The Impact of Mothers' Parenting, Involvement by Nonresidential Fathers, and Parental Conflict on the Adjustment of Adolescent Children Author(s): Ronald L. Simons, Les B. Whitbeck, Jay Beaman and Rand D. Conger Source: Journal of Marriage and Family, Vol. 56, No. 2 (May, 1994), pp. 356-374 Published by: National Council on Family Relations Stable URL: http://www.jstor.org/stable/353105 Accessed: 27-06-2016 04:14 UTC REFERENCES Linked references are available on JSTOR for this article: http://www.jstor.org/stable/353105?seq=1&cid=pdf-reference#references_tab_contents You may need to log in to JSTOR to access the linked references. Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://about.jstor.org/terms JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Wiley, National Council on Family Relations are collaborating with JSTOR to digitize, preserve and extend access to Journal of Marriage and Family This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms RONALD L. SIMONS, LES B. WHITBECK, JAY BEAMAN, AND RAND D. CONGER Iowa State University The Impact of Mothers' Parenting, Involvement by Nonresidential Fathers, and Parental Conflict on the Adjustment of Adolescent Children The present study used panel data on 207 divorced women and their children to examine the By one estimate, 44% of children born between 1970 and 1984 will live for a time in a single-par- influence of mothers' parenting practices, in- ent family (Bumpass & Sweet, 1989), and volvement of nonresidential fathers, and parental conflict on the adjustment of adolescents living in mother-headed households. In addition, the study investigated the possibility that child adjustment Furstenberg and Cherlin (1991) project that 60% of children born in the 1990s will have this experience. Unfortunately, research during the 1980s provided rather strong evidence that children living in single-parent households are at risk for a problems may be a cause, as well as a consequence, of parental behavior. Quality of parenting by nonresidential fathers was related to externalizing problems for boys and girls, although the results differed somewhat depending on the source of data used to assess father's parenting. Quality of mother's parenting showed an association with externalizing problems of boys and girls, and was also related to internalizing problems for boys. Parental conflict was associated with internalizing problems for boys but not girls. Finally, adolescent externalizing problems appeared to reduce the quality of mother's parenting for both boys and girls, and to diminish father involvement in parenting in the case of boys. Department of Sociology and Center for Family Research on Rural Mental Health, 107 East Hall, Iowa State University, Ames, IA 50011. Key Words: child adjustment, divorce, noncustodial fathers, parenting. number of negative developmental outcomes (Amato & Keith, 1991; McLanahan & Booth, 1989). These studies also reported, however, that there is much variability in adjustment among such children, and that many, if not most, avoid long-term problems (Emery, 1988). Given these findings, researchers and policy makers have been concerned with identifying factors that serve to reduce the hazards of growing up in a singleparent household. The most popular explanations for variations in child adjustment concern the parenting practices of the custodial parent, level of involvement of the noncustodial parent, parental conflict, and family economic hardship. The present study uses panel data to examine the importance of each of these variables in predicting externalizing and internalizing problems among a sample of adolescents living in mother-headed households. The study avoids the limitations of much of the past research on child adjustment in single-parent fam- 356 Journal of Marriage and the Family 56 (May 1994): 356-374 This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms Adjustment ofAdolescent Children 357 ilies by using multiple sources of data to build measures of constructs and by including controls for potentially confounding variables. Also, in contrast to most prior investigations, the present study considers the possibility that child adjustment problems may influence, as well as be influenced by, parental behavior. ASPECTS OF SINGLE-PARENT FAMILIES Parenting Practices of the Custodial Parent Although a few studies have reported an associa- tion between family structure and level of parental warmth and support, the dimension of parenting most consistently linked to number of parents in the home is that of control. There is strong evidence that single parents tend to make fewer demands on children and utilize less effec- tive disciplinary strategies than married parents (Astone & McLanahan, 1991; Furstenburg & Nord, 1985; Hetherington, Cox, & Cox, 1982). Although past studies report a tendency for single parents to display lower levels of monitoring and discipline than married parents, most of these studies find substantial variability in parent- ing practices in their samples of single parents. Thus it is probably a minority of single parents that account for discrepancies in control between single- and two-parent families. The present study examines the extent to which variations in control exerted by single mothers is related to the adjustment of their adolescent children. Differences in parental control are expected to exert a greater influence on adolescent externalizing than internalizing problems. Parenting Practices of Nonresidential Fathers Level of involvement by the nonresidential parent-usually the father-is frequently cited as an important determinant of children's adjustment to divorce. It is argued that children show better adjustment when both the custodial and noncustodi- al parent are actively involved in childrearing. Based upon this belief, researchers have initiated studies of the conditions that promote nonresiden- tial paternal involvement and many states have adopted joint custody statutes in an attempt to en- courage fathers to remain involved with their children (Furstenberg & Cherlin, 1991). Although the idea that child adjustment is facilitated by involvement by nonresidential fathers is intuitively appealing, there is actually little em- pirical evidence for the claim (Emery, 1988; Furstenberg & Cherlin, 1991). In a recent review of research regarding this issue, Amato (1993) identified 16 studies that supported the hypothesis that frequency of contact with the noncustodial father is positively related to child adjustment. However, an equal number of studies failed to support the hypothesis. Indeed, seven of the studies in the latter group found frequency of visitation with the noncustodial father to be negatively related to child adjustment. Thus, overall, the evidence suggests that frequency of visitation by fathers is not related to child adjustment. Perhaps this finding should have been anticipated as there is little reason to expect that simply having contact with the nonresidential father would significantly effect a child's development. Rather, it is most likely the quality, rather than simply quantity, of interaction with this parent that is the key to understanding his impact on child adjustment. However, contrary to this idea, studies that have examined closeness of relation- ship with the father and child outcomes have failed to find significant effects. Furstenberg, Morgan, and Allison (1987), for example, found that child reports of closeness to their father, like their reports of frequency of contact, were unrelated to child developmental outcomes. One might argue, however, that closeness per se is not the important dimension when assessing the quality of the relationship between fathers and their child. Past research suggests that nonresidential fathers often behave toward their children more as an adult friend or relative than as a parent (Arendell, 1986; Furstenberg & Nord, 1985; Hetherington, Cox, & Cox, 1976). Much of the time with their children is spent watching TV, attending movies, going out to eat, and the like. Although such interaction is apt to enhance feelings of closeness, there is little reason to believe that it would have a major impact on child adjustment. Fathers who limit their involvement to such activ- ities would not be expected to exert any more in- fluence on the developmental outcomes of their children than that exercised by other friendly adults (e.g., uncles, grandparents, mom's boyfriend). On the other hand, fathers may affect child adjustment to the extent that they continue to play the role of parent. As noted earlier, single mothers often show deficiencies in the area of monitoring and discipline. Thus nonresidential fathers might be expected to influence child development to the extent that they provide assistance with child social- This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms 358 Journal of Marriage and the Family ization and control. Thus the present study exam- ines the relationship between child adjustment and the involvement of nonresidential fathers in activities such as inductive reasoning, monitoring, and consistent discipline. Such parenting behaviors are expected to be inversely related to adolescent troubles, especially externalizing problems. Analysis of the contribution that nonresidential fathers make to child adjustment requires that the quality of the mother's parenting is controlled. Re- cently Maccoby, Buchanan, Mnookin, and Dornbusch (1993) reported a zero-order relationship between closeness to the noncustodial parent and child adjustment. However, controlling for the custodial parent's parenting practices eliminated the association. Children who have involved fa- thers also tend to have warm, supportive mothers. Therefore the impact of mother's parenting on child adjustment must be taken into account if one is to establish that the effect of involvement by nonresidential fathers is unique and not merely redundant to the contribution made by mothers. the case in samples of two-parent families. It may be that the same process operates in single-parent families. If this hypothesis is valid, controlling for quality of parenting should eliminate any relationship between parental conflict and child externalizing problems. Just as one cannot adequately assess the impact of parental conflict on child adjustment with- out controlling