Family system characteristics and parental behaviors as predictors of adolescent substance use.

In recent years the role of family factors in adolescent substance use has received increased attention (Barnes, 1990). Much of this concern emphasizes the relationship between parental and adolescent substance use (Levine, 1985). Recent studies (Barnes, Farrell, & Cairns, 1986; Simons & Robertson, 1989) indicate that while parental use places adolescents at greater risk for problems related to substance use, adolescent perceptions of family interactions are also related to variation in adolescent substance use. Research has begun to address the role of family system characteristics (Volk, Edwards, Lewis, & Sprenkle, 1989) and parental behaviors (Simons & Robertson, 1989) in adolescent substance use (Nelson, Rosenthal, Harrington, & Michelson, 1986). However, these two issues have not been examined within the same study. The purpose of this study, therefore, was to examine how adolescents' perceptions of selected family system characteristics and parental behaviors predict adolescent substance use.

Family System Characteristics and Adolescent Substance Use

Although theoretical works promote the examination of family system characteristics as predictors of adolescent substance use, there is a sparsity of research relating family systems theory to adolescent substance use (Barnes, 1990; Steinglass, 1984); Volk et al., 1989). Using a systems perspective, the behavior of family members is viewed as intertwined. Thus, individual behavior, such as adolescent substance use, is best understood in the family context (Becvar & Becvar, 1982; Levine, 1985). Within family systems, patterns of interaction regularities or redundancies may be identified that are often described in terms of family system characteristics or specific parenting behaviors (Becvar & Becvar, 1982). Since family systems develop qualities that may encourage or support substance use among adolescents, there is an interrelation between the qualities of family systems and adolescent substance use. Such family system characteristics serve as important variables in understanding the initiation, maintenance, cessation, and prevention of substance use by adolescents (Needle et al., 1986).

Several overall family system qualities appear to be related to adolescent substance use. Previous work on this relationship emphasized the level of family bonding as a critical element in the adaptation of adolescents. Family bonding refers to the extent to which families emotionally join together into a meaningful and integrated unit, combined with the degree to which the family interacts with each other or outsiders (McCubbin, Thompson, Pirner, & McCubbin, 1988). Volk et al. (1989) identified two prevailing hypotheses regarding the role of family bonding in adolescent substance use. First, the overinvolvement hypothesis suggests that the families of adolescent substance abusers are frequently characterized by one parent who is overly involved in the adolescent's life and the other who is uninvolved. Second, the functional hypothesis proposes that adolescent substance users serve a function for their families. Specifically, the youths stabilize the family by drawing the members together to focus on the substance use, freeing them from focusing on other family problems such as marital conflict.

Evident within each of these hypotheses is an emphasis upon a positive relationship between highly bonded family systems and adolescent substance use. Yet, previous research presents conflicting results regarding the role of family bonding in such cases. While some studies (Steinglass, 1984; Volk et al., 1989) indicate that strong emotional bonds among family members may reduce the risk for adolescent substance use, others (particularly by clinicians) report the overinvolvement of family members in the lives of adolescents increases the risk (Levine, 1985; Stanton, 1985). Such theoretical works propose that extremely high levels of bonding increase the risk of adolescent substance use due to the struggle between adolescents' needs for autonomy and family system needs for connection (Weidman, 1983). Thus, additional research is necessary to explore the relation between family bonding and adolescent substance use.

Flexibility is another family system characteristic that also appears to reduce the risk of adolescent substance use. Flexibility refers to the ability of families to modify their interaction patterns when they encounter situations or family developmental tasks that require change (McCubbin et al., 1988). While stability provides regularity in the activities of daily family life, flexibility promotes change and development (Simon, Stierlin, & Wynne, 1985). Family systems are challenged to develop dynamics that accommodate the developmental changes in adolescents, including the expansion of involvement outside the family system. Previous studies, however, indicate that the families of adolescent substance abusers tend to be more rigid (or less flexible) and have difficulty adapting to change (Bartle & Sabatelli, 1989).

