?Althr~rlghthe arc
a schnol with a ct*ntr,ll svrtcm Illat had t m p n ~ v c d thcir qu,lIity of dilfrrent in ccrtain ways, what they h ~ v in e common i s th,ll teacl~er\1117
cld~iiahonOVPT the <amc perltld. They might a k t , itg gut^ th;lt t h c ~c't7~tltl ~ not IIJVC L ~ ntv. ( i t i t j r f ~ ~f~rc t~fit,+
+ s!t,td7i~;. 111 oiit, C J ~ I *~t t v ~ q
f ~ m - ~ t>n
c l tlitvn
find cxan~plcsof schnnl+ whcre the lot-al .;v<tem ruaq introducrd w ~ t h withoul ~.onsuttatiim.In thc uthcr, t h ~ .prrncipal has trndi~rmini>dtl11.1r
d isnqtrnus conqequcnces for cducarional qua lie. ccmi~denccin the t'airntls o f t h system. ~ IVe might tE7c.n marl I @ rwr
Wc wnuld ntrc=d to d n more work - further casc st~ldtes(step 5 ) . revised prnprlsition to srlrndhing like: ' Whrrc tt*ilciwrshavt. confidence
Initially we rn~ghtlook for furthcr casw o f n r w t y ~ntrt>duccdTordl in thr falrnw- of a devolved ~ t a f t i n gsytern ~t will Imd to in1pr4n*cmrnts
staffing svstems to check that the initial casc studv was not ju4t good in the qualitv of rducation,'
luck. We could gn nn with additional casc st~rdicsto c h t ~ ortt k how well this
Let us imagine that we then ccrme acmss a qchool thnt o u r critics yrtrposition Ilnlds. En additicm lo looking at caw5 whcre n local q t a f f ~ n g
a l l i t d d tci - one wherr the ~ n t r o d i ~ c t i oofn the ncw .;=tern had heen a yyqiprn was inttoduced we should look at some cases whcrp the central-
disaster. I t had led to tcachcr, sturient , ~ n dparent disratisfaction, I o s ~of i7rd syqtcrn was rctainrd. We would ~ v p e c tthat r 4 would ~ not sec the
morale atrd a ctear decline in many aspccts of educational qt~altty.How s a m e Icvcl of improvement in edt~cationin thew s c h o n l ~ I. f this prr~vcd
do we make scnsr of this caw? What is dilfert-nt about this case that tn bc sn it wi7t1ld strengthen our ctlnfidrnce in our initinl throry about thc
shouId makc it behave d i f f ~ r m t l yfrom that which w r predicted pravi- inipact of local st,iffin~systcms.
sionnlly, and so diffcrentlv from t h e other caw< wc havr alrcndv 5tudicd7 We r n i ~ h look
t for cases that could di;pro~vo u r thenry. For e u ~ m y l r~f ,
Since our study should protluce a compl~trpicture OF each casc and its wi> coulcl find a school that had kept thc centraliwd syqtcrn and had
f o c i ~ and
l rultrrml: context we rvill prubahlv pick up some initial clues. shnwn an irnprcrvrmmt in educatinnal qualitv wc rn~ght2x1 a b l ~ ,to
How is the 'drviant' sd~ooldifferent from the others? W c might notice f ~ ~ r t l l eimprove
r our theory. WP might shrdy a schnol w h ~ r t we * hnvc
~ m fvaturt.q c ahnut thc pnlcrs.; by \vhlch the ~ l e v n l \ . ~scheme d ~vas idt~ntificdthat there Il,i< h e n nn Irnpravrmtnt I n cducatitln btst thc
~ntroducc*rlin the school in which it ~ 7 ~ a1 sInilurc. Fur cxampIe, tcachrrs ctntltrali;rtd ~ t a f f i n gs)rstpm had b c w rt\taincii. Thiq wt>~lldtvll a ~ a i nour ~l
report t h ~ rtt svas impn.;tld nn thcm Tlruy wPrr not c(>nsultcdand thrre r c v i ~ c dprcrp{~rh~rn. Hr~rvevcr,further in\ c..;tijiation of thc <cI?tlol rn rsllt
arc nrr p r o c r s r s 01 appcal a~ain.;! thc principal'.; decrsions. O u r linrnch ~ wlien w e Irwk a t thv qchtml in i t s widt\r contcvt ~ h e r c15 J