for quality of parenting, evaluation of the impact of the nonresidential father's parenting on child outcomes requires controlling for parental conflict. Several researchers have noted that high paternal involvement may foster persistent conflict between the former spouses (Furstenberg et al., 1987; Maccoby et al., 1993). It may be that past studies have failed to find an advantageous effect for father's involvement because any beneficial contribution is canceled by increased conflict with the mother. Family Income There is strong evidence that children of divorced parents tend to display more adjustment problems Marital disruption is usually associated with a rather dramatic loss in family income (Morgan, 1991; Raschke, 1987). Such a decline in standard of living might cause children to feel depressed and angry. Further, limited means may require when there is recurrent friction between their par- moving to neighborhoods with inadequate Parental Conflict ents (Amato, 1993; Emery, 1988). Such conflict is apt to be upsetting to children because it increases the probability that they will feel frustration over being caught in the middle, experience distress over the emotional pain that their parents are producing in each other, and experience fear that one of the parents may withdraw from the family in an attempt to avoid the altercations. Thus chronic strife between parents increases a child's risk for feelings of frustration, anxiety, and depression. This suggests that parental conflict is likely to be directly related to child internalizing problems. In contrast, one might expect the impact of parental conflict on child externalizing problems to be indirect through parental behavior. Past research has linked child conduct problems to inept parental monitoring and discipline (Maccoby & Martin, 1983; Patterson, Reid, & Dishion, 1992). Another line of research has shown that parental conflict tends to disrupt competent parenting practices (Caspi & Elder, 1988; Conger et al., 1992; Simons, Lorenz, Wu, & Conger, 1993). Hence the contribution of parental conflict to child conduct problems may be indirect through its effect on parenting practices. Caspi and Elder (1988) and Conger et al. (1992) found this to be schools, high crime rates, and a well-developed delinquent subculture (McLanahan & Booth, 1989). Such an environment is apt to increase an adolescent's risk for conduct problems. Thus both the internalizing and externalizing problems that are common to single-parent households may be a result of the economic hardship often experienced by these families. In assessing the impact of family income on child adjustment it is important that the effects of mother's parenting be controlled. Past studies of both two-parent (Elder, Van Ngugen, & Caspi, 1985; Simons, Lorenz, Conger, & Wu, 1992) and single-parent (Simons, Beaman, Conger, & Chao, 1993) families have reported that financial hardship tends to disrupt parenting practices. Thus the impact of family income on the adjustment of children of divorce may be expressed indirectly through its negative effect on the parenting of the custodial parent. Conversely, it is important that family income be controlled when evaluating the impact of non- custodial fathers on child adjustment because fathers who remain involved with their children are more apt to fulfill their child support obligations (Furstenberg & Cherlin, 1991). These payments This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms Adjustment ofAdolescent Children 359 serve to enhance the family's economic wellbeing. Thus any association between father involvement and child adjustment might be spurious due to the association of both variables with the fi- nancial status of the family. The present study addresses this issue by including both family income and level of child support in the analysis. The Impact of Child Adjustment on Parental Behavior Associations between parental behavior and child adjustment are generally interpreted as evidence for the impact of parents on child development. It may be, however, that the converse causal sequence operates as well. Over time, parents may withdraw from children who engage in unpleasant behaviors. The present study focuses on the control dimension of parenting. Parenting practices related to control are more likely to be influenced by externalizing than internalizing problems. Whereas there is little reason to anticipate that parents modify their supervision and discipline in reaction to a child's display of anxiety or depression, it would seem reasonable that they might show a tendency toward reduced monitoring and inconsistent discipline in response to a child who is persistently hostile, coercive, and defiant. Indeed, as Patterson et al. (1992) have noted, rebellious, antisocial children often punish parental efforts to monitor and discipline while reinforcing parental withdrawal and lenience. This phenomenon is especially apt to operate for noncustodial parents. Often they must invest substantial time and energy in scheduling and ar- ment problems on parenting practices, the present study examines the extent to which the level of conflict between divorced parents is influenced by the internalizing or externalizing problems of their children. Relationships between the parent variables and adolescent adjustment were tested using data from the Iowa Single Parent Project (ISPP), a three-wave study of 207 recently divorced women and their children. At initiation of the study, each of the families contained a target child who was in eighth or ninth grade. Much of the prior re- search on children's adjustment to divorce is problematic as a single reporter (either the custodial parent or the child) is used as a source of information for all of the variables included in the analysis. The ISPP data set provides multiple sources of information for several of the con- structs included in the present study, thereby re- ducing the problem of shared method variance between explanatory and criterion variables. The analytic approach used to investigate the impact of parent behavior on adolescent adjustment involved using wave-1 measures of the parent variables to predict adolescent adjustment at wave 2, and wave-2 measures of the parent variables to predict adolescent adjustment at wave 3. This approach accomplished two purposes. First, using lagged measures of adolescent adjustment provided a stronger test of the causal priorities assumed between the parent variables and child outcomes than the cross-sectional procedures utilized in most prior studies. Second, focusing on child adjustment at both waves 2 and 3 permitted examination of the possibility that the effect of cer- ranging visits with their child, particularly if there tain of the parent variables changes with time is a fair amount of distance between the child's since divorce. It is possible, for example, that the impact of paternal involvement or parental conflict declines as children adjust to the absence of their father from the household. Basically the same approach was used to examine the influence of child adjustment on parent behavior. Wave-1 measures of the adolescent internalizing and externalizing problems were used to predict parent behavior at wave 2, whereas wave-2 measures of and parent's residences. Given these costs, they may be easily discouraged from playing the role of parent if interaction with the child is unpleas- ant and offers few rewards. Besides affecting parenting practices over time, child adjustment problems might be expected to influence the level of conflict between parents. Arguments might be made for anticipating either a positive or negative impact. On the one hand, child maladjustment might encourage parents to suspend their conflict with each other in an effort to address the pressing problem of their child's emotional or behavioral problems. On the other hand, divorced parents may displace their frustration with the child onto each other, blaming one another for the child's difficulties. Thus, in addition to investigating the effect of child adjust- the adolescent adjustment were used to predict parent behavior at wave 3. METHODS AND PROCEDURES Sample A sample of 207 female-headed households was recruited through the cohort of eighth and ninth This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms 360 Journal of Marriage and the Family grade students living in approximately two-thirds selves each of the items listed on the cards and to of all counties in Iowa. Mothers were screened ac- continue talking until the interviewer returned. The family was given 25 minutes to complete the task. The items on the cards concerned family issues such as discipline and chores, and the children's friends and school performance. The second task, 15 minutes in length, also involved all three family members. For this task, the family was asked to discuss and try to resolve the issues and disagreements that they had cited in the questionnaires they had completed earlier in the visit. The third task involved only the two children and was 15 minutes in length. The youths were given a set of cards listing questions related to the way they got along, the manner in which their parents cording to the criteria that they be permanently separated from their husbands, that the separation occurred within the past 2 years, that the husband from whom they separated is the biological parent of the target child, and that they have a sibling within 3 years of age of the target child. Ninetynine percent of those eligible agreed to participate. Roughly one-third of the families lived in communities smaller than 7,500 in population, another third dwelled in towns ranging in size from 7,500 to 50,000 residents, and the remaining third resided in cities with more than 50,000 in- habitants. Median family income, including child support, was $21,521. Mean level of education for the mothers was 13 years. Only 4% had not completed high school, 42% had some post-high school training, and 16% had a college degree. Mean age for the target children was 14.3 years at wave 1. At wave 2, data was obtained from 193 of the original 207 families, for a retention rate of 93%. One hundred ninety of the families were visited again at wave 3, although in some cases not all of the members chose to participate. Thus, for wave 3, complete