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

12明星大学心理学年報2014年第32号groups.Nevertheless,attitude scores obtained werealmost comparable with each other, and there wasno significant effect ofgroup.Theseresults indicatethat the SECT does not depend on specific fillerpictures.That is,emotionallypositive and negativefiller pictures different from those used in the presentstudy could serve as appropriate filler picturesin the SECT. Thus, results indicate the practicalutility of the SECT. Next, Experiment 3 examinesthe possibility of artifacts in the SECT.Experiment 3Attitudescores obtained in theSECT in theprevioustwo experiments could significantly predictpleasantness ratings. However, it is possible thatattitude scores obtained in the two previous experimentswere caused by an artifact dependent uponthe three target pictures being specific because thetarget picturesusedintheprevioustwoexperimentswere identical. For this reason, the present experimentassigns different sets of three target picturesto two groups of participants with the four fixedfiller pictures.Thepurposeofthepresent studyistoexaminewhether theSECT appropriatelymeasuresattitudes toward target pictures that differ from theprevious two experiments.MethodParticipants. As a part of an experimental psychologylaboratory course, 39 Japanese universitystudents (17 men, 22 women, average age＝20.00years, age range:19-22)participated in the experiment.Half of the participants were selected randomlyand assigned to Group C and the remainderassigned to Group D.Materials. As target pictures, two emotionallypositive pictures, two emotionally neutral pictures,andtwoemotionallynegativepictureswereselectedfrom theIAPS (Lang et.al,2005).Thesesix picturesdid not relateto sociallydelicatematters and differedfrom the three target pictures used in the previoustwo experiments (Appendix).From the six pictures,oneeachofemotionallypositive,neutral,andnegative pictures were selected and used as thetarget pictures for Group C. The remaining threepictures were used as target pictures for Group D(Appendix). For both groups, four filler pictureswere used,identical to those in Group B of Experiment2.Procedure.Fiveto seven students participated atthe same time in the same laboratory as in theprevious experiments. Participants completed theSECT procedure and rated the pleasantness ofseven pictures as in Experiment 2.Theentireexperimentalprocedure took approximately 20 minutes.ResultsNo participants showed erroneous responses ofmore than 20% in all trials. Erroneous responseswereexcluded from theanalysis(9.56% ofall observations).Additionally,responsesthat showed shortlatencies ofbelow 300ms and long latencies ofover1500 ms were excluded as outliers (1.99% of allobservations). Table 7 shows the mean responselatency and number of errors for each of the sixmain blocks. For each participant, three attitudescores for three target pictures were obtained bysubtracting the mean response latencyfor the positiveblock from the mean response latency for thenegative block as in the previous experiments(Table 7). Magnitude relationships for threeobtained attitude scores in both groups were matchedwith the original valence of the three targetpictures. To investigate these attitude scores, a 2(group:C,D)×3(original valenceoftarget picture:positive, negative, neutral) mixed-design ANOVAwas conducted with attitudescoreas thedependentvariable. Group was the between-subjects factorand the original valence of target pictures was thewithin-subjects factor. Results revealed a significantmain effect of group (F(1, 37)＝4.46, p＜.05,η?＝.03).The main effect of the original valence oftarget pictures was also significant (F(2,74)＝32.80,p＜.001,η?＝.31).Multiple comparison using Ryan’smethod revealed significant differences between allcomparablepairs ofthreeattitudescores.Finally,aone-way interaction effect was significant (F(2,74)＝5.90, p＜.01,η?＝.06). Tests for simple maineffects revealed a significant simple main effect ofgroup on neutral targets (F(1, 111)＝12.14, p＜.001,η?＝.10). This effect indicates that participants inGroup C showed significantlylower attitude scores