Family Functioning and Child Psychopathology: Individual versus Composite Family Scores *
Jolanda J. J. P. Mathijssen,** Hans M. Koot, Frank C. Verhulst, Eric E. J. De Bruyn, and Johan H. L. Oud
This study examined the relationship of individual family members' perceptions and family mean and discrepancy scores of cohesion and adaptability with child psychopathology in a sample of 138 families, referred to Regional Mental Health Agencies. The results indicate that the family mean scores, contrary to the family discrepancy scores, explain more of the variance in parent-reported child psychopathology than individual scores. Implications for future research and clinical practice are discussed.
From different perspectives, such as sociological, psychological and family systems theory, it is assumed that the family plays an important role in the development and maintenance of psychopathology in children (Hetherington & Martin, 1986; Jacob & Tennenbaum, 1988).
A major problem in family research is to obtain information that will reflect the family as a unit and yield true family characteristics (Fisher, Kokes, Ransom, Philips, & Rudd, 1985). Although researchers recognize that responses of multiple family members are needed to obtain a more representative view of the family, collecting data from more than one family member does not automatically yield family data. Still, in the majority of studies on the relation between family functioning and child psychopathology, the individual scores of different family members are not aggregated to construct a family-based measure (Blaske, Borduin, Henggeler, & Mann, 1989; Farrell & Barnes, 1993; Friedman, Utada, & Morrisey, 1987; Kiser et al., 1988; Natakusumah et al., 1992; Prange et al., 1992; Volk, Edwards, Lewis, & Sprenkle, 1989; Watson, Henggeler, & Whelan, 1990). In these studies conclusions are drawn at the family level from data collected at the level of the individual family member. Individual perceptions of family functioning may have considerable value and may show relations with psychopathology in family members, but they are by definition not appropriate to draw conclusions about the relation between the functioning of the family as a unit and the individual's psychopathology. A challenging question is then how scores based on individual perceptions should be combined into a family score. This is not an easy task, because family members, in particular children and their parents, differ considerably in their perception of the family (Noller & Callan, 1986; Tein, Roosa, & Michaels,1994). Some researchers question aggregation because of the differences between family members (Tein et al., 1994), whereas others argue in favor of aggregation (Schwarz, Barton-Henry, & Pruzinsky, 1985). However, this lack of high agreement should not prevent us from exploring ways to treat data from different family members (cf. Wampler & Halverson, 1993). For example, Jacob and Tennenbaum (1988) made a plea for the development of composite scores from individual reports followed by a comparison of the individual and composite scores regarding their relationship with key variables.
The examination of both individual and family composite scores is important, because it provides the opportunity to investigate whether it is valuable to compute family scores. In the present study we used two different family scores, i.e., the mean of individual family members' scores and the discrepancy between scores of individual family members regarding family functioning, in order to examine their relative association with child problem behavior. Especially for clinical purposes, this information is very important. However, as far as we know, this comparison has never been addressed in previous research.
The computation of an arithmetic mean offers the possibility of locating the family on a scale relative to other …