An additional family system characteristic related to adolescent substance use is parent-adolescent communication. Watzlawick, Beavin, and Jackson (1967) postulated that every interpersonal communication is not only an exchange of information, but includes a message regarding the interpersonal relationship. Thus, communication plays an important role in interactions between adolescents and their families (Barnes & Olson, 1985). The risk for adolescent problem behaviors such as substance use appears to be reduced in families with more open communication between parents and adolescents (Peterson & Leigh, 1990; Rosenthal, Nelson, & Drake, 1986).

Based upon these ideas, it was hypothesized that adolescent perceptions of family bonding would be positively related to adolescent substance use. Further, it was hypothesized that adolescent perceptions of family flexibility and openness in parent-adolescent communication would be negatively related to adolescent substance use.

Parental Behaviors and Adolescent Substance Use

In addition to qualities of overall family systems, a substantial body of theoretical and empirical evidence supports the linkage between parental behaviors and outcomes in youth (Rollins & Thomas, 1979; Peterson & Leigh, 1990; Peterson & Rollins, 1987), including adolescent substance use (Baumrind, 1991; Coombs & Landsverk, 1988). Specific parental behaviors such as substance use, support, induction, love withdrawal, and coercion have been found to be associated with adolescent substance use (Barnes, 1990; Needle et al., 1986). Several researchers (e.g., Barnes; 1990; Steinglass, 1984) note that parental substance use places youth at greater risk for substance use. The first experience a child has with alcohol is often in the family setting, with parents setting the stage for adolescents' perceptions of the meaning of alcohol and other substances (Barnes, 1990). Many adolescent substance abusers report that their first substance experiences took place in the family home (Jurich, Polson, Jurich, & Bates, 1985). Consistent with this view, Brown, Creamer and Stetson (1987) proposed that adolescents who have a family history of substance abuse are at increased risk for substance use.

Two rival hypotheses have arisen regarding the manner in which parental substance use relates to adolescent use (Steinglass, Bennett, Wolin, & Reiss, 1987). The genetic transmission model proposes that the tendency toward substance abuse progresses to the younger generation through genetic factors that predispose youth toward low substance tolerance. This model is limited by its inability to explain how some genetically vulnerable youth develop problems with substance use, while others do not. In contrast, the family systems model posits that adolescents' perceptions and expectations regarding the use of alcohol and drugs are derived, in part, from parental expectations and perceptions of substance use (Brown et al., 1987; Barnes, 1990; Jurich et al., 1985). The latter model holds greater potential for the prevention of substance abuse by recognizing the role of family dynamics. Both the clinical and empirical literature supports the examination of parental substance use as a critical factor in adolescent substance use (Barnes et al., 1986; Steinglass et al., 1987). However, despite this strong connection, questions remain concerning the aspects of parental substance use that are associated with the increased risk of adolescent substance use.

Previous research indicates that parental substance use alone is insufficient to explain how factors in the parental subsystem relate to adolescent substance use. Another type of parental behavior shown to serve as a buffer against adolescent substance use is nurturant or supportive parental behaviors (Needle, Glynn, & Needle, 1983; Simons & Robertson, 1989). Parental support of adolescents includes praising, encouraging, physical affection, and showing approval, love, and acceptance (Rollins & Thomas, 1979; Barnes, 1990). In general, parental support is related to the positive adaptation of adolescents and lower risk for problem behaviors such as substance use (Barber, 1992; Barnes, 1990; Baumrind, 1991; Peterson & Leigh, 1990).

Another category of parental behavior relating to adolescent substance use is the style of parental control efforts (Barnes, 1990; Peterson & Rollins, 1987). Within American culture, parents are expected to retain a degree of responsibility for and control over their adolescents (Henry, Wilson, & Peterson, 1989). Yet, in the families of adolescents who are highly involved in substance use, parental subsystems within the hierarchy of family boundaries are often confused (Piercy & Nelson, 1989). That is, parental control may be limited in families with adolescent substance use problems. In contrast, parents who effectively respond to the social expectations to control their offspring, while allowing for their increasing autonomy, may be expected to have youth with lower levels of substance use.