~ I i o r *that
thcn 15 t31;11 t h r success of thrs type of ~chcrncdepends cln thr waV in l~o<titerclaticmship bctwecn thc school ~ n Ithc d Ivs,il cc,nrmr~n~tv. 1Z'c
tvhich it i q irnplemrqtp~-l I ~ a mthat thrre had b c ~ nn previous attempt b7. t h e schnnl r{n~ncil(on
M'e tllcn Inok back s t our cnrlier caws to qcr i f this c ~ ~ l a i nt ls~ c which thrrc. werv manv local cornmun~tl,rcprescntattves) tc, ~ r l t r o d ~ tar e
diftcrent outcnrnc.; Dn the s~~ccc..;.;ful cnseq ,tan4 11uta.: hal47ngdiffcr~nt Iucai stnffinq cchrmc. TcacFlcrs had sl~cce+<fullv rc>+i\trdt h i q ant1 Fc~llthat
tmplc.mt3ntationpsnccsw.; tn the clisa<trnu<ctlse? I-et 11.; s l p p o s e that the tl~tlccntrali/~dst0iihngsyqtrm was protest~ngthem.
s u r c e s ~ f u lc~weq irnplr*m~nt~,d t h r scheme nnlv affcr 3 great deal ot Mfhtli M.P It.ai-n t h a w thing.; a b n t ~ trr-hat nt firkt 5 i ~ l l t1 ( ~ k Itkc s n
.-?"-' fl'tltlc~n ,lnlj qlrr~rt-! \ . i s w~drl+~,rcn~l cnn<lbn.;llrnl,nut tht3 c l c ~ ~ r ~ b l l v h 'tltw~arilcaqr*' th,il dt+prn\rc*c:o u r propositirm, in facl it t 11rthr.r ctinl~rni.;
266 CASE STUDY DESIGNS
o u r propositinn that strfsses the notion that teachers must have con-
fidence in the system. Perhaps this case would cause us to further
broaden our proposihnn to emphasize the importance of teachers having
confidence in the system and to identify t h e factors that produce that
cnnfidcnce rather than strcss the importance of a local or centralized
systenr in itself. REFERENCES
Summary
This chapter has stressed the pi)iiint that case study analysis should he
fundamcntally a theoretically informed undertaking. This applies Aronson, t a n d Mills, j. (1954) 'me effcct of severity of initiahon on liking for a
equally to descriptive and explanatory caw studies. The use of ideal group', Joirrrlrll r l f Ablrr>r8trrllni7d Sacrol Ps!/c!t~/~gyr 5% 177-81.
types as a method of analysing descriptive case studies was di.sfrrssed. Alchley, R.C. (1976) Tlrr S o c r ~ / r ) nf ~ qXpt~r~tnenl. Cambridge, MA: Schcnkman.
Two broad methods of analysing casc studies for explanatory purposes Baumrind, U. (1964) 'Srlme thoughts on ethics of research: after reading
were outlincd. Where case studies are used to test a theory, pattern Milgram's "Behavioral study of ubedicnce"', A~ntlrfrflnP~.vc!zo~o&'~s~, I$. 421-3.
matching provides an appropriate way of analysing and comparing Becker, H.S. (1966) Orlfsi'ler~Stlrrit~,~ 01Dmiance. New York: Free
Irl tile Socii~lo~ey
Press.
cases Pattern matching can range from matching very simple predicted
Blalock, I I , (1964) Corrsal I n f e r c ~ ~ c cins Non Expi~uii~rrntalRe.~wrcli.Chapel Mill, NC:
patterns to highly complex patterns involving mrlltiple independent and
Universitv of North Carulind Press.
dcpendent variables. Where case xtudics are used to build a theory, Iqlnu, P. ( I Q h 4 ) Sxcllnrfgr nnd Pr~tlli,rit1 Sorrn! L ~ f cNew York Wlley
analytic induction provides a uscful wav of analysing and using each Rlurner, H 11956) ' S o c ~ n l o g ~ canalysis~~l. and thp "vnriable"', A m r r ~ c n ~Sorir~logicfll
i
case study. Kr'url.rtt, 2 1 683-40.
