The Journal of Genetic Psychology, 1998, 159(1), 53-67 A Longitudinal Study of the Relations Between Parent-Adolescent Conflict and Adolescent Psychological Well-Being DANIEL T. L. SHEK Department of Social Work The Chinese University of Hong Kong ABSTRACT. In this longitudinal study, the relationships between parent-adolescent con- flict and adolescent psychological well-being were examined in a sample of Chinese ado- lescents (N = 378) via children's and parents' reports of parent-adolescent conflict. The results indicate that parent-adolescent conflict based on ratings obtained from the differ- ent sources was concurrently related to hopelessness, life satisfaction, self-esteem, pur- pose in life, and general psychiatric morbidity at Time 1 and Time 2. Longitudinal and prospective analyses (Time 1 predictors of Time 2 criterion variables) suggest that the rela- tions between parent-adolescent conflict and adolescent psychological well-being are bidirectional. Although the strengths of association between parent-adolescent conflict and adolescent psychological well-being were similar for male and female adolescents, father-adolescent conflict, relative to mother-adolescent conflict, was found to exert a stronger influence on adolescent psychological well-being. ALTHOUGH PARENT-ADOLESCENT CONFLICT is widely regarded by clinicians as an etiological factor in adolescent psychopathology (e.g., Foster & Robin, 1988; Hall, 1987), empirical evidence pertinent to the links between par- ent-adolescent conflict and adolescent development is limited. In a pioneer study of conflict-resolution behavior in adolescent boys, Rubenstein and Feldman (1993) explicitly commented that “it is not known to what extent adolescent behavioral and emotional disorders are a function of the amount of conflict in the family" (p. 43). Regarding the empirical support for the link between parent-adolescent con- This work was financially supported by the Research Grants Council of the UGC (Grant CUHK155/94H). The author wishes to thank Chan Lai-kwan for her assistance in col- lecting the data. Address correspondence to Daniel T. L. Shek, Department of Social Work, The Chi- nese University of Hong Kong, Shatin, Hong Kong. E-mail: DANIELSHEK@CUHK. EDU.HK. 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 54 The Journal of Genetic Psychology flict and adolescent functioning, there are data suggesting that parent-adolescent conflict is related to adolescent maladjustment, including depression (Forehand et al., 1988); injuries (Bijur, Kurzon, Hamelsky, & Power, 1991); unacceptable behavior (Tomlinson, 1991); problem behavior at school (Forehand, Long, Brody, & Fauber, 1986; Galambos, Sears, Almeida, & Kolaric, 1995); difficulties in ado- lescent functioning (Forehand et al., 1991); and self-esteem, anxiety, and control (Slater & Haber, 1984). Past studies have also revealed that extensive parent-child conflict exists in homes of disturbed children (Reich, Earls, & Powell, 1988) and runaway adolescents (Justice & Duncan, 1976). A critical examination of the literature shows that there are several limita- tions intrinsic to the existing studies on the link between parent-adolescent con- flict and adolescent adjustment. The first limitation concerns the assessment of parent-adolescent conflict. Although some researchers have included children's and parents' perceptions in the assessment (i.e., inclusion of multiple informants), it is not a common practice to assess father-adolescent conflict (e.g., Forehand et al., 1988; Forehand et al., 1991). Even in those studies in which fathers and mothers were included, some researchers have assessed parent-adolescent con- flict without distinguishing between the father and the mother (i.e., treating the parents as a unit), therefore failing to reveal the differential nature of father-ado- lescent conflict and mother-adolescent conflict (e.g., Galambos et al., 1995; Rubenstein & Feldman, 1993). The assessment of parent-adolescent conflict separately for fathers and mothers, based on data obtained from separate sources, would probably give a more complete picture of the problem area. There are different forms of par- ent-adolescent conflict, depending on the perspective of the informant: father- adolescent conflict based on the perspective of the child, mother-adolescent con- flict based on the perspective of the child, father-adolescent conflict based on the perspective of the father, and mother-adolescent conflict based on the perspec- tive of the mother. The second limitation of the existing studies is that the related sample sizes are generally small: Forehand and colleagues recruited 69 dyads for the first study (Forehand et al., 1988) and 120 families for the second study (Forehand et al., 1991); Galambos et al. (1995) included 105 families in their study; and Ruben- stein and Feldman's (1993) study was based on the responses of 108 families. In addition, the use of multivariate statistical tests (multiple