Longitudinal Measurement Invariance of the Parenting Sense of Competence (PSoC): Evidence to Question Its Use?

Abstract

Background: This study investigated the factor structure of the parenting sense of competence (PSoC), a measure of parenting self-efficacy, in a sample of parents recruited when their infants were under 2 months old. Due to the lack of longitudinal analysis of the PSoC's factor structure over time, the study sought to establish if the published two-factor structure was consistent over an 18-month period.

Methods: Data collected from 536 parents who had participated in a randomised controlled trial of universal proportionate parenting support, delivered in five sites in England, were subject to confirmatory factor analysis (CFA).

Results: CFA revealed that a three-factor model was the best fit for the data. Longitudinal measurement invariance testing examined the stability of the three-factor model across an 18-month period. The results suggest that while the PSoC appeared to have configural variance, the metric and scalar variance were not supported. PSoC may be unstable across time and might be unreliable as a measure of parenting competence in parents of infants.

Conclusion: These findings are particularly salient for researchers and clinicians who are utilising the PSoC as a measure of change in routine practice or as part of evaluations of interventions. Further investigation of individual items is needed to refine the PSoC and improve its psychometric validity. Additional analyses are also needed to establish the invariance of the measure across different groups (age, sex, ethnicity and socioeconomic status).