Rrutvn, C;.W and Harris, 'r. (1978) TIie Socml C7rr~if1sr ~ fDt-prcssinr>:a Sfrrrly of
Fs!jrhintrrc Dl~ardrr111 M7onre1r.I.ondom: Tavistock.
Note Ruck, N., Ermisch, J a n d j~nkrns,S. (1995) Choosing n Lvngitudinal S u m q Design:
Tlic lssrt~'s,ESRC Rosearch Centre on MicrwSoclal Change, Univers~tyof Essex.
1 Of course this Iogic dssumes that our causal variable is the only factor that Campb~11,D.T. (1989) 'Foreword', in R.K. Yin (rd.), Carse 5tud.y R~rsenrch:D~sign
can produce a given imtcnme. If we find caws where the effects occur but our and MriholJs. Rcvcsly Hills and Imndon: Sage.
favoured cause is absent then we have alternative explanations of thc phenom- Cdmpb~ll,D.T. and Stanley, J.C. (196.7) E x p ~ r r t ~ ~ r t ~nnd t f l l Q~rasi-Experimentfll
enon. O u r research design should he de5igned to elim~natesuch alternative Ursr~~trs,!i~r R ~ w r r c h .Bor;tc>n'Houghton Mifflin.
explanations Cohcn, J. (1987) Statistjcnl Poxcr Annlysiclfor the Brlra7)iourrll Scrc!~c~s. Ncw York:
Academic Press.
Cook, T.D. and Campbell, D.T. (1479) Quasi-Exp~ri~tzmtat~on Dbipu und Annlysis
Issurs for Fidd Srttinp. Boston: Huughton MiMin.
Cnser, L.A. (1977) Mnslrrs of So~-inlng~~-i~l Thought: Idens Sn Historical and Socrnl
Glnl~xt New York: Flarcourt, Brace, Jo~,anovich.
Davrs. 1.A (1985) Tlre Loxic of Cnirsnl Order. Beverly I-Iills, CR: Sagc.
Dempsry, K. (1990) Smnllfuictt: n Stlrrdrt o f Snrinl I ~ ~ r q l m l r t Cohrsinri
y, nnil 6clorrapii~,q
Melbournr. Oxford Unikcrsity Press.
L1~n7rn,N . (197s) Tliv Rr..sp~truhAct- a I'l~rvr+,ticnlrtrfrr~iilrctrnnlo Rrseflrcl~Mcthods.
U e w York McGra~v-I-lill.
I l ~ , p a r t m r ~of~ tf l u m a n %;ertmicesnnd I3ralth (1995) Ynl/th S z ~ i ~ i d1t1 t ' A r ~ ~ t r n l i n0:
Br~cl;qm~/i~d Mrr~~ryrnl?h. Canberra: Australian Government Publishing Servirc.
dc V a t ~ s ,D.A, (lq9.l) Lurtlng Go: Rrl~rt~nrrsf~r~~:: b~fior'rlrAifrrlts und thcrv Pnrewls.
Melbourne: Oxfnrd Universib Press
de Vaus. [).A. (2001 ) S A L Y V C rI iIl~ Social Rt,srurcllr l..ondnn: Routlrdge.
d e Vaus, U.A. and LVolcott, r. ( e d s ) (1997) Arrsfr~lrlrrrF ~ n ~ i I'roftlr~: ly Sncrnl nnd
Drrrlnyrrtyl~rr Pnlterns- Melbnurnr: Australian Instih~teof Family Stndics.
2hH REFERENCES
I
Mjlls, C W. (1959) TIlr Socrologri-01 Imnglrraflon New York- Oxford University
in survcy analysis', Soc~nlo,~ica! M ~ t h o d snrld ResmrcIt, 4: 459-74 Press
Hochschild, A 11989) TIw S~,co?ldShift New York: Avon. Mi tchrll, J C. (1983) ' C a ~ eand situation analks~s',Socrt~/ryrrolRnllnrl, 31 : 1x7-211.