There is considerable variation in the ways parents control their adolescents. For example, although adolescents may conform when they perceive that their parents have the potential to bring about unwanted consequences for undesirable actions, the actual use of coercion (i.e., the use of direct and arbitrary force, Peterson & Leigh, 1990) is negatively related to adolescent conformity to parental expectations (Henry et al., 1989). A second type of parental control, love withdrawal, is the actual or threatened withdrawal of affection (Peterson & Rollins, 1987). Such behaviors communicate that parental love is contingent upon the behavior of the youth, increasing the risk of adolescent problem behaviors (Rohner, 1986), including substance use (Simons & Robertson, 1989). A third type of parental control, induction, is positively related to developmental outcomes and reduced risk for problem behaviors (Peterson & Rollins, 1987), including substance abuse (Baumrind, 1991; Pearson, 1989). Parental induction refers to placing rational maturity demands on children to make them aware of the consequences of their actions (Peterson & Rollins, 1987).

Based upon these ideas, it was hypothesized that adolescents' perceptions of the frequency of parental substance use and problems due to parental substance use would be positively related to adolescent substance use. Adolescent perceptions of parental support and induction were hypothesized to be negative predictors of adolescent substance use, while adolescent perceptions of love withdrawal and coercion were expected to be positive predictors.

Although family system qualities and parental behaviors were of primary importance, some studies indicated that specific demographic characteristics merited inclusion. Birth order is sometimes highlighted as a factor in adolescent substance use (Werner, 1985; Needle et al., 1986; Kaufman, 1984; Levine, 1985). Specifically, substance use among firstborns is portrayed as a way of gaining relief from pressure to achieve, while substance use among lastborns is viewed as a way of preserving their status as "baby of the family" (Barnes, 1990; Keltner, McIntyre, & Gee, 1986; Levine, 1985). Due to the conflicting views as to whether firstborns or lastborns are at greater risk for substance use, the variable of birth order was included in the present study. Other evidence suggests that the number of children in the family is an important factor that may have implications for adolescent substance use (Barnes, 1990). Finally, previous studies concluded that boys are more likely than girls to use substances (Toray, Coughlin, Vuchinich, & Patricelli, 1991; Werner, 1985).

METHOD

Sample and Procedure

This study was part of a larger research project on parent-adolescent relations. The sample of 489 adolescents was recruited through English classes in four high schools in a southwestern state. The mean age of the participants was 16.1, ranging from 13 to 20. The majority of the participants (90%) were Caucasian, 5% were Native American, 4% were African American, and 1% were other races. The mean number of children in the families was 2.77, ranging from 1 to 9. A total of 193 of the subjects reported having consumed alcohol within the past month, while an additional 93 stated that they used some form of substance at least once per month; 44% were males and 56% were females. Parental marital status was reported as follows: married (57%), divorced (29%), separated (3%), widowed (5%), single (2%), and other or not reported (4%).

Measurement

Adolescent substance use was measured using a 9-item scale, the Substance Use Indicator, developed by the first author specifically for the overall project. This scale was designed to measure the level of substance use among the subjects, based upon the DSM III-R criteria for psychoactive substance abuse and psychoactive substance dependence (American Psychiatric Association, 1987). The items assessed the frequency of substance abuse, substance tolerance, attempts to stop using substances, changes in activities, and problems stemming from substance use, including alcohol and drugs but not tobacco. Sample items were: (a) I find that I am drinking or using more alcohol/drugs now than I thought I would when I started, and (b) I have driven when I was high or intoxicated. Three response categories were used: not applicable, I do not use substances (1); no, I do not agree with the statement (2); and yes, I agree with the statement (3). Responses on the nine items were summed, resulting in scores ranging from nine (no substance was reported) to 27 (many concerns related to substance use). An internal consistency reliability coefficient (Cronbach's alpha) of .96 was established for the scale.

The measurement of family systems characteristics used previously established Likert-type scales. Bonding was measured using the Family Bonding Index (McCubbin et al., 1988), a 7-item adaptation of the cohesion scale from FACES H (Olson, Portner, & Bell, 1982) that measured adolescents' perceptions of family connectedness. Sample items were: (a) Family members go along with what the family decides to do (reverse coded); and (b) We have difficulty thinking of things to do as a family. The response choices were: almost never (5), once in a while (4), frequently (3), sometimes (2), and almost always (1). The one item that used wording which reflected high bonding was reverse coded and responses to the seven items were summed, resulting in a range of scores from 7 (low bonding), to 35 (high bonding). An internal consistency reliability coefficient (Cronbach's alpha) of .72 was established.