data was available for only 181 of the study families. Procedures Data were collected annually (each spring) from study families over a 3-year period. Essentially the same procedures and instruments were used at all three data collection points. Each wave of data collection involved two visits to each of the study families. During the first visit, each of the three family members completed a set of questionnaires focusing upon family processes, individual family member characteristics, and economic circum- stances. On average, it took approximately 2 treated them, their friends, and their future plans. The second visit lasted approximately 2 hours. The videotapes were coded by project observers using the Iowa Family Interaction Rating Scales (Melby et al., 1990). These scales focus upon the quality of behavior exchanges between family members. The project observers were staff members who had received several weeks of training on rating family interactions and specialized in coding one of the three interaction tasks. Before observing tapes, coders had to independently rate precoded interaction tasks and achieve at least 90% agreement with that standard. For purposes of assessing interobserver reliability, 25% of the tasks were randomly selected to be in- dependently observed and rated by a second observer. Reliability between observers was determined by calculating a generalizability coefficient (Suen & Ary, 1989). The magnitude of this coef- ficient varied by rating scale but on average ranged between .60 and .70. Measures Externalizing problems. Externalizing problems involve antisocial conduct such as fighting, non- hours to complete the first visit. compliance, poor performance in school, and During the second visit, which normally occurred within 2 weeks of the first, the family delinquent behavior. Scores from three instruments were standardized and summed to form a videotaped while engaging in several different structured interaction tasks. The visit began by having each individual complete a short questionnaire designed to identify issues of concern or disagreements within the family (e.g., chores, recreation, money, etc.). The family members composite measure of Externalizing Problems. The first was a 10-item, self-report school problems scale. The items focused on grades, completion of homework, attendance, and trouble with school authorities. Response format for the items ranged from 1 (strongly disagree) to 5 (strongly agree). Coefficient alpha was approximately .85 completed additional questionnaires and was were then gathered around a table and given a set at each wave. The second instrument consisted of of cards to read and discuss. The three family members were asked to discuss among them- a self-report delinquency inventory adapted from the National Youth Survey (Elliott, Huizinga, & This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms Adjustment ofAdolescent Children 361 Ageton, 1985; Elliott, Huizinga, & Millard, 1989). Respondents were asked to indicate how observational ratings, was .25 to .35, and the in- often during the preceding year that they had en- reliability averaged .60. These levels are within the range of acceptable values and suggest the gaged in each of 28 delinquent activities (0 = never, 5 = 5 or more times) . The acts varied from relatively minor offenses such as using alcohol to more serious offenses such as attacking someone with a weapon or stealing something worth over $25. The last instrument consisted of a seven-item measure of aggressive orientation adapted from Velicer, Govia, Cherico, and Corriveau's (1985) modification of the Buss-Durkee Hostility Inventory. It contained items such as: "I do whatever I have to in order to get what I want," "I don't care much about what other people think or feel," and, "When I get mad I say nasty things." The response format for the items ranged from 1 (not at all) to 5 (exactly). Coefficient alpha was approximately .80 at each wave. The correlations between the three instruments were above .45 for all three waves. Internalizing problems. Internalizing problems involve emotional distress such as depression, irri- tability, and somatic concerns. The depression, hostility, anxiety, and somatization subscales from the SCL-90-R (Derogatis, 1983) were used as indicators for this construct. This instrument requires that the respondent report how much dis- comfort (1 = not at all, 5 = extremely) he or she has experienced during the past week with regard to each of 46 symptoms. The symptoms consist of items such as crying easily, restless sleep, and uncontrollable temper outbursts. The subscales have demonstrated construct validity, internal consistency, and test-retest reliability (Derogatis, 1983). Coefficient alpha for the instrument was above .90 at each wave. traclass correlations used to measure intercoder presence of basic agreement between coders (Bakeman & Adamson, 1984; Hartmann, 1977) and between family member reports (Schwarz, Barton-Henry, & Pruzinsky, 1985). Although the measures are briefly described below, a more detailed description of the instruments and the factor analytic procedures used to generate them is presented in McGruder, Lorenz, Hoyt, Ge, and Montague (1992). Mothers reported on their monitoring using a four-item scale (e.g., "How often do you know who your ninth grader is with when he/she is away from home?"). The response format ranged from 1 (never) to 5 (always). The adolescents reported on their mothers using the same items. The observational rating of monitoring was based upon the interaction and content of the discussion in task 1. The mother reports, child reports, and observational ratings were standardized and then summed to form a measure of monitoring. Mothers were asked to rate their consistency of discipline using a four-item scale (e.g., "How often do you punish your ninth grader for something at one time and then at other times not punish him/her for the same thing?"). The response format ranged from 1 (always) to 5 (never). Adolescents rated their mothers' consistency using the same items. The observational ratings of inconsistent discipline were based upon family interaction and content of discussion in videotaped task 1. The mother reports, child reports, and observational ratings were standardized and summed to form a measure of consistent discipline. Mothers used a four-item scale to report on the extent to which they used harsh disciplinary prac- Custodial mother's parenting. Past research has tices (e.g., "When punishing your ninth grader, established that effective parents set standards for their children, monitor their behavior, are consis- how often do you hit him/her with a belt, paddle, or something else?"). Response format for this in- tent in enforcing rules, and eschew harsh punish- ments (Maccoby, 1992; Maccoby & Martin, 1983). High scores on these four dimensions of parenting were treated as indications of effective parenting. Measures for each of the dimensions were formed by aggregating mother self-reports, adolescent reports, and observer ratings of the videotaped family interaction tasks. Coefficient alpha for the several child- and mother-report scales was generally above .80. On average, the correlation between child and mother reports, or between either child and mother report and the strument ranged from 1 (always) to 5 (never). Adolescents used the same four items to report on their mothers' harsh discipline. Observational coders rated the harsh discipline of mothers toward their children based upon family interaction and content of discussion in task 1 of the video- taped interaction. The mother reports, child reports, and observational ratings were standardized and summed to form a measure of avoidance of harsh discipline. Finally, a five-item scale was used to obtain mothers' ratings of the extent to which they set This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms 362 Journal of Marriage and the Family standards in the course of enforcing rules or dis- of the 14 parenting practices. Coefficient alpha ciplining their children (e.g., "How often does was above .90 for both waves. The actual items your mother discipline you by reasoning, explain- were: ing, or talking to you?"). The response format ranged from 1 (never) to 5 (always). Adolescents reported on their mothers' behavior using the same items, and coders rated level of setting standards based upon family interaction and content of the discussion in task 1 of the videotapes. The mother reports, child reports, and observational ratings were standardized and summed to form a measure of limited setting of standards. These four composite indicators of ineffective parenting practices were summed to form an aggregate measure of Mother's Parenting. Although technically coefficient alpha is a measure of the internal consistency of scales, and not indices formed by aggregating scales, it can be used as method of summarizing the degree of intercorrelation among aggregated scales. Using alpha in this way, the coefficient for the four-item index of Mother's Parenting was approximately .70 at both waves 1 and 2. 1. How often does your dad talk with you about what is going on in your life? 2. When your dad tells you to stop doing some- thing and you don't stop, how often does he punish you? 3. How often does your dad punish you for something at one time and then at other times not punish you for the same thing? (reverse coded) 4. When your dad is punishing you, how much does the kind of punishment depend on his mood? (reverse coded) 5. How often does your dad disagree with your mom about how or when to punish you? (reverse coded) 6. How often do the same problems seem to come up again and again with your dad and never seem to get resolved. (reverse coded) 7. When you and your dad have a problem, how often can the two of your figure out how to deal with it? 8. How often do you talk to your dad about Nonresidential father's parenting. It is clear that effective parenting practices for nonresidential fathers will differ from those of fathers who live in the home. It would be difficult, for example, for fathers living outside of the home to monitor cur- fews or enforce punishments such as grounding. Indeed, noncustodial fathers who attempted to en- gage in such activities might be perceived as intrusive and the consequence might be increased conflict with the children and former spouse. Many nonresidential parents are, however, in a position to monitor their children's school performance and friendship choices, to stress the impor- tance of certain behavior standards, to enforce rules in a fair and consistent fashion, and to support the parenting efforts of the custodial parent. Neither observational ratings nor father selfreports were available as measures of father's parenting practices. The indicators for this construct were limited to mother and adolescent re- ports about the father's behavior at wave 1, while only adolescent reports were available for wave 2. An adolescent scale was formed using 14 items adapted from the scales used to measure mother's parenting. The items selected represented actions that parents might continue to perform even though they no longer lived in the same residence as the child. Adolescents were asked to indicate how much they agreed (1 = strongly disagree, 5 = strongly agree) that their father engaged in each things that bother you? 9. How often does your dad ask what you think before deciding on family matters that involve you? 10. How often does your dad give you reasons for his decisions? 