Current theory (Olson & Gorall, 2003) and practice (Cowan, Powell, & Cowan, 1998; Minuchin, 1974) advocate emphasizing the linkages between overall family system functioning and parental behaviors. In contrast, minimal research has investigated the relationship between overall family system functioning and parental behaviors. Yet, research shows that both overall family system functioning and parental behaviors are positively related to adolescent well-being (Grotevant, 1998). Additional research is needed to examine empirical relationships between overall family functioning and parental behaviors.

Overall family system functioning describes the “invisible web” of complex interaction patterns that regulate the day-to-day interactions among family members (Minuchin, 1974). Olson and Gorall (2003) noted that effective (or balanced) overall family system functioning includes moderate levels of cohesion and flexibility, a balance between closeness and individuality, egalitarian leadership, democratic approaches to discipline, and uses positive communication skills. Olson et al. (1992) identified four types of overall family functioning: (a) balanced families which tend to report moderate levels of both cohesion and adaptability, (b) moderately balanced families who report slightly higher or slightly lower than moderate levels of both cohesion and adaptability, (c) mid-range families that tend to report a slightly higher or slightly lower than moderate level of either cohesion or adaptability with an extreme score on the remaining dimension, and (d) extreme families who report extremely high or extremely low levels of both cohesion and adaptability. Each individual in the family “constructs” his or her own understanding of overall family and subsystem dynamics (Larsen & Olson, 1990).

Overall family functioning provides an emotional and organizational environment for parental behaviors in dyadic (or subsystem) relationships (Minuchin, 1974). Parent-adolescent dyadic subsystems help overall family systems carry out family functions or accomplish family goals. In parent-adolescent subsystems, parents are primarily responsible for fulfilling family functions such as providing nurturance and guidance (Small & Eastman, 1991).

Recent research supports the potential of investigating relationships between overall family system functioning and parenting variables. Using reports from mothers of first and third grade students, Mupinga, Garrison, and Pierce (2002) found that balanced and moderately balanced overall family functioning were positively related to authoritative parenting styles while balanced overall family functioning were negatively related to authoritarian parenting styles. Many parenting scholars emphasize the importance of distinguishing specific parental behaviors in research rather than using aggregate parenting styles (Peterson & Hann, 1999).

Scholarship focuses on support and control as two key dimensions of parental behavior. Parental support refers to nurturing adolescents through behaviors of parents such as encouragement, praise, general support, or physical affection (Peterson & Hann, 1999). Parental support tends to be positively related to aspects of adolescent well-being such as general competence (Amato, 1989), identity achievement (Sartor & Youniss, 2002), academic achievement and self-esteem (Bean, Bush, McKenry, & Wilson, 2003), family life satisfaction (Henry, 1994), and career self-efficacy (Turner & Lapan, 2002). Parental support is negatively related to alcohol misuse (Barnes, Reifman, Farrell, & Dintcheff, 2000), identity or peer relationship problems (Marta, 1997), eating disorders (McVey, Pepler, Davis, Flett, & Abdolell, 2002), and depressed affect (Whitbeck, Conger, & Kao, 1993).

When adolescents perceive their parents as more supportive, overall family functioning may also be higher (Barber & Buehler, 1996). Balanced families, for example, provide a family emotional climate where parents foster a balance between separateness and connectedness. In turn, adolescents might perceive their parents as providing support as they seek to maintain a balance of separateness and connectedness with their families and peers. Further, balanced family systems allow flexibility in family rules and roles as the needs of the overall family system, subsystems, or individual family members change (Olson et al., 1992). Adolescents in balanced families may see their parents as more flexible in responding to the adolescents’ needs and interpret the parental behaviors as a manifestation of parental support. In sum, balanced families tend to create a warm, flexible atmosphere in the overall family system that facilitates adolescent perceptions of parental support. In contrast, extreme families are expected to be lower in parental support since the emotional bonding in the overall system tends to be minimal or excessive and interaction patterns do not allow families flexibility to navigate change.

Overall family system functioning may also relate to parental control or behaviors parents use with the intent of influencing the behaviors, attitudes, or feelings of the young (Peterson & Hann, 1999). Previous research indicates that some forms of parental control such as punitiveness) relate to decreased adolescent well-being while other forms of parental control such as induction or monitoring relate to increased adolescent well-being.

Parental punitiveness is form of parental control where parents use or threaten to use force to gain adolescent compliance (Peterson & Hann, 1999). While punitiveness may result in short-term control over adolescents, it may encourage long term resistance to parental control. In general, punitiveness is negatively associated with family life satisfaction (Henry, 1994), general social competence, moral development, self-esteem; and increased risk of substance use and delinquency (Peterson & Hann, 1999).