Flexibility was measured using the Family Flexibility Index (McCubbin et al., 1988) that measured adolescents' perceptions of their families' ability to change roles, rules, responsibilities, and decision-making to accommodate change. Sample items were: (a) Each family member has input in major family decisions; (b) We shift household responsibilities from person to person. Response choices were: almost never (1), once in a while (2), frequently (3), sometimes (4), and almost always (5). Responses to the seven items were summed resulting in a range of scores from 7 (low flexibility) to 35 (high flexibility). An internal consistency reliability coefficient (Cronbach's alpha) of .76 was established for the scale.

Parent-adolescent communication was measured using responses to 20 items about openness in communication with fathers and mothers using a modification of the Open Family Communication Subscale of the Parent-Adolescent Communication Index (Barnes & Olson, 1982). Sample items were: (a) This parent can tell how I'm feeling without asking; and (b) I find it easy to discuss problems with this parent. The response categories were: strongly disagree (1), disagree (2), neutral (3), agree (4), and strongly agree (5). The responses for items about fathers and mothers were combined and summed, yielding a range of scores from 20 (low openness in communication) to 100 (high openness in communication). Internal consistency reliability (Cronbach's alpha) was .92.

Two dimensions of parental substance use were assessed: adolescent perceptions of the frequency of parental substance use and adolescent perceptions of problems associated with parental substance use. To assess the frequency of parental substance use, the subjects were asked to respond to the following question with respect to both their fathers and mothers: How frequently does your mother/stepmother (or father/stepfather) use alcohol or drugs? The seven response choices were: never or not applicable (1), tried alcohol/drugs but not has not used them regularly (2), regularly used alcohol/drugs in the past but not now (3), only at parties or with friends and less than once a month (4), once or twice a month (5), about once a week (6), and daily (7). A second item assessed the extent to which the subjects perceived parental substance use to be problematic for their families: My mother/stepmother's (father/step father's) use of alcohol or drugs has been a problem for our family. The four response choices ranged from strongly disagree (1) to strongly agree (4). On both the problems with parental substance use and the frequency of parental substance use scales, scores resulted from summing responses about mothers and fathers on each item.

The other parental behaviors (i.e., support, induction, love withdrawal, and coercion) were measured on Likert-type scales utilizing subscales of the Parental Behavior Measure (Peterson, 1982). These scales involved the use of items with the highest factor loadings on previously established scales measuring parental support, induction, coercion, and love withdrawal utilized by Peterson, Rollins, and Thomas (1985) (also see Henry et al., 1989). The total number of items on each scale included responses to the same items about the parenting behaviors of mothers and fathers. Response choices were: strongly disagree (1), disagree (2), neutral (3), agree (4), and strongly agree (5). Scores on each scale resulted from the summation of responses to the items regarding mothers' and fathers' parental behaviors.

Adolescents' perceptions of parental support were measured on an 8-item Likert-type scale assessing the extent to which adolescents saw their mothers and fathers as providing emotional and resource support. Perception of parental induction was measured on a 10-item Likert-type scale which assessed adolescent views of parental control attempts based on logical reasoning. Adolescents' perceptions of parental coercion were measured by a 10-item scale which assessed their perceptions of parental control attempts based on punitiveness. Parental love withdrawal was measured on a 4-item Likert-type scale which assessed perceptions of parents' attempts to control their behavior through avoiding contact with them until cooperation was gained. Sample items were as follows: (a) This parent seems to approve of me and the things I do (support); (b) This parent punished me by not letting me do things I really enjoy (coercion); (c) This parent explains to me how good I should feel when I do right (induction); (d) This parent will not talk to me when I displease him/her (love withdrawal). The following internal consistency reliability coefficients (Cronbach's alphas) were established for support, induction, love withdrawal, and coercion as .86, .86, .78, and .86, respectively. The number of children in the families, birth order, and gender of the subjects were assessed using standard fact sheet items.