11. How often does you dad ask you what you think before making a decision that involves you? 12. When you don't understand why your dad makes a rule for you to follow, how often does he explain the reason? 13. How often does your dad discipline you by reasoning, explaining, or talking to you? 14. When you do something your dad likes or approves of, how often does he let you know he is pleased about it? The mother-report scale for wave 1 focused on similar dimensions of parenting. The mothers were asked to indicate the extent to which (1 = strongly disagree, 5 = strongly agree) each of seven statements described the current parenting practices of their former spouse. Coefficient alpha for the scale was .86. The items were: 1. Knows your children's teachers and how well they are doing in school. 2. Explains to your children the reasons for the rules he expects them to follow. 3. Encourages your children to be responsible in the things they do. 4. Sometimes punishes your children for something and other times doesn't punish them for This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms Adjustment ofAdolescent Children 363 the same thing. (reverse coded) 5. Is clear about what he expects your children to do. 6. Agrees with you about how the children should be raised. 7. Lets your children do things they shouldn't do. (reverse coded) 8. Explains to your children the reasons for the rules he expects them to follow. Divorced women usually harbor feelings of anger, frustration, and resentment regarding their former husbands. Similarly, children of divorce often possess strong feelings about their nonresidential father, although these emotions may be either positive or negative. One would expect these emotional states to distort the objectivity of mother and adolescent reports of the father's parenting practices. To reduce this problem, we planned to standardize the two types of information and sum them to form a composite measure of father's parenting for wave 1. Unfortunately, the correlation between the two scales was rather low. For girls and their mothers the correlation was .29 whereas (never) to 7 (always), with a middle category of 4 (about half of the time). She then used the same items and response format to report on her actions toward her former spouse. Coefficient alpha for this 16-item instrument was above .90 for both waves. The adolescents reported on the amount of conflict between their parents using a three-item scale. The items were: 1. Sometimes when people get divorced or separated they are still able to get along pretty well. How well would you say your parents get along with each other? (1 = very well, 4 = not at all well) 2. How often would you say they argue or disagree with each other? (1 = never, 4 = often) 3. How often do they show affection or say nice things to each other? (1 = often, 4 = never) The three items were summed to form an adoles- cent-report scale. Coefficient alpha was above .70 for both waves. Correlations between mother and child reports were above .45. The mother and adolescent instruments were standardized and for boys the association was .01. Given these low associations, the scales could not be assumed to be indicators of the same construct. Hence, in the re- conflict. gression analyses reported below, the models first use mother reports of father's parenting practices, and then are repeated using adolescent reports. Family income. The mothers were asked to report the amount of money that they had received dur- Contact with father. Adolescents were asked about how often they had contact with their father, either by phone or in person. The response format was: (1) no contact, (2) two or three times a year, (3) occasionally, less than once a month, (4) one weekend or day a month, (5) every other week, (6) one or two times a week, (7) three or more times a week, and (8) every day. Parental conflict. The process of marital disruption almost always involves a high degree of conflict and emotional turmoil for both parents and children. Given their strong feelings about these events, family members are apt to have difficulty providing unbiased reports about the quality of interaction that currently exists between the former spouses. In an effort to reduce this bias, mother and adolescent reports were combined to form an indicator of parental conflict. Mothers completed an eight-item scale concerned with how often during recent interactions her former spouse had engaged in behaviors such as shouting, criticizing, arguing, insulting, and calling names. The response format ranged from 1 summed to form a composite indicator of parental ing the previous year from employment, child support, alimony, government payments, their children's earning, etc. These amounts were summed to form a measure of total family income. This total was divided by number of family members to obtain a per capita measure. Child support payments. The mothers responded to the following question: During the past 6 months, how much child support has your former spouse paid per month on average? The women wrote in the dollar amount or "0" if they had received no financial assistance from their former husband. RESULTS In general, the adolescents in the sample had rather high contact with their fathers. Approximately 60% indicated that their father visited at least twice a month, and over half reported that they had weekly contact. For boys, frequency of father visitation had a bivariate correlation of -.07 with Externalizing Problems and .01 with Internalizing Problems. The correlations for girls were -.01 and .01, respectively. Thus, consistent with This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms 364 Journal of Marriage and the Family prior studies, there was no evidence that frequency of interaction with father influenced child adjustment. lems at both waves 1 and 2, but the coefficients are not statistically significant. Parental Conflict The variable for Father's Parenting assumed dren must have had some interaction with their child reports of Father's Parenting, but is not related to adolescent problems. Table 1 shows a similar pattern of findings for girls with two exceptions. First, Mother's Parent- father if they and their mother were to answer questions concerning the father's parenting prac- ing (TI) is related to Externalizing Problems at waves 1 and 2 but not Internalizing Problems. tices. The same was true for Parental Conflict. Second, the correlations between the indicators of father's parenting and child adjustment are just the reverse of those found for boys: Father's Parenting-CR (TI) is negatively related to Externalizing and Internalizing Problems, whereas Father's Parenting-MR (TI) is not related to adoles- that the adolescent had had at least one contact with his or her father in recent months. The chil- The former spouses must have had at least some contact in recent months if the mother and adoles- cent were to rate the level of conflict. Thus analy- ses involving these variables necessarily excluded those cases where either the child or the mother had not seen the father in the prior 3 months. There was almost complete overlap between these conditions; if the child had had contact with the father so had the mother, and vice versa. There was no evidence that fathers were more likely to terminate contact with children of a particular gender. Excluding the no-contact cases resulted in a deletion of approximately 20% of both the boys and girls at wave 1, leaving 74 boys and 83 girls. At wave 2 there were 63 boys and 71 girls who had had some contact with their father. Given that several of the variables involved multiple indicators, structural equation modeling with latent variables was an obvious method for assessing associations between constructs. Unfortunately the small number of cases precluded the use of this procedure. Therefore, as noted in the Measures section, multiple indicators were summed to form composite measures and analyses were performed using Pearson correlations and ordinary least squares regression. (TI) is negatively associated with mother and cent adjustment at either wave. Table 2 presents the bivariate correlations between the parent variables at wave 2 and adolescent externalizing and internalizing problems at waves 2 and 3. The table differs from Table 1 in that it does not contain the variable Father's Par- enting-MR because this data was not available at wave 2. With one exception, the pattern of correlations for boys largely parallels that reported in Table 1. The exception involves Father's Parenting-CR (T2) which is negatively related to Externalizing Problems at waves 2 and 3 and Internalizing Problems (T2). Consistent with the findings reported in Table 1, Mother's Parenting (T2) shows an inverse association with both External- izing and Internalizing Problems at waves 2 and 3. Family Income (T2) and Child Support (T2) show small negative associations with Internalizing and Externalizing Problems at waves 2 and 3, although only a few of the coefficients are statisti- cally significant. Parental Conflict (T2) is nega- tively associated with Father's Parenting-CR