A family therapy approach, combining the conceptual frameworks of the Bowen model and Olson's Circumplex model, was used to restructure family relationships in order to prevent further runaway behavior of a 15-year-old Mexican-American female. A structured family interview was conducted to identify problems, a genogram was used to obtain the family's history, and therapy (family, individual, and mother-daughter) was provided to change the family system. The Family Satisfaction Scale was administered to evaluate the effectiveness of this therapeutic approach.
Runaway behavior for youths in the 1990s is usually not the result of a wish to have a Huckleberry Finn experience. It is often their dramatic way of dealing with longstanding problems or conflicts with family (Sharlin & Mor-Barak, 1992).
It is believed that between 1 and 1.3 million youths in the United States live in emergency shelters or on the streets (Feitel, Margeston, Chamas, & Lipman, 1992; Crespi & Sabatelli, 1993). Research indicates that the problem is more prevalent for adolescent females (Cohen, MacKenzie, & Yates, 1991; Kurtz, Jarvis, & Kurtz, 1991; National Network of Runaway and Youth Services, 1985; Sharlin & Mor-Barak, 1992). Cohen et al. (1991) found that homeless teens tended to be younger, female, and white. Further, these young females engaged in problematic behaviors, such as vagrancy, sexual promiscuity, prostitution, suicide attempts, and becoming pregnant (Sharlin & MorBarak, 1992).
Designing interventions that reduce or eliminate runaway behavior is a challenge for family therapists (Lappin & Covelman. 1985). The Bowen model of family therapy (Goldenberg & Goldenberg, 1991) and the Circumplex model (Olson, Russell, & Sprenkle, 1983) provide a complementary framework from which effective interventions can be designed.
The literature uses the concepts of homelessness and runaway interchangeably. However, for the purpose of this study, the definition of runaway is "a youth who is away from home without the permission of his or her parent(s) or legal guardian; a situation in which a youth is absent from his/her home or place of legal residence at least overnight without permission" (General Accounting Office, 1989). Running away may be a phenomenon related to the search for self-identity, or it can be due to personal, family, or environmental problems. There is a correlation between runaway behavior and family instability (Sharlin & Mor-Barak, 1992).
The Context of Runaway Behavior
The types of runaway behavior were initially viewed dichotomously as "running from" or "running to" something (Greene & Esselstyn, 1972; Homer, 1973). Later, Barth (1986) elaborated on these to include three categories: the child who runs away from family strain caused by a crisis; the child who runs away from excessive parental expectations and control; and the child who runs away from a physically or sexually abusive situation. Assessment includes determining whether the child's behavior is new and amenable to change or habitual and well-entrenched. An additional category is the "thrown out" youth (Zide & Cherry, 1992).
The "running to" youth may have experienced some problems with family or in school or the community, but is often searching for excitement. This type of runaway usually comes from a more normal family situation and runs away for existential reasons.
The "running from" youth is unhappy about one or more major areas of life (e.g., conflict with, or alienation from, family). The family situation is often pathological, such as that involving an alcoholic parent, physical or sexual abuse, or extreme financial difficulties. Running away in these instances can be viewed as a rational decision to escape harm (Speck, Ginther, & Helton, 1988; Zide & Cherry, 1992).
The "thrown out" youth may have been forced to leave home, because of extreme family alienation or premature dissolution caused by chronic poverty, family conflict, substance abuse, or the death of caretakers (Feitel, Margetson, Chamas, & Lipman, 1992; Zide & Cherry, 1992).
The Framework for Family Treatment
There are several models from which a framework for family intervention can be built. Lappin and Covelman (1985) have formulated a developmental/structural model for examining runaway behaviors. They view running away as a symptom of the family's difficulty in dealing with the adolescent's separation and individuation. According to Crespi and Sabatelli (1993), family dynamics, coupled with the adolescent's developmental processes, account for the attempt to individuate and gain autonomy by running away from the family unit.