ANALYSIS AND RESULTS

Bivariate correlations and simple regression were used for data analysis. Prior to the analysis, a dummy variable was developed for the gender of adolescent variable (male coded 0, female coded 1; Pedhazur, 1983). Means and standard deviations for the variables are reported in Table 1. The correlation coefficients were examined to determine the extent to which there were significant relationships between the predictor and criterion variables to be used in the multiple regression equation. The bivariate correlations revealed a significant negative correlation between adolescent substance use and adolescent perceptions of openness in parent-adolescent communication, family flexibility, family bonding, parental support, and parental induction. In contrast, the bivariate correlations demonstrated a significant positive relationship between adolescent substance use and love withdrawal, coercion, frequency of parental substance use, and problems with parental substance use. The only sociodemographic variable that correlated significantly with adolescent substance use was the gender of the adolescent, with greater substance use reported by adolescent boys.

Table 1

Means and Standard Deviations

                                        Mean         SD

Communication                           69.30      19.18
Flexibility                             21.65       5.59
Bonding                                 23.16       5.52
Support                                 30.32       7.92
Induction                               32.55       9.15
Love withdrawal                          9.19       4.15
Coercion                                34.54      13.42
Frequency of parental substance use      4.53       3.19
Problems with parental substance use     2.46       1.04
Birth order                              2.14       1.24
Number of children                       2.78       1.27
Adolescent substance use                13.73       5.27

n = 489; SD = Standard deviation
Table 2

Bivariate Correlations and Multiple Regression Analysis of Family
Characteristics and Parental Behaviors as Predictors of Adolescent Substance
Use

                               r          b      B       t      sig t

Communication               -.27(**)    -.02   -.07    -.90     .37
Flexibility                 -.19(**)     .03    .03     .58     .56
Bonding                     -.30(**)    -.14   -.15   -2.72     .01(**)
Support                     -.27(**)    -.12   -.19   -2.31     .02(*)
Induction                   -.13(*)      .05    .09    1.33     .19
Love withdrawal              .23(**)     .10    .08    1.31     .19
Coercion                     .18(**)     .02    .06    1.00     .32
Frequency of parental
substance use                .26(**)     .34    .21    4.64     .00(***)
Problems with parental
substance use                .17(**)     .06    .01     .26     .79
Number of children           .00        -.20   -.05   -1.01     .32
Birth order                  .04         .35    .08    1.66     .10
Gender                      -.08(*)     -.74   -.07   -1.65     .10

Multiple Correlation (R)                                        .44
Multiple Correlation Squared ([R.sup.2])                        .19
F-Value                                                        9.45(***)

n = 489, * p [is less than] .05, ** p [is less than] .01, *** p [is less than]
.0001

Since each of the family system and parental behaviors yielded significant bivariate correlations with adolescent substance use, multiple regression analysis was conducted. Specifically, the family system characteristics (openness in parent-adolescent communication, bonding, and flexibility), parental behaviors (frequency of parental substance use, problems with parental substance use, support, love withdrawal, coercion, and induction), and sociodemographic variables (number of children, birth order, and gender) were entered as predictor variables into a multiple regression equation with adolescent substance use as the criterion variable.

Results of the multiple regression analysis provided partial support for the research hypothesis. Among the family system characteristics, adolescent perceptions of family bonding yielded a significant negative beta, indicating that adolescents who perceived their families to be highly bonded reported fewer problems with substance use. In contrast, adolescent perceptions of openness in parent-adolescent communication and family flexibility failed to yield significant negative beta coefficients.

The frequency of parental substance use manifested a significant positive beta coefficient in relation to problems with adolescent substance use while parental support yielded a negative beta coefficient. Parental induction, coercion, love withdrawal, and problems with parental substance use demonstrated nonsignificant beta coefficients in relation to adolescent substance use.

Number of children in the family, birth order, and gender of the adolescent yielded nonsignificant beta coefficients in relation to adolescent substance use. Tolerance tests using the value of .10 indicated that multicollinearity was not sufficient to create a problem among the predictor variables in the regression equation. The overall model achieved significance, accounting for 19% of the variance in adolescent substance use.

