The Case Study Method in Social Inquiry

Author(s): Robert E. Stake

Source: Educational Researcher, Vol. 7, No. 2 (Feb., 1978), pp. 5-8
Published by: American Educational Research Association
Stable URL: http://www.jstor.org/stable/1174340
Accessed: 06-12-2015 20:18 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.

Sage Publications, Inc. and American Educational Research Association are collaborating with JSTOR to digitize, preserve
and extend access to Educational Researcher.

http://www.jstor.org

This content downloaded from 189.125.96.10 on Sun, 06 Dec 2015 20:18:01 UTC
All use subject to JSTOR Terms and Conditions

The

Case

Study

Method

in

Social

Inquiry,

ROBERTE. STAKE
Universityof Illinois, Urbana-Champaign

It iswidelybelievedthatcasestudies
are useful in the studyof humanaffairs
because they are down-to-earth and
attention-holdingbut thatthey arenot a
suitablebasis for generalization.In this
paper, I claim that case studies will
often be the preferredmethod of research because they may be epistemologically in harmonywith the reader's
experience and thus to that person a
naturalbasis for generalization.

Experience. We expect an inquiryto

be carriedout so thatcertainaudiences
will benefit - not just to swell the
archives, but to help persons toward
furtherunderstandings.If thereadersof
our reportsare the persons who populate our houses, schools, governments,
and industries;and if we are to help
them understandsocial problems and
social programs,we must perceive and
communicate(see Bohm, 1974; Schin,
1977) in a way thataccommodatestheir
presentunderstandings.2Those people
have arrived at their understandings
mostly throughdirectandvicariousexperience.
And those readers who are most
learned and specialized in their disciplines are little different. Though they
write and talk with special languages,
their own understandingsof humanaffairs are for the most partattainedand

amendedthroughpersonalexperience.
I believe that it is reasonable to conclude that one of the more effective
means of adding to understandingfor
all readers will be by approximating
throughthe words and illustrationsof
our reports, the naturalexperience acquired in ordinary personal involvement.
At the turn of the century, German
philosopher Wilhelm Dilthey (1910)
claimed that more objective and "scientific" studiesdid not do the best job
of acquaintingman with himself.
Onlyfromhis actions,his fixed utterances,hiseffectsuponothers,can
man learn abouthimself;thus he
learnsto knowhimselfonly by the
round-about
way of understanding.
Whatwe once were, how we developedandbecamewhatwe are,we
learn from the way in which we
acted, the plans which we once
adopted,thewayin whichwe made
ourselvesfelt in ourvocation,from
old deadletters,fromjudgmentson
which were spokenlong ago. .

we

ourselvesandothersonly
understand
whenwe transferourown livedexperienceinto everykindof expression of ourown andotherpeople's
lives.
He distinguished between the human
studies and other kinds of studies.

Thehumanstudiesarethusfounded
on thisrelationbetweenlivedexperience, expression,and understanding. Hereforthefirsttimewereacha
quiteclearcriterion
by whichthedelimitationof the humanstudiescan
be definitivelycarriedout. A study
belongsto thehumanstudiesonlyif
its objectbecomesaccessibleto us
theattitude
whichis founded
through
on therelationbetweenlife, expression, andunderstanding.
Dilthey was not urgingus merely to
pay moreattentionto humanisticvalues
or to put more affective variablesinto
our equations. He was saying that our
methods of studying human affairs
need to capitalize upon the natural
powers of people to experienceandunderstand.
Knowledge. In statements fundamental to the epistemology of social
inquiry, Polanyi3 distinguished between propositional and tacit knowledge. Propositionalknowledge - the
knowledge of both reason and gossip
- was seen to be composed of all interpersonallysharablestatements,most
of which for most people are observations of objects and events. Tacit
knowledge may also dwell on objects
and events, but it is knowledge gained
from experiencewith them, experience
with propositions about them, and
rumination.
5

February1978

This content downloaded from 189.125.96.10 on Sun, 06 Dec 2015 20:18:01 UTC
All use subject to JSTOR Terms and Conditions

Throughreasonmanobserveshimself; but he knows himself only

throughconsciousness. (Tolstoy,
Warand Peace)
Tacit knowledge is all that is remembered somehow, minus that which is
remembered in the form of words,
symbols, or otherrhetoricalforms. It is
that which permits us to recognize
faces, to comprehendmetaphors, and
to "know ourselves." Tacitknowledge
includes a multitude of unexpressible
associations which give rise to new
meanings, new ideas, andnew applications of the old. Polanyirecognizedthat
each person, expertor novice, has great
stores of tacit knowledge with which to
build new understandings.

