The Sleeper Effect. Myth or Reality?

This article presents results from a 10-year Australian study of the effects of parental divorce at adolescence. It examines whether a has occurred, whereby young people, and especially young women, initially found to be coping well experience a significant reversal in adjustment in early adulthood as reported by Wallerstein and Blakeslee (1989), Wallerstein and Corbin (1989), and Wallerstein, Corbin, and Lewis (1988). Over the last 20 years, recognition of the many factors contributing to children's well-being following divorce has led to increasing methodological sophistication. Even so, a number of writers (e.g., Zaslow, 1989) have called for greater specificity in examining differing patterns of response. Nowhere is this more needed than in adolescent research, where age is frequently confounded with time since divorce, few samples being contacted at divorce itself. Developmental differences within adolescence are also often ignored, with the term defined so broadly in differing studies that it embraces young people from as young as nine to those in their early twenties. Known gender differences in socioemotional development during adolescence are not always allowed for in drawing conclusions as to divorce-specific outcomes. And finally, longitudinal studies tracking patterns of change and their interactions are rare indeed in the adolescent literature. A study by Hetherington and Clingempeel (1992) examining children between the ages of 9 and 13 over 26 months is a notable exception, but again divorce had been experienced approximately 4 years previously when the children were, on average, 7 years old. The longitudinal California study of 60 divorcing families initiated by Wallerstein and Kelly in 1971 (Wallerstein, Corbin, & Lewis, 1988; Wallerstein & Kelly, 1980) meets the criterion of distinguishing the age of the child from time since divorce, in that the sample comprised families with children of different ages who were recruited near the time of divorce, and findings are presented separately for each age group. The study also meets the criterion of longitudinality, tracking patterns of change over 10 years, in four waves of interviews. It is less successful in allowing for developmental effects unrelated to divorce, because it includes no comparison group of intact families, and the age range of those who experienced divorce at adolescence is rather wide--13 to 18 in early reports (Wallerstein & Kelly, 1980) and 9 to 19 in later reports (Wallerstein et al., 1988). The study also presents a problem in interpretation, because the researchers used clinical assessments rather than replicable standard measures, and the sample seems somewhat unrepresentative. Families were offered divorce counseling as an incentive to participation, and, while children previously in therapy were excluded, two-thirds of the parents had a history of moderate to severe psychopathology. Despite these shortcomings, the California study has produced much material that has provoked interest and indeed controversy (Hetherington & Furstenberg, 1989; Kelly & Emery, 1989). One reported finding is that divorce has long-term effects on children not anticipated on the basis of their earlier adjustment. Wallerstein, Corbin, and Lewis (1988) reported that at 10 years, a significant number of those aged 9 to 19 at divorce were troubled and drifting. A third of the women were wary of commitment and fearful of betrayal and were caught up in a web of short-lived relationships (p. 210). Elsewhere Wallerstein and Corbin (1989) reported more specific results. Describing the phenomenon as a effect, they reported that 60% of girls aged 9 to 13 at divorce, and coping well then, had declined in adjustment 10 years later. There were data for only four girls in the 14 to 19 age group, two of whom were doing poorly. Thus the sleeper effect is largely based on data from those who were in early adolescence or preadolescence at divorce. …