DISCUSSION

The results provide support for the proposal that family system characteristics and parental behaviors would predict adolescent substance use. In this study, the family system characteristic that was related to reduced risk for adolescent substance use was adolescents' perceptions of family bonding. In contrast to the hypothesis, which predicted that perceptions of high levels of family bonding would be related to increased problems with adolescent substance use, these results challenge a view prevalent in some of the clinically based research on adolescent substance use. Stanton (1985), for example, proposed that families with extremely high levels of bonding are at increased risk for problem behaviors such as substance use due to the decreased opportunity for adolescents to develop autonomy. Yet, these results provide further support for Volk et al.'s (1989) research which found family cohesion (or bonding) to be negatively related to substance abuse in adolescents. Thus, family bonding may serve as a buffer against the risk of problems with adolescent substance use. Adolescents, therefore, who perceive their families to have high levels of bonding see them as providing a secure foundation from which to explore the world and develop a sense of self apart from the family system (Barber, 1992).

A related finding is that parental support was decreased related to problems with adolescent substance use. Parental support, or conveying warmth, acceptance, and personal value, is consistently linked to positive developmental outcomes in youth (Peterson & Rollins, 1987; Steinberg & Levine, 1990). The current results support previous findings that parental support encourages the development of adolescent social competence including positive parent-adolescent relations, greater identification of youth with their parents, and the development of adolescent autonomy (Peterson & Leigh, 1990). Thus, the use supporting parenting behaviors including praise, approving, encouraging, assisting, companionship, and physical affection not only enhances overall parent-adolescent relations (Peterson & Rollins, 1987), it is related to decreased adolescent substance use (Needle et al., 1983). Consequently, parenting education programs and therapeutic interventions that encourage supportive behaviors by parents has the potential for reducing the risk of adolescent substance use.

None of the parental control behaviors were significantly related to adolescent substance use within the multiple regression model. These results are in contrast to the proposal by Piercy and Nelson (1989) that in families where adolescents are substance abusers, appropriate parental influence is frequently lacking and adolescents often appear to control their parents. The lack of significant findings related to parental control highlights the importance of additional research that would explore more complex models of parental control approaches in relation to adolescent substance use (Barber, 1992).

Consistent with previous research (Barnes, 1990; Levine, 1985), adolescent perceptions of the frequency of parental substance use served as a significant predictor of adolescent substance use. The prevention and treatment of adolescent substance use, therefore, may benefit from assessing the extent to which substance use is prevalent in parents. Using a systems approach, when family dysfunction occurs in any part of the system, symptoms may occur within any family member (Minuchin, 1974). Thus, in cases where parents are heavily involved in substance use and adolescents are experiencing the same difficulty, intervention in the overall family system is critical.

Additional comment is required on the finding that adolescent perceptions of problems with parental substance use had a significant positive relationship with adolescent substance use in the bivariate correlation, but not in the overall regression model. Since previous research indicates a positive relationship between parental and adolescent substance use, it may be that factors such as communication, support, or love withdrawal modify the relationship between parental and adolescent substance use. Further research that would examine both the direct and indirect relationships between parental and adolescent substance use may provide further insights on this issue.

An alternative interpretation may lie in the use of adolescent perceptions of parental substance use. It is possible that parental substance use may be assessed differently when based on adolescents', parents', and clinicians' perspectives (Volk et al., 1989). Since the present study assessed adolescents' reports of family problems with parental substance use, different results might be forthcoming if clinical assessments of parental substance use were utilized. Thus, therapists and researchers may find it beneficial to use multiple perceptions of adolescents, parents, and clinicians in the assessment of parental substance use when using a systems approach to treatment problems with adolescent substance use.

Despite the logical pattern of these results, there are certain methodological limitations. First, the use of correlational analyses in the absence of longitudinal data within a systems perspective precludes definitive interpretations regarding the direction of influence. That is, while the authors offered the model of family system qualities and parenting behaviors as predictors of adolescent use, the opposite direction of relationship is also possible. For example, although the results indicated that parental support predicted decreased adolescent substance use, it is also possible that decreased adolescent substance use predicts increased parental support. The results are further limited by the purposive nature of the sample. Additional studies are required to consider how geographic and greater ethnic diversity might modify the results.