Parental induction is a form of control that communicates behavioral expectations to adolescents while using logical reasoning and consequences to guide behavior (Hoffman, 1970). Induction encourages adolescents to develop the capacity for self-regulation of behavior based upon principles or logic. Parental induction is positively related to adolescent general competence (Amato, 1989), aspects of moral development, self-esteem (Peterson & Hann, 1999), and empathic perspective taking (Henry et al., 1996).

Another control behavior is parental monitoring, or guidance based upon parents’ attention to and knowledge of their adolescents’ schedules, friends, activities, and interests (Dishion & McMahon, 1998). Monitoring allows parents to be involved and provide guidance while promoting healthy adolescent development. Parental monitoring is related to reduced risk of adolescent problem behaviors (Crouter & Head, 2002) and is positively related to psychosocial well-being (Salem, Zimmerman, & Notaro, 1998), identity achievement (Sartor & Youniss, 2002), health behaviors (Markey, Ericksen, Markey, & Tinsley, 2001), and greater parental enjoyment of the parent-adolescent relationships (Laird, Pettit, Dodge, & Bates, 2003).

Parents face the challenge of identifying and implementing control techniques that meet expectations with both the broader society and overall family system (Small & Eastman, 1991). The control approaches used by parents toward their adolescents are, to a large extent, reflections of interaction patterns in the overall family system. For example, adolescents in balanced families might be expected to report different parental behaviors than those in extreme types due to the variations in emotional bonds and time spent together. In addition, parents in balanced families might be more likely to use time-intensive control techniques such as induction or monitoring that show interest in and concern for adolescents’ development and well-being. In contrast, parents in extreme families might use less child-oriented control techniques such as punitiveness, which emphasize youthful compliance based upon parental authority.

Since balanced families allow some flexibility in family roles and responsibilities, parents in these families might be perceived as using individualized approaches to providing parental guidance or control. For example, parents who are seen as setting clear expectations and emphasizing the logical consequences of behavior are described as using parental induction as a control attempt focused on influencing the attitudes and thinking of adolescents. Parental monitoring allows adolescents to engage in normative developmental activities that often involve expanding the social space to allow for greater involvement with individual interests, friends, school, or work while maintaining sufficient overall parental control to meet societal and family expectations. Parental punitiveness might be more frequent in extreme families as parents emphasize compliance over negotiation and communication. In sum, adolescents who see their overall family functioning as balanced can be expected to perceive their parents as using control techniques such as induction or monitoring that show interest in the activities, friends and interests of the adolescents or setting clear expectations for behavior based upon logic and communication of logical consequences. In turn, adolescents in extreme families might tend to perceive their parents as using more authority-focused or punitive control techniques.

Demographic characteristics may explain variation in the strength of the relationship between adolescent reports of overall family system functioning and parental behaviors. Specifically, overall family functioning or parental behaviors may vary according to family form, sex of the adolescent, or sex of the parent (Peterson & Hann, 1999). Based on these ideas, the present study investigated the extent to which adolescents in intact families reports’ of parental behaviors distinguished among the types of reported overall family functioning, while controlling for the sex of the adolescent and sex of the parent.

Method

Participants were a subsample of 160 students from a larger study of 324 ninth and tenth grade students who responded to self-report questionnaires in the only public high school in three rural communities in a southwestern state. Subsample (i.e., 160 subjects who reported on 320 parents – 160 mothers, 160 fathers) selection was based upon reports of residing in intact family households with two biological or adopted parents. The subsample composition follows: mean age = 14.8 (range 14–17); 94 (58.8%) ninth grade students and 66 (41.2%) tenth grade students; 70 (43.8%) males and 90 (56.2%) females; 124 (76.3%) Caucasian, 17 (10.6%) Native American, 8 (2.5%) Hispanic, 4 (2.5%) African American, 4 (2.5%) Asian American, and 18 (5.6%) other race or not stated.

Measurement

Existing self-report instruments and standard fact sheet items (i.e., to assess the demographic characteristics of the participants) were used to collect the data.

