Construct Validity and Cross Validity of a Test Battery Modeling Autism Spectrum Disorder (ASD) in Mice.

Modeling in other organism species is one of the crucial stages in ascertaining the association between gene and psychiatric disorder. Testing Autism Spectrum Disorder (ASD) in mice is very popular but construct validity of the batteries is not available. We presented here the first factor analysis of a behavioral model of ASD-like in mice coupled with empirical validation. We defined fourteen measures aligning mouse-behavior measures with the criteria defined by DSM-5 for the diagnostic of ASD. Sixty-five mice belonging to a heterogeneous pool of genotypes were tested. Reliability coefficients vary from .68 to .81. The factor analysis resulted in a three- factor solution in line with DSM criteria: social behavior, stereotypy and narrowness of the field of interest. The empirical validation with mice sharing a haplo-insufficiency of the zinc-finger transcription factor TSHZ3/Tshz3 associated with ASD shows the discriminant power of the highly loaded items.