Of particular interest here, Mirkin, Raskin, and Antognini (1984) have described a pattern in which the runaway adolescent female may be seeking to maintain or protect the parents' marriage by uniting them around her dysfunctional behavior. They recommend that treatment be focused on structural/strategic interventions where the daughter is removed from marital conflict. Others (for example, Schwartzman & Restivo, 1985) contend that disagreement among authority figures within the family or across systems (e.g., courts, juvenile probation, and school) contributes to delinquent behavior, such as running away. Early intervention with these authority figures would help to prevent the rift between the adolescent and family from deepening, thus increasing the prospects for an early return home or reducing recidivism (Kurtz, Jarvis, & Kurtz, 1991).
Bowen's family systems approach is relevant to the structural context of pathology noted by Lappin and Covelman (1985). Bowen proposed that all families can be characterized on a continuum of differentiation levels. Differentiation, in conjunction with triangles, the nuclear family emotional system, the family projection process, emotional cutoff, the multigenerational transmission process, sibling position, and societal regression, shapes family functioning (Goldenberg & Goldenberg, 1991). Triangulation, the multigenerational transmission process, and the differentiation of individuals within the family system form the basis for the family interventions described here.
The triangulation process is the result of an unstable two-person system under stress. In an attempt to reduce the increasing tension of their relationship, a third person is drawn in to form a triangle, or triad. For example, an adolescent female in the normal developmental stage of individuation and differentiation may be drawn into the parents' conflict, thus forming a triangle. This creates more tension for the adolescent, with running away being an attempt to decrease stress and increase individuation.
The multigenerational transmission process refers to family functioning over several generations. A genogram can be used as a means of gathering a family history as well as an intervention technique (Guerin & Pendagast, 1976). A genogram is a visual representation of multigenerational patterns and influences determining nuclear family functioning. Bowen contends that the transmission of pathology transcends generations and affects the patterns of behavior in the family. In dysfunctional families, each generation produces individuals with progressively poorer differentiation, who are increasingly vulnerable to anxiety and fusion (Goldenberg & Goldenberg, 1991). A genogram reveals this process.
The structured family interview is a technique used to assess the family's ongoing transaction patterns. It is concerned with family members' roles, with particular attention to hierarchy and the ability of subsystems to carry out their functions, family members' alignments and coalitions, and the ease with which individuals can cross subsystem boundaries.
To bridge the gap between research, theory, and practice, Olson (1989) developed the Circumplex Model of Marital and Family Systems. The model analyzes the family's cohesion and adaptability. When there is too much loyalty to the family and too little independence, the family members are enmeshed. The disengaged family is at the other extreme--members feel little or no commitment to each other (Olson, 1989). Runaway behavior could result from either extreme. For example, an enmeshed family system does not allow the adolescent to develop an independent set of values. A balanced family system accommodates the adolescent's growing need for independence and differentiation.
Adaptability is defined as the ability of the family to adjust to change or stress. This dimension was extracted from a variety of sociological theories regarding family power (assertiveness, control, discipline), negotiation styles, role relationships, and relationship rules (Olson, 1989). There are four levels of adaptability: rigid, structured, flexible, and chaotic. As with cohesion, more central levels of adaptability are compatible with healthy family and marital functioning.
Critics of family systems theory suggest that, since it developed from theory related to nonhuman systems, it may be limited in its ability to account for family dysfunction and to guide intervention (Searight & Merkel, 1991). Others suggest that it is not a theory but a metatheory which is not translatable to clinical work (L'Abate & Colondier, 1987). Other difficulties include deciding which technique to apply and when to implement it, as well as adapting it to the cultural context (Searight & Merkel, 1991).
The purpose of this study was to investigate the usefulness of a family systems approach with families of adolescent females who have experienced at least one runaway episode. Bowen's concepts of differentiation of self, triangles, and the multigenerational transmission process guided the strategies to improve family functioning and cohesion, redefine the parental dyad, and assist the adolescent in appropriate individuation. The Circumplex model's method of diagnosing a family's level of cohesion based on constructs theoretically similar to the Bowenian model strengthened the ability to evaluate change. A pre/posttest design was selected due to its ability to produce acceptable outcome data without being intrusive. This design, however, limited the ability of the therapist to evaluate the success of each therapy session or which techniques were most helpful for family members.