t is a common belief thatthese ordi-

naryunderstandings,bothnew andold,
are merely the pieces from which
mighty explanations are made. And
that explanationis the grandestof understandings.But explanationand understanding are perhaps not so intimately interwoven.
Practicallyeveryexplanation,be it
causal or teleologicalor of some
otherkind,canbe saidto furtherour
of things. But "ununderstanding
alsohas a psychologiderstanding"
cal ring which "explanation"has
not. Thispsychologicalfeaturewas
emphasized by several of the
nineteenth-centuryantipositivist
perhapsmostforcemethodologists,
fully by Simmelwho thoughtthat
as a methodcharacunderstanding
is a formof
teristicof thehumanities
in themindof
empathyorre-creation
thescholarof thementalatmosphere,
thethoughtsandfeelingsandmotivations,of the objectsof his study.
is alsoconnected
... Understanding
in a way thatexwithintentionality
the
is not.Oneunderstands
planation
aims andpurposesof an agent,the
meaningof a signorsymbol,andthe
or
significanceof a socialinstitution
religiousrite. This intentionalistic
. . dimensionof understanding
has
come to play a prominentrole in
morerecentmethodological
discussion. (VonWright,1971)

Explanationbelongs more to propositional knowledge, understandingmore

to tacit.
Philosophersof the positivist school,
Carl Hempel and Karl Popperparticu-

larly, have posited that propositional

statements of lawful relationship are
the closest approximationsof Truthwhether we are talking about physical
matteror human. They would have us
speak of attributesandconstructs,such
as energy and mass or work-ethic and
masculinity, and the relationships
among them. Antipositivists such as
Dilthey, Von Wright, and William
Dray have claimed that Truth in the
fields of humanaffairsis betterapproximatedby statementsthat are rich with
the sense of humanencounter:To speak
not of underlyingattributes,objective
observables, and universal forces, but
of perceptions and understandingthat
come from immersion in and holistic
regardfor the phenomena.
In American research circles most
methodologistshave been of positivistic persuasion.The moreepisodic, subjective procedures,commonto the case
study, have been considered weaker
than the experimentalor correlational
studies for explaining things.
When explanation, propositional
knowledge, and law are the aims of an
inquiry,the case studywill often be at a
disadvantage. When the aims are understanding,extension of experience,
andincreasein convictionin thatwhich
is known, the disadvantagedisappears.
Generalizations. The scientist and
the humanist scholar alike search for
laws that tell of order in their disciplines. But so do all otherpersonslook
for regularityand system in their experience. Predictablecovariationis to
be found in all phenomena. In 1620
Francis Bacon said:
Thereareandcanbe onlytwo ways
of searchingand discoveringtruth.
The one flies fromthe senses and
to the most generalaxparticulars
ioms. . . thisis nowthefashion.The
otherderivesaxiomsfromthesenses
andparticulars,
risingby a gradual
and unbrokenascent,so thatit arrivesat themostgeneralaxiomslast
of all. Thisis thetrueway,butas yet
untried.
He claimed that Truthlies in the most
general of axioms, a far and labored
trek from experience.4
Another point of view holds that
Truthlies in particulars.WilliamBlake
offered these intemperatewords:
To generalizeis to be an idiot. To
is thelonedistinction
of
particularize
merit.Generalknowledgesarethose
thatidiotspossess.

