A Brief Summary of Cultual Bias in Educational Assessments
Ste!anie "an#en ED$S %&' Uni(ersity of Cal)ary Dr* Meado+ Sc!roeder CULTURAL BIAS IN EDUCATIONAL ASSESSMENTS ' Cultural assessments may aear to ,e cultually un,iased in t!eir content- and indeed muc! effort !as ,een made in creatin) norm referenced tests as fair and e.uita,le for all indi(iduals* /et desite our ,et efforts- no test can ,e created- adated or mar0ed +it!out escain) t!e effects of culture* T!us +e can assume t!at e(ery test +ill ,e ,iased in fa(our of t!e eole from t!e dominant culture +!o desi)ned it* Because +e can1t escae it- +e must t!en ma0e our ,est efforts in adatin) t!ese tests for different cultures as +ell as reco)ni#in) our o+n ,iases so as to curtail su,2ecti(e results* T!is issue is esecially ertinent in countries li0e Canada and t!e United States +!ere minority )rous ma0e u muc! of our cultural landscae* T!e I3 test created ,y Binet +as t!e first norm4referenced test of its 0ind- and t!us +as used as a standard for many years* /et ,ecause t!e test +as created +it!in one culture- it +as in!erently ,iased to+ards t!e dominant )rou- t!at ,ein) ,ein) +!ite- male- and middle class* T!is test !as ,een used !ea(ily o(er t!e years to esta,lis! intelli)ence le(els- and +as t!ou)!t to ,e (ery relia,le desite its o,(ious fla+* In fact- rat!er t!an reco)ni#in) t!is as a ,ias- it +as instead used to esta,lis! racial- and socio4cultural sueriority* It +asn1t until 1565 t!at t!is test +as c!allen)ed and t!en ,anned in California for discriminatin) a)ainst ,lac0s and minority )rous 78is- 1555- )* 9%:;* Today- t!e I3 test is criti.ued as ,ein) inaroriate for assessin) a sectrum of cultures and lan)ua)es- and yet it is still +idely used* Ot!er tests !a(e since emer)ed t!at are more sensiti(e to cultural differences- yet t!e fact remains t!at e(ery asect of intelli)ence testin) is imacted ,y cultures* Accordin) to Mus!.uas! and Bo(a 7'::6; t!is includes <test content- test materials- !rasin)- directions and conditions- scorin)- test administrator- and test ,e!a(iour= 7)* &%;* T!us- t!ere is no test free from culture ,ias* >o+ t!en do +e attemt to try and assess indi(iduals in suc! multicultural laces suc! as Britis! Colum,ia +!ere some sc!ools !a(e an East Indian ma2ority- +!ile ot!ers are C!inese- or ?irst Nations@ T!e first ste is to reco)ni#e t!at cultural differences do eAist from one )rou to t!e neAt and t!at t!e concet of stnadardi#ed tests and muc! of t!e content may not ,e CULTURAL BIAS IN EDUCATIONAL ASSESSMENTS 9 aroriate for all )rous +!en measurin) for intelli)ence* 8is demonstrates t!is in statin) t!at 7for instance;- standardi#ed situations may not ,e otimal for some minority )rous* T!us t!ey may erform oorly +!ic! +ould comromise t!e results of t!e test and render it in(alid* Ot!er )rous demonstrate cultural (ariations in (alue 2ud)ements on t!in)s suc! as seed or accuracy +!ic! +ould imact t!eir results* Mean+!ile- some )rous may (alue as0in) for !el- +!ile +e +ould consider t!is c!eatin) 7'::6- )* 96%;* Anot!er eAamle is t!e Latino immi)rant community +!o laces em!asis on t!e concet of educacBon- yet t!eir interretation of t!is +ord is t!e ac.uisition of social s0ills and resectful