Holnnn, R. (19Y1) T l t ~Ftllrr~nf Sllcrnl Rrsenrch. London I.ongrnan. Moser, C A and Kaltun, C (1 971) S I I M~Hrodlr ~ ~ 111 SOCIIIII~f~~strxnt1011 London
i
I
t!clnlans, C;. (1961) Sncrnl Rrltavici~rr:[IFElrmtbnhory l ortrr?. Yew York Harcorirt,
Brdce, luvanovich.
Hrlnemann
Mlreller, I H., Schuessler K r and Cosmcr, H.L. (1977') Strrhsf>cn/ K I ' I ~ S O I I I I!II ~
Tnhnsnn, h, (1977) Sac:,ll Sfnt~strc5r t < r f / ~ c ~f r Tt3~?rs.
r N e w Y n r k McGraw-Hill.
:I Kalton, G. (19833) 'Compensating for miss~nfis u n e y data', Research Report
Srrc~olo~~r Ibston Houghtrjn h'liifl~n
n t i v c In N Denzin ntld
II Scr~cs,lnstih~teof Soc1a1 Rcscarch, University ol kllchzqan
I'unch, M (1W4) 'l%I~tlc+and etll~csIn q ~ ~ a ~ ~ t ~ trr<enrch',
Y. L~ncoIn(cds), Hiri~iflmoF.ot Qtralifoti~rReqarch Thou5and Oaks, CA S d g ~
Kalton, C; (1983b) ltrtrod~tcfiurtto SItn~e!/So)npli?r,g.Beverl' HIITS,CA: Sage. pp 83-97
Kalton, G. (1086) 'Ciandling wave notl-response in pancl surveys', Jorrrrrfll ~ r f Rrdftcld, R. (1'130) Tl~puztlnn,o M1,rrcolr Yrllyiy 17 5tlrti1, of Frdk l @ Chicago
C?fir>nlStof7sf1rs, 2: 303-14 Unrvi.r<~tyrrf Ch~cagoP r e s
Kt'llehear, A. (1947) 7 . l ~ililclbtr~r.iuc Rmei~rt-lwr: Grr~dcto ~ u f h d d hSt . Leonnrds: Riwthlrsbcrgrr F.l and Dicksnn, W T (1919) M ~ r r f a ~ ~ ~ rrurd ~ l ~ iIht'
i t Wnrhrr.
Allrn nrtd Unwin Carnbridgp, MA: Harvard Universlh, Prrs5.
IZEFERENCES
tl.1wthon1c ctfcct. 77 Intcr.~ct~rln cfftrt.;, 42, 55. 65-6. 73, 74. 103, M,i>kin~relationfhips. I<>(>-7
c r r w wrtionaE ct t.xp~.rimcnts,IK4 History, 132 1114 Ma(rlicd fluup<, 44-5, 75
rrt~)rssccti(tnaE c f pallel dudips, 184 ~mxpvrimcnts,73 clabr~ration~ n ~ l y s i20.1-5
s, r x p m t fdctn, 44-5
casc studl: clesip, 237-411 CJ\C s t u ~ l ~ t236
s. Internal vnltdity, 27-8 cross sectional d ~ r ~ g n171; s,
cross scctlcmnl design.;, IH4-5 c r w s scctir~naI d~sikm5,171 ad hr>c t*xplanaho~rf,182 ~labnratlnnanalysis, 203-1 R
cxperimmt5,7h-8 [ongittdinal designs, 132 cause and correlation, 234 matchcd block d e s l g c , 47-6
longit~rdin~rl rl~~sigps, 135-K Mnli?Iir units o l annl\rsis, 22U-1 uar~ableanalvsis, 3 - 6 post test, 33, 4H.hl
inln~gratlcmand outrnigratron, 137 crlntext in cace studies, 234-h p r p - k t , 14
panvl attrition, 135-h cross sectiorral dcsibm, 177 sandomisa tion, 45
renctiviiy, 7h pffectof no randoln~7eJcontrol group, 131 statislical controls, 202-3
repwx.nt;ltivcnL%s, 1 P4-5 cqtahlishtng rauw, 177-81 Matur.it~nn
5t;l trstrcal cpnrralixalirm, 237, 2111, 249 pft.lbl~shlngmvaning, 18 1-4 wsc sfudics, 23h
iesting for bins cffcctl;, 77 !11510l y, 77-4, 132, 177, 276 cross seciiunal des~jins,177
thcorr~ticnl~ m r m l i f a l i n n227,