Generalizationmay not be all that despicable, but particularizationdoes deserve praise. To know particularsfleetingly of course is to know next to nothing. What becomes useful understanding is a full andthoroughknowledge of
the particular, recognizing it also in
new and foreign contexts.
That knowledge is a form of
generalizationtoo, not scientific induction but naturalistic generalization,
arrived at by recognizing the
similaritiesof objects and issues in and
out of context and by sensing the
naturalcovariationsof happenings.To
generalize this way is to be both intuitive and empirical, and not idiotic.
Naturalisticgeneralizationsdevelop
within a person as a productof experience. They derivefromthe tacitknowledge of how things are, why they are,
how people feel about them, and how
these things are likely to be lateror in
other places with which this person is
familiar.They seldom take the form of
predictionsbut lead regularlyto expectation. They guide action, in fact they
are inseparablefrom action (Kemmis,
1974). These generalizationsmay become verbalized, passing of course
from tacit knowledge to propositional;
butthey have not yet passedthe empirical and logical tests that characterize
formal (scholarly, scientific) generalizations.
Sociologist Howard Becker5 spoke
of an irreducible conflict between
sociological perspective and the perspective of everyday life. Which is
superior? It depends on the circumstance, of course. For publishingin the
sociological journals, the scientific
perspective is better;but for reporting
to lay audiences and for studying lay
problems,the lay perspectivewill often
be superior. And frequently that
everyday-life perspective will be
superiorfor discourse among scholars
for they too often share among themselves more of ordinary experience
than of special conceptualization.The
special is often too special. It is foolish
to presumethata more scholarlyreport
will be the more effective.
The other generalizations, i.e.,
rationalistic, propositional, law-like
generalizations,can be useful for understandinga particularsituation.And
they can be hurtful. Obviously, bad
laws foster misunderstandings. And
abstractstatementsof law distractattention from direct experience. Good
generalizations aid the understanding

ER

This content downloaded from 189.125.96.10 on Sun, 06 Dec 2015 20:18:01 UTC
All use subject to JSTOR Terms and Conditions

of general conditions, but good

generalizations can lead one to see
phenomena more simplistically than
one should.

t is the legitimateaim of many scho-

larly studies to discover or validate

laws. But the aim of the practicalartsis
to get things done. The bettergeneralizations often are those more parochial,
those more personal. In fields such as
education and social work, where few
laws have been validated and where
inquirycan be directedtowardgathering informationthat has use other than
for the cultivationof laws, a persistent
attentionto laws is pedantic.
Cases. The object (target)of a social
inquiry is seldom an individualperson
or enterprise.Unfortunately,it is such
single objects that are usually thought
of as "cases." A case is often thought
of as a constituentmember of a target
population. And since single members
poorly represent whole populations,
the case study is seen to be a poorbasis
for generalization.
Often, however, the situationis one
in which there is need for generalization about that particular case or
generalizationto a similar case rather
than generalizationto a populationof
cases. Then the demandsfor typicality
and representativenessyield to needs
for assurance that the target case is
properly described. As readersrecognize essential similarities to cases of
interestto them, they establishthe basis
for naturalisticgeneralization.
The case need not be a person or
enterprise. It can be whatever
"bounded system" (to use Louis
Smith's term)is of interest.An institution, a program,a responsibility,a collection, or a populationcan be the case.
This is not to trivialize the notion of
"case" but to note the generalityof the
case study method in preparationfor
noting its distinctiveness.
It is distinctive in the first place by
giving greatprominenceto what is and
what is not "the case" - the boundariesarekeptin focus. Whatis happening and deemed importantwithin those
boundaries (the emic) is considered
vital and usually determineswhat the
studyis about, as contrastedwith other
kinds of studies where hypotheses or
issues previouslytargetedby the investigators(the etic) usually determinethe
content of the study.

Case studies can be used to test

hypotheses, particularlyto examine a
single exception that shows the
hypothesisto be false. Case studiescan
be highly statistical; institutional research and vocational counseling case
studies often are. But in the social science literature,most case studies feature: descriptions that are complex,
holistic, and involving a myriadof not
highly isolated variables;,data that are
likely to be gatheredat least partlyby
personalisticobservation;anda writing
style that is informal, perhaps narrative, possibly with verbatimquotation,
illustration, and even allusion and
metaphor. Comparisons are implicit
rather than explicit. Themes and
hypotheses may be important,but they
remain subordinateto the understanding of the case.6
Although case studies have been
used by anthropologists, psychoanalysts, and many others as a method of
exploration preliminary to theory
development,'the characteristicsof the
method are usually more suited to expansionist than reductionist pursuits.
Theory building is the search for essences, pervasive and determiningingredients, and the makings of laws.
The case study, however, proliferates
ratherthan narrows. One is left with
more to pay attention to rather than
less. The case study attends to the
idiosyncratic more than to the
pervasive.8 The fact that it has been
useful in theorybuildingdoes not mean
that that is its best use.
Its best use appearsto me to be for
adding to existing experience and
humanistic understanding.Its characteristics match the "readinesses" people have for addedexperience. As Von
Wrightandothersstressed,intentionality and empathyare centralto the comprehensionof social problems, but so
also is informationthat is holistic and
episodic. The discourse of persons
strugglingto increasetheirunderstanding of social mattersfeatures and solicits these qualities. And these qualities matchnicely the characteristicsof
the case study.9
The study of humanproblemsis the
work of scientists, novelists, journalists, everybody of course - but
especially historians. The historian
Howard Butterfield(1951) recognized
the centralityof experientialdata and
said:
we ever
?. the onlyunderstanding
reachin historyis buta refinement,