Correlations Between Parental Behavior and Child Adjustment Table 1 presents the means and standard devia- tions and the bivariate correlations between the parent variables at wave 1 and adolescent externalizing and internalizing problems at waves 1 and 2. For boys, the table shows that the Mother' s Parenting at time 1 (TI) and mother's report of father's parenting-Father's Parenting-MR (Ti)-are inversely related to Internalizing and Externalizing Problems at waves 1 and 2. Child's report of father's parenting-Father's ParentingCR (T1)-on the other hand, is not related to the adjustment variables. Family Income (T1) and Child Support (T1) show small negative associations with Internalizing and Externalizing Prob- (T2), but is not related to adolescent problems. Table 2 shows a pattern of correlations for girls that closely resembles that for boys, and is similar to the findings reported for girls in Table 1. Both Mother's Parenting (T2) and Father's ParentingCR (T2) are negatively related to Externalizing Problems at waves 2 and 3. Income (T2) is negatively related to Internalizing Problems (T2). Parental Conflict (T2) shows an inverse association with Father's Parenting-CR (T3), but is not significantly related to any of the adolescent outcomes. Child Adjustment Regressed on Parental Behavior The intercorrelation between the parent variables shown in Tables 1 and 2 demonstrates the impos- This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms TABLE 1. CORRELATIONS BETWEEN PARENTAL VARIABLES AT WAVE 1 AND ADOLESCENT OUTCOMES AT WAVES 1 AND 2 BY G 1 2 3 4 5 6 7 8 9 10 1. Externalizing problems (Ti) - .82** .20 .28* -.31* .03 .38** .06 2. Externalizing problems (T2) .65** - .24 .43** -.21 .01 .42** .07 3. Internalizing problems (TI) .57** .47** - .49** .01 -.12 .26* .18 -. 4. Internalizing problems (T2) .59** .56** .65** - .01 -.04 .12 -.03 -. 5. Mother's parenting (TI) -.38** -.39** -.39** -.47** - .09 .15 .09 . 6. Father's par-MR (TI) -.29* -.30* -.13 -.30* .21 - .29* -.45** .1 7. Father's par-CR (T1) -.11 -.06 -.14 -.10 .08 .01 - -.33** -.1 8. Parental conflict (TI) .14 -.06 -.07 .15 .04 -.21 -.25* - .1 9. Family income (T1) -.14 -.16 .03 -.06 -.03 .03 -.13 .05 - .69 10. Child support (TI) -.13 -.16 -.01 -.07 .17 .21 .01 -.00 M .31a SD .71a 1.96 74.00 2.47 70.51 24.14 -.823a 23.25 12.40 3.53 3.81 43.67 6.62 .166a 1.56 28,161 16,552 Note: Correlations above the diagonal are for girls whereas those below the diagonal are for boys (n - 67 boys and 72 girls). aMean based on composite measure formed by summing standardized scores. *p < .05. **p < .O1. TABLE 2. CORRELATIONS BETWEEN PARENTAL VARIABLES AT WAVE 2 AND ADOLESCENT OUTCOMES AT WAVES 2 AND 3 BY GEN 1 2 3 4 5 6 7 8 9 1. Externalizing problems (T2) - .85** .43** .19 -.31** -.31* -.08 2. Externalizing problems (T3) .85** - .40** .40** -.36** -.32** .02 3. Internalizing problems (T2) .54** .42** - .45** -.19 -.18 .10 -.24 4. Internalizing problems (T3) .65** .58** .70** - -.08 -.13 .07 -.1 5. Mother's parenting (T2) -.50** -.39** -.48** -.33* - .13 .02 6. Father's par-CR (T2) -.30* -.35** -.27* -.18 .35* - -.27* .1 7. Parental conflict (T2) -.03 -.06 .27* .24 -.02 .02 - .11 8. Family income (T2) -.25 -.29* -.24 -.29* .11 -.22 -.19 - 75 9. Child support (T2) -.24 -.31* -.14 -.23 .01 .03 .04 .4 M SD .64a .69a 2.48 72.06 2.82 70.18 26.48 -.31a 22.36 43.83 3.52 6.22 -.42a 1.43 Note: Correlations above the diagonal are for girls and those below the diagonal are for boys (n = 54 boys and 67 girls). aMean based on composite measure formed by summing standardized scores. *p? .05. *p < .01. This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms 23,066 12,367 366 Journal of Marriage and the Family sibility of assessing the contribution of any one of the factors to child adjustment without controlling for the effects of the others. Multiple regression procedures were used to perform this multivariate analysis. Adolescent adjustment at wave 2 was re- gressed on wave-1 measures of the explanatory variables, and adolescent adjustment at wave 3 was regressed on wave-2 measures of the explanatory variables. The regressions were run twice, once without controlling for earlier levels of adolescent adjustment and a second time controlling for prior levels. Controlling for earlier levels of adjustment would seem to provide the strongest evaluation of the extent to which the explanatory variables predict change in adolescent adjustment. However, it was unlikely that this approach would produce many, if any, significant effects. First, the impact of parental behavior on child development is usually assumed to be immediate rather than lagged. Therapeutic interventions to enhance parental monitoring and discipline, for example, tend to table presents standardized regression coefficients (3) and their associated p values. The left half of the table reports the findings for externalizing problems. Without controlling for Externalizing Problems (TI), the association between Mother's Parenting (TI) and Externalizing Problems (T2) is significant for boys (-.38, p ? .01) and approaches significance for girls (-.18, p = .09). Father's Parenting-MR (Ti) shows a significant effect for boys (-.25), but not for girls. None of the other explanatory variables approaches signifi- cance for either gender. When Externalizing Problems (Ti) is added to the regression equation, none of the parent variables are significant for girls. The same is true for boys except that Fa- ther's Parenting-MR (TI) and Parental Conflict (T1) approach significance (p = .10). The right side of Table 3 displays the results obtained for Internalizing Problems. Without In- ternalizing Problems (T1) in the model, both produce rather instantaneous changes in child behavior (Kazdin, 1984; Patterson, 1982). Therefore Mother's Parenting (T1) and Parental Conflict (T1) are related to Internalizing Problems (T2) for boys (-.48 and .18, respectively). Controlling for Internalizing Problems (T1) does not eliminate one would expect parental behavior at one time these associations, and has the effect of increasing period to have little effect on child adjustment at a later point beyond its impact on adjustment during the earlier period. Second, as shown in Tables 1 and 2, Internalizing and Externalizing Problems are quite stable over time. Indeed the stabilities in is not in the model. No other factors show an ef- some cases are greater than .80. Thus removing fect. the effect of earlier levels of adjustment leaves lit- tle variance to be explained by the parent variables. Finally, analyses (not shown) demonstrate that the explanatory variables are also rather stable between waves 1 and 2 (e.g., correlations of .70 and .62 for mother's parenting for boys and girls, respectively). Thus, while parenting practices, income, parental conflict, and child adjustment undoubtedly changed dramatically during the weeks immediately following marital breakup, neither the parent variables nor the child the significance of Father's Parenting-MR (T1) from a probability of .40 to .09. For girls, the ef- fect of Family Income (TI) approaches significance whether Externalizing Problems (Ti) is or Table 4 shows the results obtained when child reports of father's parenting are substituted for mother reports. The findings contrast with those reported in Table 3 in that Father's Parenting-CR (Ti) is not related to Externalizing Problems (T2) for boys but is for girls. Indeed, the effect of this variable approaches significance for girls even after controlling for Externalizing Problems (Ti). The rest of the findings largely parallel those in Table 3. Although the results for the regression analyses adjustment variables show extreme fluctuations reported in Tables 3 and 4 differ somewhat de- over the postdivorce period that is the focus of the pending on the source of information (mother ver- study. Rather, both parental behavior and child adjustment appear to be rather slowly evolving processes. Given this situation, the explanatory sus child) used to measure father's parenting, analysis showed that this was not the case for the other factors were not expected to have a significant effect after controlling for prior scores on the outcome variables. Table 3 shows the results obtained by regressing adolescent problems at wave 2 on the parent variables assessed at wave 1. The analysis used father's parenting as reported by mothers. The variables for which multiple sources of data were available. Although not shown for the sake of par- simony, the effect of Mother's Parenting and Parental Conflict on adolescent adjustment remained basically the same when the different sources of information used to form these composite measures were treated as separate indicators. This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms TABLE 3. ADOLESCENT ADJUSTMENT AT WAVE 2 REGRESSED ON PARENTAL BEHAVIOR AT WAVE 1 USING MOTHER REPORTS OF Externalizing Problems (T2) Internalizing Prob Boys Girls Boys G Variables 3 p Value 3 p Value 3 p Value 3 p Value 3 p Value [3 p Value 3 Mother's parenting (T1) -.38 .01 .11 .21 -.18 .09 -.09 .20 -.48 .01 -.20 .0 Father's parenting-MR (Ti1) -.25 .01 -.13 .10 -.05 .68 -.04 .62 -.08 .40 -.13 .0 Parental conflict (T1) -.12 .21 -.12 .10 .13 .28 -.03 .75 .18 .06 .18 .01 Family income (T1) -.07 .50 -.01 .88 -.15 .27 -.06 .50 -.07 .49 -.04 .57 Child support (T1) -.02 .85 -.02 .77 -.01 .98 -.01 .87 .03 .79 .01 .98 Child adjustment (T1) - - .65 .01 - - .82 .01 - - .60 .0 TABLE 4. ADOLESCENT ADJUSTMENT AT WAVE 2 REGRESSED ON PARENTAL BEHAVIOR AT WAVE 1 USING ADOLESCENT REPORTS Externalizing Problems (T2) Internalizing Prob Boys Girls Boys G Variables 3 p Value [3 p Value [3 p Value 3 p Value [3 p Value [3 p Value 3 Mother's parenting (T1) -.41 .01 .16 .13 -.13 .14 -.08 .33 -.48 .01 -.25 .0 Father's parenting-CR (TI) -.06 .61 .09 .33 -.40 .01 -.14 .07 -.04 .72 .03 .7 Parental conflict (T1) -.02 .85 -.09 .36 -.03 .76 -.04 .63 .18 .10 .22 .0 Family income (T1) -.13 .27 -.06 .51 .04 .75 -.01 .89 -.08 .46 -.08 .38 Child support (TI) -.06 .63 -.03 .74 -.07 .64 -.02 .82 .04 .70 .01 .94 Child adjustment (TI) - .60 .01 - - .80 .01 - - .57 .0 TABLE 5. ADOLESCENT ADJUSTMENT AT WAVE 3 REGRESSED ON PARENTAL BEHAVIOR AT WAVE 2 USING ADOLESCENT REPORTS OF F Externalizing Problems (T2) Internalizing Prob Boys Girls Boys G Variables p Value [3 p Value 3 p Value [3 p Value 3 p Value ) p Value f Mother's parenting (T2) -.25 .05 -.08 .37 -.29 .01 -.07 .28 -.27 .05 -.01 .8 Father's parenting-CR (T2) -.33 .01 -.17 .04 -.27 .02 -.03 .68 -.12 .38 -.01 .9 Parental conflict (T2) -.09 .43 -.05 .46 -.03 .77 .09 .22 .22 .07 .06 .5 Family income (T2) -.25 .07 -.12 .18 -.21 .24 -.19 .06 -.16 .30 -.08 .50 Child support (T2) -.19 .16 -.06 .42 .19 .27 .20 .05 -.19 .21 -.11 .32 Child adjustment (T2) - .79 .01 - .80 .01 - - .62 .0 This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms 368 Journal of Marriage and the Family Table 5 presents the results obtained by regressing adolescent problems at wave 3 on the parent variables assessed at wave 2. Without con- the interaction terms showed a significant effect, suggesting that the impact of the parent variables is limited to their main effects. trolling for Externalizing Problems (T2), the asso- ciation between Mother's Parenting (T2) and Externalizing Problems (T3) is significant for boys (-.25, p = .05) and girls (-.29, p = .01). Father's Parenting-CR (Ti) also shows a significant effect for both boys (-.33, p = .01) and girls (-.27, p = .02). Family Income approaches significance for boys (-.25, p = .07). None of the other explanatory variables approaches significance for either gender. Father's Parenting-CR (T2) continues to have a significant effect for boys even after Exter- nalizing Problems (T2) is added to the regression equation. For girls, Family Income (T2) and Child Support (T2) emerge as significant once the effect of Externalizing Problems (T2) is controlled. However, the effect of Family Income (T2) is negative while the coefficient for Child Support (T2) is positive. Table 2 shows that there is a high degree of multicollinearity for these two variables (r = .75), and hence it was possible that the regression coefficients for these variables are misleading given this high association. Indeed, when the model was run with one or the other of the variables deleted neither showed a significant effect. Turning to boys' Internalizing Problems, Mother's Parenting (T2) has a significant effect (-.27, p = .05) and the coefficient for Parental Conflict (T2) approaches significance (.22, p = .07) when Internalizing Problems (T2) is not in the model. None of the explanatory variables shows a significant association once Internalizing Problems (T2) is added to the regression equation. For girls, none of the explanatory variables has a significant effect regardless of whether In- ternalizing Problems (T2) is included in the model. The regressions reported in Tables 3, 4, and 5 examine the main effects of the parent variables. It is possible that these factors interact with each other in influencing adolescent problems. For example, involvement by the father may serve to compensate for low control by the mother, or father's parenting may only have a positive effect when conflict with the mother is low. These hypotheses were tested by forming multiplicative interaction terms. Mother's Parenting was multiplied by Father's Parenting, Father's Parenting by Parental Conflict, and Mother's Parenting by Parental Conflict. These interaction terms were entered into the regressions one at a time. None of Parental Behavior Regressed on Child Adjustment In an attempt to assess the extent to which child maladjustment influences parental behavior, each of the parent variables (viz., Mother's Parenting, Father's Parenting, and Parental Conflict) as assessed at wave 2 was, in turn, regressed on wave1 measures of Adolescent Internalizing and Ado- lescent Externalizing Problems as well as the other parent variables measured at wave 1. Thus Mother's Parenting at wave 2, for example, was regressed on wave-1 assessments of Adolescent Internalizing Problems, Adolescent Externalizing Problems, Father's Parenting, Parental Conflict, Family Income, Child Support, and Mother's Par- enting. These regressions were then repeated using wave-2 measures to predict parental behavior at wave 3. The results of these analyses are presented in Table 6. For the sake of brevity, the only regression coefficients presented are those representing relationships between the three de- pendent variables--Mother's Parenting, Father's Parenting, and Parental Conflict-and both Adolescent Internalizing and Externalizing Problems. Table 6 shows, as expected, that externalizing problems has more of an effect on parental behavior than internalizing problems. None of the coef- ficients reported in Table 6 between Adolescent Internalizing Problems and either Mother's or Father's Parenting is statistically significant. On the other hand, there is a significant inverse relation- ship between Externalizing Problems (Ti) and Mother's Parenting (T2) for both boys (-.27) and girls (-.22). For boys, Externalizing Problems (TI) also shows a strong negative association with Father's Parenting (T2) (-.48). There are no significant relationships, however, between Externalizing Problems (T2) and Mother's or Father's Parenting measured at wave 3, suggesting that the coercive impact of child conduct problems on parental behavior may be most evident during the transition period immediately following divorce. Internalizing Problems is not related to subsequent levels of Parental Conflict in any of the regressions, and for girls, Externalizing Problems has no effect at either wave. For boys, Externalizing Problems (T2) shows a significant association with Parental Conflict, although Externalizing Problems (T1) does not. The significant coeffi- This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms TABLE 6. STANDARDIZED REGRESSION COEFFICIENTS FOR THE PARENTAL BEHAVIOR CONSTRUCTS REGRESSED ON PRIOR ASSESSMENTS OF ADOLESCENT EXTERNALIZING AND INTERNALIZING PROBLEMS Mother's Parenting Father's Parenting Paren T1-* T2a T2---- T3b T1---- T2a T2---- T3b T1--- Adolescent Outcomes p p Value p p Value p Value p Value p p Boys Externalizing problems -.29 .04 -.20 .14 -.48 .01 -.18 .22 Internalizing problems .10 .42 .03 .86 .22 .17 .03 .84 Girls Externalizing problems -.22 .05 .05 .66 -.02 .89 .03 .80 Internalizing problems .01 .94 .23 .11 -.10 .42 -.06 .60 aParental behavior at Wave 2 regressed on externalizing and internalizing problems at Wave 1. bParental behavior at Wave 3 regressed on externalizing and internalizing problems at Wave 2. This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms 370 Journal of Marriage and the Family cient is negative (-.38), suggesting that divorced test of the causal priorities assumed between the parents reduce their conflict with each other in re- parent variables and child outcomes than the sponse to antisocial behavior on the part of their sons. It should be noted, however, that Externalizing Problems is related to Parental Conflict in only one regression, and that this regression involves Externalizing Problems measured at wave 2, whereas it was Externalizing Problems mea- cross-sectional procedures utilized in most studies. The parenting variables tended to predict subsequent adolescent problems as long as measures of prior level of adolescent adjustment were not in the model. In most instances controlling for sured at wave 1 that is associated with Mother's ent variables. Other studies have reported similar findings (Agnew, 1991; Anderson, Lindner, & Bennion, 1992). This finding was anticipated and is considered to be a function of two factors, one and Father's Parenting. Thus it may be that this coefficient is a function of random error. SUMMARY OF FINDINGS Overall, five findings are suggested by the regression analyses. First, it appears that quality of father's parenting is related to externalizing problems for boys and girls, although the results differ somewhat depending on the source of data used to assess father's parenting. Second, quality of mother's parenting shows an association with externalizing problems for boys and girls, and is also related to internalizing problems for boys. Third, parental conflict is associated with internalizing problems for boys but not girls. Fourth, family income and child support do not appear to have an effect on child adjustment once the influ- ence of parenting practices and parental conflict are taken into account. Finally, adolescent externalizing problems appear to reduce the quality of mother's parenting for both boys and girls, and to diminish father's parenting in the case of boys. However, these effects may be limited to the transition period immediately following the divorce. DIScusSION The present study avoided the limitations inherent in many prior studies of children's adjustment to divorce by using longitudinal data, utilizing mul- tiple sources of information to build measures, and including rarely investigated factors such as the parenting practices of nonresidential fathers. The results provided no evidence for the view that the impact of parental behavior on child adjustment either increases or wanes with time since di- vorce. The results were very similar for both time periods: mother's parenting, father's parenting, and, to a lesser degree, parental conflict showed significant effects at both waves, whereas the other factors did not. The regression analyses involved regressing child adjustment on the parent variables measured a year earlier. This approach provided a stronger earlier adjustment eliminated the effect of the par- theoretical and the other statistical. Theoretically, the impact of parental behavior on child development is usually assumed to be immediate rather than delayed. This assumption is supported by intervention studies reporting that modifications in parenting practices produce changes in child behavior within a matter of days or weeks (Kazdin, 1984; Patterson, 1982). Therefore one would expect parental behavior at a particular point in time to have little effect on child adjustment a year later beyond its impact on adjustment during the earlier period. Statistically, both internalizing and externalizing problems are quite stable over time. Thus removing the effect of earlier levels of adjustment leaves little variance to be explained by the parent variables. The variable that was most strongly and consistently related to adolescent adjustment was mother's parenting. This finding is consonant with past research. Studies have consistently shown low parental monitoring and inept discipline to be