Selection and Description of Client System
Setting. A large, multiservice, nonprofit child welfare agency provided assistance to runaway youths. The project was made possible by a state-funded grant under the Texas Department of Protective and Regulatory Services, Services to Truant and Runaway Youth. Cases were referred to project staff by school personnel, school nurses, juvenile probation officers, and directly from families. Youths who were seen in the emergency shelter were also referred. The project office was located in a rural community within a county whose population exceeded 250,000.
The project was designed with an evaluation component to measure the effectiveness of the intervention. For the purpose of the present study, data from a single family were employed to evaluate results and to more closely link the changes in the family to the techniques used. The family was randomly selected from among the 32 that participated. The Family Satisfaction Scale (Pratt & Hansen, 1987), an adaptation of Olson's Family Adaptability and Cohesion Evaluation Scales (FACES III), was used to measure success. This scale provided a good fit with the intervention strategies, which were based on Bowen's family systems approach.
"Healthy" families demonstrate balanced degrees of cohesion and adaptability, whereas problem families evidence extreme degrees of these constructs (Rodick, Henggeler, & Hanson, 1986). The structural therapist's task, then, is to use Bowen's concepts to design interventions that will rebalance the family system. Here, the intervention needed to address family problems quickly to reduce the risk of another runaway episode.
Project staff. The project was placed within a larger program of intensive home-based therapy. The four staff members were supervised by a master's level social worker with advanced clinical certification. Project staff who provided services were trained in social work, marriage and family therapy, or clinical psychology. They spent two weeks in intensive training in the family systems approach and in the use of the structured family interview, the genogram, and other intervention techniques. Therapists were often teamed to work with families and to assist with cotherapy when it was indicated for an individual family member, usually the adolescent.
Families in the project. As stated previously, families were referred from several sources in the community as well as self-referred. The families were contacted for a first interview, usually in the home, where the structured family interview was completed. The contract stipulated that families could receive services for only six months. Services beyond six months required preauthorization.
The selected family. The C. family consisted of a mother, age 35, a stepfather, age 38, who refused services, Melinda, age 15, who had a one-year-old son living with the family, and two brothers, ages 13 and 16. The siblings were from Mrs. C.'s first marriage. Melinda was referred by her teacher. She was in the 9th grade and an average to low-average student. She did not have any learning problems, but missed school frequently and thought about dropping out. Two runaway episodes, which lasted several days each, and sexual promiscuity were Melinda's presenting problems. Mrs. C. reported poor communication between Melinda and herself, that Mr. C. was an alcoholic, and that she had used drugs at various times in her life. Mrs. C. and Melinda both reported that Mr. C. had many expectations regarding Melinda's helping with housework and other chores, whereas Mrs. C. was more lenient, causing conflict between the parents. When services began, Mrs. C. was unemployed and Mr. C. was employed as a construction laborer . He and Mrs. C. were on probation for selling marijuana.
Implementation of the Intervention
The variables selected for measurement were family cohesion and adaptability as defined by Olson and Wilson (1982). Cohesion is the emotional bond that family members feel toward one another (Green, 1989). Family adaptability relates to the extent to which the family system is able to change. Further, it is the ability of the marital or family system to alter its power structure and role relationships in response to situational or developmental stress (Olson, Russell, & Sprenkle, 1983). It was hypothesized that the intervention would produce a more balanced degree of both family cohesion and adaptability.
The family had a choice of office visits or having the therapist visit them in their home. There were fifteen sessions over a four-month period, with three family sessions in which everyone was present except the stepfather, six sessions with mother and daughter, three sessions with Mrs. C., and three sessions with Melinda.
The first family session was used to explain the reason for the therapist's visits and to gather family data through the structured interview process. The data collected in the interview was used to assist the family in formulating the goals of treatment. The goals were to: (1) have Melinda remain in school, (2) raise the quality of communication between Melinda and Mrs. C., (3) have Melinda stop running away when she experienced conflict with family members, (4) have Melinda make a decision about how to parent Henry, Jr. (her 1-year-old son), and reduce the conflict regarding Melinda's helping with housework.