Perceptions of overall family system functioning

Adolescent perceptions of overall family system functioning were assessed using FACES II, a 30 item Likert-type self-report scale (Olson et al., 1992). FACES II provides a linear assessment of overall family system functioning that combines the cohesion (16-item) and adaptability (14-item) subscale scores to categorize families into types of overall family functioning. FACES II assessed adolescent perceptions of the emotional bonding among family members and ability to balance regularity in fulfilling family responsibilities while modifying interaction patterns to accommodate specific family situations. Prior to the development of FACES IV (Olson & Gorall, 2003), FACES II was recommended over FACES III due to the greater number of items yielding a higher Cronbach's alpha, fewer problems with social desirability, and higher concurrent validity with other instruments measuring similar constructs (Olson et al., 1992). The participants were directed to respond to the items regarding the household that served as their primary residence. Sample items are: (a) “Family members feel closer to people outside the family than to other family members” (cohesion) and (b) “When problems arise, we compromise” (adaptability). Response choices ranged from 1 = strongly disagree to 5 = strongly agree. Olson et al.'s (1992) scoring was used to identify the types of overall family functioning, resulting in overall family functioning scores as follows: 1 = extreme, 2 = mid-range, 3 = moderately balanced, and 4 = balanced. Current data yielded a Cronbach's alpha of .85.

Perceptions of parental behaviors

Adolescent perceptions of parental behaviors were assessed by utilizing the support (4 items), induction (5 items), punitiveness (7 items), and monitoring (6 items) subscales of the Parental Behavior Measure (Peterson, 1982). This measure has been extensively refined through the perspectives of numerous scholars over three decades (Bush, Peterson, Cobas, & Supple, 2002), including the addition of parental monitoring items based on the work of Small (1990) and Barber, Olson, and Shagle (1994).

Participants responded to each parental behavior item twice, once for fathers and once for mothers. The subscales assessed the extent to which adolescents viewed each parent as providing emotional and affectional support (support subscale), aware of the adolescent's activities, interests, friends, and schedule (monitoring), providing guidance by using logical reasoning and explanation (induction), and using threats or force to exact compliance (punitiveness). Sample items include: (a) “This parent seems to approve of me and the things I do” (support); (b) “This parent tells me how good others feel when I do what is right” (induction); (c) “I tell this parent who I am going to be with when I go out” (monitoring); and (d) “This parent punishes me by not letting me do things that I really enjoy” (punitiveness). Response choices ranged from 1 = strongly disagree to 5 = strongly agree. The responses about each parental behavior were summed and scored separately for fathers and mothers and mean scores were calculated, yielding 320 scores from 160 adolescents on each parental behavior variable. Using the current data, the internal consistency reliability coefficients (Cronbach's alphas) were: .80 for support, .77 for induction, .72 for punitiveness, and .82 for monitoring.

Results

The goal of our study was to examine the extent to which adolescents’ perceptions of parental behaviors (support, monitoring, induction, and punitiveness) explained variation in adolescent perceptions of the type of overall family system functioning (balanced, moderately balanced, mid-range, extreme) after considering sex of adolescent and sex of parent. Dummy variables were created for sex of adolescent (0 = males, 1 = females) and sex of the parent (0 = fathers, 1 = mothers) about whom the adolescent reported (Stevens, 2002). The significance level for the analyses was set at p≤.05.

Bivariate correlations between the parental behaviors and overall family functioning showed significant relationships between adolescent perceptions of parental behaviors and overall family functioning (see Table 1). Overall family functioning was positively related to parental support, induction, and monitoring; while negatively related to parental punitiveness. Thus, the highest parental support, induction, and monitoring were reported in balanced families, whereas the highest punitiveness was found in extreme families.

Table 1 Bivariate correlations, means, and standard deviations (N=320)a
Full size table

Preliminary analyses of sex of parent and sex of adolescent

Preliminary analyses examined possible sex of adolescent or sex of parent differences in reports of the parental behaviors. Means and standard deviations for the overall sample are reported in Table 1. Bivariate correlations were examined to evaluate how sex of the parent and sex of the adolescent varied according to the parental behaviors (see Table 1). Two significant bivariate relationships between parental behaviors and sex of parent (i.e., adolescents reported higher support and monitoring by mothers than by fathers) were revealed. One significant bivariate relationship was evident between parental behaviors and sex of adolescent: girls reported greater parental punitiveness than boys.

Next, a series of hierarchical multiple regression analyses were conducted. Four analyses were conducted to see if sex of adolescent moderated overall family functioning as a predictor of any the four parental behaviors (e.g., sex of adolescent was entered in Step 1, overall family functioning was entered in Step 2, and sex of adolescent × overall family functioning was entered in step 3). The procedure was repeated using (a) sex of parent × overall family functioning and (b) possible three-way interaction of sex of parent × sex of adolescent × overall family functioning using an additional step in the regression equation. None of the interaction terms were significantly related to any of the parental behaviors in any of the 12 equations. Thus, sex of parent and sex of adolescent were examined as covariates rather than independent variables.