In summary, support was provided for the use of both family system characteristics and parental behaviors in the study of adolescent use. While research has previously been limited to a focus on either family system characteristics or parental behaviors in relation to adolescent substance use, the present results support Barnes' (1990) proposal that the two be considered simultaneously.

REFERENCES

American Psychiatric Association. (1987). Diagnostic and Statistical Manual of Mental Disorders (3rd ed.). Washington, DC: American Psychiatric Association.

Barber, B. K. (1992). Family, personality, and adolescent problem behaviors. Journal of Marriage and the Family, 54, 69-79.

Barnes, G. M. (1990). Impact of the family on adolescent drinking patterns. In R. L. Collins, K. E. Leonard, & J. S. Searles (Eds.), Alcohol and the family: Research and clinical perspectives (pp. 137-162). New York: Guilford Press.

Barnes, G. M., Farrell, M. P., & Cairns, A. (1986). Parental socialization factors and adolescent drinking behaviors. Journal of Marriage and the Family, 48, 27-36.

Barnes, H. L., & Olson, D. H. (1982). Parent-adolescent communication. In D. H. Olson, H. L. McCubbin, H. Barnes, A. Larson, M. Muxen, & M. Wilson (Eds.), Family inventories (pp. 33-45). St. Paul: University of Minnesota Press.

Barnes, H. L., & Olson, D. H. (1985). Parent-adolescent communication and the circumplex model. Child Development, 56, 438-447.

Bartle, S. E., & Sabatelli, R. M. (1989). Family system dynamics, identity development, and adolescent alcohol use: Implications for family treatment. Family Relations, 38, 258-265.

Baumrind, D. (1991). The influence of parenting style on adolescent competence and substance use. Journal of Early Adolescence, 11, 56-95.

Becvar, R. J., & Becvar, D. S. (1982). Systems theory and family therapy. Washington, DC: University Press of America.

Brown, S., Creamer, V. A., & Stetson, B. A. (1987). Adolescent alcohol expectancies in relation to personal and parental drinking patterns. Journal of Abnormal Psychology, 96, 117-121.

Coombs, R. H., & Landsverk, J. (1988). Parenting styles and substance use during childhood and adolescence. Journal of Marriage and the Family, 50, 473-482.

Henry, C. S., Wilson, S. M., & Peterson, G. W. (1989). Parental power bases and processes as predictors of adolescent conformity. Journal of Adolescent Research, 4(1), 15-32.

Jurich, A. P., Polson, C. J., Jurich, J. A., & Bates, R. A. (1985). Family factors in the lives of drug users and abusers. Adolescence, 20(77), 144-159.

Kaufman, E. (1984). Family systems variables in alcoholism. Alcoholism: Clinical and Experimental Research, 8(1), 4-8.

Keltner, N. L., McIntyre, C. W., & Gee, R. (1986). Birth order effects in second-generation alcoholics. Journal of Studies on Alcoholism, 47(6), 495-497.

Levine, B. L. (1985). Adolescent substance use: Toward an integration of family systems and individual adaptation theories. The American Journal of Family Therapy, 13(2), 3-16.

McCubbin, H. I., Thompson, A. I., Pirner, P. A., & McCubbin, M. A. (1988). Family types and strengths: A life cycle and ecological perspective. Edina, MN: Burgess International Group.

Minuchin, S. (1974). Families and family therapy. Cambridge, MA: Harvard University Press.

Needle, R. H., Glynn, T. J., & Needle, M. P. (1983). Drug abuse: Adolescent addictions and the family. In C. R. Figley, & H. I. McCubbin (Eds.), Stress and the family: Volume II, Coping with catastrophe (pp. 37-52). New York: Brunner/Mazel.

Needle, R., McCubbin, H., Wilson, M., Reineck, R., Lazar, A., & Mederer, H. (1986). Interpersonal influences in adolescent drug use: The role of older siblings, parents, and peers. The International Journal of the Addictions, 21(7), 739-766.

Nelson, T., Rosenthal, D., Harrington, R. G., & Mitchelson, D. (1986). Assessment of adolescent substance use. In R. G. Harrington (Ed.), Testing adolescents: A reference guide for comprehensive psychological assessment (pp. 178-199). Kansas City, MO: Test Corporation of America.