moreor less subtleandsensitive,of

thedifficult- andsometimesdeceptive- processof imaginingoneself
in anotherperson'splace.
Case studies are likely to continueto
be popularbecause of their style andto
be useful for explorationfor those who
search for explanatory laws. And,
moreover, because of the universality
and importanceof experientialunderstanding, and because of their compatability with such understanding,
case studies can be expected to continueto have anepistemologicaladvantage over other inquiry methods as a
basis for naturalistic generalization.
Unlike Bacon's "true way" of discovering Truth,this methodhas been tried
and found to be a direct and satisfying
way of adding to experience and improving understanding.
Notes
'Writtenat the Centrefor Applied Researchin
Education, University of East Anglia, as partof
an assignmentfor the Organizationfor Economic
Cooperationand Development, Paris.
2In this paper I am writing about the formal
inquiryto be done by people, on or off the campus, who aresubjectto greaterrewardsfor scholarly workandknowledgeproductionandto lesser rewardsfor professionalsupportand problem
solving. In the USA therearefew civil service or
applied research agencies which validate their
inquiriesaccordingto its service value more than
to its "internal and external validities," as defined by Campbelland Stanley(1966). I see it as
unfortunatelynecessary to overstatethe distinction between academicresearchand practicalinquiryas a step towardimprovingandlegitimizing
inquiries that are needed for understandingand
problem solving but which are unlikely to produce vouchsafed generalizations.
3Iam indebtedto statementsby HarryBroudy
(1972) andAndrewOrtony(1975) for helpingme
understandthe educationalrelevanceof the writing of Polanyi.
4Buthe notedthatat least before 1620 that was
not the way humansreachedunderstanding.
Becker(1964). Importantideasabout
"5Howard
the special use of case study as precursor to
theoreticalstudy are found in his "Problems of
InferenceandProofin ParticipantObservation,"
(1958).
6This is not to say that all case studies are as
describedhere. Medical "write-ups" for example are very different. But these characteristics
are commonly expected and little differentthan
those specifiedby LouisSmith (1973), for example, to be: credible, holistic, particularistic,individualizable, process-oriented, ego-involving,
and blending of behavioraland phenomenological methodologies.
7In Julian Simon, (1969), for example.
MacDonaldandRobWalkerhavemade
"8Barry
the strongestcase I know for using idiosyncratic
instancesto createunderstandingof moregeneral
matters, as in "Case Study and the Social
Philosophy of EducationalResearch" (1975).
91twould be of interestto get empiricaldataon
the perceived utility of case studies. It can be

February1978

This content downloaded from 189.125.96.10 on Sun, 06 Dec 2015 20:18:01 UTC
All use subject to JSTOR Terms and Conditions

presumed,I fear,thatsomerespondents,
having
heardobjectionsto the case studymethodfrom
such authoritiesas JulianStanleyand Donald
Campbellandthinkingmoreof politicalvalue
thaninformational
their
value,wouldunderrate
andgeneralization.
utilityfor understanding
References
Bacon,SirFrances.NovumOrganum.1620.
Becker,HowardS. Problemsin the publication field studies.In ArthurJ. Vidich,Joseph
Bensman, and Maurice R. Stein (Eds.)
Studies.New York:
Reflectionson Community
JohnWiley, 1964, p. 273.
of inferenceand
Becker,HowardS. Problems
proof in participantobservation.American
SociologicalReview,1958,59, 652-660.
Blake, William.Annotationsto Sir Joshua
Reynold's"Disclosures"1808.
- comBohm,David.Scienceas perception
In F. Suppe(Ed.),TheStructure
munication.
of
ScientificTheories.Urbana:Univ. Ill. Press,
1974.

Broudy, HarryS. The life uses of schooling as

a field for research. In L. G. Thomas (Ed.),
Philosophical redirection of educational research. 71st Yearbookof NSSE, PartI, Chapter

IX, 1972.

