ブックタイトル明星大学 心理学年報 第32号

概要

明星大学 心理学年報 第32号

2明星大学心理学年報2014年第32号“insect/pleasant”on the left and“flower/unpleasant”onthe right. The difference between the twoaverage response latencies represents the participant’srelative attitude toward flowers and insects.The prominent feature of the IAT is that participantsdo not report their attitudes directly; therefore,itis difficult for them to distort their attitudeswith the IAT. For example, Banse, Seise, andZerbes (2001) reported that participants instructedto disguise themselves as having positive attitudestoward homosexuality showed more positive attitudescores toward homosexuality on a self-reportmeasure relative to a control group, but theyshowed attitudes toward homosexuality almostequal to the control group in the IAT. Similarly,Kim (2003) reported that participants instructed todisguise their attitudes on two IAT measuresshowed attitude scores almost identical to those ofcontrol participants. These reports imply that theIAT is resistant to social desirability bias to someextent.The IAT is limited in that it is used to measurerelative attitudes toward two target concepts, nottowardspecificandconcreteattitudinal objects.Forexample, the IAT measures attitudes toward conceptssuchas“flower,”butcannot measureattitudestoward a picture of a flower appearing before aparticipant. In this regard, an emotional primingparadigm has been used to measure participants’attitudes toward specific words and small pictures.In an experimental trial of typical emotional priming,aprimestimulusappearsonascreenfor200ms.Following an interval of 100 ms, an emotionallypositive or negative word appears on a monitor.Participants judge the emotional valence of thetarget (positive or negative) and press the correctkey as quickly and precisely as possible; theirresponse latencies for the target stimuli are recordedin ms. Many researchers have reported thataverage response latencies on trials in which theprime and the target have the same emotionalvalence are significantly shorter than on trials inwhich the prime and the target have inconsistentemotional valence (Bargh, Chaiken, Govender, &Pratto, 1992; Fazio, Sanbonmatsu, Powell, &Kardes,1986;Hermans,DeHouwer,& Eelen,1994).Therefore, by assigning a picture of a concreteattitudinal object as the prime stimulus, the paradigmcould be used as a measure of indirect attitudestoward that picture. That is, the primingscore,obtainedbysubtracting responselatenciesfora positive target following a prime stimulus fromresponse latencies for a negative target followingthe same prime stimulus,is considered to be representativeof indirect attitudes toward the primestimulus(Fazio,Jackson,Dunton,& Williams,1995;Hayashi,2007;Hermans,Vansteenwegen,Crombez,& Baeyens, 2002).However, it would be difficult to measure attitudesusing an emotional priming paradigm whentheattitudinal object is a complicated picturecoveringa largeproportion ofa participant’s visual field.In such cases,theentirepicturemaynot beprojectedonto theparticipant’s central fovea under a shortpresentation duration of200ms.In particular,whenimportant components are located in a peripheralarea of the picture, measuring attitudes using anemotional priming paradigm would be difficult.Therefore, prime pictures used in the emotionalpriming paradigm have been limited to relativelysmall pictures offaces and words.Inadditiontothisproblem,the emotional priming paradigm is knownto demonstrate extremely low reliability scores.Olson and Fazio (2003)reported that the emotionalpriming paradigm as an indirect attitude measureshowed r＝.06on a split-halftest.This is apparentlylower than the reliability of IAT tests, whichhave an internal consistencycoefficient ofnearly.7(e.g.,Bosson,Swann,& Pennebaker, 2000).Thus,it isdifficult toclaim that social psychologyhas developed an effective indirect method to measureattitudes toward various pictorial stimuli.Meanwhile, measuring attitudes toward pictorialstimuli using a reliablemethod that is unaffected bysocial desirabilitybiaseswouldbeuseful,particularlywhen research is related to delicate social matters.