The Establishment Of Empathy Semi-projective Testing Tool（ESP-MTTI）

Objective:This study intends to establish a semi-projective empathy testing tool with Chinese textbooks’ illustrations of elementary school as raw materials.Methods:First we evaluated the 78 pictures’ empathy valences(emotional,narrative and arousal). Second we got each picture’s characteristics(background environment, active contents, emotion) by open investigation and formed specific items on the basis of those characteristics. Third we got the final version of semi-projective empathy testing tool(ESP-MTTI) after 3 rounds of testing and screening. Last we tested the reliability and validity of the final version, calculating its test-retest reliability, coefficient of internal consistency, split-half reliability and employing Interpersonal reactivity index-C(IRI-C) and Basic empathy scale(BES), Adolescent prosocial behavior scale(PTM), Toronto alexithymia scale(TAS-20) and General social desirability questionnaire(SDR) as the criterion.Results: This scale is consisted of 64 items. Its test-retest reliability, split-half reliability and Cronbach ’ s αcoefficient are 0.857, 0.910 and 0.968 respectively. Confirmatory factor analysis with Amos17.0 shows there existsonly one dimension, named “implicit attention bias of others’ emotions”, of the 64 items.The χ2/df, RMSEA and PGFI are 2.201,0.041 and 0.736 respectively; This scale significantly correlates with IRI-C’s total score, dimension PT, EC at 0.01 level and correlation indexes come to be 0.270、0.260 and 0.300, respectively. However it doesn’t significantly correlate with dimension FS and PD; This scale significantly correlates with BES ’ s total score, dimension AE, CE at 0.01 level and correlation indexes come to be 0.381,0.345 and 0.243, respectively;This scale significantly correlates with PTM’s total score, dimension “anonymous”, “altruistic”, “urgent” at 0.01 level and correlation indexes come to be 0.204, 0.171, 0.187 and 0.198,respectively.However it doesn’t significantly correlate with dimension “ public ”, “ compliable ”, and “emotional”;This scale doesn’t significantly correlate with TAS-20’s total score, dimension “lack of ability to identify the emotion ”, “ lack of ability to describe emotion ”, “ externally oriented thinking ”;This scale doesn ’ t significantly correlate with SDR’s total score, dimension IM, SDE; Female’s score(M±SD=240.52±32.723) is significantly higher than male’s(M±SD=227.55±36.708), t=4.079, p<0.001. Normal male undergraduates’ score(M±SD=227.39±37.929)issignificantly higher than male violent crime’s(M±SD=217.64±31.318),t=2.178, p<0.001. Empathy score got from this scale can significantly predict prosocial behavior tendency. The initial regression equation is Y(predicts prosocial behavior tendency)=0.065X(empathy)+77.059. Standardized regression equation is Y(predicts prosocial behavior tendency)=0.203X(empathy),R2=0.041,F=10.122,t=3.182,ps<0.01.Trait empathy influences prosocial behavior tendency by complete mediation effect of state empathy. Mediation effect of state empathy share 51.8% cumulative proportion of total effect and 13.7% variance variation of prosocial behavior tendency.Conclusion:(1)empathy semi-projective testing tool(ESP-MTTI) consists of 64 items, and only one dimension, named “implicit attention bias of others’ emotions”;(2)The scale has good reliability and validity and reaches the basic requirements of psychometrics;(3)Trait empathy can predict prosocial behavior tendency and influence prosocial behavior tendency by complete mediation effect of state empathy.