Fiabilité et validité de l’Échelle de compréhension sociale des enfants

Abstract

This manuscript presents data on the reliability and validity of a French translation of the Children's Social Understanding Scale (CSUS; Tahiroglu et al., 2014), a parent-report scale assessing theory of mind in young children. In a first study, parents of 382 typically developing children (34–82 months; M=55 months; 193 males) completed an online questionnaire (215 in English and 167 in French) including the CSUS or its French translation, l’Échelle de compréhension sociale des enfants (ÉCSE), as well as subscales of the short-form Children's Behavior Questionnaire (CBQ; Putnam & Rothbart, 2006) and demographic questions. Internal consistency of the CSUS/ÉCSE was high and similar between French and original English versions (French: long form α=.929, ρ=.933, n=99, short form α=.896, ρ=.901, n=128; English: long form α=.935, ρ=.936, n=100, short form α=.860, ρ=.869, n=146). Correlations between the mean CSUS/ÉCSE score and children's age in months were significantly positive and similar across languages, English: r=.478, P<.001, French: r=.569, P<.001, difference: Z=1.21, P=.226, ns. In both languages, the CSUS/ÉCSE correlated positively with CBQ subscales of attention (English: r(205)=.298, P<.001; French: r(165)=.353, P<.001) and inhibition (English: r(205)=.538, P<.001; French: r(165)=.421, P<.001). A confirmatory factor analysis constraining variances and covariances of the six CSUS/ÉCSE subscales to be equal between languages yielded a non-significant Chi2, χ2(33)=6.204, ns. In a second study, French-speaking parents completed the ÉCSE twice one week apart for 63 children (M child age: 50 months; 31 females). Test-retest reliability was high, r=.967. In a third study, 68 children (33–78 months; M=54 months; 30 males) were administered the four easiest tasks of a widely used behavioral theory of mind scale (Wellman & Liu, 2004) and their French-speaking parents completed the ÉCSE. Behavioral theory of mind from the Wellman & Liu (2004) scale and parent-report theory of mind from the ÉCSE were positively correlated (long form: r(68)=.350, P=.003; short form: r(68)=.364, P=.002), and these correlations held even after controlling for age (long form: r=.281, P=.021; short form: r(68)=.242, P=.049). Together, these findings support the reliability and validity of the French-language ÉCSE. Translated items are provided in appendix for use by researchers working with francophone populations.