, 240, i d e o ~ a p h i cexplanations, 233-4 cxpcrirnents, 73-4
Z4Y race s t u d ~ e s233-7
, Iurigrtud~naldesigns, 132, 133
tlirents tu. 76-7 cxpcnmcnk, 72-6 Meaning, 5, 11
unrr-pr~wntativfsanlplecl, 76-7 long~tudlnaldeslgna. 131-5 establish~ngmeanins, 181- 4
Ertrancoiir variahlcs, 19-,20, 37-4 arv nlslo 0 p t - rtirlnali~atinn
~ Instrument d ~ c a y ,7-1-5,1334, 177' Facesl~wtscxrolt>gy, 182
lndlrect causal relationships. 36-9, 71 rnnttrrat~on,73-4, 132 case snldv annlyhis, 2 3 4 4 , 250
Factor ,~nalysis,lL)B t.laboration ~nalyi.;,204, 2Dh-7 n?r*asurt,ment frror, 134 cross swllnnal designs, 1814
1:;lclnriat a n ~ l y s i ui i vnrianr~ lndlvidual level chc~nge. 122, 129-9.134. 175, mnrtnlitv, 75-6, 1.14-5, 177 pro\~dinfimcaning, 1H2
~ L > L -I r w - \ v a ~ an;llyc~k
- of vari~ncc 157-54 nnmntlietic explanat~nns,273-4 variable analvs~s.181
r:;lctori,>[designs, fji-H, 103 grapt15, Ib3 sclc~hr>n, 73 Mmsr~rt.menlprnlr, 24-71, 84, I DJ, I Sb, 157
I',~llacy o f aiiii-minfi tlw ron.;rqutqt, 1npthr3d.; of nnalvsic, 1 5 1 4 stalislical regressron, 75, 134, 177 conqtant error. 31
13-14 ~nulttdircctionalct~~angi-, 151 tcstlng/pand cnnditionlng. 71, 17.7. 177 corrt4a led PrIvr 72
F;rl.;iiira lion, 14- 15, 267, 2h.5 panel d r s ~ g n s 157 , unlly of casw, 234-5 term* of, 31 -2
closed thcr~rieu,14- IF; t<lt)lt.h,163, 1M Interval estlmntrs, 2tH Inn~~tudunnl drsipis, I34
Fcminlst rprrarch ethics, 217 Induction Itjtewals random crrrlr, 71
]:loor and cciljnfi cffccts ,inalytic, 2411,263 -6 attrition, 79 Measuring changt,
.,itrtruni-alim ~ i l m - l s mumeratlvc, 240 hetwwn t.xpcrimt.nta1 s f a ~ e s79 , bcr rhange mmPa5uwrnent
I:OCUSW~ s,?mpling, 240 lndtlcti vc typologies, 225-6 memory rflert~,80 Micro c l ~ a r ~ g e
Framcb nf refcrenw, 3 9 4 7 Inductive reafonlng, 5-h ~ n ~ 19,3h
I n t e r ~ ~ e nvnriablf, qt-r indlv~dualc l i a r ~ ~ r
5m lrlst~cornpiris~~ns st? nlso nnalvhc ~nducticln, mdwctlo~i s , 172
I n t ~ n ~ e n t r o n47. Missing data, 117-53
F r c q u ~ n c yiiistril~\itiont;,145 Inference nctivc, natural, 47 alprage o l span rmprltatlon, 150
srr causal rrlatii)nsh~ps,thcnry types ol, 48 bias, 148
htt~td~n~ uneven, 82-3 dealing with, 14s
Inferential stati~tich,89-90, 99 Ilem non-response, 147 group means ~mputatinn,150
Informed consent, 7H,P7, Zlh ignoring missrng data, 151
rasp studleq, 24h I-adder of abstractton, 24 imputatirm Fmm earllr=rwaves, 151
d e c ~ p t i o n 7,R Level of rnpacuremrnl, 40-2 ltem non-response, 147
how much infin-n~ntiun,R.4 interv,~l,91 prrwise, 151
erperunents, 64-6 nnminnf, 91-2 randnm wstgnmcnt within ~ O I I P F
l(m3rttldinal dc~ikms,145 ordin.11, 91 i r n p i i t a t ~ o ~150