related to child conduct problems such as difficulties at school and delinquency (Macco- by, 1992; Maccoby & Martin, 1983). Further, there is strong evidence that single parents tend to make fewer demands on children and utilize less effective disciplinary strategies than married par- ents (Astone & McLanahan, 1991; Furstenburg & Nord, 1985; Hetherington et al., 1982). Together these two lines of research suggest that differences in parental control may explain much of the variation in externalizing problems among children living in single-parent families. Findings from the present study corroborate this idea in that quality of mother's monitoring and discipline was related to externalizing problems for both boys and girls. For boys, mother's monitoring and discipline was also related to internalizing problems. Indeed, the magnitude of the coefficients was quite large. While there is an intuitively obvious connection between parental discipline and conduct This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms Adjustment ofAdolescent Children 371 problems, it is less clear why quality of parental discipline would be related to psychological distress, and only for boys. Past research has estab- ther's parenting was not related to internalization problems for either sons or daughters. Thus the primary consequence of involvement by nonresi- lished that boys have a greater propensity for con- dential father appeared to be a reduction in the adolescent's risk for conduct problems. This suggests that nonresidential fathers who continue to duct problems than girls (Loeber & LeBlanc, 1990). It may be that young males understand the dangers that low parental control poses for them, and hence are more likely than adolescent females to interpret low control as an absence of parental concern and caring. Or it may be, as past studies have reported, that adolescent males, whether living in a single- or two-parent family, tend to resist control attempts by their mothers (Baumrind, 1991). Thus it may be that many of the single mothers in the present study who scored low on effective discipline had reduced their level of monitoring and discipline in response to the anger and resistance shown by their sons. If this is the case, the relationship between low maternal control and internalizing problems fill the role of parent may serve to counter the low control and inept discipline manifested by some single mothers. While overall the findings indicate that involvement by nonresidential fathers diminishes an adolescent's involvement in conduct problems, this conclusion is tempered by the fact that the re- sults varied depending on the reporter used to assess father's parenting. In the case of girls, adolescent reports about father's parenting at either waves 1 or 2 predicted externalization problems, whereas mother reports at wave 1 did not. For boys, on the other hand, mother reports at wave 1 for boys may be spurious due to their common as- and adolescent reports at wave 2 were related to externalization problems whereas adolescent re- sociation with mother-son conflict. The fact that ports at wave 1 were not. It is difficult to interpret mother's control is associated with internalizing problems for boys but not girls is intriguing because it suggests that the functions or consequences of various parental behaviors may differ by gender of the adolescent. This finding needs to be replicated with other samples, however, since this pattern of findings. One possibility is that mothers are more accurate reporters of fathers' parenting when the child is male. Single mothers often have a hard time controlling adolescent males (Hetherington, 1987) and hence they may be very aware (and appreciative) of any assistance provided by the father. Unfortunately, mothers' reports about fathers' parenting were not available for wave 2 so analysis could not be per- it may merely represent sampling error. Many past studies of children of divorce have been based on the simple assumption that children do better when they have regular contact with their nonresidential father. Research has failed to support this idea (Amato, 1993; Emery, 1988). A few studies have gone beyond focusing on frequency of visitation to consider the affective quality of the father-child relationship. However, these studies have also failed to find an association between father involvement and child adjustment (e.g., Furstenberg et al., 1987; Macco- by et al., 1993). In response to these negative findings, Furstenberg and Cherlin (1991) have speculated that noncustodial fathers only exert an influence on child development to the extent that they remain actively involved in the role of par- ent. Findings from the present study provide some support for this idea. Father's involvement in the role of parent was negatively related to externalizing problems in two out of three tests for both boys and girls. In- deed, the coefficient for this variable was often significant (p < .05), or at least approached significance (p < .10), even after controlling for prior scores on externalizing problems. In contrast, fa- formed to determine whether mothers' accounts of fathers' parenting practices continue to predict the behavior of sons at subsequent waves. If for no other reason than shared method vari- ance, one would expect adolescent reports of paternal behavior to be a better predictor of adolescent self-reports of conduct problems than mother reports of paternal behavior. Curiously this was the case only for girls. Perhaps in the months immediately following marital breakup, adolescent males are so disturbed by feelings of abandonment by their father that they cannot objectively report on the quality of his parenting practices. Hence their wave-i reports about their father's behavior are not related to conduct problems. As time passes they may become more objective, with the result being that boys' reports about their father's parenting obtained at wave 2 are strongly associated with conduct problems. Such ideas are, of course, mere speculation. The finding that the effect of father's parenting varied depending on the reporter used to assess father's parenting is disconcerting and precludes This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms 372 Journal of Marriage and the Family drawing firm conclusions from the analyses concerning the effects of paternal involvement. However, at a minimum, the fact that in four out of six tests father's involvement in the role of parent had an effect on conduct problems suggests that it would be fruitful for future studies to focus on the parental behaviors of nonresidential parents rather than merely considering frequency of contact or warmth of the parent-child relationship. It is often argued that recurrent friction be- tween parents causes psychological distress among children of divorce. Consistent with this contention, several studies have found that persistent conflict between former spouses is related to depression and anxiety among children (Amato, 1993; Emery, 1988). Most of these studies, however, do not control for the parenting practices of the parents. This is an important omission since research on two-parent families indicates that the effect of parental conflict on child adjustment may be indirect through its disruptive impact on parenting practices (Conger et al., 1992). The present study found that after controlling for qual- ity of parenting by both mothers and noncustodial fathers, parental conflict was related to internaliz- ing problems for adolescent males but not females. It is not clear why this gender difference might exist. Perhaps daughters are able to cope with parental conflict by siding with the mother and reducing emotional investment in the father, whereas sons continue to identify with their nonresidential fathers and therefore are troubled by parental conflict that threatens the father-son relationship. Past studies of two-parent families have shown that economic hardship has a coercive effect on parental behavior (Elder & Caspi, 1988; Simons et al., 1992). Findings from the present study suggest that similar processes may operate in single- parent families as well. Although there was a bivariate association between family income and adolescent adjustment, the two variables were not significantly related once the effects of mother's parenting were controlled. This result suggests that the impact of financial strain on children of divorce may be indirect through its disruptive effect on the parenting of the custodial parent. Finally, although the study was primarily con- cerned with the impact of parental behavior on child adjustment, consideration was also given to the possibility that child adjustment problems influence parental behavior. There was no evidence that internalizing problems have an impact on the dimensions of parental behavior included in the analysis. Externalizing problems, on the other hand, had a coercive effect on mother's monitoring and discipline of both sons and daughters, and on noncustodial father's parenting of sons. Interestingly enough, this effect was found for wave-1, but not wave-2, assessments of conduct problems. This suggests that child antisocial behavior may exert its greatest influence during the transi- tion period immediately following divorce. This is a time when parents are adjusting to new parental roles. The custodial parent needs to accommodate his or her childrearing approach to life as a single parent whereas the nonresidential parent must define the role of parent given his or her more limited access to the child. Both parents are likely to find that the realities of their new living situation make parenting a more difficult en- deavor. During this period of transition and role redefinition, mothers and fathers may be particu- larly vulnerable to aversive child behaviors that discourage parental involvement. Combining this idea with the findings discussed earlier, it appears that divorced families may often encounter an ugly circular pattern where children develop conduct problems as a result of disruptions in parent- ing occurring during and immediately following parental divorce. These child behavior problems, in turn, may operate to further discourage effective parenting, with the result being an escalation of child antisocial behavior. Although the present study avoided