The structure of behavior and the eye of the beholder : rater-gender by ratee-gender interactions

Abstract

Interactions between mothers and offspring is viewed to differ significantly when compared with the father's. Present literature on parenting differences comprise few studies statistically comparing these disparities, although replete with inferences that there are differences. The purpose of this study was to lay a foundation and determine if statistically-based differences between the latent structures of behaviors for offspring, when rated by mothers and by fathers, exist. The following research questions were proposed: Does a mother's or a father's report of a child's behavior produce a different latent structure, based upon the child's gender? Does a mother's or a father's report of an adolescent's behavior produce a different latent structure, based upon the adolescent's gender? And, is the report of the latent structure of behavior different between that reported by mothers and fathers for their child or adolescent, regardless of gender? Using the national standardization sample from the BASC's Parenting Rating Scales (n=4385), the sample was split into child and adolescent groups, then divided into eight subgroups based upon both child's and parent's gender. Eight principal axis factor analyses with varimax rotation were then completed. Three separate factor invariance techniques were applied to determine the similarity of the resulting factor matrices. The results from the comparison of the underlying structures demonstrated that parents agree upon the concept of positive social skills. Fathers and m others agree on what constitutes undesirable behaviors with their daughters, but disagree upon this concept with their sons. In most other areas unique underlying structures of behaviors were found. These impact a test's construct validity and question the relevance of the scales when the parent's gender is disregarded. This research concludes that behavioral reports should add a third dimension (parent's gender) to the existing structure using only a child's gender and developmental group.