Butterfield, Sir Herbert.History and human

relations. London: Collins, 1951.
Campbell, Donald T., and Stanley, Julian C.
Experimentaland quasi-experimentaldesignsfor
research. Chicago: Rand McNally, 1966.
Dilthey, Wilhelm. The constructionof the historical world of the human studies (Der Aufbauder Welt in den Geisteswissenschaften.
1910)GesammelteSchriftenI-VIILeipzig:B. G.
Teubner, 1914-1927.
Dray, William H. Laws and explanation in
history. Oxford Univ. Press,1957.
Hampel, Carl G. The functionof generallaws
in history. Journal of Philosophy, 39, 1942.
Kemmis, Stephen. An ecological perspective
on innovation. Urbana:Univ. of Ill. College of
Education, 1974. (mimeo)
MacDonald, Barry, and Walker, Rob. Case
study and the social philosophy of educational

Research

for

research. Cambridge Journal of Education, 5,

No. 1, 1975.
Ortony, Andrew. Knowledge, language and
thinking.Urbana:Univ. of Ill. College of Education, 1975. (mimeo)
Polanyi, Michael. Personal knowledge. New
York: Harper& Row, 1958.
Popper, Karl. The poverty of historicism.
London: Routledge and Kegan Paul, 1957.
Schon, Donald A. Metaphor and the social
conscience. Paperdeliveredat the Conferenceon
Metaphorand Thought,Univ. of Ill., September
1977.
Simon, Julian L. Basic research methods in
vocialscience: Theartof empiricalinvestigation.
New York: RandomHouse, 1969.
Smith, Louis. An aesthetic education workshop for administrators:Some implicationsfor a
theoryof case studies. Paperpresentedat AERA
Chicago, 1974.
Tolstoy, Leo. War and Peace. 1869.
Von Wright, Georg Henrik.Explanationand
understanding. London: Routledge & Kegan
Paul, 1971.

Action'

GERARD PIEL
Publisher, ScientificAmerican

the pecking order of academia,

Ipuren research
out-ranks applied. Re-

search into education must fall, therefore, intothe secondor lowerclass. Not
a disciplinein its own right- I hope, at
least, it does not pretendto such status
- education researchis the object of
the attentionsof the establisheddisciplines: psychology in all its varieties,
sociology, anthropology,demography,
statistics and, nowadays, economics.
Because our universities are departmentalized by discipline and the
scholar's career in each department
turns on contributionto its parochial
interests, it is not surprisingthat much
educationresearchis addressedto the
archivesof the disciplinesandnot to the
improvementof education.
The notionthattherearetwo ways to
do science, andthe unequaldistribution
of rewardsbetweenthem, thustogether
work a disservice to education and,

more often than not, to the disciplines

as well. This outcomewe mustdeplore.
Educationis the routeto salvationfor a
secularsociety. It should be carriedon
with deeper understandingand with
nagging concern for its improvement.
To this end there is but one way to do
research, and that is to do it well.
Having now used the value-laden
word "well," I mustdeclaremy bias. I
should do so because, in addition, I
shall commenthere aboutthe biases of
others. My romanceis with education
in the 18th Centurysense of this enterprise. By education, society promotes
the perfectibilityof the individual.It is
the mission of our schools not to shape
the citizen to the uses of society, but to
nurturecitizens capableof shapingsociety to betterpurposes. My test of the
worthof a piece of educationresearch,
therefore, is whether it promotes or
hindersthatenterprise.If it is done well
by that criterion,then'researchfor ac-

tion will also, by definition, pass muster in the discipline that has been
brought to bear upon the illumination
and improvementof education.
It is difficult to do social science
well, whetherpureor applied, in a culture as dominated as ours is by the
pragmaticsuccesses of the naturalsciences. The social sciences did not retain for long the innocent confidence
with which they were launchedin the
19th Centuryby Auguste Comte. The
Founderplacedthem at the pinnacleof
the scientificendeavor.In our country,
social scientists began to be elected to
the NationalAcademyof Sciences only
a little morethana decade ago; they are
still defendingtheirstatusas scientists.
The second class citizenship of social scientists causes them to try to
make social researchlook as much as
possible like the work of the physical
and biological sciences. The extreme
case is given by the distinguished

ER

This content downloaded from 189.125.96.10 on Sun, 06 Dec 2015 20:18:01 UTC
All use subject to JSTOR Terms and Conditions