[Psychometric properties of the French version of the How I Think Questionnaire].

UNLABELLED Certain research has pointed out the relative importance of cognitive distortions in the development of antisocial behavior. Distortions of social cognitions that facilitate aggression and other types of antisocial behavior have been described as self-serving cognitive distortions (Barriga et al., 2001 [2]). Considering the importance of the assessment of delinquent cognitive distortions, an instrument has been validated to measure self-serving cognitive distortions: the How I Think Questionnaire (HIT-Q, Barriga et al., 2001 [2]). Thus, the aim of the present study was to evaluate the convergent, discriminant and concurrent validity of a French version of the HIT-Q and its four dimensions (self-centered, blaming other, minimizing/mislabeling and assuming the worst). METHOD A sample of 972 French high-school students completed the Youth Psychopathic traits Inventory (YPI ; Andershed et al., 2002 ; Andershed et al., 2007 [26,27]) used to evaluate the convergent validity of the HIT-Q. To investigate discriminant validity, participants also completed the Sociomoral Reflection Measure-Short Form (SMR-SF ; Gibbs et al., 1992 [28]) and the Interpersonal Reactivity Index (IRI ; Davis, 1983 [29]) used to assess cognitive and affective empathy. To measure antisocial behavior, the French versions of the Self-Reported Delinquency Questionnaire (SRDQ ; Le Blanc and Frechette, 1989 [22]), the Antisocial Behavior Scale (ABS ; Schawb-Stone et al., 1996 [23]), the Self-Reported Delinquency Behavior (SRDB ; Elliott and Menard, 1996 [24]) and three items of the Sexual Experiences Survey (SES ; Koss et al., 2007 [25]) were used. Two samples were composed based on the same socio-demographic (age and gender) and socio-economic characteristics and ethnic background. French males scoring in the upper quartile on the antisocial behavior total score were classified in the antisocial group (n=135) and the rest of the sample in the non-antisocial group (n=306). Convergent and discriminant validity was determined using Pearson coefficients of correlation. One-way analyses of variance were used for mean scores comparisons. Regression analysis was used to evaluate the relative contribution of self-serving cognitive distortions to the prediction of antisocial behavior. RESULTS The French version of the HIT-Q showed acceptable reliability and validity and also a satisfying convergent and discriminant validity. The HIT-Q and its dimensions were positively and highly correlated to psychopathic traits (r=0.50 to r=0.61, p<0.05) ; these findings suggested a good convergent validity. The correlation between self-serving cognitive distortions and social moral reasoning (r=-0.23 to r=-0.28, p<0.05) and empathy (r=-0.13 to r=-0.20, p<0.05) appeared to be negative and low ; this relationship was conclusive and showed a satisfying discriminant validity. Mean comparison showed that participants in the antisocial group reported higher scores of self-serving cognitive distortions and its dimensions than the non-antisocial group. Self-serving cognitive distortions significantly predicted antisocial behavior (β=0.58, SE=0.02, p<0.001) and explained 34% of the variance. DISCUSSION The study of the convergent validity of the HIT-Q with the YPI as external criteria showed they were closely linked, which suggests self-serving cognitive distortion may be the cognitive expression of psychopathic traits. The discriminant validity of the HIT-Q with moral reasoning and empathy was satisfying and similar to previous results (Lardén et al., 2006 [20]). In concordance with previous studies, delinquent cognitive distortions appeared to be a significant predictor of antisocial behavior and moreover, participants in the antisocial group reported higher scores of self-serving cognitive distortions (Capuano, 2007 ; Barriga and Gibbs, 1996 ; Nas et al., 2008 [10-12]). CONCLUSION The present study showed that the French version of the HIT-Q presents good psychometric properties.