~,
IirnlnK, H? Lifc ct,ur\r analvsi.;, I lt; regrrssion haqecl i r n p ~ t ~ l b o151 n,
who frrlnl, 8.5-6 I .~teraEr~pTlcatior1,262-3 wmplp mean< Irnpul.ltion, 1.19
lnm~~rntiom, 120, 1:7 Lon~itudlnaldc\irm\, 4g-50, 1 1 3-hq sample s v e , 148
I t ~ ~ t r u n i r nc tl ~ r a v c f cvpcrlrnrnral d e s i ~ n 113
, scalc ct?n.;tructinn, 14X
ckpvrrmcntal, 71-5 dlnieiisions rrf, I l ? l Gnurrpg, 14;
SFPSS srrtronal d ~ s ~ p nI77 s, p n f l cf trend, 113 unit IIU~ rt3sponsc,I47
I a ~ n p t t t d ~dnc~slr p \ , 1W,131-4 prospectwe cf r c t r r ~ ~ p c c ~113-14
rc, w a v r nnn rpsponse. 137
purpr~o r f, 11;-lH CIY R / W nnn resptmfe
Mortdlih, 134-i
Macro changr cross x ~ t i ~ n idl ~l s i ~ m177 5,
srr aggreaatt? change e k p t i m c n t s . 75-6
Mnun effcvtb, h5 l o ~ l g i t ~ ~ dd+l+it;ns,
i ~ ~ . ~ t134-5
INDEX 277
Multiple jiroups s~thst<>nkial arrJ theoretical types, 22-4 RPactivity hard kn frnd groups, 123
multiple c a w study design, 22h tedinical and rnctlio~icllu~,ical, 21 case studies, 236 pooling panels, 127
multtple coI>orl dcstgn. 124 Post test only wirh conlrul group Recall purpose, 122
multtplc comparison test?, 101 e x ~ ~ r i n t n ~MI
ts, anchor events, 128
multtple group cnpcr~rnentald w p , Prcdictton, 3-4 memosahle events, 128 %mplc size
62-3 I'r~vacy problems with, I40 attrition, 143
Multiple time points 5r.r' confidcntiality recall problems, 127 crms h~chonaldeslgns, 187-90
multiplc potnt panel d e s g n , 119-20 Probability sampl~ng,79, 90 rel~ospectiwdesips, 140 dctcrrninants of, 187-188
multipkr yre test experimental dfijgn, I'roof, 15 r c v e s c telescoping, 140 elaboration analysis. 210
62 Provisronill nature of suppnrt, 15 telescoping, 140 experiments, 78-9
Multivar~nteanaTvsi5, 211 Proportions, 95 Rpcoding, 196 missing data, I 4 8
mu1 tiplc. regrcsslon. 104 Propositions. bH dishlbutionaI approach, 197-8 oversampling, 143, 187-8
Pruspcrti\ c d c r i ~ m masktng relabonships, 196-7 random, 79
Net change cnw studv dec~gnq,227-8 substantive approaches, 196-7, 198 single and multlple case stttdy dcsignq,
sw aggregate change disndvantagcs, 127 Rword I i n k a ~dcsips, 128 22h-7
Nomlnal d ~ f ~ n i t i n24-7
n, rnult~plrpoitll panel, 119-20 confidet~tialitv,128 statistical power, 78
Nomothctic explanation, 22,22 1, 233-4 simple pinrl, 118-19 Jteltabllity, 29, 30-1, fK subgroups, 187-8
Yon-pararnetr~cstatist~cs.1U2 changc scares, 156. 157 Sampling, 14
Yon-probabilily snmplmg, 90 QuaI!latlrre, IZepeated cmsy-sectional d~siw, 129, 137, access to cases, 242
Yorrnal distrrbutiun, 1UZ-3 cf quanhtativc. 10 21 1 caw s c w n i n g , 24 1
