Measuring mothers' representations of their infants: Psychometric properties of the clinical scales of the working model of the child interview in a low- to moderate-risk sample.

The Working Model of the Child Interview (WMCI) is frequently used to measure parents' representations. Beyond the global categories (balanced, disengaged, distorted), the reliability, factor structure, and validity of all the 15 clinical scales have not previously been studied. The WMCI was administered to 152 Norwegian mothers of infants (mean age = 7.3 months) recruited from community well-baby clinics. Interrater reliability was adequate for the global categories and moderate for the clinical scales. Exploratory factor analysis and confirmatory factor analysis yielded three factors with evidence of factorial validity: Factor 1 balanced; factor 2 resentful; factor 3 apprehensive. Factor 1 corresponded with the original category balanced, while factor 2 and factor 3 corresponded with the original category distorted. Concurrent validity was supported as mothers with balanced representation (factor 1) were less stressed and the mother-infant interaction was more positive than that of mothers with resentful representation. Mothers with resentful or apprehensive representations (factor 2 and factor 3) reported more stress. The extracted factors and demographic variables correlated weakly or not at all, confirming discriminant validity. Our findings show that the clinical scales of the WMCI can be used in research with low- to moderate-risk samples.