Amato, P.R. & Keith, B. (1991). Parental divorce and the well-being of children: A meta-analysis. Psychological Bulletin, 110(1), 26-46.
Hypotheses/Research Questions - This fairly complex meta-analysis examined 92 different studies of differences in children's well-being as a function of whether the family is divorced or intact. While no specific hypotheses were tested, the study's basic research questions involved an attempt to estimate the degree of differences in children's well-being across the various studies as well as to examine whether differences in research design (study characteristics) and in different measures of well-being contributed to response variance.
Also, the study attempted to assess which of three theoretical perspectives (Parental Absence Perspective; Economic Disadvantage Perspective; Family Conflict Perspective) best explained the overall findings of the meta-analysis.
Methods. Study methods consisted of meta-analysis techniques. In brief, 92 studies were selected from four behavioral science databases.
All selected studies were comparative (divorce vs. intact families) with at least one quantitative measurement of children's well-being yielding data that allowed for the calculation of a size of impact effect. Effect size was calculated in several different ways depending upon the nature of provided data (i.e. whether data was provided in the forms of means, standard deviations, F rations, t-tests, chi square analyses, etc.)
Measures of well-being were divided into the following categories: (1) academic achievement; (2) psychological adjustment; (3) conduct problems; (4) self-concept; (5) social adjustment; (6) mother-child relations; and (7) father-child relations. Variance in design (study characteristics) included differences in sample selection methods, size of sample, univariate versus multivariate approach, presence of statistical controls, and source of data.
Results. Findin...