and correct ,e!a(iour 78reenfield- 1556- )*1116;* It is unfortunate t!en t!at rat!er t!en ac0no+led)e and utili#e t!is interretation- +e instead eAect t!em to c!an)e and conform to our standard* In t!eir article Cross-cultural Assessment and Measurement Issues- Mus!.uas! and Bo(a 7'::6; eAamine cultural ,ias secifically +it!in t!e Canadian ?irst Nations oulation* Alt!ou)! t!is oulation is Canadian- t!ey do still come from a culture t!at is searate and uni.ue from t!e mainstream T!is t!en uts t!em at a disad(anta)e +!en ta0in) norm4referenced test* In fact- it is documented t!at ?irst Nations students score lo+er t!an non4?irst Nations students on t!e CISC ?ull Scale and Der,al Scale 7'::6- )* &6;* Alt!ou)! it mi)!t ,e easy to assume t!at t!ese students are deficient in t!ese areas- t!e test must ,e eAamined for any trace of cultural ,ias* Indeed- <t!is test re.uires secific 0no+led)e or eAerience +it! mainstream culture and does not measure t!e entire ran)e of cultural content of t!e indi(idual= 7'::6- )* &6;* T!us t!is test s!ould ,e edited for )rous suc! as t!eirs- or t!eir ans+ers must ,e +ei)!ed and interreted correctly in accordance +it! t!eir cultural )rou* In addition to t!ese o,stacles- findin)s su))est t!at it is !ard to assess ?irst Nations in an un,iased manner 7Mus!.uas! and Bo(a- '::6- )* &E;* Most li0ely +e !a(e already made 2ud)ements a,out t!em su,consciously +!ic! !as comromised our imartiality* T!erefore it is imerati(e t!at t!e CULTURAL BIAS IN EDUCATIONAL ASSESSMENTS F clinicians s!ould conduct a self4assessment of t!eir o+n ,iases and re2udices ,efore conductin) a test* Ot!er efforts at eliminatin) ,ias include determinin) a referred lan)ua)e and esta,lis!in) raort* T!is is esecially imortant as test erformance may suffer if t!e student feels uncomforta,le or can identify attitudes of distrust or re2udice from t!e administrator* Raort mi)!t also soot! ne)ati(e attitudes ,rou)!t ,y t!e student* ?or t!e student too could ,rin) +it! t!em attitudes of resentment and distrust from t!eir communities stemmin) from t!eir lon) !istory of su,2ection and assimilation under Caucasian rule 7'::6- )* &5;* T!e most effecti(e measure to eliminatin) ,ias for ?irst Nations +ould ,e to esta,lis! norms for t!e ?irst Nations oulations* Unfortunately t!is +ould ,e a (ery time consumin) and eAensi(e underta0in) and so is not li0ely to !aen soon* T!erefore- se(eral ot!er stes can ,e ta0en to lo+er t!e ,ias of t!ese test* T!e first otion is to select t!ose tests t!at !a(e t!e lo+est cultural loadin)s and lin)uistic demands* As t!is ro(ed to ,e t!eir +ea0ness in standardi#ed tests- it needs to ,e reco)ni#ed as a ,ias and eliminated as a criteria or relaced +it! somet!in) more aroriate 7Mus!.uas! and Bo(a- '::6- )* %1;* Anot!er otion mi)!t ,e to !a(e t!e assessment translated into t!eir dominant lan)ua)e* Alt!ou)! t!is seems easy enou)!