regression analyses have been widely used) in small samples increases the possibility that the data obtained are attributable to chance effect (Tabachnick & Fidell, 1989). The third limitation is that most of the existing studies have been conducted in Western societies. Actually, an examination of the available databases (Ho, Spinks, & Yeung, 1989; Shek, 1995a; Spinks & Ho, 1993) showed that, to date, no scien- tific study has been conducted to examine the relations of parent-adolescent con- flict to adolescent adjustment in Chinese society. Yang (1981) argued that social- ization practice is parent centered in traditional Chinese culture, where parent-child Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Shek 55 conflict is basically not encouraged, children are socialized to be submissive to the parents, and conflict with parents is regarded as an unfilial act. Within such a cul- tural context, parent-adolescent conflict might constitute a source of stress for the adolescent involved, and this stress might influence his or her adjustment. The fourth limitation is that although most of the existing studies have been carried out to study parent-adolescent conflict and depression (Forehand et al., 1988), problem behavior (Forehand et al., 1991; Galambos et al., 1995), and dis- tress and psychosomatic complaints (Rubenstein & Feldman, 1993), few studies have addressed the link between parent-adolescent conflict and adolescent posi- tive mental health. Conceptually, psychological well-being can be defined in terms of the absence of manifested psychiatric symptoms or presence of positive mental health attributes (Bradburn, 1969; Diener, 1984) or coping resources (Folkman, Schaefer, & Lazarus, 1979). According to the latter view, indicators such as life satisfaction, meaning in life, hope, or self-worth should be the focus. Folkman, Schaefer, and Lazarus (1979) regarded such qualities as coping resources that would help an individual develop positive coping repertoires. Thus, it would be theoretically illuminating if measures of psychiatric morbidity and positive mental health could be included in a single study. The final limitation of the existing studies is a conceptual one. Regarding the link between parent-adolescent conflict and adolescent psychological well- being, there are three possibilities: (a) parent-adolescent conflict influences ado- lescent psychological well-being, (b) adolescent psychological well-being influ- ences parent-adolescent conflict, and (c) parent-adolescent conflict and adolescent psychological well-being do not influence each other. It is obvious that the existing studies have been guided primarily by the first possibility, assuming that increased parent-adolescent conflict is conducive to impaired adolescent functioning. In contrast, few studies have been conducted to examine the second possibility (i.e., that adolescent adjustment is an antecedent of parent-adolescent conflict), and existing data have seldom been interpreted in light of this hypoth- esis. For example, Forehand et al. (1988) collected parent-adolescent conflict and adolescent depression data at two time periods (i.e., at Time 1 and Time 2). How- ever, the data were interpreted in terms of the influence of parent-adolescent con- flict on adolescent depression, and data on the association between Time 1 ado- lescent depression and Time 2 parent-child conflict were simply not presented. Theoretically, there are some bases for the hypothesis that adolescent psy- chological well-being influences parent-adolescent conflict. According to the assertions of systems theories, functioning at the individual subsystem and at the parent-child subsystem is interdependent (e.g., Goldenberg & Goldenberg, 1980; Hinde & Stevenson-Hinde, 1988). Borrowing ideas from the sociological litera- ture on family stress and role strain (e.g., Margolin, 1981), one can also argue that poor adolescent mental health is stressful, which would limit an adolescent's sensitivity and response to parents' demands and desires, therefore contributing to increased parent-adolescent conflict. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 56 The Journal of Genetic Psychology The data reported in this article are based on a longitudinal study that was designed to investigate the relations between family environment (including par- ent-adolescent conflict) and adolescent adjustment in Hong Kong. Data were col- lected at two time periods (Time 1 and Time 2) separated by 1 year. A large vol- ume of data has been generated from this study (Shek, 1995b, 1997). In this article, the primary focus is on the relations between children's and parents' reports of parent-adolescent conflict and adolescent psychological well-being. Method Instruments: Assessment of Parent-Adolescent Conflict Child version of the Father-Adolescent Conflict Scale (FAC-C) and Mother- Adolescent Conflict Scale (MAC-C). Robin and Foster (1989) demonstrated that the child version of the