Psychometric evaluation of the nursing child assessment teaching (NCAT) scale in a community-based sample.

The effect of the mother-child relationship on long-term child development has received research attention for decades. Because the quality and quantity of mother-infant interactions have been established as important predictors of the child's development, early identification of areas in the relationship requiring support and intervention is essential for promoting positive child outcomes. The Nursing Child Assessment Teaching Scale (NCAT) is an instrument long used to quantify the quality of interaction in the mother-child relationship during the first 36 months of a child's life. While the NCAT has been shown to be a reliable and valid instrument, limited evidence exists of the theoretical congruence between the Barnard Model it is based on and the NCAT scale. The psychometric properties of the NCAT scale and subscales were examined using item response theory in relation to characteristics of interactions in the Barnard Model using data collected during a clinic visit at 12 months of infant age in a sample of mothers and children (N = 1,121 dyads) from a community-based sample in Shelby County, TN. In this secondary analysis, descriptive statistics, reliabilities, and factor loadings for the NCAT were obtained using confirmatory factor analysis and augmented to form multiple indicators, multiple causes models, linking demographic predictors of the mothers and children to the NCAT subscales. Results supported scale abbreviation and theoretical congruence with the Barnard Model, which may provide researchers and practitioners with a more concise, reliable way of measuring maternal-child interaction in community settings.