The second family session focused on the development, of the family genogram (see Figure 1). This activity elicited enthusiasm from family members and resulted in Mrs. C. noting that there appeared to be "lots of family problems" in her father's family. The genogram achieved its purpose of revealing family history and identifying significant events in the nuclear family and the family of origin.
The next two sessions, held in the therapist's office, focused on the mother/daughter dyad and dealt with such issues as communication, Mrs. C.'s expectations that Melinda would complete school, and making a decision to parent Henry, Jr., who Melinda had been moving from one aunt to another for several weeks at a time. Mother and daughter also agreed upon housework Melinda could do to help out and that she would take care of her own clothes and bedroom. During the first and second mother/daughter sessions, a decision was made for Melinda to move to an emergency shelter for runaway and homeless youth, as conflict had escalated. Melinda remained in the shelter for two weeks and attended school, group therapy with other youths, one family session in the home, and one mother/daughter session.
Individual sessions with Mrs. C. focused on her marital relationship and Mr. C.'s problems with alcohol. She also talked about her need to reduce the conflict between Mr. C. and Melinda, and came to understand that she needed to support Mr. C.'s attempts to have Melinda help with household chores. She became employed at a local library after the third family session. She began to focus on her need to feel supported by her family so that she could work, and to have Melinda more stable and responsible at home. Individual sessions with Melinda focused on her long-range plans for completing high school, ambivalence over parenting Henry, Jr., and communication problems with her mother and stepfather.
The role of females in the Hispanic family was an important element, as much of the conflict involved discordance between the value systems of Mr. and Mrs. C. and their highly assimilated teenage daughter. The majority culture's permissiveness conflicted with this Hispanic family's attempts to protect their daughter and structure her environment. Melinda used running away and sexual promiscuity to distance herself from her family. She was attempting to individuate, which accounted for the increasing levels of mother/daughter conflict.
Initially each family was assessed using the structured family interview technique. In addition, the parent(s) and the identified client completed the Family Satisfaction Scale (FSS). The FSS was administered at one of the first two sessions with the family and again at the last visit.
The FSS indicates whether a person has become more or less satisfied with his or her family (higher scores denote greater satisfaction). It is constructed to tap two dimensions of family functioning: cohesion and adaptability. Investigating the validity and reliability of this instrument, Olson and Wilson (1982) found that the total score proved best. Cronbach's alpha coefficients for the pilot sample (N = 433) were .82 for the cohesion subscale, .86 for the adaptability subscale, and .90 for the total scale. For the purposes of this study, the total scores for the parent(s) and the adolescent were used.
Although the validity of the Family Satisfaction Scale has not been determined for different ethnic groups, Flores and Sprenkle (1989) did use FACES III with a large sample of Mexican-American families. Since the FSS is a reformulation of FACES III, it is interesting to note that the researchers found that though there were statistically fewer balanced Mexican-American families when compared with Anglo-American families, their FACES III scores were similar. For the family adaptability dimension, the largest difference showed up on the "rigid" end of the continuum, with a higher percentage of Mexican-American families scoring in the rigidly disengaged area of the Circumplex model. This is significant in that therapists should have realistic expectations about shifting adaptability scores toward the center for Mexican-American families.
A simple AB design, which lends itself to the minimization of the effects of autocorrelation, was used for this single-system study. The use of a formal evaluation design, with explicit analytic procedures, met the requirements of the single-system approach. It was hypothesized that the use of a family therapy intervention would increase the level of family satisfaction as measured by the FSS. A one-way analysis of variance was conducted to determine whether there were statistically significant differences in the pre/post scores of mother and daughter.
The effect size (ES) was computed to compare the amount of change from one phase to the other (pretest and posttest scores on the FSS). The basic delta index was used, as there were fewer than 20 data points in each phase. The larger the difference in the ES, the larger the magnitude of the effect. Since the desired direction was toward higher scores, a positive or negative effect of the intervention could be inferred from the results of the ES.
