How cluster B-related personality traits in mothers shape their ratings of their adolescents' personality: a multisource approach using the Trait-Reputation-Identity model.

BACKGROUND

Parents are often considered reliable reporters of their child’s personality, and are typically the primary informants in youth assessments. However, little is known about how parents’ own personality trait pathology influences their evaluations, raising concerns about potential bias in these assessments. This study investigates the impact of Cluster B DSM-5 maternal borderline, antisocial, and narcissistic trait constellations on the agreement between late adolescent self-reports and maternal reports of adolescent personality.

METHOD

Relying on a Belgian sample of 597 adolescent-parent dyads (i.e.,  = 1,194 participants), who both filled out the NEO PI-R to evaluate the adolescent’s personality, the Trait-Reputation-Identity (TRI) model was applied to disentangle shared and unique perspectives on adolescent personality. Bifactor modeling was used to partition variance into shared perceptions between mothers and their adolescent (trait component), unique maternal perspectives (reputation component), and unique adolescent self-perceptions of personality (identity component). Maternal trait configurations were assessed using the PID-5, and modeled as predictors of the TRI components to examine their impact on rater agreement and discrepancy patterns.

RESULTS

The findings showed a significant role of maternal personality trait constellations in parent-child agreement, although results diverged across configurations. With increasing maternal antisocial and borderline trait elevations, the personality information on the adolescent provided by mothers and not shared by the adolescent’s self-view becomes larger. In addition, adolescents who grow up with a mother higher in borderline traits, appear to demonstrate a larger component of unique personality information that is not observed by their mothers. In mothers with higher narcissistic trait elevations, no effects were found on unique informant variance, although the overall consensus on the adolescent’s personality generally decreases.

CONCLUSION

These findings offer key points of attention regarding sources of informant bias related to the informant’s own personality, and will be discussed from an applied perspective.