Multivariate analysis of covariance and discriminant analysis

Multivariate analysis of covariance (MANCOVA) was used to test the hypothesis that significant differences exist in adolescent perceptions of parental behaviors among the categories of overall family functioning. Overall family functioning was the fixed factor, parental behaviors were the dependent variables, and sex of the adolescent and sex of the parent about whom the adolescent reported were covariates. Results of the MANCOVA revealed significant relationships between the parental behaviors depending on overall family functioning while reducing the error variance introduced by the sex of the adolescent or the sex of the parent (Stevens, 2002). Further examination of which family types contributed to the significant discrimination in parental behaviors involved discriminant function analysis.

Means and standard deviations for groups associated with the MANCOVA are reported in Table 2. Results of the MANCOVA revealed a significant effect of the parental behaviors (support, monitoring, induction, and punitiveness) on the fixed variable (overall family functioning) (F(4,12)=17.35, p≤.05) yielding a Wilk's lambda of .55. However, neither sex of the adolescent nor sex of the parent demonstrated significant effects in relation to the variables of interest. Significant differences among the four categories of overall family system functioning (balanced, moderately balanced, mid-range, extreme) when all four parental behaviors (support, punitiveness, induction, monitoring) were considered simultaneously in the MANCOVA.

Table 2 Means and standard deviations of perceptions of overall family functioning types and perceptions of parental behaviors for adolescent girls and boysa
Full size table

To distinguish which of the parental behaviors were more influential in distinguishing among the types of overall family functioning, standardized discriminant function coefficients were employed (Stevens, 2002). Two significant discriminant functions were obtained that maximize differences among the categories of overall family functioning on perceived parental behaviors. The first discriminant function yielded an eigenvalue of .72, explaining 93% of the variance in overall family functioning. The second discriminant function yielded an eigenvalue of .06 and accounted for 7% of the variance in overall family functioning. The standardized coefficients and the discriminant function-variable correlations (see Table 3) of the first function indicate that parental support contributed the most substantively to the differences in overall family functioning. In the second function, parental monitoring maximized group differences. Considering the functions at group centroids (see Table 3), perceptions of higher parental support were associated with reports of balanced or moderately balanced overall family functioning, while perceptions of higher parental monitoring were associated with perceptions of balanced or extreme overall family functioning.

Table 3 Standardized canonical discriminant function coefficients, discriminant function-variable correlations, and unstandardized canonical discriminant functions (N=320)a
Full size table

Discussion

The results of our study provide substantial support for the expectation that adolescent reports of overall family system functioning are related to adolescent reports of parental behaviors. After controlling for sex of parent and sex of adolescent, results of the multivariate analyses use data from adolescents revealed (a) significant differences in parental behaviors were related to the type of overall family functioning, and (b) two parental behaviors (support and monitoring) were significantly associated with the type of overall family function-ing.

Adolescent reports of parental behaviors distinguished among three of four types of reported overall family functioning. Balanced overall family functioning was significantly related to higher parental support and monitoring. Moderately balanced overall family functioning related to greater parental support. Mid-range overall family functioning was not significantly related to any differences in the parental behaviors. Extreme overall family functioning related to high levels of parental monitoring. These findings are consistent with Minuchin's (1974) theoretical works that suggest that interaction patterns in the overall family systems are intertwined with those in parent-adolescent subsystems.

Among the parental behavior variables, support explained the most difference in overall family functioning. Adolescents who saw their families as balanced (moderate levels of both cohesion and flexibility) or moderately balanced (moderate levels of either cohesion or flexibility) reported their parents as more supportive than adolescents who reported other types of overall family functioning. Thus, adolescents who perceive their overall family functioning to be characterized by warmth, closeness, and flexibility also perceived their parents as supportive. These findings are consistent with a small body of research which found relationships between parenting and family functioning (Mupinga et al., 2002) and research showing both overall family system and parental qualities to relate to adolescent well being (e.g., Henry, 1994). Thus, parental support may be a key element of the emotional atmosphere of overall family system functioning (Barber & Buehler, 1996). Since Peterson and Hann (1999) concluded that parental support is the parental behavior most consistently associated with positive outcomes in offspring, additional research is needed to more fully explore how balanced and moderately balanced overall family functioning may interface with parental support.