Olson, D. H., Portner, J., & Bell, R. (1982). FACES II: Family adaptability and cohesion evaluation scales. St. Paul, MN: Family Social Science, University of Minnesota.

Pearson, J. C. (1989). Communication in the family: Seeking satisfaction in changing times. New York: Harper & Row.

Pedhazur, E. J. (1983). Multiple regression in behavioral research. Fort Worth, TX: Holt, Rinehart & Winston.

Peterson, G. W. (1982). Parental Behavior Measure. Unpublished manuscript. Knoxville, TN: The University of Tennessee.

Peterson, G. W., & Leigh, G. K. (1990). The family and social competence in adolescence. In T. Gullota, G. R. Adams, & R. Montemayor (Eds.), Advances in adolescent development: Social competence (pp. 97-138). New-bury Park, CA: Sage.

Peterson, G. W., & Rollins, B. C. (1987). Parent-child socialization as symbolic interaction. In M. Sussman, & S. K. Steinmetz (Eds.), Handbook of marriage and the family (pp. 471-508). New York: Plenum Press.

Peterson, G. W., Rollins, B. C., & Thomas, D. L. (1985). Parental influence and adolescent conformity: Compliance and internalization. Youth and Society, 16, 397-420.

Piercy, F. P., & Nelson, T. S. (1989). Adolescent adolescent abuse. In C. R. Figley (Ed.), Treating stress in families (pp. 209-230). New York: Brunner/Mazel.

Rohner, R. P. (1986). The warmth dimension: Foundations of parental acceptance-rejection theory. Beverly Hills, CA: Sage.

Rollins, B. C., & Thomas, D. L. (1979). Parental support, power, and control techniques in the socialization of children. In W. R. Burr, R. Hill, F. I. Nye, & I. L. Reiss (Eds), Contemporary theories about the family: Research-based theories (Vol. 1, pp. 317-364). New York: Free Press.

Rosenthal, D., Nelson, T., & Drake, N. (1986). Adolescent substance use and abuse: A family context. In G. K. Leigh, & G. W. Peterson (Eds.), Adolescents in families (pp. 337-357). Cincinnati, OH: South-Western.

Simon, F., B., Stierlin, H., & Wynne, L. C. (1985). The language of family therapy: A systemic vocabulary and sourcebook. New York: Family Process.

Simons, R. L., & Robertson, J. F. (1989). The impact of parenting factors, deviant peers, and coping style upon adolescent drug use. Family Relations, 38, 273-281.

Stanton, M. D. (1985). The family and drug abuse. In T. E. Bratter, & G. G. Forrest (Eds.), Alcoholism and substance abuse: Strategies for clinical intervention (pp. 398-430). New York: Free Press.

Steinberg, L., & Levine, A. (1990). You and your adolescent: A parents guide for ages 10 to 20. New York: Harper & Row.

Steinglass, P. (1984). Family systems theory and therapy: A clinical application of general systems theory. Psychiatric Annals, 14(8), 582-586.

Steinglass, P., Bennett, L. A., Wolin, S. J., & Reiss, D. (1987). The alcoholic family. New York: Basic Books.

Toray, T., Coughlin, C., Vuchinich, S., & Patricelli, P. (1991). Gender differences associated with adolescent substance abuse: Comparisons and implications for treatment. Family Relations, 40, 338-344.

Volk, R. J., Edwards, D. W., Lewis, R. A., & Sprenkle, D. H. (1989). Family systems of adolescent substance abusers. Family Relations, 38, 266-272.

Watzlawick, P., Beavin, J. H., & Jackson, D. D. (1967). Pragmatics of human communication: A study of interactional patterns, pathologies and paradoxes. New York: W. W. Norton.

Weidman, A. (1983). Adolescent substance abuse: Family dynamics. Family Therapy, 10, 47-55.

Werner, E. E. (1985). Resilient offspring of alcoholics: A longitudinal study from birth to age 18. Journal of Studies on Alcohol, 47(1), 34-40.

Allan R. Anderson is a doctoral student at Oklahoma State University.