An empirical examination of contemporary methods for assessing social competence in urban Head Start children
Abstract
The cornerstone of Head Start's continued development is the establishment of a social competence assessment technology which is sensitive to low-income minority children. The purpose of the present study was to empirically examine the usefulness of four widely-used measures of social competence, representing major assessment methods, with these children. A large sample of children was recruited from an urban Head Start program over a two-year period. Three measures, the Pictorial Scale of Perceived Competence and Social Acceptance (PSPCSA) and the Social Skills Rating System (SSRS), parent and teacher versions, were evaluated to determine the extent to which they are reliable and valid for this sample. Factor analyses did not identify meaningful constructs for the PSPCSA and therefore, the use of this measure with this population is not recommended. For both versions of the SSRS, factor analyses yielded psychometrically adequate and meaningful dimensions. However, these dimensions differed from those published by the test authors. The fourth measure, sociometric procedures, was examined to determine the relationships among peer ratings and nominations. Correlations showed a strong relationship between highest and lowest possible rating variables but did not demonstrate anticipated relationships with highest and lowest peer nomination scores or between the two sociometric procedures. The relationships among measures evidencing construct validity for this population were examined by three statistical processes: zero-order correlations, factor analyses, and canonical analyses. Measures included the SSRS, parent and teacher versions, and sociometric procedures. Findings from the factor analyses show four distinct dimensions of social competence: Teacher-Perceived Social Skills, Teacher-Perceived Behavior Problems, Parent-Perceived Behavior Problems and Popularity. Parent ratings of social skills and peer nominations did not contribute to this final solution. Supporting the final factor structure, the correlation matrix and canonical analyses showed few significant, but negligible, relationships between measures. Many of the expected relationships were not found. Results are related to previous research and implications for Head Start assessment are discussed.
Subject Area
Developmental psychology|Psychological tests
Recommended Citation
Manz, Patricia Holliday, "An empirical examination of contemporary methods for assessing social competence in urban Head Start children" (1994). Dissertations available from ProQuest. AAI9427576.
https://repository.upenn.edu/dissertations/AAI9427576