Assessing Whether Negative Parenting Cognitions Bias Parent Report of Preschoolers' Externalizing Symptoms: A Regularized Moderated Non-Linear Factor Analysis Approach.

Abstract

Parent report is frequently used to assess children's psychopathology, however, researchers have expressed concerns about the validity of parent reports. Some parental characteristics, attitudes, or beliefs may systematically bias a parent's report of their child's behaviors and functioning. Informed by social information processing models of parenting risk, parents with more difficulties reflecting on and interpreting their children's behavior may be at risk for less accurate reports. The present study conducted two regularized moderated non-linear factor analyses with LONGSCAN data to examine how parents' self-reported negative parenting cognitions were associated with structural parameters of parent-reported child aggression and attention problems. While differential item functioning (DIF) was present on the aggression and attention problems subscales as a function of negative parenting cognitions, the DIF was small in magnitude, inconsistent in directionality and did not significantly alter factor-level parameters. Negative parenting cognitions did demonstrate a small but significant negative impact on all latent externalizing factors (aggression and attention problems), such that caregivers with fewer negative parenting cognitions endorsed fewer items, and this was associated with a lower mean of each latent variable. Given that accounting for DIF did not contribute to meaningful differences in impact parameters or improve criterion validity, findings suggest that the aggression and attention subscales are functionally invariant to negative parenting cognitions, suggesting that externalizing symptoms can be reliability compared across parents of varying parenting cognitions.