Although parental support was related to greater variation in family functioning than any of the parental control behaviors, one of the three parental control variables (monitoring) was significantly related to differences in family functioning. The results showed that adolescents reported higher parental monitoring when they perceive their family systems as functioning at high (balanced overall family functioning) or low levels (i.e., extreme overall family functioning) rather than moderate levels. This finding challenges researchers and theorists to refine ideas about parental monitoring and adolescent well-being which typically using professional or parent reports (Crouter & Head, 2002). One interpretation of these results is that adolescent reports of parental monitoring may represent different family emotional atmospheres in balanced families when compared to extreme families. In balanced families, perceived parental monitoring might represent a preventive function as parents seek an awareness of adolescents’ lives as youth interact with others outside the family (i.e., monitoring) while providing a balance of connection to and freedom from the overall family system. However, in extreme families, parental monitoring may be a form of intervention as parents seek to emphasize limiting rather than maintaining an awareness of adolescent activities and friends. Further, the relationships of overall family functioning and parental behavior may vary depending upon which family member or outsider reports on the family system or parenting. Thus, additional research is needed to further investigate whether similar or different patterns are found when using different perceptions of these variables. Further investigation is needed about other parental behaviors that may mediate the relations between overall family functioning and parental monitoring. Longitudinal investigations during early, middle, and late adolescence might reveal how patterns of parental control evolve among the four types of overall family functioning. Future research with multiple family members and professionals using multiple methodologies will provide more definitive conclusions.

Caution is advised in considering these results due to the use of self-report questionnaires from ninth and tenth grade students residing with both their fathers and mothers in three rural communities in a southwestern state. Replication of the results with more diverse samples (e.g., diverse geography, ethnic/racial groups, or family forms) is needed before generalization to other populations. Despite these limitations, the present results provide support for an increased focus on the relationship between overall family system functioning and parental behavior as an area for strengthening and refining parent education and family therapy. By building a solid theoretical and empirical foundation for understanding both overall family systems and parental behaviors, the range of solid family interventions and preventions increases. Implications for parenting programs involve the continued efforts to integrate systems perspectives to strengthen both the content and process of parent education. Content can be broadened by strengthening the conceptualization of parent education to include concepts such as how adolescents’ perceptions of overall family system functioning may describe the emotional atmosphere in which specific parental behaviors such as monitoring or support are experienced. Adolescents, for example, who perceived their families as balanced (or providing moderate levels of both cohesion and adaptability) also perceived their parents as supportive and engaging in monitoring. Families may benefit from openly acknowledging that type of family system functioning may limit or enhance the potential for effective parenting. Thus, parent education programs are needed that focus on how to improve overall family functioning to provide a solid base for parenting.

Recognizing that overall family system functioning relate to aspects of parent-adolescent subsystems provides guidance about who participates in parent education and how parent education programs are conducted. Traditional approaches to parent education often assume that change in parental guidance or communication techniques often miss the opportunity for indirect change in parenting based upon change in the overall family system. In contrast, parent education programs might include adolescents or other members of family systems to provide opportunities for enhancing areas of overall family system functioning such as family decision making processes where families learn to work together toward shared goals that recognize the increasing competence of adolescents. A goal in family therapy can be to move families to the next higher level of functioning (e.g., an initial goal would be to move the extreme family to become a mid-range family; Olson & Gorall, 2003). Thus, incremental recognition of improved overall family interaction might be promising.

Because overall family functioning and parent-adolescent dyadic interaction are intertwined in a bidirectional manner, it is important to consider how parental behaviors such as support or monitoring might be targeted to enhance overall family well-being in family life education or therapy. Families that are moderately balanced, mid-range, or extreme may be able to move to a more balanced level of functioning through enhanced levels of parental support and monitoring. Engaging family systems in the process developing these qualities more effectively at the subsystem level might maximize the likelihood of change in the overall system. Families may need guidance in finding healthy ways of maintaining their connection to one another as they realign their family boundaries to afford a greater level of independence for adolescents. Parents and adolescents also may need guidance in understanding the function of parental monitoring or how to negotiate boundaries relating to monitoring.

Family assessments should be examined to identify whether overall family functioning or parent-adolescent subsystem behaviors are measured. Although systems perspectives often clearly distinguish between overall family functioning and parenting behaviors within parent-adolescent subsystems, parental research often lacks clear distinctions between overall family systems and parenting (Johnson, 2001). Practitioners are advised to conduct assessments of adolescents and families using both overall family functioning and parenting measures.