assuniption of. l(12 change, 154 Replacement for attrillon, 137 case qtudy designs, 240
crttical ratio, 1W data, 191 analysis irnplicatirlns, 122 confidence tnlerval, 189
t-tPStS. 1110 casc stud ips, 230- 1 strategies, 121 cnnfidenrp Ievc.1, 189
Y u r m a l l ~ e ddistributions. 102 ~ ~ ~ e r i m m83t s , hplicahtnl, 77, 78, 226 external valrditv, 29
Quantitative case studies, 226, 237-8 focuswd, 2-11
One grrl~tpcornparirons with known value, ct qual!lativp, 111 elabora t ~ r manalpcis, 203-6 number of cases, 2411-1
97.99 change, I54 exper~menh,90 oversampltng, 143, 187-8
One shot case s k d y design, 230-1 data. 191 literal, 279, 262-3 prvcislon of estimates, 188
Operational definlt~on,24, 27 rnsc stltdtcs, 249-50 theoretical, 239,262-3 s a m p l ~ n gPrror, I #
Operationalization, 24-27 methods, 1U Rppnrting case stud~es,2 4 - 5 selection in case studies, US-40
Clutmigration, 120,137 Quahr longitudinal designs, 12H-4 Research design theoretical, Z40
Oversampt~ng,143, 187-8 aggregate change, 12R-9 cf method, 9-10 typical caws, 240
indivtdual change, 128-9 cf work plan, 9 unrepresentative, 29
P a i d sample design, 102 Qwstronnaire length, 190-1 definition, 8-9 variance, 190
Panel agcing, 124-5 dimension5 of, 47-8 weighting, 188
Panel condit~oning,143 Random error, 31 Research question, Sampl lng error, H9,92,99, 188,
assessing cffi~k,123 I<andrjm sampling- descriphve, 17-19 longh~cllnaldeslgm, 143
Panel data, analysing, 155 scc prob,ttrll~tysampling explanatory, 19-21 Scale constr~rction,148. 151
Parallel casc study d c s i p s , 227,242-3 Iiandomizcul assignment, 45 focuss~ng,clarifying, 17-21 Stepticism, 11-15
investigators. 242 attrition, 8l,1(2 Rcsidualizcd change scorns, 156 Scheffe test. 101
research protwds, 242 control v o u p resentment, 81 Respondent burden, 13940,148, 140-1 %,lection of cases, 2313-41
Parametr~cstat~stics,102 gatekeeperp, HI lictrmpectivc design4 access to cases, 242
Part~clpantobccrvat~tm,246 group S ~ J P ,81 anchor ~ v e n h 128, case screening, 239, 241 -2
case s h l d i ~246
~, s random alltwation, 81
I ~ m j t of casc studics, 227-8 case studies, 238-40
Pattern matching, 2553-60 problems, 71 -2, 80-2 experiments, h0-1 number of cases, 240-1
P?r~Wit;lg~ change, l i h - 7 r c f t r ~ l s 81-2
, memorable cvenh, 128 sdrnpling, 240
calculal~ng,156-7 tinderestiinatin~~ , i u s , cffprt,
~l panelf, 126-H I;electmn, 75
rnPnsuremcnt crioc, I57 71 -2 recall problems, 127 %;eft-fulhll~ngpropl~ecy.83
probpems, 157 R a w diange Ccrjrc5, 155-6 re%erse telpscnp~ng,140 '+qucntial case study dcs~gns.227
reliabil~v,157 ~alcul~itin)?;. 155-h telescoping, I40 Shape nf distribution$, 94-5