Preintervention FSS scores were 43 for the mother and 50 for the daughter, indicating moderate levels of family satisfaction. Postintervention scores were 48 for the mother and 68 for the daughter, revealing improvement in their levels of family satisfaction. These scores were higher than the national means (47.0 for parents and 45.0 for adolescents).
FSS scores were tested for autocorrelation. Computation of the auto-correlation statistic indicated that the pre and post scores were independent of one another ([r.sub.k] = 0); thus, the data did not need to be transformed so as to be suitable for statistical analysis.
Using the formula to compute the delta index, the value of the effect size (Z) was 2.32. Therefore, the effect of the intervention was positive, with a 49% improvement over the baseline score on the Family Satisfaction Scale.
The analysis of variance (F = 1.178) indicated that the pre/post differences were not statistically significant. This may have resulted from the large difference in the estimate of variance of the within- and between-group means, which violates the assumption of ANOVA that there be equal variance among the groups.
The focus of this single-system design was to evaluate the usefulness of a family systems approach with a family that was experiencing problems regarding cohesion and adaptability as evidenced by the runaway behavior of an adolescent daughter. The findings confirm that use of the structured family interview, a genogram, and the combined modalities of family and individual therapy increased family satisfaction.
The literature reveals that most adolescents who run away do so because of family problems, and that this behavior is more prevalent for females (Barth, 1986). The literature also supports the use of a developmental/structural perspective to examine runaway behavior (Lappin & Covelman, 1985). From this framework, running away is seen as a symptom of the family's difficulty in dealing with the developmental processes of adolescence, such as individuation/differentiation. Restructuring relationships so that there is a more optimal commitment to the family improves members' ability to deal with change and stress (Olson & Wilson, 1982). This, in turn, should facilitate the adolescent's move toward independence and reduce runaway behavior (Crespi & Sabatelli, 1993).
The results of this study suggest that the use of a single-system design to evaluate interventions with adolescent runaways and their families will produce valuable information for clinicians working with this difficult population. However, this design, though unobtrusive and simple to use, has several limitations. It can provide clear information only between baseline and intervention. It appeared to furnish strong evidence that the intervention caused the observed changes, but without collecting data at multiple points, it was impossible to determine whether some elements were more important than others. Without measurement after each technique, it is difficult to tease out which was more effective (Bloom, Fischer, & Orme, 1995).
The fact that the stepfather chose not to participate in family therapy, and was reported to have a substance abuse problem, limited the amount of total family change that could be effected by the intervention. The mother and daughter, however, reported improved satisfaction with their relationship. The daughter's running away behavior had stopped by the termination of services and she had decided to let a maternal aunt and uncle permanently parent her one-year-old son. The mother had obtained a job and appeared to be making progress in dealing with the difficult marital situation.
Overall, the family systems approach showed promise in assisting clinicians in the difficult task of treating families with runaway adolescents. The Family Satisfaction Scale proved useful in measuring the success of the intervention.
Linda J. Courtney, Ph.D., LMSW-ACP, DePelchin Children's Center, Houston, Texas.
Reprint requests to E. Lane Coco, Ph.D., ACSW, LMSW-ACP, DePelchin Children's Center, 100 Sandman, Houston, Texas 77007.
Barth, R. P. (1986). Social and cognitive treatment of children and adolescents. San Francisco: Jossey-Bass.
Bloom, M., Fischer, J., & Orme, J. G. (1995). Evaluating practice: Guidelines for the accountable professional (2nd ed.). Massachusetts: Allyn and Bacon.
Cohen, E., Mackenzie, R. G., & Yates, G. L. (1991). HEADS, a psychosocial risk assessment instrument: Implications for designing effective intervention programs for runaway youth. Journal of Adolescent Medicine, 12, 539-544.
Crespi, T. D., & Sabatelli, R. M. (1993). Adolescent runaways and family strife: A conflict-induced differentiation framework. Adolescence, 28, 867-878.
Feitel, B., Margetson, N., Chamas, J., & Lipman, C. (1992). Psychosocial background and behavioral and emotional disorders of homeless and runaway youth. Hospital and Community Psychiatry, 43, 155-159.