some of the weaknesses inherent in much of the past research on children in single-parent families, it also contained several limitations. Three are particularly noteworthy. First, the sample was small and consisted largely of families living in small towns. Although there is no reason to believe that the effects of the parent variables would be differ- ent for families living in large metropolitan areas, the findings need to be replicated using a larger, more urban sample. Second, measures of the parenting practices of the noncustodial father were based on reports by his children and former spouse. Both of these parties are apt to possess strong feelings that might serve to bias their perceptions of his behavior. Future research needs to examine the extent to which father self-reports are related to mother and child reports of his parental involvement and to assessments of child adjustment. Finally, the age range of the children in the sample was restricted to middle adolescence. There is a need to replicate the study using samples containing younger children. The effect of This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms Adjustment ofAdolescent Children 373 some of the variables may well differ by age of the child. It may be, for example, that involvement by nonresidential fathers is most critical during the turbulent adolescent period. On the other hand, parenting by nonresidential fathers may have its most potent effect during the highly formative preschool and early elementary school years. The issue of the extent to which the impact of the parent variables interacts with the developmental level of the child could not be addressed with the sample used in this study. NOTE This article is part of the Iowa Single Parent Project, a panel study supported by the National Institute of Men- tal Health (MH48165) and the National Institute of Child Health and Human Development (HD27724). Journal Paper No. 15281 of the Iowa Agriculture and Home Economic Experiment Station, Ames, IA. Project No. 2931. Correspondence concerning this article should be addressed to Ronald L. Simons. REFERENCES Agnew, R. (1991). A longitudinal test of social control theory and delinquency. Journal of Research in Crime and Delinquency, 28, 126-156. Amato, P. R. (1993). Children's adjustment to divorce: Theories, hypotheses, and empirical support. Journal of Marriage and the Family, 55, 23-38. Amato, P. R., & Keith, B. (1991). Parental divorce and the well-being of children: A meta-analysis. Psychological Bulletin, 110, 26-46. Anderson, E. R., Lindner, M. S., & Bennion, L. D. (1992). The effect of family relationships on adoles- cent development during family reorganization. Monographs of the Society for Research in Child Development, 57, 178-199. Arendell, T. (1986). Mothers and divorce: Legal, economic, and social dilemmas. Berkeley, CA: University of California Press. Astone, N. M., & McLanahan, S. S. (1991). Family structure, parenting practices, and high school completion. American Sociological Review, 56, 309-320. Bakeman, R., & Adamson, L. B. (1984). Coordinating attention to people and objects in mother-infant and lationships: Mutual influences (pp. 218-240). New York: Oxford University Press. Conger, R. D., Conger, K., Elder, G. H., Lorenz, F. O., Simons, R. L., & Whitbeck, L. B. (1992). A family process model of economic hardship and influences on adjustment of early adolescent boys. Child Development, 63, 526-541. Derogatis, L. R. (1983). SCL-90-R: Administration, scoring, and procedures manual-II. Towson, MD: Clinical Psychometric Research. Elder, G. H., & Caspi, A. (1988). Economic stress in lives: Developmental perspectives. Journal of Social Issues, 44, 25-45. Elder, G. H., Van Nguyen, T., & Caspi, A. (1985). Linking family hardship to children's lives. Child Development, 56, 361-375. Elliott, D. S., Huizinga, D., & Ageton, S. S. (1985). Explaining delinquency and drug use. Beverly Hills, CA: Sage. Elliott, D. S., Huizinga, D., & Millard, S. (1989). Multiple problem youth: Delinquency, substance use, and mental health problems. New York: Springer-Verlag. Emery, R. E. (1988). Marriage, divorce, and children's adjustment. Newbury Park, CA: Sage. Furstenberg, F. F., & Cherlin, A. J. (1991). Divided families: What happens to children when parents part. Cambridge, MA: Harvard University Press. Furstenberg, F. F., Morgan, S. P., & Allison, P. D. (1987). Paternal participation and children's wellbeing after marital dissolution. American Sociological Review, 52, 695-701. Furstenberg, F. F., & Nord, C. W. (1985). Parenting apart: Patterns of childrearing after marital disruption. Journal of Marriage and the Family, 47, 893- 904. Hartmann, D. P. (1977). Considerations in the choice of interobserver estimates. Journal of Applied Behavioral Analysis, 10, 103-119. Hetherington, E. M. (1987). Family relations six years after divorce. In K. Pasley & M. Ihinger-Tallman (Eds.), Remarriage and stepparenting: Current research and theory (pp. 185-205). New York: Guil- ford. Hetherington, E. M., Cox, M., & Cox, R. (1976). Divorced fathers. Family Coordinator, 25, 417-428. Hetherington, E. M., Cox, M., & Cox, R. (1982). Effects of divorce on parents and children. In M. E. Lamb (Ed.), Nontraditional families: Parenting and child development (pp. 233-285). Hillsdale, NJ: Lawrence Erlbaum Associates. peer-infant interaction. Child Development, 55, Kazdin, A. E. (1984). Treatment of conduct disorders. In Baumrind, D. (1991). Effective parenting during the early adolescent transition. In P. A. Cowan & M. Hetherington (Eds.), Family Transitions (pp. 111-164). Hillsdale, NJ: Lawrence Erlbaum Associ- chotherapy research: Where are we and where 1278-1289. ates. Bumpass, L. L., & Sweet, J. A. (1989). Children's experience in single-parent families: Implications of cohabitation and marital transitions. Family Planning Perspectives, 21, 256-260. Caspi, A., & Elder, G. H. (1988). Emergent family patterns: The intergenerational construction of problem behavior and relationships. In R.A. Hinde & J. Stevenson-Hinde (Eds.), Relationships within re- J. B. W. Williams & R. L. Spitzer (Eds.), Psy- should we go? (pp. 3-27). New York: Guilford. Loeber, R., & LeBlanc, M. (1990). Toward a developmental criminology. In Michael Tonry & Norval Morris (Eds.), Crime and justice: A review of research (Vol. 12, pp. 375-473). Chicago: University of Chicago Press. Maccoby, E. E. (1992). The role of parents in the socialization of children: An historical overview. Developmental Psychology, 28, 1006-1017. Maccoby, E. E., Buchanan, C. M., Mnookin, R. H., & Dornbusch, S. M. (1993). Postdivorce roles of moth- This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms 374 Journal of Marriage and the Family ers and fathers in the lives of their children. Journal of Family Psychology, 7, 24-38. Maccoby, E. E., & Martin, J. A. (1983). Socialization in Student the context of the family: Parent-child interaction. In P. Mussen (Ed.), Handbook of Child Psychology (pp. 1-101). New York: Wiley. Subscription Rates for NCFR journals! McGruder, B., Lorenz, F. 0., Hoyt, D., Ge, X. J., & Montague, R. (1992). Dimensions of parenting: A technical report. Iowa State University, Center for Family Research in Rural Mental Health, Ames, IA. McLanahan, S. S., & Booth, K. (1989). Mother-only families: Problems, prospects, and policies. Journal of Marriage and the Family, 51, 557-580. Melby, J., Conger, R., Book, R., Rueter, M., Lucy, L., Repinski, D., Ahrens, K., Black, D., Brown, D., Huck, S., Mutchler, L., Rogers, S., Ross, J., & Journal of Marriage and the Family (JMF) and Family Relations: Journal of Applied Family and Child Studies (FR) are now available to students at special low rates. Stavros, T. (1990). The Iowa family interaction cod- Students save up ing manual. Ames, IA: Iowa Youth and Families Project. to 50% Morgan, L. A. (1991). After marriage ends: Economic consequences for mid-life women. Newbury Park: Sage. Patterson, G. R. (1982). Coercive family process. Eugene, OR: Castalia. Patterson, G. R., Reid, J. B., & Dishion, T. J. (1992). Antisocial boys. Eugene, OR: Castalia. Raschke, H. J. (1987). Divorce. In M. B. Sussman & S. K. Steinmetz (Eds.), Handbook of Marriage and the Family (pp. 597-624). New York: Plenum. Schwarz, J. C., Barton-Henry, M. L., & Pruzinsky, T. (1985). Assessing childrearing behaviors: A comparison of ratings made by mother, father, child, and sibling on the CRPBI. Child Development, 56, 462- over individual rates. Student Rates: JMF only $25/year FR only $20/year These discounted rates allow you to save time, too! Have the journals delivered directly to your home or work. You no longer have to wait for the library copy to become 479. Simons, R. L., Beaman, J., Conger, R. D., & Chao, W. (1993). Stress, support, and antisocial behavior trait as determinants of emotional well-being and parenting practices among single mothers. Journal of Marriage and the Family, 55, 385-398. available. Simons, R. L., Lorenz, F. O., Conger, R. D., & Wu, C. (1992). Support from spouse as mediator and moderator of the disruptive influence of economic strain on parenting. Child Development, 63, 1282-1301. Simons, R. L., Lorenz, F. O., Wu, C., & Conger, R. D. (1993). Marital and spouse support as mediator and moderator of the impact of ecomonic strain upon parenting. Developmental Psychology, 29, 368-381. Suen, H. K., & Ary, D. (1989). Analyzing quantitative behavioral observation data. Hillsdale, NJ: Lawrence Erlbaum Associates. Velicer, W. R., Govia, J. M., Cherico, N. P., and Corriveau, D. P. (1985). Item format and the structure of the Buss-Durkee hostility inventory. Aggressive Behavior, 11, 65-82. Prfesos Many of you already use JMF and FR as supplemental texts. Now they are more affordable for your students. Please share this special rate information with your students! Encourage them to subscribe. They will reap the benefits of exposure to the latest original theory and applied research in the family field. If the journals are not currently a part of your courses, consider their value to your students. The new student rates make it possible to integrate journals into various family courses. Students receive 4 issues per year of JMF for only $25 and FR for only $20. This content downloaded from 131.247.112.3 on Mon, 27 Jun 2016 04:14:18 UTC All use subject to http://about.jstor.org/terms