Conflict Behavior Questionnaire (CBQ-20) is a valid and reliable measure of parent-adolescent conflict from the perspective of the child. The items in the child version of the CBQ-20 were translated into Chinese by Shek, Lee, Ngai, Law, and Chan (1995). Based on the translated items, the Father- Adolescent Conflict Scale (FAC-C) and Mother-Adolescent Conflict Scale (MAC-C) were formed to assess father-adolescent and mother-adolescent con- flict, respectively. Shek et al. (1995) found that the FAC-C and MAC-C have high internal consistency and temporal stability. A higher FAC-C scale score (α = .87 at Time 1 and .89 at Time 2) or MAC-C scale score (α = .85 at Time 1 and .86 at Time 2) indicates a higher level of parent-adolescent conflict in this study. Parent version of the Father-Adolescent Conflict Scale (FAC-P) and Mother- Adolescent Conflict Scale (MAC-P). Robin and Foster (1989) demonstrated that the parent version of the Conflict Behavior Questionnaire (CBQ-20) is a valid and reli- able measure of parent-adolescent conflict from the perspective of a parent. The items in the parent version of the CBQ-20 were translated into Chinese by Shek, Lee, Ngai, Law, and Chan (1995). Based on the translated items, the Father- Adolescent Conflict Scale (FAC-P) and Mother-Adolescent Conflict Scale (MAC- P) were formed to assess father-adolescent and mother-adolescent conflict, respec- tively, with identical items in each scale. A higher FAC-P scale score (α = .88 at Time 1 and .88 at Time 2) or MAC-P scale score (α = .89 at Time 1 and .90 at Time 2) indicates a higher level of parent-adolescent conflict in this study. Instruments: Assessment of Psychological Well-Being Chinese Hopelessness Scale (HOPEL). The Hopelessness Scale was devised by Beck, Weissman, Lester, and Trexler (1974) to measure the respondent's sense of hope. The Chinese Hopelessness Scale was developed by Shek; evidence sup- porting its reliability and validity has been reported (Shek, 1993b). A higher Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Shek 57 HOPEL scale score (α = .81 at Time 1 and .82 at Time 2) indicates a lower sense of hope in this study. Satisfaction With Life Scale (LIFE). The Satisfaction With Life Scale was devel- oped by Diener, Emmons, Larsen, and Griffin (1985) to assess the respondent's own global judgment of his or her quality of life. The Chinese version of this scale was translated by Shek; adequate reliability of this scale has been reported (Shek, 1992a). A higher LIFE scale score (α .71 at Time 1 and .73 at Time 2) indi- cates a higher level of life satisfaction in this study. ===== Chinese Rosenberg Self-Esteem Scale (ESTEEM). The Rosenberg Self-Esteem Scale was designed to assess the self-esteem of high school students (Rosenberg, 1979). The Chinese Rosenberg Self-Esteem Scale was developed by Shek; acceptable reliability of this scale has been reported (Shek, 1992a). A higher ESTEEM scale score (α = .70 at Time 1 and .72 at Time 2) indicates a higher level of self-esteem in this study. Chinese Purpose in Life Questionnaire (PIL). The Purpose in Life Questionnaire was designed to assess the respondent's perceived existential meaning of life (Crumbaugh, 1968). Evidence has been accumulated on its reliability and valid- ity (Shek, 1988, 1992b). The Chinese PIL was translated by Shek (1988), who found that the Chinese PIL has good psychometric properties. A higher PIL scale score (α = .87 at Time 1 and .88 at Time 2) indicates a higher level of life pur- pose in this study. The Chinese version of the 30-item General Health Questionnaire (GHQ). The General Health Questionnaire was developed to measure current nonpsychotic disturbances (Goldberg, 1972). Chan (1985) found that the Chinese GHQ com- pared favorably with the English version at the scale level, and there is evidence that the GHQ possesses acceptable psychometric properties (Shek, 1987, 1989, 1993a). A higher GHQ scale score (α = .88 at Time 1 and .88 at Time 2) indi- cates a higher level of psychiatric symptoms in this study. Although the GHQ can be treated as an instrument assessing mental health in terms of manifested psychiatric symptoms, the other scales can be regarded as tools measuring mental health in terms of positive mental health attributes (Diener, 1984) or coping resources (Folkman, Schaefer, & Lazarus, 1979). On the basis of the existing literature, I predicted that if parent-adolescent conflict impairs psychological well-being in Chinese adolescents, a higher level of parent-adolescent conflict (higher FAC-C, MAC-C, FAC-P, or MAC-P scores) at Time 1 should be associated with poorer adolescent well-being (higher GHQ and HOPEL scores, and lower LIFE, PIL, and ESTEEM scores) at Time 2. Alter- natively, if adolescent psychological well-being intensifies parent-adolescent conflict, poorer adolescent well-being (higher GHQ and HOPEL scores, and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission./nPsychology in Adolescent || 7 pages, APA

Fig: 1