Generalization of the Child Observation Record: a validity study for diverse samples of urban, low-income preschool children

The present investigation addressed the construct validity of the Child Observation Record (COR) with low-income urban preschool children. From two separate samples representing low-income preschool children, COR ratings were analyzed using multivariate techniques. Independent analyses from these two urban sites yielded a three-dimensional structure: Cognitive Skills, Social Engagement, and Coordinated Movement. Further analyses cross-validated this structure for males and females and across ethnic groups. Concurrent assessments provided convergent and discriminant validity for the Social Engagement dimension and convergent validity for Cognitive Skills dimension. Analyses of item distributions of the 5-point developmental sequences represented by the 30 COR items were used to examine the assumption that all the distributions were continuous unimodal distributions. Findings did not universally support this assumption revealing some irregular distributions with troughs in the mid-range of continua. Implications of the findings for early childhood assessment of vulnerable children and future research were discussed. (C) 2002 Published by Elsevier Science Inc.