Inthisstudy,theobjectoflearning (whatwastobelearned) wascomparing (high) scoresondifferentscalesfromdifferentdistributions,:“whichoftwostudents,whowereindifferentbiologyclassesandtookdifferenttests,didbetterontheirrespectivetest?” Suchproblemsarecommonlysolvedinstatisticsbyusingstandardizedscores (i.e.,bydividingthedifferencebetweentheactualscoreandthemeanforthedistributionthatthescorebelongstobythestandarddeviationforthesamedistribution).Inordertograsptherationaleunderlyingthistransformation,theninthgradestudentsparticipatinginthestudyhadtonoticetwocriticalrespectsinwhichdistributionsmightdifferfromeachother,namelycentraltendencyandvariability.［TherearesimilaritieswiththechildreninJudd's (1908) experiment,whohadtonoticetheangleofrefractionandthedepthofwaterastwocriticaldifferencesbetweenthetwotaskstheyweredealingwith.］
Afterhavingdealtwithstatisticsfor2weeks (allinthesameway),thestudentswerepresentedscenarios,suchasthefollowing:“whobroketheworldrecordbythemostimpressiveamount—JohninhighjumporMikeinjavelinthrow?” Inoneoftheinstructionaltreatments (SituationA1), (SituationA2),-twoworked-outexamples (SituationB). (e.g.,“IsBettybetteratassistsorsteals?”).Finally,allofthestudentsreceivedthetargettransferproblem,namelytocomparescoresontwodifferentbiologytests (SituationC).Inthiscasenotalltherawdatawereprovided,butonlythemeansandstandarddeviationsforthetwotestsalongwithrawscoresforthetwostudentstocompare.