Pcnod efferts. 121, 115-6, 221, 21 1-14,216, cxtwrne scorph, 156 Rotating panel design, 122-5, 133 S ~ m p l epanel design
Placebo, 81,83, mpasurpment frrnr, 1 % attntlun, 122, 12.3 attrlhon, 120.121
l'!ous~blc rival hypotheses, 9, 1 1-14. pmblems. 156 avording burnout, 122 h a s , 120
case shldies, 224 condihoning effects, 123 inmigmhun and outrnigratiun. 120
sourcc nf* 21,22-4 external vnltdity. 137 samplc size, 120
S i n ~ p t tp~iriel
* tlcc1i:n ( t - m ~ l , ) llir.or!., 5 Validllv (cr~rt) ektcrn.~lv a l ~ ~ i i t34
v,
wit h ~ C ~ I J C P T I I N ~ I , 1 3 - 2 cu po.;L fnrlo, h crit'ric~n groups, 30 intcmal i7alidit\.%4
$viIlinul rrplacrmm~t.120 post tnclun1, h rrit~rinnv a l i d i l ~2Y
, lon~iludinaldcsijinc, 145
S~rn~rlatcd bcforr-.liter d r s i ~ n ,l?Q c~se study .~nalvsis,240-2, 753-hl) rxterndl, 28-'1 probleml: frnm, %.I
i i n ~ l ca5p
r shldy dcsi~m.2 2 ~ case study drsihm, 121 -h internal, 27-H
S i n ~ l cculinrl cle~ign,177-4 ~ t . 1 . nl.40 tl1ecv-y h~tildrng,thmry testing natui-c of, 28 W a w non rc\pon\r, 147
a ~ t - i ncif~~crt.;,
fi 124 Thcorv building, 5-6 thwats, 24 U'3vr'l
pcr~ud~ j i l ' c l ~124 . annlylic induction, ?6D, 263-6 sirbnlslr internal validity, external validity inh?nf;rl\ htween, 142-3
124
pr~ihlcn~>, 227,227, 2h3-3
crlht' ~iudivs, Variability, mcasurcs r)i, 43-4 nilmber In long~tudrnnlstudzes, 1 4 2
Solomon .I-grtwp desijin, h3-5, HI). Theory testing, 6-3, 11, 253 Variable analys~s,1H1,234-6 pinpl r o n d ~ h o n l n143
~
103 rase sludy tlesisms, 221 -3, 227, Variahlc. bv casc matrix, 185-6 rcspclndent hurdcn, 143
lestin~ ell<brls, 71,fh 253-hll Variance, samplc, 100 clletghts, 7h.3 7h. kHH
Sprri ficatitw, 2114-5, Tl~ick~lcscripti~~n,
2.711 iQoluntay p>rticipilivn mi>sing ci,?t,? bid., 151-2
Spur~oul;rcl,ilinnshipx. 33, 37-'1. 44 'I'irnr orderetl an,~lysis,35, 11: bins, 84
vlahnwt~nnnnatysis, 2LI-I. ZU7 biogmyhy, 251 experiments, K - 4
Sti~ffi~ig cnsrnl;t~rdivs, 752-.3
lnn~itudlnaldesigns, 144 chmnnln~ical,~nnl ysis, 261-2
Standariiiratltin, 157-hl ct~ronoIr)girs,2.52
inllalinr~adl~~slmcnt. 157-0 hislories. 252
inhtruincnth, 17s i ~ i t r - r n i ~ t tinw
c d s~ries,Zhl
~ * ~ c ~ mnk, ~ l t I~V l- hc 1 linir ~ t ~ r iannl\sis.
~'. 2M1-2
prtiynrlion.;, Ih(V-1 twnd onalysi\, 2MI
SC.I I IIS, 157 -h 1 Tmc klng paricl rncrnbl,crs. 13S-? I - l t ~
vnriablt,~. ILK-s cvht, 1111
~-5cnrcs,159 - h l dccl~r%tcrrng, 14 t
Statistical cnntrol5. 162, IW Tr.insk~rntt~d distrjhut~on';, 102
atlju5tinp ftlr mi.~hingdata bias, 152 nor~nnlizcrlcli<hibutiom, 102
cC ~xpeririieritalcontrols, 201 -2 Translorrnrrl variables, 1'17-8
1~1,lhr)ralionanalysis, 313-1 [I Trvilhn~nb
crms scctiunal cicsihms. 1 7 7 4 s r t 7 intervenkions