Flores, M. T., & Sprenkle, D. H. (1989). Can therapists use FACES III with Mexican Americans? A preliminary analysis. In D. H. Olson, C. S. Russell, & D. H. Sprenkle (Eds.), Circumplex model: Systemic assessment and treatment of families. New York: Haworth Press.
General Accounting Office. (1989). Homelessness: Homeless and runaway youth receiving services at federally funded shelters (GAO Publication No. HRD-90-45). Washington, D.C: Author
Goldenberg, I., & Goldenberg, H. (1991) Family therapy: An overview. Pacific Grove, CA: Brooks/Cole.
Green, R. G. (1989). Choosing family measurement devices for practice and research: SFI and FACES III. Social Service Review, June, 304-320.
Greene, N. B., & Esselstyn, T. C. (1972). The beyond-control girl. Juvenile Justice, 23, 13-19.
Guerin, P. J., & Pendagast, E. G. (1976). Evaluation of family system and genogram. In P. J. Guerin, Jr. (Ed.), Family therapy: Theory and practice. New York: Gardner Press.
Homer, L. E. (1973). Community-based resource for runaway girls. Social Casework, 54, 473-479.
Kurtz, P. D., Jarvis, S. V., & Kurtz, G. L. (1991). Problems of homeless youths: Empirical findings and human services issues. Social Work, 36, 311-314.
L'Abate, L., & Colondier, G. (1987). The emperor has no clothes! Long live the emperor! A critique of family systems thinking and a reductionistic proposal. American Journal of Family Therapy, 15, 19-33.
Lappin, J., & Covelman, D. W. (1985). Adolescent runaways: A structural family therapy perspective. In M. P. Mirkin & S. L. Koman (Eds.), Handbook of adolescent and family therapy. New York: Gardner Press.
Mirkin, M. P., Raskin, P. A., & Antognini, F. C. (1984). Parenting, protecting, preserving: Mission of the adolescent female runaway. Family Process, 23, 63-74.
National Network of Runaway and Youth Services. (1985). To whom do they belong? A profile of America's runaways and homeless youth and the programs that help them. Washington, D.C: Author.
Olson, D. H. (1989). Circumplex model of family systems: FACES III. Family assessment and intervention. In D. H. Olson, C. S. Russell, & D. H. Sprenkle (Eds.), Circumplex model: Systemic assessment and treatment of families New York: Haworth Press.
Olson, D. H., Russell, C. S., & Sprenkle, D. H. (1983). Circumplex model VI: Theoretical update. Family Process, 22, 69-83.
Olson, D. H., & Wilson, M. (1982). Family satisfaction. In D. Olson et al. (Eds.), Family inventories. St. Paul, MN: Family Social Science, University of Minnesota.
Pratt, D. M., & Hansen, J. C. (1987). A test of the curvilinear hypothesis with FACES II and III. Journal of Marital and Family Therapy, 13, 387-392.
Rodick, J. D., Henggeler, S. W., & Hanson, C. L. (1986). An evaluation of the Family Adaptability and Cohesion Evaluation Scales and the Circumplex model. Journal of Abnormal Child Psychology, 14, 77-87.
Searight, H. R., & Merkel, W. T. (1991). Systems theory and its discontents: Clinical and ethical issues. American Journal of Family Therapy, 19, 19-31.
Sharlin, S. A., & Mor-Barak, M. (1992). Runaway girls in distress: Motivation, background, and personality. Adolescence, 27, 387-405.
Schwartzman, J., & Restivo, R. J. (1985). Acting out and staying in: Juvenile probation and the family. In J. Schwartzman (Ed.), Families and other systems: The macrosystemic context of family therapy. New York: Guilford Press.
Speck, N. B., Ginther, D. W., & Helton, J. R. (1988). Runaways: Who will run away again? Adolescence, 23, 881-888.
Zide, M. R., & Cherry, A. L. (1992). A typology of runaway youths: An empirically based definition. Child and Adolescent Social Work Journal, 9, 155-168.…