The use of Piers -Harris Children's Self -Concept Scale to measure the multidimensional structural model of self-concept for children in second grade
Research in the past three decades has resulted the consensus in that self-concept is multidimensional for middle-childhood children and adolescents. However, the multidimensionality of self-concept has not been not well-studied and accepted for young children. Some researchers argue that young children do not develop general self-concept and are unable to differentiate domains of self-concept while other researchers believe that young children can differentiate different aspects of self-concept (Harter, 1983; Marsh, Craven & Debus, 1991). The controversy is partly due to a lack of research on young children and a lack of theoretically based self-concept instruments appropriate for young children. This study investigated the use of the recently published Piers-Harris Children's Self-Concept Scale-Second Edition (Piers & Herzburg, 2002) with second graders and explored the multidimensional structural model of self-concept for second graders.
The study involved 644 second graders and 586 third to sixth graders. Results of this study indicated excellent reliability for the Piers-Harris 2 Total Score and good reliability for the subscales for both the second graders and the older children. The existence of a general self-concept was suggested by the high correlations between the subscale scores and the Total Score. The multidimensionality of self-concept was strongly supported by correlations between the subscales and the confirmatory factor analyses results. The second graders had slightly better goodness-of-fit indexes for the six-factor model of the Piers-Harris 2 than the older children had. Correlations between academic and behavioral measures and the Piers-Harris 2 scores also provided evidence to support the multidimensional structure of self-concept. Academic achievement measures such as Math and Reading grades and the ITBS Composite were not correlated with non-academic self-concept measures such as Physical Appearance and Attributes. The Classroom Behavior Rating Scale scores were significantly correlated with the Behavioral Adjustment and the Intellectual and School Status subscales, and not significantly correlated with other subscales. Home Behavior Rating Scale scores, however, were not significantly correlated with the Behavioral Adjustment scores, which was not expected. Reading grade and ITBS Composite score were not significantly correlated with the Intellectual and School Status subscale. Possible explanations were discussed.
0620: Developmental psychology
0632: Psychological tests