abstract
- This study uses information collected on two occasions from a probability sample of families with 5- to 12-year-old children (N = 1151) participating in a general population study in 1983 and follow-up in 1987. It evaluated the importance of maternal bias in the assessment of child behavior by comparing the relative strengths of association between maternal depression and childhood behavior and between maternal depression and mother reporting errors. Conduct problems and hyperactivity were measured as latent criterion variables constructed from mother, teacher, and youth (aged 12 to 16 years) ratings and their associations with maternal depression were modeled using covariance structure analysis. The analyses revealed that maternal depression was associated significantly with conduct problems (phi = .35) and hyperactivity (phi = .38) among 5- to 7-years-olds in 1983 but not 4 years later in 1987. None of the associations between maternal depression and mother reporting errors were significant. Among 8- to 12-year-olds in 1983, maternal depression was associated significantly not only with conduct problems (phi = .17) and hyperactivity (phi = .15) but also with mother rating errors of these behaviors (psi = .13 and .17, respectively). Four years later in 1987 when this cohort was 12 to 16 years old, the only significant association was between maternal depression and conduct disorder (phi = .25). Although evidence exists for associations between maternal depressed mood and mother rating errors, there also appears to be a substantive association between maternal depression and childhood behavior.