- Email: [email protected]
Decision-making in interdisciplinary treatment teams
Child Abuse & Negkcr, Vol. 12, pp. 231-239. Printed in the U.S.A. All rights reserved.
1988 Copyright
DECISION-MAKING
0145-2134188 $3.00 + .otl Q 1988 Pergamon Press plc
IN INTERDISCIPLINARY
TREATMENT JANE F. GILGUN,
TEAMS PH.D.
Assistant Professor, School of Social Work, University of Minnesota, Twin Cities Abstract-Interdisciplinary teams for the treatment of child abuse and neglect are becoming more common. Studies have shown that decisions made by groups who have had the opportunity to discuss their perspectives are more accurate than judgments made by individuals. Why this may be true is not clear. The purpose of the present study was to discover the procedures an interdisciplinary treatment team uses in making decisions. A single interdisciplinary incest treatment team was observed over a 15-month period. Open-ended interviews with team members also were conducted. Findings show that the interdisciplinary treatment team made its decisions using procedures analogous to procedures used in social research to establish reliability and validity. The decision-making process of the team was characterized by multiple observations of family members by multiple observers in multiple settings over time. This decision-making process is similar to processes used by many other treatment teams. The findings of the present research, then, are likely to be generalizable to other teams whose decision-making processes are similar.
INTERDISCIPLINARY
TEAMS for the treatment of child abuse and neglect are becoming increasingly common in settings such as hospitals, mental health centers, prisons, and public child welfare and child protection agencies [ 1, 21. These teams have developed out of the assumption that members of one discipline alone are not able to assess and treat the complex issues with which clients contend [l-3]. Composition of teams varies by setting, with nurses, social workers, medical doctors, and psychiatrists heavily represented on hospital-based teams, while social workers, psychologists, and sometimes police, home aides, and teachers are more common on community-based teams. The decisions these teams make range from deciding whether abuse or neglect has occurred to making a judgment about whether a perpetrator of child sexual abuse is likely to re-offend. Team decisions, therefore, have serious implications for individual and social wellbeing. Laboratory studies have shown that decisions made by groups whose members have the opportunity to interact and to discuss their perspectives are more accurate than judgments made by individuals [4]. Why group decisions are more accurate than individual decisions is not clear. Procedures by which groups come to decisions have not yet been delineated. Criteria by which to judge the quality of group decision-making procedures are lacking. Finally, evaluative criteria for judging the quality of group decisions for real world problems have not been developed [4]. The present study sought to address some of the gaps in the understanding of the team decision-making process.
This research was sponsored through a grant from the Minnesota Agricultural Experiment Station, University of Minnesota, Twin Cities. This paper is a revised version of a paper presented at the annual meeting of the Groves Conference on Marriage and the Family held in London, England, July 16-20, 1986. Submitted 1987.
for publication March 30, 1987; final draft received September
25, 1987; accepted
Reprint requests to Jane F. Gilgun, Ph.D., School of Social Work, University of Minnesota, S.E., Minneapolis, MN 55455. 231
September
30,
224 Church St.
232
Jane F. Gilgun
The purpose of the study was to answer the following research question: What procedures does an interdisciplinary treatment team use in its decision-making process? This question is focused on procedure and process. It does not address issues of outcome of treatment nor does it investigate the specific content of a treatment team’s deliberation. The goal of the study was to begin to develop criteria by which to judge the quality of group decisions for real world problems.
METHOD The interdisciplinary team of the Central Minnesota Incest Treatment Program located in St. Cloud, Minnesota, was the subject of this study. St. Cloud is a small city with a population of about 43,000 located 70 miles west of Minneapolis-St. Paul. Servicing a predominantly rural, four-county area, the program provides treatment to all family members on the same day, and the method is group treatment [5]. The program site is the Central Minnesota Mental Health Center, which also provides mental health services to the four counties. Personnel from the Central Minnesota Mental Health Center and from the social service agencies of the counties which the treatment program serves staff the treatment program. Administrators from the counties and the mental health center release personnel from their regular duties in order to work on the incest treatment team. At the time of data collection, the team was composed of 14 members: two M.S.W.s, four B.S.W.s, one Ph.D. clinical psychologist, two Ph.D.-level school psychology interns, one master’s level counseling intern, one master’s level practitioner in general counseling, and three volunteer foster grandparents. Clients During the time of data collection, the number of clients ranged from a low of 35 to a high of 55. Numbers of families ranged from 9 to 14. When data collection began, 55 clients who were members of 14 families were in treatment. Eight of these families were nuclear; three other families were mothers with children. Two perpetrators attended without their families, and one victim attended without her family. The victims ranged in age from 12 to 18. Nonvictimized siblings also were in treatment. They ranged in age from 5 to 18. Program Structure The program used three different types of group treatment: peer groups, multi-family groups, and a didactic group called the family lab. Peer groups met weekly, and there were six: one for perpetrators, one for mothers, two for victims, and two for siblings. The sibling and victim groups were divided by age. Multi-family group and family lab met on alternate weeks. In multi-family groups, two or more families were in treatment together. In family lab, information on parenting, sexuality, family communication, social support, and other topics intended to build on family strength were presented. Each group had two leaders from the professional staff. The mothers’ group also had a foster grandmother, and one foster grandmother was in each of the two sibling groups. Families also were “staffed” every three months. At this staffing, individualized treatment objectives and strategies were developed and progress in treatment was assessed. The entire family was present, as were at least two members of the treatment team. The goals, objectives, strategies, and progress made were recorded by the treatment team members. Thus, two independent written records were kept of the families’ progress in
Interdisciplinary
treatment teams
treatment. Special family meetings called focus groups were convened were in crisis or if one or more family members were having difficulty treatment.
233
when families progressing in
Criteria for Graduation
In order to graduate from the program, clients had to meet program objectives. Some of the objectives were similar for mothers, perpetrators, victims, and siblings. These goals were taking appropriate responsibility for one’s own behavior, demonstrating an understanding of the dynamics of incest, forgiving oneself for any hurtful behavior, developing communication skills including assertiveness, expressing and managing the emotions engendered by the incest, and developing an understanding of human sexuality. A common task was writing and sharing an autobiography. Objectives specific to the perpetrator included apologizing to the victim, developing a plan for the control of incestuous impulses, and demonstrating progress in working out an equitable marital relationship. For mothers, the more specific objectives were confronting husbands on their incestuous behavior, demonstrating ability to listen to and communicate with the daughter on the daughter’s feelings and thoughts on the mother’s role in the incest, demonstrating ability to protect their children from further incest, and working out an equitable marital relationship. The more specific objectives for victims were communicating their feelings about the incest to their mothers, fathers, and siblings, developing a self-protection plan if they were confronted again with a possible sexual victimization, and demonstrating an ability to develop equitable relationships with peers. Sibling objectives included demonstrating an understanding of how the incest affected them, developing an appreciation of the difficult role thrust on the victim because of the incest, and forgiving parents for their roles in the incest. Families entered the program in August and January. Graduations were in June and December, although some families completed treatment before the regularly scheduled graduation and thus graduated throughout the program year. Length of treatment varied from nine months to two years. Average length of treatment was 15 months. Treatment
Day
The treatment day provided the team with a minimum of three opportunities to discuss treatment strategies and clients’ progress in treatment. The first opportunity for team discussion was at the two-hour staff meeting which opened the treatment day. The second opportunity came at dinner break, which was also time out between peer group and either multi-family group or family lab. Over the half-hour dinner, the team continued their discussion of the families’ treatment progress. The team ate separately from the families. The third opportunity for team discussion was at the end of the program day, when the team did a one-hour summary session. Data
Collection
and Procedure
Participant observation and open-ended interviews with team members were the methods used to answer the research question. The treatment team was observed by a single researcher a total of ten times between February and July 1985. Each professionally trained team member (11 members) was interviewed during June, July, and August 1985. Three follow-up observations were conducted in January, April, and May 1986. A group interview with the treatment team was held in May 1986. The total observation period was 15 months.
234
Jane F. Gilgun
The researcher observed the team members as they interacted with each other over the treatment day. Following Bogdan and Biklen, the researcher took no notes during the observations, a procedure intended to decrease observation intrusiveness [6]. No observations were made during the treatment group sessions, except of the family lab. While the team was involved in groups, the researcher audiotaped observations just made of the team’s interactions. Thus, observations were recorded immediately after they were made. The audiotapes were transcribed and content analyzed. For the open-ended interview, the researcher asked each subject individually the research question: How do you see the team making decisions about a family’s readiness to graduate? Graduation was chosen as the focal point for the question since graduation was the focal point and over-all goal of treatment. The question generated a conversation which the researcher guided so as to cover the same general areas in each interview and to test the hypotheses generated in the interviews and through the observations. The researcher took notes for these interviews. In addition, four of the interviews were also audiotaped and transcribed. The notes were expanded into narrative form within 24 hours. These data were content analyzed. RESULTS The major finding of this study is the following: The treatment team of the Central Minnesota Incest Treatment Program makes its decisions using procedures analogous to procedures used in social research to establish reliability and validity. This statement is based on the finding that the decision-making process of the treatment team was characterized by multiple observations of family members by multiple observers in multiple settings over time. Under varying conditions, family members and whole families were observed by the team members. The behavior of each client was continually noted and observed by members of the team. In the discussion which follows, data supporting this finding are reported, and concepts of reliability and validity are applied to the decisionmaking procedures underlying the team’s interactions. The decision-making process of the treatment team has many similarities to the procedures used to establish reliability and validity of social research findings. Reliability has to do with consistency and repeatability of findings [7]. Findings are considered reliable when they are consistent across subjects, observers, and settings. For example, findings are judged reliable when the same subjects provide similar responses in different settings with observations by the same and different observers, when different subjects provide similar responses in the same settings with observations by the same researcher and in any number of other combinations of subjects, settings, and researchers. Time can be an important dimension in establishing reliability, where similar observations made at different times and thus in different settings lead to the conclusion that findings probably are reliable. The repeatability of findings is called replication, the hallmark of findings judged reliable. Validity is judged by the thoroughness and accuracy with which phenomena are examined and by the convergence of related phenomena and the divergence of unrelated phenomena, Findings are considered valid when there are multiple sources of data and when the data from divergent sources can be grouped into categories 171.The grouping depends on a judgment that certain phenomena are related and certain phenomena are not. A procedure for promoting validity of research findings is called triangulation of sources of data. Examples of data sources are observations, self-report, and written documentation. Patterns which emerge from data collected from multiple sources can be judged to be more valid than data collected from a single source [S, 91. Neither validity nor reliability can be established de~nitively, but they can be established in a logical and probabilistic
Interdisciplinary
treatment teams
235
sense. A finding can be considered probably reliable and valid under the conditions discussed above. Team members did not consciously structure their decision-making process in order to follow procedures used to establish reliability and validity. Rather, these procedures developed out of the structure of the program day. That the team did follow these procedures is supported by the following data. Multiple
Observations
and Data Sources
The opportunities for observation of family members were multiple. Team members observed families in four different settings during the treatment day: in peer group, in either multi-family group or family lab, and informally at dinner and in the hallway. Each family also was seen for the quarterly staftings. The settings in which family members were observed and the observers varied across the treatment day. Not only were the leadership pairs varied across groups, but there were three different groups in which the family members were observed. The result of this variation in leadership and group composition was that some clients were observed by the same team members in more than one group and some clients were observed in these settings by different team leaders. Through sharing of observations made in various settings week after week, a detailed longitudinal picture of individual and family functioning emerged. This process fulfills the repeatability and consistency criteria for the establishment of reliability. Reliability by itself is relatively meaningless unless it is coupled with validity. The use of multiple data sources is a major criterion for the establishment of validity. Not only did the team have multiple first-hand observations of family and individual functioning, the team’s observations were supplemented by data from other sources. These sources included weekly process notes made by at least one leader of each treatment group, the written records of the quarterly staffing, pre-sentence reports supplied by the department of probation and parole, social histories from county social workers, and conferences and informal contacts with other service providers who had contact with family members. Team members sometimes had collateral contacts with teachers, foster parents when children in treatment were in foster care, and with prison officials when perpetrators were in prison. In addition, county social workers for four of the families were also members of the treatment team. These team members supplied the team with detailed information on family functioning outside of the treatment program. This array of data, gathered both from their observations of and interactions with family members and from many other sources, provided the team with an in-depth description of individual and family status over time [lo]. Given the varied source of data on families, observations of family members likely were valid. Data supporting these logical arguments are presented below. The findings are grouped under the following headings: same observer of same client in different settings, different observers of same client in different settings, different observers of members of same families in different settings, and multiple observers of family members in interaction with others. The findings are organized this way to demonstrate that the decision-making procedures follow the logic of establishing the reliability and validity of social research findings. Same
Observer
of Same
Client in Different
Settings
When team members had opportunities to observe the same client in more than one setting, they were able to assess the consistency of the client’s behavior. An interview of a siblings group leader illustrates how this process worked. She reported:
236
Jane E Gilgun
The client, an adolescent boy, brother of an incest victim, operated in two such different ways. The boy was definitely one of the most active members of the peer group and was extremely insightful and could be very responsive to others’ feelings. I would, in that group, describe him as a very caring person, Then we would go into family group, and not always but on occasion he would be a pure snot, typically to his mother. And he would tind it very difficult to verbalize his own feelings. I’m not sure if he could even get in touch with his feelings, he was so upset, but I know for sure that he had trouble verbalizing what was happening for him in muIti-family group.
Such data provided the team with a basis on which to assess the boy’s progress in treatment. Had the team only observed this client in one setting, they would have had an incomplete picture of his capabilities and how his behavior related to the behavior of other family members, Decisions based on incomplete pictures of client functioning run a risk of not being reliable and valid. Different Observers of Same Client in Different Settings Consistency or inconsistency of client behaviors also can be evaluated when there are different observers of the same client in different settings. The following description of a perpetrator’s behavior in peer group and in multi-family group on the same treatment day illustrates how this process works. A perpetrator who had been highly resistant to admitting abusing one of his daughters admitted to touching her breasts. He said to her in the multi-family group, “I’m really sorry I did this to you. I’m really sorry I touched your breasts. And I will do all in my power to make sure that it doesn’t happen again. I’m sorry I hurt you.” At the team’s final summary session of the day, the perpetrator’s peer group leader reported that the perpetrator had shown more change, being empathic yet gently confrontive to a man new to the group that night. When the new man began to minimize what he had done to his daughter, the perpetrator paid close attention and told the new man that his behavior may have hurt his daughter a great deal. Several team members expressed surprise at the changes the perpetrator was showing, and the team agreed that he was making some major breakthroughs. Different observers of the same client in various settings were able to contribute to a reliable and valid picture of the client’s progress through treatment. Different Observers of Members of Same Famiiies in Different Settings Reliability was also assessed when different team leaders observed different members of the same family. Sometimes clients made false reports of the status of the marriage or family life. For example, one man said he had been having great sex with his wife. When the leader of his peer group reported this at the team meeting, the leader of the mother’s group reported, “His wife said she hasn’t had sex with him for six months and she’s divorcing him.” The reliability of the husband’s statement about the couple’s sexual life was challenged by the wife’s statement. Again, having the statement of only one of the partners would have provided an incomplete picture. Another example of inconsistency of reports across different family members was the divergent viewpoint of a father and his sons. The father said that things were much better with his sons. When the father’s peer group leader reported this at a team meeting, a leader of the older siblings group indicated, “That’s funny. The sons said they’re ready to kill their father.” In such ways a more complete evaluation of family relationships was done. Often several team members made contributions to an emerging picture of a family,
Interdisciplinary
237
treatment teams
One example is the following: The team coordinator out of jail. The following team dialogue ensued:
announced
that a father was getting
Sibting Group Leader
I: Why? For what he did he should stay in jail for the rest of his life. He sexually abused his son, made him scratch himself until he bled, burned him with cigarettes. Coordinator: In muIti-famiIy group, you can’t get a piece of paper between them. In the last three groups, he sat on his father’s lap. Victim Group Leader: His daughter hates the father. If looks could kill, he’d be dead. SiMng Group Leader 2: Another son is spaced out. He could walk out in the middle of traffic, and he wouldn’t know what was going on.
Team leaders, then, commenting on the functioning of one family, were able to give a developing picture of family interaction and progress through treatment. Multiple Observers of the Family Members ip1Interaction
with Others
The structure of the treatment program allowed the entire family to be observed as they interacted with peers and with members of their own and others’ families. For perpetrators whose victims were not participating in the treatment program, their ability to take responsibility for their own behavior was assessed in multi-family group or family lab by their interactions with daughters of other perpetrators. One man told another man’s daughter: “I hope you are not blaming yourself for the abuse, I knew a lot more than my own daughter, and I took advantage of her. It is very important for you to understand that. I took advantage of her.” This clear statement of responsibility was an important factor in the team’s decision that this man was ready to graduate. This statement and others which were similar were reported to the treatment team. Over time, a picture of a man capable of dealing with incest emerged.
DISCUSSION Observation and open-ended interviews led to the finding that the incest treatment team followed procedures for establishing reliability and validity. The repeated nature of their observations of clients, observations characterized by variation in observer and setting, combined with multiple data sources, all are fundamental procedures used to establish validity and reliability of the findings of social research. The decision-making process observed for the purpose of this study would be similar to the decision-making processes used by other interdisciplinary teams who structure their time in the manner of the present team. Sharing and discussion of observations over time are crucial elements in the present team’s decision-making. The findings of the present research, then, can be hypothesized as generalizable in other teams whose decision-making processes are similar. Applying the concepts of reliability and validity to team decision making may help to solve a problem discussed above, namely, the lack of evaluative criteria on which to judge the quality of group decisions which have implications for real world problems. The present study suggests some evaluative criteria. This study provides a basis for concluding that decisions about complex phenomena, such as child abuse and neglect, require the knowledge and observations an interdisciplinary team can provide. A dimension of team decision-making which this study highlighted is the dimension of time. The establishment of reliability and validity rests on repeated observations from multiple observers over time.
Jane F. Gilgun
238
Money often is a barrier to the development of interdisciplinary teams. Administrators and boards of directors can be reluctant to set aside capital to fund a team. Yet, the team in the present research operated with no budget at all. The mental health center and the county agencies supporting the team provided release time for team members. The mental health center donated the space, which would have gone unused during the time the team operated. Of course, releasing employees for about seven hours per week represents a cost to each of the agencies, but it was a cost offset by the benefits of providing intensive treatment which would not have been available otherwise. Interdisciplinary treatment teams provide clear advantages over practitioners working alone. Unfortunately, however, many practitioners do operate alone. For individuals to expand the reliability and validity of their decision-making processes, they would need to gather information on client functioning from as many sources as possible. Peer supervision, case conferences, and the development of an informal group of consultants are ways the individual practitioner can improve the decision-making process. Consultants would include persons knowledgeable about the family or about child abuse and neglect, and they could include medical doctors, psychiatrists, psychologists, and social workers. Child abuse and neglect are complex phenomena. Decisions practitioners make about families where abuse and neglect occur can affect the course of the family members’ lives. Ethics require that these decisions be as reliable and valid as possible. Multiple perspectives of multiple observers over time are ways of ensuring ethical practice. Acknowledgment-The Central Minnesota
Incest
author thanks Dr. Rosalie A. Kane, the anonymous reviewers, and the team Treatment Program for their comments on earlier drafts of this report.
of the
NOTES AND REFERENCES 1. STEELE, B. F. Experience with an interdisciplinary concept. In: Child Abuse and Neg/ect: The Family and the Community, R. E. Helfer and C. H. Kempe (Eds.), pp. 163-168. Ballinger, Cambridge, MA (1976). 2. FALLER, K. C. (Ed.). Social Work with Abused and Neglected Children. Collier Macmillan, London (1981). 3. PICKETT, J. and MATION, A. The multi-disciplinary team in an urban setting: The special unit concept. In: The Abused Child in the Family and in the Community, C. H. Kempe, A. W. Franklin, and C. Cooper (Eds.), pp. 115-121. Pergamon, Oxford (1980). 4. HART, S. L. Toward quality criteria for collective judgments. Organizational Behavior and Decision Process 36:209-228 (1986). 5. The Central Minnesota Incest Treatment Program is based on a model developed by the Range Mental Health Center of Virginia, Minnesota. See. RANDALL, H., BELL, C. M. and CADY, K. Treuting Incest in Rurul Families. Pange Mental Health Center, Virginia, MN (1982). 6. BOGDAN, R. C. and BIKLEN, S. K. Qualitative Researchfor Education. Allyn and Bacon, Boston (1982). 7. KERLINGER, F. N. The Foundutions of Behuvioral Research. Holt, Rinehart and Winston, New York (1983). 8. COOK, T. D. and CAMPBELL, D. T. Quasi-Experimenfution: Design and Analysis for Field Settings. Houghton Mifflin, Boston (1979). 9. DENZIN, N. K. The Research Act. Basic Books, New York (1978). 10. GERTZ, C. The Inrerprefafion ofCultures. Basic Books, New York (1973).
Resume-Les Cquipes interdisciplinaires destinees a assumer le traitement des enfants victimes de s&ices ou de negligences se sont multiplies. Des etudes ont bien demontre que les decisions prises par des groupes dont les membres discutent entre eux selon differents point de vue sont meilleurs que les jugements port& par des individus isolts. Le pourquoi de cet avantage n’est pas evident. L’idee de la presente etude Ctait d’essayer de trouver par quel moyen une Cquipe interdisciplinaire arrive a ses decisions. Les auteurs ont done observe pendant 15 mois, une telle Cquipe chargee de traitement de cas d’incestes. On a interviewe des membres de Les auteurs ont trouve que l’equipe interdisciplinaire I’equipe separement selon la technique “open-ended.” therapeutique arrivait a ses decisions en utilisant des cheminements analogues ?t ceux dont on se set-t en recherche sociologique pour etablir la fiabilite et la validitt. Le processus decisionnaire Ctait fonde sur l’observation rep&e des membres de la families par de multiples observateurs dans des situations variees au tours du
Interdisciplinary
treatment teams
239
temps. Le processus decisionnaire ressemble a ceux qui sont utilises par beaucoup d’autres Cquipes therapeutiques. Les resultats de la presente etude sont probablement generalisables a I’activite d’autres Cquipes utilisant les memes processus decisionnaires. Resumen -El ndmero de equipos interdisciplinarios para el tratamiento de1 abuso y la negligencia de1 nitio esta creciendo. Las investigaciones han demonstrado que ias decisiones hechas por grupos que han tenido la oportunidad de intercambi~ perspectivas son mas exactas que 10sjuicios hechos solamente par individuos. La razon no es conocida. El objective de la presente investigaci6n fuf: el de describir 10s procedimientos que un equip0 interdisciplinario usa en llegar a decisiones. Un equipo interdisciplinario para el tratamiento de1 incest0 fue observado por un period0 de 15 meses. Tambien se llevaron a cabo entrevistas no estructuradas con miembros de1 equipo. Los resultados indican que el equip0 interdisciplinario de tratamiento llego a sus decisiones usando procedimientos analogos a 10s usados en las investigaciones de las ciencias sociales para establecer la formalidad y la validez. El proceso para llegar a decisiones de1 equip0 se caracteriz6 por observaciones multiples de 10s miembros de las familias llevadas a cabo por observadot-es multiples en locaciones multiples durante el curso del tiempo. Esta manera de llegar a decisiones es similar a la usada por muchos equipos de tratamiento. Los resultados de la presente investigation, por tanto, parecen ser aplicables a otros equipos cuyas formas de llegar a decisiones son similares.
1988 Copyright
DECISION-MAKING
0145-2134188 $3.00 + .otl Q 1988 Pergamon Press plc
IN INTERDISCIPLINARY
TREATMENT JANE F. GILGUN,
TEAMS PH.D.
Assistant Professor, School of Social Work, University of Minnesota, Twin Cities Abstract-Interdisciplinary teams for the treatment of child abuse and neglect are becoming more common. Studies have shown that decisions made by groups who have had the opportunity to discuss their perspectives are more accurate than judgments made by individuals. Why this may be true is not clear. The purpose of the present study was to discover the procedures an interdisciplinary treatment team uses in making decisions. A single interdisciplinary incest treatment team was observed over a 15-month period. Open-ended interviews with team members also were conducted. Findings show that the interdisciplinary treatment team made its decisions using procedures analogous to procedures used in social research to establish reliability and validity. The decision-making process of the team was characterized by multiple observations of family members by multiple observers in multiple settings over time. This decision-making process is similar to processes used by many other treatment teams. The findings of the present research, then, are likely to be generalizable to other teams whose decision-making processes are similar.
INTERDISCIPLINARY
TEAMS for the treatment of child abuse and neglect are becoming increasingly common in settings such as hospitals, mental health centers, prisons, and public child welfare and child protection agencies [ 1, 21. These teams have developed out of the assumption that members of one discipline alone are not able to assess and treat the complex issues with which clients contend [l-3]. Composition of teams varies by setting, with nurses, social workers, medical doctors, and psychiatrists heavily represented on hospital-based teams, while social workers, psychologists, and sometimes police, home aides, and teachers are more common on community-based teams. The decisions these teams make range from deciding whether abuse or neglect has occurred to making a judgment about whether a perpetrator of child sexual abuse is likely to re-offend. Team decisions, therefore, have serious implications for individual and social wellbeing. Laboratory studies have shown that decisions made by groups whose members have the opportunity to interact and to discuss their perspectives are more accurate than judgments made by individuals [4]. Why group decisions are more accurate than individual decisions is not clear. Procedures by which groups come to decisions have not yet been delineated. Criteria by which to judge the quality of group decision-making procedures are lacking. Finally, evaluative criteria for judging the quality of group decisions for real world problems have not been developed [4]. The present study sought to address some of the gaps in the understanding of the team decision-making process.
This research was sponsored through a grant from the Minnesota Agricultural Experiment Station, University of Minnesota, Twin Cities. This paper is a revised version of a paper presented at the annual meeting of the Groves Conference on Marriage and the Family held in London, England, July 16-20, 1986. Submitted 1987.
for publication March 30, 1987; final draft received September
25, 1987; accepted
Reprint requests to Jane F. Gilgun, Ph.D., School of Social Work, University of Minnesota, S.E., Minneapolis, MN 55455. 231
September
30,
224 Church St.
232
Jane F. Gilgun
The purpose of the study was to answer the following research question: What procedures does an interdisciplinary treatment team use in its decision-making process? This question is focused on procedure and process. It does not address issues of outcome of treatment nor does it investigate the specific content of a treatment team’s deliberation. The goal of the study was to begin to develop criteria by which to judge the quality of group decisions for real world problems.
METHOD The interdisciplinary team of the Central Minnesota Incest Treatment Program located in St. Cloud, Minnesota, was the subject of this study. St. Cloud is a small city with a population of about 43,000 located 70 miles west of Minneapolis-St. Paul. Servicing a predominantly rural, four-county area, the program provides treatment to all family members on the same day, and the method is group treatment [5]. The program site is the Central Minnesota Mental Health Center, which also provides mental health services to the four counties. Personnel from the Central Minnesota Mental Health Center and from the social service agencies of the counties which the treatment program serves staff the treatment program. Administrators from the counties and the mental health center release personnel from their regular duties in order to work on the incest treatment team. At the time of data collection, the team was composed of 14 members: two M.S.W.s, four B.S.W.s, one Ph.D. clinical psychologist, two Ph.D.-level school psychology interns, one master’s level counseling intern, one master’s level practitioner in general counseling, and three volunteer foster grandparents. Clients During the time of data collection, the number of clients ranged from a low of 35 to a high of 55. Numbers of families ranged from 9 to 14. When data collection began, 55 clients who were members of 14 families were in treatment. Eight of these families were nuclear; three other families were mothers with children. Two perpetrators attended without their families, and one victim attended without her family. The victims ranged in age from 12 to 18. Nonvictimized siblings also were in treatment. They ranged in age from 5 to 18. Program Structure The program used three different types of group treatment: peer groups, multi-family groups, and a didactic group called the family lab. Peer groups met weekly, and there were six: one for perpetrators, one for mothers, two for victims, and two for siblings. The sibling and victim groups were divided by age. Multi-family group and family lab met on alternate weeks. In multi-family groups, two or more families were in treatment together. In family lab, information on parenting, sexuality, family communication, social support, and other topics intended to build on family strength were presented. Each group had two leaders from the professional staff. The mothers’ group also had a foster grandmother, and one foster grandmother was in each of the two sibling groups. Families also were “staffed” every three months. At this staffing, individualized treatment objectives and strategies were developed and progress in treatment was assessed. The entire family was present, as were at least two members of the treatment team. The goals, objectives, strategies, and progress made were recorded by the treatment team members. Thus, two independent written records were kept of the families’ progress in
Interdisciplinary
treatment teams
treatment. Special family meetings called focus groups were convened were in crisis or if one or more family members were having difficulty treatment.
233
when families progressing in
Criteria for Graduation
In order to graduate from the program, clients had to meet program objectives. Some of the objectives were similar for mothers, perpetrators, victims, and siblings. These goals were taking appropriate responsibility for one’s own behavior, demonstrating an understanding of the dynamics of incest, forgiving oneself for any hurtful behavior, developing communication skills including assertiveness, expressing and managing the emotions engendered by the incest, and developing an understanding of human sexuality. A common task was writing and sharing an autobiography. Objectives specific to the perpetrator included apologizing to the victim, developing a plan for the control of incestuous impulses, and demonstrating progress in working out an equitable marital relationship. For mothers, the more specific objectives were confronting husbands on their incestuous behavior, demonstrating ability to listen to and communicate with the daughter on the daughter’s feelings and thoughts on the mother’s role in the incest, demonstrating ability to protect their children from further incest, and working out an equitable marital relationship. The more specific objectives for victims were communicating their feelings about the incest to their mothers, fathers, and siblings, developing a self-protection plan if they were confronted again with a possible sexual victimization, and demonstrating an ability to develop equitable relationships with peers. Sibling objectives included demonstrating an understanding of how the incest affected them, developing an appreciation of the difficult role thrust on the victim because of the incest, and forgiving parents for their roles in the incest. Families entered the program in August and January. Graduations were in June and December, although some families completed treatment before the regularly scheduled graduation and thus graduated throughout the program year. Length of treatment varied from nine months to two years. Average length of treatment was 15 months. Treatment
Day
The treatment day provided the team with a minimum of three opportunities to discuss treatment strategies and clients’ progress in treatment. The first opportunity for team discussion was at the two-hour staff meeting which opened the treatment day. The second opportunity came at dinner break, which was also time out between peer group and either multi-family group or family lab. Over the half-hour dinner, the team continued their discussion of the families’ treatment progress. The team ate separately from the families. The third opportunity for team discussion was at the end of the program day, when the team did a one-hour summary session. Data
Collection
and Procedure
Participant observation and open-ended interviews with team members were the methods used to answer the research question. The treatment team was observed by a single researcher a total of ten times between February and July 1985. Each professionally trained team member (11 members) was interviewed during June, July, and August 1985. Three follow-up observations were conducted in January, April, and May 1986. A group interview with the treatment team was held in May 1986. The total observation period was 15 months.
234
Jane F. Gilgun
The researcher observed the team members as they interacted with each other over the treatment day. Following Bogdan and Biklen, the researcher took no notes during the observations, a procedure intended to decrease observation intrusiveness [6]. No observations were made during the treatment group sessions, except of the family lab. While the team was involved in groups, the researcher audiotaped observations just made of the team’s interactions. Thus, observations were recorded immediately after they were made. The audiotapes were transcribed and content analyzed. For the open-ended interview, the researcher asked each subject individually the research question: How do you see the team making decisions about a family’s readiness to graduate? Graduation was chosen as the focal point for the question since graduation was the focal point and over-all goal of treatment. The question generated a conversation which the researcher guided so as to cover the same general areas in each interview and to test the hypotheses generated in the interviews and through the observations. The researcher took notes for these interviews. In addition, four of the interviews were also audiotaped and transcribed. The notes were expanded into narrative form within 24 hours. These data were content analyzed. RESULTS The major finding of this study is the following: The treatment team of the Central Minnesota Incest Treatment Program makes its decisions using procedures analogous to procedures used in social research to establish reliability and validity. This statement is based on the finding that the decision-making process of the treatment team was characterized by multiple observations of family members by multiple observers in multiple settings over time. Under varying conditions, family members and whole families were observed by the team members. The behavior of each client was continually noted and observed by members of the team. In the discussion which follows, data supporting this finding are reported, and concepts of reliability and validity are applied to the decisionmaking procedures underlying the team’s interactions. The decision-making process of the treatment team has many similarities to the procedures used to establish reliability and validity of social research findings. Reliability has to do with consistency and repeatability of findings [7]. Findings are considered reliable when they are consistent across subjects, observers, and settings. For example, findings are judged reliable when the same subjects provide similar responses in different settings with observations by the same and different observers, when different subjects provide similar responses in the same settings with observations by the same researcher and in any number of other combinations of subjects, settings, and researchers. Time can be an important dimension in establishing reliability, where similar observations made at different times and thus in different settings lead to the conclusion that findings probably are reliable. The repeatability of findings is called replication, the hallmark of findings judged reliable. Validity is judged by the thoroughness and accuracy with which phenomena are examined and by the convergence of related phenomena and the divergence of unrelated phenomena, Findings are considered valid when there are multiple sources of data and when the data from divergent sources can be grouped into categories 171.The grouping depends on a judgment that certain phenomena are related and certain phenomena are not. A procedure for promoting validity of research findings is called triangulation of sources of data. Examples of data sources are observations, self-report, and written documentation. Patterns which emerge from data collected from multiple sources can be judged to be more valid than data collected from a single source [S, 91. Neither validity nor reliability can be established de~nitively, but they can be established in a logical and probabilistic
Interdisciplinary
treatment teams
235
sense. A finding can be considered probably reliable and valid under the conditions discussed above. Team members did not consciously structure their decision-making process in order to follow procedures used to establish reliability and validity. Rather, these procedures developed out of the structure of the program day. That the team did follow these procedures is supported by the following data. Multiple
Observations
and Data Sources
The opportunities for observation of family members were multiple. Team members observed families in four different settings during the treatment day: in peer group, in either multi-family group or family lab, and informally at dinner and in the hallway. Each family also was seen for the quarterly staftings. The settings in which family members were observed and the observers varied across the treatment day. Not only were the leadership pairs varied across groups, but there were three different groups in which the family members were observed. The result of this variation in leadership and group composition was that some clients were observed by the same team members in more than one group and some clients were observed in these settings by different team leaders. Through sharing of observations made in various settings week after week, a detailed longitudinal picture of individual and family functioning emerged. This process fulfills the repeatability and consistency criteria for the establishment of reliability. Reliability by itself is relatively meaningless unless it is coupled with validity. The use of multiple data sources is a major criterion for the establishment of validity. Not only did the team have multiple first-hand observations of family and individual functioning, the team’s observations were supplemented by data from other sources. These sources included weekly process notes made by at least one leader of each treatment group, the written records of the quarterly staffing, pre-sentence reports supplied by the department of probation and parole, social histories from county social workers, and conferences and informal contacts with other service providers who had contact with family members. Team members sometimes had collateral contacts with teachers, foster parents when children in treatment were in foster care, and with prison officials when perpetrators were in prison. In addition, county social workers for four of the families were also members of the treatment team. These team members supplied the team with detailed information on family functioning outside of the treatment program. This array of data, gathered both from their observations of and interactions with family members and from many other sources, provided the team with an in-depth description of individual and family status over time [lo]. Given the varied source of data on families, observations of family members likely were valid. Data supporting these logical arguments are presented below. The findings are grouped under the following headings: same observer of same client in different settings, different observers of same client in different settings, different observers of members of same families in different settings, and multiple observers of family members in interaction with others. The findings are organized this way to demonstrate that the decision-making procedures follow the logic of establishing the reliability and validity of social research findings. Same
Observer
of Same
Client in Different
Settings
When team members had opportunities to observe the same client in more than one setting, they were able to assess the consistency of the client’s behavior. An interview of a siblings group leader illustrates how this process worked. She reported:
236
Jane E Gilgun
The client, an adolescent boy, brother of an incest victim, operated in two such different ways. The boy was definitely one of the most active members of the peer group and was extremely insightful and could be very responsive to others’ feelings. I would, in that group, describe him as a very caring person, Then we would go into family group, and not always but on occasion he would be a pure snot, typically to his mother. And he would tind it very difficult to verbalize his own feelings. I’m not sure if he could even get in touch with his feelings, he was so upset, but I know for sure that he had trouble verbalizing what was happening for him in muIti-family group.
Such data provided the team with a basis on which to assess the boy’s progress in treatment. Had the team only observed this client in one setting, they would have had an incomplete picture of his capabilities and how his behavior related to the behavior of other family members, Decisions based on incomplete pictures of client functioning run a risk of not being reliable and valid. Different Observers of Same Client in Different Settings Consistency or inconsistency of client behaviors also can be evaluated when there are different observers of the same client in different settings. The following description of a perpetrator’s behavior in peer group and in multi-family group on the same treatment day illustrates how this process works. A perpetrator who had been highly resistant to admitting abusing one of his daughters admitted to touching her breasts. He said to her in the multi-family group, “I’m really sorry I did this to you. I’m really sorry I touched your breasts. And I will do all in my power to make sure that it doesn’t happen again. I’m sorry I hurt you.” At the team’s final summary session of the day, the perpetrator’s peer group leader reported that the perpetrator had shown more change, being empathic yet gently confrontive to a man new to the group that night. When the new man began to minimize what he had done to his daughter, the perpetrator paid close attention and told the new man that his behavior may have hurt his daughter a great deal. Several team members expressed surprise at the changes the perpetrator was showing, and the team agreed that he was making some major breakthroughs. Different observers of the same client in various settings were able to contribute to a reliable and valid picture of the client’s progress through treatment. Different Observers of Members of Same Famiiies in Different Settings Reliability was also assessed when different team leaders observed different members of the same family. Sometimes clients made false reports of the status of the marriage or family life. For example, one man said he had been having great sex with his wife. When the leader of his peer group reported this at the team meeting, the leader of the mother’s group reported, “His wife said she hasn’t had sex with him for six months and she’s divorcing him.” The reliability of the husband’s statement about the couple’s sexual life was challenged by the wife’s statement. Again, having the statement of only one of the partners would have provided an incomplete picture. Another example of inconsistency of reports across different family members was the divergent viewpoint of a father and his sons. The father said that things were much better with his sons. When the father’s peer group leader reported this at a team meeting, a leader of the older siblings group indicated, “That’s funny. The sons said they’re ready to kill their father.” In such ways a more complete evaluation of family relationships was done. Often several team members made contributions to an emerging picture of a family,
Interdisciplinary
237
treatment teams
One example is the following: The team coordinator out of jail. The following team dialogue ensued:
announced
that a father was getting
Sibting Group Leader
I: Why? For what he did he should stay in jail for the rest of his life. He sexually abused his son, made him scratch himself until he bled, burned him with cigarettes. Coordinator: In muIti-famiIy group, you can’t get a piece of paper between them. In the last three groups, he sat on his father’s lap. Victim Group Leader: His daughter hates the father. If looks could kill, he’d be dead. SiMng Group Leader 2: Another son is spaced out. He could walk out in the middle of traffic, and he wouldn’t know what was going on.
Team leaders, then, commenting on the functioning of one family, were able to give a developing picture of family interaction and progress through treatment. Multiple Observers of the Family Members ip1Interaction
with Others
The structure of the treatment program allowed the entire family to be observed as they interacted with peers and with members of their own and others’ families. For perpetrators whose victims were not participating in the treatment program, their ability to take responsibility for their own behavior was assessed in multi-family group or family lab by their interactions with daughters of other perpetrators. One man told another man’s daughter: “I hope you are not blaming yourself for the abuse, I knew a lot more than my own daughter, and I took advantage of her. It is very important for you to understand that. I took advantage of her.” This clear statement of responsibility was an important factor in the team’s decision that this man was ready to graduate. This statement and others which were similar were reported to the treatment team. Over time, a picture of a man capable of dealing with incest emerged.
DISCUSSION Observation and open-ended interviews led to the finding that the incest treatment team followed procedures for establishing reliability and validity. The repeated nature of their observations of clients, observations characterized by variation in observer and setting, combined with multiple data sources, all are fundamental procedures used to establish validity and reliability of the findings of social research. The decision-making process observed for the purpose of this study would be similar to the decision-making processes used by other interdisciplinary teams who structure their time in the manner of the present team. Sharing and discussion of observations over time are crucial elements in the present team’s decision-making. The findings of the present research, then, can be hypothesized as generalizable in other teams whose decision-making processes are similar. Applying the concepts of reliability and validity to team decision making may help to solve a problem discussed above, namely, the lack of evaluative criteria on which to judge the quality of group decisions which have implications for real world problems. The present study suggests some evaluative criteria. This study provides a basis for concluding that decisions about complex phenomena, such as child abuse and neglect, require the knowledge and observations an interdisciplinary team can provide. A dimension of team decision-making which this study highlighted is the dimension of time. The establishment of reliability and validity rests on repeated observations from multiple observers over time.
Jane F. Gilgun
238
Money often is a barrier to the development of interdisciplinary teams. Administrators and boards of directors can be reluctant to set aside capital to fund a team. Yet, the team in the present research operated with no budget at all. The mental health center and the county agencies supporting the team provided release time for team members. The mental health center donated the space, which would have gone unused during the time the team operated. Of course, releasing employees for about seven hours per week represents a cost to each of the agencies, but it was a cost offset by the benefits of providing intensive treatment which would not have been available otherwise. Interdisciplinary treatment teams provide clear advantages over practitioners working alone. Unfortunately, however, many practitioners do operate alone. For individuals to expand the reliability and validity of their decision-making processes, they would need to gather information on client functioning from as many sources as possible. Peer supervision, case conferences, and the development of an informal group of consultants are ways the individual practitioner can improve the decision-making process. Consultants would include persons knowledgeable about the family or about child abuse and neglect, and they could include medical doctors, psychiatrists, psychologists, and social workers. Child abuse and neglect are complex phenomena. Decisions practitioners make about families where abuse and neglect occur can affect the course of the family members’ lives. Ethics require that these decisions be as reliable and valid as possible. Multiple perspectives of multiple observers over time are ways of ensuring ethical practice. Acknowledgment-The Central Minnesota
Incest
author thanks Dr. Rosalie A. Kane, the anonymous reviewers, and the team Treatment Program for their comments on earlier drafts of this report.
of the
NOTES AND REFERENCES 1. STEELE, B. F. Experience with an interdisciplinary concept. In: Child Abuse and Neg/ect: The Family and the Community, R. E. Helfer and C. H. Kempe (Eds.), pp. 163-168. Ballinger, Cambridge, MA (1976). 2. FALLER, K. C. (Ed.). Social Work with Abused and Neglected Children. Collier Macmillan, London (1981). 3. PICKETT, J. and MATION, A. The multi-disciplinary team in an urban setting: The special unit concept. In: The Abused Child in the Family and in the Community, C. H. Kempe, A. W. Franklin, and C. Cooper (Eds.), pp. 115-121. Pergamon, Oxford (1980). 4. HART, S. L. Toward quality criteria for collective judgments. Organizational Behavior and Decision Process 36:209-228 (1986). 5. The Central Minnesota Incest Treatment Program is based on a model developed by the Range Mental Health Center of Virginia, Minnesota. See. RANDALL, H., BELL, C. M. and CADY, K. Treuting Incest in Rurul Families. Pange Mental Health Center, Virginia, MN (1982). 6. BOGDAN, R. C. and BIKLEN, S. K. Qualitative Researchfor Education. Allyn and Bacon, Boston (1982). 7. KERLINGER, F. N. The Foundutions of Behuvioral Research. Holt, Rinehart and Winston, New York (1983). 8. COOK, T. D. and CAMPBELL, D. T. Quasi-Experimenfution: Design and Analysis for Field Settings. Houghton Mifflin, Boston (1979). 9. DENZIN, N. K. The Research Act. Basic Books, New York (1978). 10. GERTZ, C. The Inrerprefafion ofCultures. Basic Books, New York (1973).
Resume-Les Cquipes interdisciplinaires destinees a assumer le traitement des enfants victimes de s&ices ou de negligences se sont multiplies. Des etudes ont bien demontre que les decisions prises par des groupes dont les membres discutent entre eux selon differents point de vue sont meilleurs que les jugements port& par des individus isolts. Le pourquoi de cet avantage n’est pas evident. L’idee de la presente etude Ctait d’essayer de trouver par quel moyen une Cquipe interdisciplinaire arrive a ses decisions. Les auteurs ont done observe pendant 15 mois, une telle Cquipe chargee de traitement de cas d’incestes. On a interviewe des membres de Les auteurs ont trouve que l’equipe interdisciplinaire I’equipe separement selon la technique “open-ended.” therapeutique arrivait a ses decisions en utilisant des cheminements analogues ?t ceux dont on se set-t en recherche sociologique pour etablir la fiabilite et la validitt. Le processus decisionnaire Ctait fonde sur l’observation rep&e des membres de la families par de multiples observateurs dans des situations variees au tours du
Interdisciplinary
treatment teams
239
temps. Le processus decisionnaire ressemble a ceux qui sont utilises par beaucoup d’autres Cquipes therapeutiques. Les resultats de la presente etude sont probablement generalisables a I’activite d’autres Cquipes utilisant les memes processus decisionnaires. Resumen -El ndmero de equipos interdisciplinarios para el tratamiento de1 abuso y la negligencia de1 nitio esta creciendo. Las investigaciones han demonstrado que ias decisiones hechas por grupos que han tenido la oportunidad de intercambi~ perspectivas son mas exactas que 10sjuicios hechos solamente par individuos. La razon no es conocida. El objective de la presente investigaci6n fuf: el de describir 10s procedimientos que un equip0 interdisciplinario usa en llegar a decisiones. Un equipo interdisciplinario para el tratamiento de1 incest0 fue observado por un period0 de 15 meses. Tambien se llevaron a cabo entrevistas no estructuradas con miembros de1 equipo. Los resultados indican que el equip0 interdisciplinario de tratamiento llego a sus decisiones usando procedimientos analogos a 10s usados en las investigaciones de las ciencias sociales para establecer la formalidad y la validez. El proceso para llegar a decisiones de1 equip0 se caracteriz6 por observaciones multiples de 10s miembros de las familias llevadas a cabo por observadot-es multiples en locaciones multiples durante el curso del tiempo. Esta manera de llegar a decisiones es similar a la usada por muchos equipos de tratamiento. Los resultados de la presente investigation, por tanto, parecen ser aplicables a otros equipos cuyas formas de llegar a decisiones son similares.