The Association Between Parent-Reported and Observed Parenting: A Multi-Level Meta-Analysis

The purpose of the present meta-analysis was to investigate the strength of the association between parent-reported and observed parenting, and to investigate which specific characteristics of participants, questionnaires, or observational procedures moderate this association. A systematic search of relevant peer-reviewed articles published between January 2000 and December 2014 yielded 36 articles (N = 8,510) and 89 effect sizes. Results from a 3-level random-effects meta-analysis demonstrated a weak, yet significant, overall association of r = .17 between parent-reported and observed parenting. The magnitude of the effect size depended on questionnaire length (larger effect for more items) and the type of parenting investigated (largest effects for negative parent behaviors, next largest effects for positive parent behaviors, and smallest effect for controlling parent behaviors). In conclusion, this study shows that the strength of the association between parent-reported and observed parenting is small but significant. Public Significance Statement Parenting is often measured using self-reports or observations. This study has shown that the association between parenting measured through self-report and through observation is small, but significant. The association is larger for negative and positive parenting compared with controlling parenting and the association is larger for studies using questionnaires with more items.