Effects of nonmatemal care on child development: A meta-analysis of published research

A meta-analysis of published research on the effects of nonmatemal care on child development was undertaken. One hundred and one studies published between 1957 and 1995 involving 32,271 children met the inclusion criteria. Most of the children in these studies came from middle class homes (63 studies; 62.3%), were Caucasian (77 studies; 76.2%), and were American (78 studies; 77.2%). Dependent variables in four domains (attachment, social-emotional, behavioral, and cognitive) were coded and effect sizes (d) between maternal and nonmatemal care (independent variable) were computed for both unweighted and weighted effects size. Results of unweighted effect size analysis indicated that there was a small effect and negative effects of nonmatemal care in the cognitive (d = .14) and social-emotional (d = .26) domains, and larger negative consequences for nonmatemal care for behavioral outcomes (d = .38), and attachment to mother (d = .39). Weighted effects size analysis decreased the magnitude of effect sizes in the social-emotional (d = .16) and behavioral domains but not in the cognitive and attachment domains. Moreover, males tended to fare more poorly with nonmatemal care than did females in all domains. A number of potentially mediating family, quality of care, and study characteristic variables were assessed and analyzed. The results are discussed within the context of attachment theory. We conclude that extensive nonmatemal care of infants and children results in negative developmental outcomes for children's development.