- translatin) suc! i(otal information is not an easy tas0* In t!eir article Increasing the Validity of Cross-cultural Assessment- >am,leton and Gan2ee 7155&; note t!at translatin) suc! tests is a oular otion ,ecause it is more cost effecti(e and efficient t!an )eneratin) ne+ ones* /est t!ere are o,stacles and imortant factors in(ol(ed t!at need to ,e ta0en into account ,efore translatin) norm4referenced assessments* ?irstly- t!e assessment in .uestion must ,e tested so t!at t!ey can ,e sure t!at different )rous +ould ro(ide t!e same ans+er as t!e norm* T!at ,ein) said- it is ossi,le t!at a )rou could roduce a !i)!er or lo+er score t!an ot!er )rous* But +e need to ensure t!at any differences are not due to t!e failure of t!e test to ro(ide e.ui(alent scores 7)* 1F5;* Secondly- it is imortant t!at t!e translator is a rofessional rat!er t!an a relati(e or ac.uaintance* T!is is ,ecause t!ere is a s0ill in +ordin) a test- and not ro(idin) t!e ri)!t (oca,ulary CULTURAL BIAS IN EDUCATIONAL ASSESSMENTS & or usin) casual !rasin) could confuse t!e results* T!e translator also needs to 0no+ t!e culture in .uestion so t!at t!ey 0no+ t!e ,est translations to use rat!er t!an literal ones* ?urt!ermore- t!ey need to !a(e an understandin) of test construction* ?or instance- someone unfamiliar +it! standardi#ed tests could em,ellis! t!e ri)!t ans+er- t!us imactin) t!e (alidity 7155&- )* 1F5;* As so many sc!ools today are made u of many cultural )rous- it is imortant to reco)ni#e t!at standardi#ed tests are in!erently culturally ,iased* T!us if you c!oose to use suc! tests +it!out adation or modification- you must reco)ni#e t!at t!e (alidity may ,e comromised* Ot!ers mean+!ile recommend t!at +e sacrifice relia,ility for en!ancin) (alidity ,y usin) a more ,road and multi4met!od aroac! in )at!erin) data* ?or instance- rat!er t!an relyin) on one form of standardi#ed test only- it could also ,e accomanied +it! o,ser(ations- inter(ie+s- )rou +or0- and so fort!* T!e mo(e in t!is direction +ill ,e difficult- as t!e idea of !i)! sta0es norm4referenced testin) is deely en)rained in our culture- yet ,y usin) t!em +e are only uttin) our students at a disad(anta)e* E(en countries suc! as Sin)aore !a(e made a mo(e a+ay from !i)! sta0es testin) to more ,road4 ,ased analysis 78is- 1555- )* 9EF;* T!erefore- alt!ou)! inroads !a(e ,een made- t!ere is still muc! t!at can ,e done in uttin) all cultures on e.ual footin) re)ardin) norm4referenced tests* CULTURAL BIAS IN EDUCATIONAL ASSESSMENTS % References Cam,ell- T*- Dolla)!an- C*- Needleman- >*- H "anos0y- "* 71556;* Reducin) ,ias in lan)ua)e assessmentI rocessin)4deendent measures* "ournal Of Seec!- Lan)ua)e- And >earin) Researc!I "SL>R- F:79;- &154&'&* 8is- C* 71555;* Socio4Cultural Asects of Assessment* Review of Research in Education- 247:;- 9&&495'* doiI 1:*91:'J::5169'K:'F::19&& 8reenfield- $* M* 71556;* /ou can1t ta0e it +it! youI C!y a,ility assessments don1t cross cultures* American $syc!olo)ist- &'71:;- 111&411'F* doiI1:*1:96J:::94:%%K*&'*1:*111& >am,leton- R* G*- H Gan2ee- A* 7155&;* Increasin) t!e (alidity of cross4cultural assessmentsI Use of imro(ed met!ods for test adatations* Euroean "ournal Of $syc!olo)ical Assessment- 1179;- 1F641&6* doiI1:*1:'6J1:1&4&6&5*11*9*1F6 Mus!.uas!- C* "*- H Bo(a- D* L* 7'::6;* Cross4Cultural Assessment and Measurement Issues* Journal on evelo!mental isa"ilities- #$71;- &94%%* $oortin)a- /* >* 7155&;* Cultural ,ias in assessmentI >istorical and t!ematic issues* Euroean "ournal Of $syc!olo)ical Assessment- 1179;- 1F:41F%* doiI1:*1:'6J1:1&4&6&5*11*9*1F: