6.2m 3w! .18.)...” . Ll 1.41.1.1 .J !. w. qslt» . ‘ V £13.“me A. V V , . 1. “a: .V ~< - ‘ ‘ . . V . r . JV. . 4 ll ICDJD ~ MICH HOANS TA ”$532th IIII IIII II III IIIIIIII III III University This is to certify that the dissertation entitled Companionship Programs and Nonprofessional Volunteers: Are They a Viable Supplement } to Professional Therapy for the Mentally Ill presented by Linda Springs has been accepted towards fulfillment of the requirements for Ph.D degree in Social Science Major professor Datewso; ’7 ’C MSU is an Affirmative Action/Equal Opportunity Institution 042771 PLACE II RETURN BOX to remove thle checkout from your record. TO AVOID FINES return on or bdore dete due. DAIEDUE DATE DUE DATE DUE BL: I |_I' Il L MSU le An Mmettve ActionEquel Opportunity lnetltulon L COMPANIONSHIP PROGRAMS AND NONPROFESSIONAL VOLUNTEERS: ARE THEY A VIABLE SUPPLEMENT TO PROFESSIONAL THERAPY FOR THE MENTALLY ILL BY Linda Springs A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Social Science 1996 ABSTRACT COMPANIONSHIP PROGRAMS AND NONPROFESSIONAL VOLUNTEERS: ARE THEY A VIABLE SUPPLEMENT TO PROFESSIONAL THERAPY FOR THE MENTALLY ILL BY Linda Springs The dissertation is both a case study of the Compeer companionship program for adults receiving professional treatment for a variety of mental illnesses and a cross— sectional comparison of three groups: Icurrent Compeer clients with volunteers, former Compeer clients no longer having volunteers, and individuals on the waiting list who have never had volunteers (the baseline). The investigation tested the effectiveness of nonprofessional volunteers that supplemented professional therapy with one-on-one friendships with clients . The subjects were referred to Compeer by their mental health professionals. Most were referred to improve their social skills. Each subject responded to a questionnaire and responses were analyzed quantitatively. The subjects' mental health professionals and Compeer volunteers were interviewed as informants. Their responses were analyzed qualitatively. The dependent variable was quality of life, measured by constructed mental health indicators and computed based on responses. Scores on mental health indicators were computed for the Community Activity Index with Self-care and Pleasant Activity subscales, a School and Therapy questionnaire, and Likert adjectivejpairing'scaleS‘which.called.for introspection about self-concept, concept of therapist and concept of others. Analysis of variance tested the effect of Compeer participation status on quality of life indicators. Two—way analysis of variance assessed the impact on quality of life of the independent variable, Compeer participation status, and.of seven additional independent variables in turn — employment status, education, self-supporting earnings, time spent with family and friends, psychiatric hospitalizations and gender. The quantitative results did not support the more positive qualitative assessments. However, there were significant 2-way interactions in regard to gender and Compeer participation status. Compeer was more successful with female participants than it was with its male participants. Copyright by LINDA SPRINGS 1996 DED I CAT I ON In loving memory of my father, Robert Springs, whose fortitude, boundless energy and daily prayers, gave me the inspiration to believe in myself, to work hard and to make ideas grow into realities. ACKNOWLEDGEMENTS Special thanks to my Social Work committee members Dr. Margaret Nielsen (my Chairperson) and Dr. Victor Whiteman for their patience and guidance during this long dissertation process. Also, thanks to my committee members, Drs. Harry Perlstadt, Mark Wilson and David Wiley for their valuable insights and suggestions. Many thanks to Betty Hayes, Michigan’s Compeer coordinator who, in addition to her busy schedule, gave so much extra time and effort to obtain subjects' consents for data collection, assist me in 'various interviews, offer suggestions On.hOW'tO approach certain clients and.her overall coordination between all of the subject and informant groups during my research. Thanks to Anne Meier, Colorado’s Compeer coordinator, for providing me with the means to gather the information.I needed for successful data collection. I want to especially express my sincere thanks and gratitude to my family. Thanks to my father and mother, Robert and Margaret Springs, who stood by me in this, and all of my endeavors. Sadly, my father’s death ten short months before my graduation prevented him from seeing the culmination of my most important endeavor, completing a Ph.D. vi r9? eov- a-~v‘v‘ ubk . u i | I . . rt: vb “it. . e . :_ 2. w .H .a C. v. .3 PC. ”a : v. ... Va .3 l. 2. no 3. 3-». .54 .. a. C’Nv- ‘4‘. I am so grateful for my daughter, Paula Rodriguez, and her husband, Robert, and for their unconditional love and support. Thanks for putting a roof over my head, for sustaining me with life’s necessities and for the invaluable computer assistance during this past year: .Also, thanks to my grandson, Joey, for an ample supply of most-welcomed hugs and kisses. Also, here’s to my daughter, Mari Brown. She was born during the Christmas break eleven years ago when I was in the Masters of Social Work program and she is now about to experience, for the first time, having a mother who is not a graduate student. Finally, my deep appreciation to my very good friend, Robert Blake, who was always there to support me whether it be in-person, by telephone or by letter. Thank you for your confidence in me. vii V. .; b TABLE OF CONTENTS Page List of Tables .............................................. x I. Introduction .......................................... 1 II. Literature Review .................................... 17 III. Methods .............................................. 43 A. Subjects ......................................... 43 B. Informants ....................................... 45 C. Instruments ...................................... 46 J” CommunityhActivity Index ..................... 47 2. School and.Therapy Questionnaire ............. 49 3. .Adjective Pairing Scales ..................... SO 4. Therapist and‘Volunteer Instruments .......... 50 D. Procedures ....................................... 51 IV. Results .............................................. 59 V. Discussion ........................................... 98 VI. Conclusions ......................................... 107 VII. Appendices A” Community Activity Index (Revised) .............. 118 B. School and Therapy Questionnaire ................ 124 C. Semantic Differential for Patients .............. 126 IL Client Consent Form ............................. 130 E. Therapist Questionnaire ......................... 131 F. Volunteer Questionnaire ......................... 138 G. Consent Form for Volunteers and Therapists ...... 142 viii Page H. Reliability Coefficients of Concept, Community Activity Index, Self-care and Pleasant Activity Scale Scores - Coefficient Alpha ................ 143 I. Guidelines for Incorporating Volunteers into a Companionship Program .......................... 144 J. Listing of Mental Health and Quality of Life Indicators With Most Positive Scores in Each Compeer Participation Status Group (P: <.05 Significance) ................................... 149 K. Personal Narrative .............................. 150 VIII. Bibliography ....................................... 161 ix ‘Is «ls fi‘ 7+ Table IO. 11. 12. 13. LIST OF TABLES Page Characteristics Within All Subject Groups ............ 61 Mental Health Treatments of Current, Former and Waiting List Groups .................................. 62 Summary of Community Activity Index, Self-care and Pleasant Activity Scale Scores by State by ANOVA ..... 64 ANOVA Summary of Community Activity Index, Self-care and Pleasant Activity Scores by Compeer Participation Status Group ......................................... 66 ANOVA Summary of Concept Scales by Compeer Participation Status Group ........................... 68 Community Activity Index by Gender and Compeer Participation Status in 2-Way ANOVA .................. 72 Self-care by Gender and Compeer Participation Status in 2—Way ANOVA ....................................... 74 Self-care by Compeer Participation Status and Time Spent with Friends in 2-Way ANOVA .................... 76 Pleasant Activity Scales by Compeer Participation Status and Psychiatric Hospitalizations During Previous 6 Months in 2-Way ANOVA ..................... 78 Concept of Therapist by Compeer Participation Status and Education in 2—Way ANOVA ........................ 80 Concept of Others by Compeer Participation Status and Educational Level in 2-Way ANOVA ..................... 82 Concept of Others by Compeer Participation Status and Time Spent with Friends During Previous 6 Months in 2-way ANOVA .......................................... 84 Concept of Others by Compeer Participation Status and Psychiatric Hospitalizations During the Previous 6 Months in 2-Way ANOVA .............................. 86 X Table Page 14. Types of Client Improvement since Compeer Participation as Perceived by Mental Health Professionals ........................................ 91 15. Reasons Why Volunteers Joined Compeer ................ 94 16. The Most Important Ways Compeer Friends Benefitted from their Friendships with Volunteers as Reported by Volunteers ........................................... 95 xi I . I NTRODUCT I ON This dissertation examines the usefulness and practicality of using "nonprofessional" volunteers as a supplement tx: professional mental health therapy; Mental illness is a major social and economic issue. According to the.American.Psychiatric.Association (1994), during aiqne-year period, up to 50 million Americans - more than 22% - suffer from a clearly diagnosable mental disorder involving a degree of incapacity that interferes with employment, attendance at school or daily life. The direct costs for support and medical treatment of Rental illness total $55.4 billion a year, excluding the costs for substance abuse disorders which is another $11.4 billion a year. The.American.Psychiatric.Association.further reports that The National Institute of Mental Illness has shown that one out of five adults suffer from a diagnosable and treatable mental disorder in any 6-month period. One reason that only one out of five of those suffering from mental illness seeks treatment is that most insurance plans, although they have mental illness coverage, do not pay the same for treatment of mental illness as they do for other medical conditions. For example, only seven percent of insurance plans pay for visits to a psychiatrist's office. :_ 1 . . , v ._ p .re v . . ’5‘ .> r, l E E H “A t . r. 1 ... t v. t n. C. 5 5 r1 v. a p. .. .r~ . .5 u D. C e S Y. c. a; .d . .aa ”A a: C a: C S a 1 a at t Pl. .3 a Q... «J ya Q. 8 .st «C A» Q» 2 Not only is there a lack of insurance protection for the treatment.of mental illness, mental health.agencies themselves are finding it more difficult to provide adequate services. Funding cutbacks, managed care treatment limitations and.staff downsizing have necessitated mental health agencies to do more with less. .Administrators in the social services need to stay abreast of ways to provide continued effective treatment for clients, not just take steps to cut costs. Orfield (1991) gave a vivid account of social service programs’ severe cutbacks that put workers in the system and those needing services in role conflicts, limited possibilities oerffective action and decreased standards and controls to substandard levels. Volunteers can stretch organizational budgets. They are especially valuable in human services, which is the most traditional place to volunteer (Abdennur, 1987). (Also see health and human services (Lotz, 1982); programs for the elderly (Diana, Lawrence and.Draine, 1985; McCroskyy Brown and Greene, 1983), and financial savings (Hawrylyshyn, 1978). Improving and managing the mental health care process plays a crucial role in the effective delivery of services. With the growing urgency to increase efficiency and control costs, it is imperative to optimize the use of volunteers. Social service systems One way agencies can effectively treat clients is through an interagency treatment plan using open systems theory. According to Davies (1977), "Systems theory applies to more ‘0‘ 5 3 than aspects of social work but of validity throughout society" (p. 82) . An open system is interdisciplinary and wide—ranging. It is conceived as a general science of ’wholeness’. Whatever the nature of the components or the relationship between the components, systems are interrelated. A social service professional rarely, if ever, meets a client who has a clearly defined problem which falls exactly within the professional's range of skills and the range of services provided by one particular social service agency. Clients usually have a cluster of problems which call for a wide range of services. Basic problems of scarcity of certain services can cause service gaps in which some human needs can be missed entirely. To prevent service scarcity, social workers utilize interagency cooperation and referrals. Agency interdependency and interaction helps ensure successful intervention for the client. Leiter and Webb (1983) refer to interagency groups that sometimes ignore agencies that are too unconventional or innovative even though those agencies are relevant to their clients' treatments. Ignoring the agencies cost interagency groups important sources of support because diversity in treatment approaches provides more options to individuals needing services. Bertalanffy (1973) contends that systems theory’ has relevance to the psycho-social context of the clients’ life. Parts of one system interact with other systems in an ever- unique way. Likewise, Raiff & Shore (1993) describe advanced 4 case management as both a practice and as a system of care to meet the needs of a diverse clientele. Advanced. case management provides "wrap around" service plans that use flexible dollar and strategies in order to be more responsive to client uniqueness. Meyer (1973) suggests that casework goals recognize the relationship between the client and the client’s environment. Social work cannot be viewed in isolation from the social conditions which foster the problems the social worker is attempting to treat. A systems approach allows the social worker to operate in whatever fashion is appropriate to the client’s person-in-situation needs. Linear causality over- looks the importance of interacting relationships and the complex pattern of causal relationships as a whole. There is not an accurate reflection of the whole situation. According to leiter and Webb (1983), community mental health agencies are using prevention techniques through "resource cycling". The community mental health movement seeks to decrease emotional disorders by making community agencies more responsive to human needs. "Resource cycling" considers the full range of human needs, ways that can meet these needs and potential consequences should these needs not be met. Mental illness and the quality of life Studies about mental illness and.the quality of life show the amount and quality of social networks are significant to life satisfaction, psychological well-being and social 5 function (Baker, Jodrey and Intagliata, 1992; Tantam, 1981; Greenblatt, Becerra and Seraetinides, 1982). In this investigation, quality of life means the sense of mastery and personal control over one's own life in both physical and mental adjustment in society as measured by objective and subjective mental health indicators. Objective mental health indicators are observable factors like independent living. Independent living emphasizes self-care such as driving a vehicle, using public transportation, or paying bills. Independent living also includes pleasant activities such as going to church, seeing a movie, or visiting friends. (Also see Bigelow, McFarland and Olson, 1991). Other objective indicators are frequencies of psychiatric hospitalizations or changes in the degree of therapy. Subjective mental health indicators measuring the quality of life include self-concept and concept of others. Mentally ill individuals frequently think of themselves as general failures in many facets of daily living like inadequate finances, poor relationships, or the lack of employment opportunities, indicating a generalized poor quality of life. The sense of mastery and personal control over one’s own life is critical for chronically mentally ill individuals. Without resources to positively develop and reinforce this control, up to 5095 of patients return to psychiatric hospitals within a year of their last hospital discharge (Rosenfield, 1992). A study of the importance of client support and "{ LU rn 6 representation emphasizes that although.mentally ill patients might be discharged from an institution because of reduced, or controlled, symptoms, stresses of everyday living without adequate resources or the knowledge about how to help themselves frequently leads to relapses (Freddolino, Moxley, and Fleishman, 1988). Increased physical and mental health is linked with supportive personal relationships (Taylor, Lam, Roppel and Barter, 1984). Informal and personal social networks create a nonthreatening environment for the mentally ill that a professional setting cannot offer. Volunteers in social networks appear' toI be 'more accepting of the situation, expectations more flexible, and relationships more reciprocal. They can readily act as role models and citizen advocates while providing a link between mentally ill individuals and the community. Treatment programs become more "humanized" (Mitchell, 1986). This type of environment is vital, especially during times of stress, and increases the chances for successful intervention (Grusky, Tierney, Manderscheid and Grusky, 1985). A criticism of professional mental health delivery is that it represents the providers' theories and self-interests over the needs and values of the clients (Szasz, 1961). In the past, therapists were more concerned with the patient’s clinical state and symptomatology than social issues (Platt, 1981; Platt, Weyman, Hirsch and Hewett, 1980). "Best interests" of a client as perceived by clinical and.diagnostic 7 judgements can become more important than the client’s own wishes and goals (Moxley and Freddolino, 1990). The research question was whether or not there are positive changes in mentally ill clients’ current quality of life due to their participation a companionship program. To investigate this question, the effectiveness of nonprofessional volunteers was analyzed to determine if the volunteers, using only one-on-one friendship with mentally ill clients, can be a viable supplement to professional mental health treatment. The impact of Compeer participation was tested noting significant differences and/or interactions among current client, former client and individuals on Compeer’s waiting list groups regarding the degree of objective and subjective quality of life indicators. Statistical significance was measured using an analysis of variance with an P: <.05 significance level. The nationwide Compeer companionship program was chosen by this researcher because of the differences between Compeer agencies (one is a non-profit agency and one is part of a larger mental health organization) and, although its volunteers follow basic guidelines, they are given a great deal of independence. The volunteers of Compeer provide vital social interaction for their Compeer "friends". Rook (1987) underscores that companionship buffers stresses of daily living and helps sustain emotional well-being. He separates the terms "social support" and "companionship". Social 1 e o .s l v. v. .5 a. 4w m: A s a Y. C ca .1 r. ... ... cc A» _ a. at r. a: a . C I .,,. T. 1 I a. S B. v . a... F. E at, no. 1 ME > C. S E. C. \h a C C e a p .1 e C c a -l .3 r S {No .3. mt. . A 8 support tries to alleviate problems and companionship is social interaction. Compeer is an international companionship program which serves the mentally ill. In 1973, Compeer was founded by Bernice Skirboll in Rochester, New York. Winner of the 1989 President’s Volunteer" Action. Award, the National Compeer Network is comprised.of over 120 programs in 35 states, Canada and.The Netherlands. Its staff is subject to all requirements of Compeer Headquarters based in Rochester, New York, which oversees quality control of all Compeer agencies. The word compeer means "a companion who is a peer or equal". Compeer volunteers, who are responsible adults from all walks of life, are matched in one-to-one friendships with persons in their community who face the stigma and social isolation of a mental disability. One-to-one relationships ensure a personalized friendship between the volunteer and the friend. Although Compeer has at least two social events annually for clients and volunteers, initially, dynamics involved in a "normal" group situation might hinder the development of the relationship and cause the friend to retreat. Compeer friends get together at least one hour a week and socialize as friends would normally do. For example, they go for walks or shopping, go to a movie, a restaurant, a sporting event, or sometimes just talk over coffee. Usually without prior mental health training or experience, each volunteer receives both training specific to 9 the friend’s disability and, ideally, ongoing support from the friend’s therapist as well as Compeer's professional staff. Frequently seen mental illnesses in clients include clinical depression, manic-depression, borderline personality disorders, early stage dementias, and phobias or other forms of extreme anxiety. Some clients come from chaotic backgrounds that have left them emotionally disturbed. However, schizophrenia is the most typical mental illness encountered in Compeer (see Table 1). Mental health.professionals refer clients to Compeer for a variety of reasons. The most common reasons, based on the 1987 Compeer Program Evaluation Survey for Rochester, New York, are to (1) provide a positive role model, (2) improve social and communication skills, (3) decrease isolation, (4) boost self-image and (5) experience a consistent and caring relationship with someone. Often therapists choose patients for the program who are most in need of a friend. Some chronically'mentally'ill need.extra help recovering from their illnesses but there are not enough volunteers to go around. A client is not referred if the mental health professional thinks the individual does not have the capacity to benefit from the program or if the client is judged to be dangerous to self or others. Two Compeer programs are used in this study. (hue is based in Colorado and one is in Michigan. Each of these states has only one active Compeer program. Compeer of Colorado, Incorporated is an.agency that offers its service to . .- wee-A F] -,J A“,— A V5,- lb 0 10 all local mental health agencies in the Denver area. It is a charter member of the National Compeer Network. The first client/volunteer match was made in 1983. Presently, there are 17 adult matches, 12 individuals on the 1992-1995 waiting list for volunteer matches (there were no former clients who had been returned to the waiting list) and one former client on file. Although services are available to both children and adults, this study includes only the adult population. Michigan’s Compeer is located in Grand Haven, Ottawa County, Michigan and it is one of the services offered by the Ottawa County Community Mental Health Agency. It is also a charter member of the National Compeer Network. It made its first client/volunteer match in June, 1982. In 1991, it was the first Compeer in the United States to also serve people who have developmental disabilities. On July 27, 1992, the prograniwas recognized, in-person, by President George Bush.as the 843rd Point of Light for outstanding volunteer service. There are currently 19 active adult matches. Sixteen individuals are on the waiting list (nine who have never been in Compeer and seven who were returned to the 1992 - 1995 waiting list after the loss of their volunteers), and 20 former clients from 1992 - 1995 listings. Today, mental health.professionals are supplemented more and more by volunteers to assist mentally ill clients and to relieve agency manpower shortages. The definition of a nonprofessional mental health volunteer in this study is a volunteer who acts as a friend and companion of the mentally . i I... . x w“ ... .3 .S .. r. . v ,. . y.. .H. .L rk. It A. . .... .A E» r» . . C r. . . z . o. a A c c. E . .A. a. v... M. c . .4 T. E .71 .d C u .5 . . C CI 5 C v. .5 r“ .C 15 C C S E S S .. I I LC. T. U... C f 11 ill person. Although aware of mental health issues, the condition of his/her "friend" and accessibility to the friend’s mental health professional or Compeer’s professional staff, the volunteer does not provide mental health treatment of any kind. Researchers like Karlsruher (1974) called for increased research efforts to study effects nonprofessional volunteers have on the mentally ill. Although Karlsruher observed that nonprofessionals demonstrate a definite and positive role in helping psychiatric patients, he felt more conclusive evidence between the professional and nonprofessional effects on the client was necessary. Shipley (1976) also realized the need for objective outcome measures rather than reports based solely on subjective evaluations of companionship programs. (For examples, see Skirboll and Pavelsky, 1984; Kovnat, 1990; Tulumello, 1990). After a two-year quantitative, objective study, Shipley’s results were not as glowing as results from subjective studies. Up to the time of Shipley’s research, for instance, one companionship program for the mentally ill received very positive responses on questionnaires completed by therapists, volunteers and.olients. Clients appeared.to be making great strides toward higher functioning capabilities. However, little statistical data was done to verify that outcome. Shipley’s statistical findings significantly differed from the qualitative study findings of the above literature. F"" Cu“ I...” e.‘ -v--\ 51..., PF" 5 F‘, \J 12 Mentally ill with companionship program volunteers were in Shipley's experimental group. They were compared to the control group, mentally ill without companionship program volunteers. In a two-year follow-up, he found the group that had companionship volunteers had a smaller decrease in the number of hospitalizations and more variability in improvement and in the number of days hospitalized than the group without companionship volunteers. This finding was significant because it questioned.the effectiveness of the volunteers that had. been found. to be effective in qualitative studies. Quantitative results based on statistics can challenge, or even disprove, qualitative results based on perception. Quantitative research relies on statistical analysis. After data is collected, the analysis is done and the statistical information reported” Qualitative research is more narrative and explanatory. The researcher describes characteristics and looks for patterns and themes of the findings. The present study was a methodological triangulation research, synthesizing both the quantitative and qualitative methodologies. The statistical information of quantitative methods and the explanatory information of qualitative methods strengthen the research (Patton, 1986). The research subjects were mentally ill persons categorized into three groups. The first group were current Compeer participants. The second group were former Compeer participants. The third group were referred to Compeer by mental health professionals and placed on the program’s 13 waiting list, but had not yet participated. The hypotheses were based on the literature about the benefits of ‘using' nonprofessional 'volunteers in informal social networks for the mentally ill. The literature suggested that when professional therapy is supplemented with a reciprocal friendship with a volunteer, the mentally ill client will progress both objectively and subjectively in his/her quality of life. (For examples see Anttinen, Jokinen and Ojansen, 1985, Riley, 1981; Oei and Tan, 1981; Goldberg, Evans and Cole, 1973; Gartner, 1966; and Ellsworth, 1968.) (1) Compared to mentally ill adult clients receiving only professional therapy and not in a companionship program, mentally ill adult clients currently receiving ongoing professional mental health therapy and participating in reciprocal friendships with nonprofessional companionship program volunteers will improve their degree of objective and subjective quality of life as measured by mental health indicators. Comparisons will be made by the statistical analysis of responses from client questionnaires and by independent measures and comparing measures as seen by each client's mental health professional and the client's volunteer. (2) Compared to mentally ill adults receiving only professional therapy and never having participated in a companionship program, former Compeer participants will display sustainability of positive (3) (4) 14 changes in their degrees of objective and subjective quality of life as measured by mental health indicators. Comparisons will be made by the statistical analysis of responses from client questionnaires and. by independent measures and comparing measures as seen by each client’s mental health professional. On the average, mental health professionals in this study will note changes in their clients’ degree of objective and subjective quality of life measured by mental health indicators when the clients are participating in a companionship program using nonprofessional volunteers. These changes are based on mental health professionals’ perceptions. Findings will be measured by each mental health professional's qualitative questionnaire responses and by the statistical analysis of responses from his/her client's questionnaire. On the average, mental health professionals in this study will note some sustainability in their clients’ changes in the degree of objective and subjective quality of life as measured by mental health indicators when the clients are former companionship program participants. These changes are based on mental health professionals' perceptions. Findings will be measured by each mental health professional's qualitative 15 questionnaire responses and by the statistical analysis of responses from his/her client’s questionnaire. (5) Companionship volunteers will consider their interactions with their friends helpful in the friends’ mental health progress. This will be measured by each volunteer’s qualitative questionnaire responses. (6) On the average, companionship volunteers will maintain positive contact with their friends’ mental health professional, or the Compeer professional staff, on an "as needed" basis throughout their companionship volunteer experience. This will be measured by each volunteer’s and each mental health professional's qualitative questionnaire responses and based on their perceptions. Dependent variables were measured by two types of indicators. One type was objective quality of life (e.g. employment, self-care, amount.of pleasant activities and.other items indicating community activity, the degree of independence, and the amount of psychiatric hospitalizations and/or' mental health. services). Subjective indicators of quality of life included self-concept and concept of others. The concept instruments were bipolar Likert scales. Answers were introspective and based solely on the subjects' perceptions. The independent variable was the extent of participation in the organized program with a companion yvf‘ t»- in; any blil. n;- I I»- (1' (I) u) (I) T'e i l6 volunteer, Compeer. Mental health. professionals completed questionnaires which focused on their attitudes, viewpoints, perceptions and experiences with Compeer and working with nonprofessional volunteers, as well as client improvement, stagnation or deterioration. All mental health professionals in this study have, or have had, at least one client in the Compeer program or on the waiting list. This questionnaire was analyzed qualitatively because of the open-ended perception questions. Compeer volunteers were also informants. Cmalitative methods were used to analyze their questionnaires. They were asked open—ended, perception questions about experiences as volunteers and about their interactions with their Compeer friends, with the referring mental health professionals with whom the friends received ongoing treatment, and with Compeer’s professional staff. .. . . . ldl A APV a» Y. ¢ » v C. A. r... a C . v- .u a: S. l C ray. Aev‘.. 2.. ’44 u a v I r». .d O II . LITERATURE REVIEW This is a study of the effectiveness of nonprofessional volunteers in the treatment of the mentally ill when they are used as a supplement to ongoing professional treatment. Social service agencies need volunteers for many reasons. For instance, rather than diminish services to the mentally ill or increase mental health professionals’ workloads because of agency downsizing, therapy efforts can.be enhanced.by using nonprofessional volunteers. Volunteers help agencies spread services over wider areas and reach a greater number of people. They also add a new dimension of "caring" to agency services and are an important link to the community (Brudney, 1990). Their involvement can be a salient approach for providing and enhancing services and helping organizations achieve policy goals. The literature reviewed focuses on mental health professionals and nonprofessional volunteers, results of friendship interactions between the client and nonprofessional volunteer, mental illness and a profile of the volunteer and volunteerism. Mental health professionals and nonprofessional volunteers In their book, Jones and Herrick (1976) wrote about the emergence of social work and volunteerism from 1900 to 1941. Historically, social work in America began as volunteerism. Most social work volunteers provided relief for the poor while others, like Bertha Reynolds, brought about societal reforms. 17 18 During the early'part of the 1900’s, trained, disciplined and salaried social workers began to emerge. Social service organizations that employed social workers began to frown on other social service organizations that still used untrained volunteers to do tasks suitable only for professionals. In times oficrises volunteers were more readily accepted. During World War I, volunteers were widely accepted by the social work profession to help relieve manpower shortages. This acceptance was temporary and conditional on professional control. After the wary the need for citizen.involvement decreased substantially and many volunteers were dismissed or their worth disregarded. In New York settlement houses, volunteers became irregular and turnover rates were high. Some settle— ment houses had turnover rates of nearly 100% annuallyu Other settlement houses had no volunteers at all and relied on the residents to assist in the day-to-day operations. By the 1920's, volunteer participation in the social services was ambiguous. They were used for three reasons: 1) manpower shortages, 2) public opinion and political attitudes toward government assistance and 3) the job market itself. Shortly before the Great Depression of the 1930’s, family caseworkers were complaining about their ever-growing caseloads. Social service again began to look to volunteers for assistance. By 1931, many family relief agencies had a larger force of volunteers than paid staff (Johnson, 1933). Even so, at congressional hearings, social work leaders l9 denounced the overemphasis.of volunteer service. Instead, the leaders approved national planning and government intervention to provide relief to the unemployed. Jones and Herrick further wrote that the entry of the federal government into social welfare attracted many volunteers into public agencies. Professionals provided professional service and volunteers were used for ancillary tasks such as relief work and friendly visiting. As in the past, the main purposes of volunteers were to broaden the area of service where the volume of work was the greatest and to develop grass-root support for the proliferation of public programs for the needy. New Deal programs of the 1930’s wanted to use volunteers on social service boards and in direct service areas. However, the social workers did not encourage volunteers to participate in these ways, thus protecting their own professional status. The nature of the service was, instead, friendly visiting, reverting back to the earlier pattern for volunteer service. Volunteers also performed routine tasks, largely unimportant, serving the agency, not the client. In the late 1930's, the threat of war meant the threat of manpower for social service agencies. Again, volunteers would be sought but, again, social workers would hold positions of authority and volunteers would broaden the area of service and promote new welfare agencies. Mencher (1959) summed up the role of voluntary activity in the social services this wayu It is "...strongly connected 20 with the rise of government responsibility for social welfare and the growth of social work as a profession" (p. 291). As social work grew into an accredited profession, requiring a social work degree and certification or licensure, the resistance to use volunteers lessened. Volunteers are still recruited by agencies during times of crisis, manpower shortages and funding cuts. Changing trends in the U.S. political and economic environments with the Reagan Administration and managed care systems of the 1980's and 1990’s have led to drastic curtailments of the service network. Volunteers, both trained and untrained, are welcomed in many social service and mental health programs. Numerous volunteers are now entrusted with more responsibility and direct client contact such as in tertiary prevention programs. Tertiary' prevention programs are not really prevention programs but services attempting to reintegrate persons suffering from mental illness into the community. The preventive function is to reduce relapses (Leiter & Webb, 1983). Studies of volunteer effectiveness Durlak (1979) did a comprehensive comparison between the effectiveness of professional therapists and paraprofessional helpers of the mentally ill. After analyzing 42 studies, he concluded that paraprofessionals were at least as effective as professionals. In some cases, paraprofessionals had better results than professionals. He demonstrated being an effective helper was, for the most part, an intrinsic fI‘F'. —o y..-r" vs.‘ 21 phenomena no matter the professional status. Posner (1966) attributed volunteers’ effectiveness to their "naive enthusiasm" and "lack of professional stance". Glasser stated, " Companionship does not require professionals . . . .best performed by warm, interested, responsive volunteers" (1955, p. 9). After a meta-analysis, Gartner (1971) discovered that various researchers, each using different methods and sources and each interpreting his/her own results, combine to validate that nonprofessionals do contribute to mentally ill patients’ improvement. Holme and Maizels (1978) conducted a study on the involvement of volunteers by social workers in Great Britain. They found only 51% of the social workers in the study enlisted the help of volunteers. Volunteers were assigned to interactions with clients such as befriending, visiting, shopping and transporting. Professional agency employees were unable to undertake these activities because of other responsibilities. Volunteers often enjoy the latitude to place the needs of the clients before the needs of the organization. Of the social workers using volunteers, 55% of them saw noticeable benefits for their clients. Although 14% of the social workers using volunteers thought volunteers lacked skill and experience, the majority perceived no disadvantages in utilizing volunteers. Mentally ill can and do benefit from their contacts with nonprofessional volunteers, as well as with mental health -v’ R v-— v--v. . ‘ wwvh ‘5‘. v y‘fli'rr- .A-‘I..A. ‘lr‘h’n fie.» , f‘f‘p. V“. (I) C r. (I? 1 (D 22 professionals. Volunteers from the community are best suited for social activism, community involvement, "grassroots" for client assistance, social.networks and.advocacy (Riley, 1981). Goldberg, Evans and Cole (1973) conducted evaluations using adults in the community'as volunteers to be a supportive network for mentally ill patients. They found more community involvement with the mental hospital and a decrease in the number of re—hospitalizations whenever patients were moved into community placements. These programs can be carried out in a hospital psychiatric ward, at the patient’s home or in a community setting. (Also see Froland, Bradsky, Olson and Stewart, 1979; Andrews, Tennant, Hewson and Vaillant, 1978; Kennedy, 1989; Henderson, 1980; Miller and Ingham, 1976). Further research points out that interpersonal environment is a consequence of psychiatric illness. Oei and Tan (1981) studied a companionship program by university students and their impact on inpatient chronic schizophrenic women” iFor seven weeks, untrained" but.psychologically aware, students visited one group of women once a week, the second group twice a week and a third group three times a week. Results showed only the group visited three times a week made significant sustainable functional and behavioral improvement . Companionship programs using college student volunteers interacting with mentally ill clients for short time periods each week illustrated positive client changes (Spoerl, 1968) . Anttinen, Jokinen and Ojansen (1985) described an integrative rehabilitation model for schizophrenics. The 23 rehabilitation program combined a therapeutic community environment, experiential learning, enhancement of the patients’ self-esteem and integration into a social network. The psychiatric care system and the support of friends and volunteers, both.college students and.adults in the community, played a prominent role in the patients’ progression. Tracking the program during its first 14 years, research results indicated the majority of schizophrenic patients could be rehabilitated.to live as:fairly'independent and responsible persons. Buckley, Muench and Sjoberg (1970) also found significant "general improvement in personality integration" in their research of companionship programs. However, as Davis, Dinitz and Passamanick (1972) demonstrated when their schizophrenic subjects’ support systems were removed, the patients had relapses. Clinical assessments had no predictive value. Chartier and Ainley (1979) observed 32 adult chronic psychotic state hospital residents of both sexes. Results suggested that chronic psychotics could acquire new behaviors through observation and demonstration of the behaviors. Copying these behaviors could be enhanced either by a previous positive relationship or, in the absence of prior interaction, by sufficiently strong incentives to reproduce the modeled responses. Unfortunately, models can be positive or negative and can cause adaptive or maladaptive behaviors. u-..» '— ,..,..- ”I Q-v-V ’d ”27.. ~— Ave.- .— e-b'OO-o Ow... Var” 6"“! we. I' AA” Uvr. U) ’PJ . ( f m r? V“ A.‘ ,1)! I" x‘ 24 Mental illness Subjects in the present study represent a number of mental illnesses. Therefore, there are many variations of mental illness with which companionship volunteers come in contact. It is important to understand some of the dynamics of mental illness to better grasp volunteers’ challenges when they assume their friendship roles. We are all subject to illness, both.physical and mental. Mental illness can occur through heredity, social factors, environmental conditions, or through a combination of these components. There are significant environmental and situational factors with which the mentally ill must deal. "...clients face very real environmental challenges, barriers and resource problems ...because of the discrimination, stigmatization, and lack of support suffered by many people who are labeled as mentally ill" (Moxley and Freddolino, 1990, p. 72). Genetic factors play a role in mental illness susceptibility, especially in the affective disorders, schizophrenia, anxiety disorders and dementias. Individuals with a family history of mental illness are more prone to develop it, two to three times higher than the general population (Papolos, 1988; Andreasen, 1984; Kiev, 1979; Snyder, 1974). A high genetic mental illness relationship was found in the studies of identical twins although they were raised apart. In schizophrenia, for example, if one twin developed U) h‘ :l. ”A ...v 51 l l) 7.; (All (T I D 7. ”L51: 'v “I” . ()1 IT) (n 25 schizophrenia, chances were very high (up to 50%) that the other twin would develop the disorder. The occurrence rate dropped to about 10% in fraternal twins and even lower among other relatives (Papolos, 1988; Andreasen, 1984). To test whether personalities and social behavior could be, in part, genetically programmed into one's brain from birth, children of criminal mothers were followed during a longitudinal study. Although adopted immediately after birth, these children had notably higher rates of antisocial behavior and criminal activity than children of law abiding birth mothers (Andreasen, 1984). Mental illness is often caused by organic conditions. Interconnected areas within the brain govern both bodily and mental activities. Neurotransmitters in the brain relay messages to the rest of the body. This flow of impulses must be steady both in the amount and timing or the brain cannot make the proper connections. For instance, Papolos (1988) wrote, " . . .neurotransmitters in the limbic-diencephalic system may play a critical role in the regulation of mood. A change in the neurotransmitter activity, through a deficiency or excess of norepinephrine or serotonin, is associated with depression or mania, respectively" (p. 66). Papolos (1988) pointed.out that hormone secretion is also influenced by neurotransmitters and can be a biological trait marker in various affective disorders. For example, the suprachiasmatic nucleus, which. is localized in the hypothalamus, stimulates the pineal gland to transform run» I ‘v‘ r~~ ..-U. v »\\ LL V S C. ‘d Wu. 5. S ‘3. Q» «I\ .l 26 serotonin into the melatonin hormone. He noted that patients with affective disorders have disturbances in their normal nighttime increase of the melatonin hormone. The hormone was absent in three out of four depressed.patients and in patients with bipolar disorder, its rhythm was desynchronized. Dewan and Spaulding (1958) wrote a book about organic psychoses to guide medical doctors in making diagnoses because symptoms similar to mental illness often are organically based” For instance, people with endocrine disorders, such as thyroid or adrenal gland diseases, often experience mood changes. Dewan and Spaulding asserted mental illness can be caused by factors like metabolic disorders, disordered blood supply of cerebral cells, obstructions or other stresses interfering' ‘with. cerebral cell function, infections, intoxications (both exogenous and endogenous), altered functioning of the brain tissue and degenerations of cerebral tissue. Andreasen (1984) noted the movement away from traditional psychotherapy and into the "mainstream biological traditions of medicine". Neuroscientific breakthroughs continue to increase understanding' how the Ibrain functions and malfunctions. "Medical science is now more convinced that the serious forms of mental illness, such as schizophrenia and severe depression, are due mainly to abnormalities in brain structure and chemistry rather than to emotional traumas in early development or crises in later life. Further, these illnesses are best treated by medical means. . . " (book jacket) . 27 In their correlational analysis of predictors of premorbid adjustment in 152 psychiatric patients, Flics and Herron (1991) suggested the strongest relationship of all demographic variables were gender and premorbid adjustment. For instance, they found females had a higher premorbid adjustment and increased premorbid competence. Females also had a better prognosis than males because they were more social, more help-seeking and had a greater ability for intimacy and verbal expression. Males, on the other hand, suffered more debilitating illnesses, like schizophrenia, at earlier ages than females“ IMales were more withdrawn and less inclined to seek help. The tendency to isolate is the nature of schizophrenia. Torrey (1983) reported that studies of schizophrenics living in the community show 25 % are described as very isolated, 50% as moderately isolated and only 25 % as leading active social lives. .Almost half have no recreational activities other than watching television. Torrey (1983) pointed out that an analysis of a group of 17 and 18 year old.individuals with schizophrenia would reveal there are four or five males for every female. Schizophrenia is also a more serious disease in men than it is in women. "Men do not respond as well to antipsychotic drugs, they require higher doses of the drugs, they have a higher relapse rate, and their long-term adjustment...is not nearly as good as women’s" (p.83). The majority of companionship volunteers in the present study were female and the majority of mentally ill with companionship volunteers were female. U AI .14 -4; I—v c.‘ "‘ at .C 28 Genetic factors only partly account for mental illness. Social and environmental factors also play a major role. Precipitating circumstances can affect one’s mental health. For instance, researchers discovered factors like physical unattractiveness could lead to a breakdown in mental health and adjustment. "Unattractive individuals are rejected in work, dating, and marriage. People forget them soon after meeting them, attributewmore evil characteristics to them, and are less likely to be helpful to them. Less attractive individuals also have less influence on other people, and they are likely to receive worse treatment even in a court of law" (Farina, Austad, Burns, Bugglin and Fischer, 1986, p. 139). The world for "ugly" individuals is often difficult, lonely, stressful and depressing, placing them at a higher risk for emotional problems (Farina, Fischer, Sherman, Smith, Groh and Mermin, 1977; Fischer, Farina, Council, Pitts, Eastman and Millard, 1982). Napoleon, Chassin and Young (1980) concurred with these findings. They compared how psychiatric patients looked at the time of the study and how they looked in a yearbook picture before the onset of the illness. All were substantially less attractive than their peers. After becoming ill, physical attractiveness decreased even more. Schramski, Beutler, Launer and Arizmendi (1984) noted that socioeconomic class was a potent predictor of sustainability of therapy outcomes. Combined effects of low socioeconomic status and negative life events caused clients to either deteriorate or unable to progress. Silberfield (1978) wrote 29 that low social support was a characteristic of many lower— social-class environments, particularly ixl'urban settings. Persistent mental illness and factors such as social class, ethnicity, stress, marginality and distorted communication patterns could be "medicated" by the quantity and quality of social bonding (Hammer, Makiesky-Barrow and Gutwirth, 1978). Close family ties could also discourage personal adjustment for the mentally ill. Clausen and Huffine (1975) suggested that close family ties could isolate or overprotect the individual, discourage independent living' and. hinder personal adjustment. lkianother study, schizophrenic subjects reported few close ties with a social network but many were heavily dominated by family ties. Patients in the most intrusive and conflictual family environments were at the greatest risk of relapse (Tolsdorf, 1976). After researching the labeling theory, Warner, Taylor, Powers and Hyman (1989), believed that mentally ill individuals who accepted a mental disorder diagnosis assumed they lacked.mastery over their lives and did not have positive treatment outcomes. They tended to lose self-control, became unable to trust their own judgment, became indecisive and ultimately chose to adopt a label of mental disorder to avoid responsibility for their actions (Chamberlin, 1978; Ludwig, 1971). Earlier, Ellsworth (1968) also theorized that labeling assumptions implied that when patients were not accountable for behaviors associated with particular mental illnesses; it became accepted and expected by all parties involved. This 30 created expectancy for enduring behavior patterns and the patient was regarded as a passive subject and a recipient of treatment from others. Ellsworth found that untrained volunteers reacted with more spontaneous and "normal" responses to these behaviors than professionals, helping the patient identify thoughts and feelings as his/her own. This taught skills in the differences between "self" and "others". Good.social support networks that provide empowerment and a sense of mastery lead to increased self-control and a more positive outcome in psychosis treatment. Validation plays a key role in empowering the chronically mentally ill (Tobias, 1990). Likewise, social skills training leads toward a sense of empowerment which, according’ to iBenton and. Schroeder (1990), appears sustainable. Kiev (1979) supported the importance of empowerment. He pointed out that past conditioning contributes to the way a person functions. For instance, if a child lacks loving reassurances, feelings of hopelessness and self-blame may develop into a self-defeating cycle that continues into adulthood» Rejection, not approval, is the expected response in any given situation. Kiev found that when chronically depressed people responded to frustration they tended to (1) continually seek approval and support, constantly testing the responsiveness of others, (2) lean on others to the point where others are forced to reject them, (3) be afraid to do independently what would give them a positive sense of self because of the excessive need for others’ approval. Hence, 31 the need for mutual friendships and support systems is very important to help break this self—defeating cycle. Other factors affecting mental health are one’s level of social awareness (Boise, 1983), social skill deficits (Fingeret, Monti and Paxson, 1983; Erickson, Beiser, Iacono, Fleming and Lin, 1989; Luborsky, Mintz and Christoph, 1979; Monti, Curran, Cooriveau and DeLancy, 1980; Morell, Levine and Perkins, 1982; Sullivan, Marder, Liberman, Donahoe and Mintz, 1990), posttraumatic stress disorder (Keane and Wolfe, 1990; Ramchandani, 1989; Robins, 1990; Watson, Kucala, Juba, Manifold and Anderson, 1991); premorbid maladjustment (Flic and.Herron, 1991; Glick and Zigler, 1986; Zigler and Phillips, 1962; Platt, Weyman and Hirsch, 1978), depression (Coyne, 1976; Johnson, 1991), and loneliness (Sullivan and Poertner, 1989; Tessler, Bernstein, Rosen and Goldman, 1982). Profile of the nonprofessional volunteer and volunteerism Nonprofessional volunteers are found in organizations that are :religious, educational, political, governmental, professional, medical and social service in nature within communities across the United States and worldwide. Between 1965 and 1975, active volunteerism increased nearly 60%. By 1981, there were approximately 37 million volunteers in the United States representing a broad cross— section of society. The majority of volunteers are middle- class females from urban areas who hold white collar jobs, have a higher than average educational level, between 30 and 40 years old and married (Abdennur, 1987). . . ,. .. . .2 c S E C. a men I . . J V». T. a I a, a at I“,- 32 Abdennur investigated. motives behind social service volunteers. Through review of the literature about volunteers, although it was quite conflicting, he was able to gather dominant generalizations that most volunteers, especially those in social services areas, exhibited conflict avoidance orientations and behaviors. Through his own research, he analyzed the psychological, social and political aspects of volunteers. His questionnaires, all established in reliability and validity by previous researchers, were designed to assess preferences and attitudes at psychodynamic, perceptual, cognitive and social—political levels. He found support for his theory. Social service volunteer responses clustered around low- conflict types of interests compared to responses of nonvolunteers. Abdennur asserted that ..."although all conflict involves the experiencing of psychic tension, individuals vary in their tolerance or endurance of such tension" (p. 9). He concluded that social service volunteers responded to conflict in our society by doing service to those on the "losing" side (e.g. poor, mentally ill, etc.). (Also see Pearce, 1983; Bradner, 1993). "Participation gives you the feeling you are doing something about something. . . " (Glasser, 1955, p. 15). Abdennur’s profile of the social service volunteer is: Volunteers generally appear to be well socialized individuals who view themselves as sensitive to other people, and as sympathetic, compassionate, nurturant, and benevolent. They appear to be rather conservative in their social and political views, and tend to accept 33 with little analytical thought or criticism the conventional and established views of their community. They also appear to be unusually flexible in their attitudes, tending to be tolerant of other people’s views. Their attitudes appear to be significantly influenced by the values they are exposed to in their volunteer work. Their ideological positions seem to be neither strong nor clearly thought out (p. 41). Sociological factors Ix) longer constitute au1 adequate explanation for volunteering. In the past, the most common reasons for volunteerism were thought to be the tradition of mutual.helpfulness, increasing leisure time, theidisappearance of the self-sufficient and self-contained family, the need to belong, to serve, to gain special knowledge or put one's own talents to work, and for recognition in the community (Glasser, 1955; Aves, 1969). Many volunteers have become an integral part of agencies; their "life-blood" . There are opportunities for volunteers of widely differing skills and abilities. According to the 1985 International City Management Association (ICMA) survey, the estimation of volunteers used in at least one service domain in cities with over 4,500 population was 72.6 percent (Duncombe, 1985). Literature suggests that persons with positive attitudes toward a particular organization are led by those feelings to volunteer there. Volunteers work for rewards of social interaction and service to others and their work is more praiseworthy. If they are satisfied with their functions, they are less likely to leave their organizations (Pearce, 1993; Smith and Freedman, 1972; Mulford, Klonglan, Beal and 34 Bohler, 1968; Barker, 1968). Individuals who have a strong personal interest in achieving the organization’s goals or see the organization.as the only likely vehicle for their personal goal attainments make good volunteers and will likely stay with the organization (Pearce, 1983). Knoke and Prensky (1982) wrote about threats to volunteerism. Volunteers may be strongly committed to the goals of their organizations but have weak ties to that organization. Building organizational commitment is of serious practical importance. Etzioni (1975) perceived that employee "calculated" involvement as, "..Ji partisan, affective role in relation to goals and values, and to the organization for its own sake, apart from its purely instrumental worth" (p. 533). In Pearce’s opinion, "Volunteers’ attitudes are, in general, more positive than comparable employees’ attitudes" (1993, p. 92). Volunteers usually saw themselves as friendly, flexible and spontaneous. When asked to compare themselves with social workers they indicated social workers were rigid, inhumane, close-minded, apathetic and ’official' in their attitudes. Social work was simply a job. One-fifth of the volunteers Aves (1969) surveyed said they had little or no contact with professional social workers because the social workers were inaccessible to give advice or guidance. The volunteers who had more contact with social workers, however, took a name positive viewpoint of them. Aves suggests that the struggle for recognition of social work as a profession might lead 35 social workers to deny nonprofessionals can be effective. Staff members may sometimes be reluctant to relinquish part of their jobs but volunteers can free the professional from many tasks to make fuller use of his/her expertise (Glasser, 1955). Professionals are often hindered by office confines and professional boundaries when treating their patients (Arthur, 1978). IDual relationships, professional and. friendship, between therapist and client are forbidden by professional ethics requirements. The practitioner’s influence and the client’s vulnerability carries over into the friendship and is detrimental to the client (Kaygle and Giebelhauser, 1986; Argyle and Henderson, 1984; Schultz, 1991; Wiseman, 1986). Friendships encourage openness, loyalty, comfort, trust, confidentiality, support and.psychological growth, similar to a therapeutic relationship. However, friendships differ from a therapeutic relationship because friendships are between peers and are voluntary and reciprocal for both parties. However, Aves (1969) asserts, "...volunteers should not be regarded as substitutes for professional workers" (p. 86). She further states that functions of decision making, report writing and social control activities are reserved for paid employees. However, generalized client support is not distinct between professional and volunteer but between different skill levels and abilities which are derived from learning and experience. (Also see Davies, 1977.) Volunteers can react strongly to their "unpaid" status. They take pride in its symbolism of sacrifice and service and 36 resent suggestions they might be "unprofessional" or their labor is worth nothing (Pearce, 1993). Brudney (1990) warned that when volunteers are used as "tokens", it can lead to serious deficiencies in volunteer morale, reliability and retention, ultimately jeopardizing the effectiveness of the volunteers and the working relationship between the professional and the volunteer. A cohesive work group can operate a potentially powerful control system for volunteers but it must be normative, or value-based, control to be effective (Shaw, 1976; Pearce, 1993). The greater the feelings of importance to the organization and greater social involvement with other organizational members lead to a higher volunteer organizational commitment (Mowday, Porter and Steers, 1982). Volunteers are able partners with professionals in their productivity assessment of volunteer programs in not-for— profit human service organizations (Gamm and Kassab, 1983). Conversely, Young (1987) reasoned the impact of volunteers on an organization could be quite negative. Their presence emphasized the importance of service motive, making performance incentive for staff more difficult. Volunteers promoted "patronage awarding" of paid positions among themselves rather than hiring based on merit. The "clubbiness" atmosphere detracted from professional service to clients and, because volunteers were not employees, they could bypass supervisors and go directly to board members with their complaints. 37 Brudney (1990) believed that the most enduring obstacle to the implementation and operation of a productive volunteer program was the often antagonistic reaction from employees. He admitted that volunteers were largely'unreliable, balked.at paperwork and resisted supervision. However, agencies, especially those with high financial constraints, could hardly turn down citizens who wanted to help out, regardless of their qualifications. Without the leverage of a paycheck, organizations had no quality control over their volunteers nor could hold them accountable for performance. Therefore, when a mental health professional uses a volunteer to supplement and/or enhance client treatment, there is a certain amount of risk-taking. The professional is taking the chance the volunteer is suited for the task and will indeed help, not hinder, the client. The client must have trust and confidence in the therapist to set the necessary foundation for therapeutic benefit in the helping relationship (Reamer, 1982). Aves (1969) observed difficulties between the volunteer and the client in her study. If volunteers were in a companionship program, sometimes their mentally ill "friends" were rude, disagreeable or took too much for granted. There were often personality conflicts. Additionally, some volun- teers found after the "friendfl had been.improving functionally and/or mentally, they felt frustrated and helpless watching periodic regression. Schilling (1987) wrote about the limitations of social 38 support and the potential of harming the client. Put in a situation that is incongruous with his/her own coping skills or expectations, psychological.disturbances can.resultu .Also, if the client perceives rejection or betrayal in the volunteer, mental health crisis can occur. This rejection perception can be very real in a companionship program when volunteers fail to fulfill commitments or leave the program. Lessons learned from projects using volunteers Davies (1977) reports on a three-year project in England in which volunteers provided support to help facilitate the coping skills of selected children with learning disabilities from special schools and their families. Volunteers extended support on a regular basis. They befriended the children and their parents, offered them guidance and helped them through times of crisis. Duties given to the volunteers in this project were largely without adequate training or resources. Efforts were doomed to fail through resentment and frustration on both the part of the client and volunteer. Most volunteers who came "under fire" from families had over-involved themselves. They did too much in the house, took the children out too frequently and gave too many presents. Overall, most families felt volunteers overstepped boundaries, becoming intrusive and interfering into their private lives and going beyond limits of privacy and independence. Friendship was not reciprocal. Volunteers were more like the classical "friendly visitor". The best volunteers first sought to establish a 39 relationship. If a need arose, they offered advice and/or material aid but reminded the family that they were not under obligation to accept it. This type of volunteer was a leading partner but not a dominant one. Wolf (1985) described what can happen when some organizations treat nonprofessional volunteers as employees, paying them a stipend and assigning duties similar to employees’ duties. The volunteers were txn'visit selected neighbors on a routine, scheduled basis, assisting them as needed. Recordkeeping, written reports of the visits and recommending various professional services were part of the work. Volunteers were encouraged to act like professionals and before long their neighbors became as clients. The spirit and effectiveness of volunteerism was lost when a professional boundary formed between the volunteers and neighbors. Compeer recruitment and traininq To encourage and maintain volunteers, Brudney (1990) pointed out there must be adequate funding to recruit, screen, orientate, train, provide materials, facilities, publicity, recognition and feedback. Compeer, the companionship program in this study, does all of these things to make sure their mentally ill clients have the best volunteers possible. Primarily, the Compeer volunteer is to be a friend. The volunteer is not to be a social worker, parent, taxi cab service, probation officer or rehabilitator. It is more than enough to be a friend, role model and advocate. The following information about Compeer volunteer 4O recruitment and how the selected volunteers are trained to be effective "friends" to Compeer clients was obtained from Compeer’s training handbooks and brochures and from personal interviews with one former and two current Compeer directors from Colorado and Michigan. Compeer volunteers are actively recruited in the spring and fall. This study found recruitment by word of mouth (21%), own research, unspecified (12%), newspapers (27%), local churches (21%), other programs (6%), flyers (6%) and a Compeer booth at a fair (3%). The potential volunteer typically responds to an advertisement by calling the Compeer office and is sent an information packet containing the volunteer job description and an application. After reviewing the completed question- naire, the Compeer coordinator schedules an interviewrwith the applicant. During the interview, the applicant’s background, interests, geographic location, etc. are discussed. Strengths and weaknesses are noted. The most common elements for matching are geographical location and mutual interests. According to the current Compeer of Colorado’s coordinator, Anne Meier, factors such as age, severity of the mental illness and incapacitation do not seem to be significant concerns in the matching process. The coordinator then meets with the mental health professional of the Compeer waiting list client who might make the best match for the applicant and produce the most productive, as well as compatible, relationship. If the mental health professional agrees, a 41 meeting is scheduled between the professional and the applicant. The applicant is educated about the client and the mental illness involved. Upon the professional’s approval, the Compeer coordinator releases the client’s name, address and telephone number to the new Compeer volunteer. The volunteer makes the initial contact with the new friend and the friendship begins. Consistency is important because the mentally ill often deal with rejection issues. Having someone they can trust to positively interact with them regularly is a very important factor in the healing process. After the first few years of operation, a study of Compeer showed about 60% of previously hospitalized mental patients were readmitted to hospitals. Among Compeer clients the number dropped to 15% (Kovnat, 1988). In Michigan’s Compeer, each volunteer receives training specific to the friend’s disability. Mental health professionals, volunteers and clients also attend in-service training sessions quarterly. :n1 Colorado’s Compeer, volunteers receive approximately five hours of initial group training. .Although methods differ, volunteers are taught how to meet their friends, realistic expectations and limitations of the relationship, communicating, handling silences, effectively handling inappropriate behavior, what to do should a crisis arise, advocacy; changes that could.4affect the relationship and how to end the relationship. If the friend is hospitalized due to the mental illness, Compeer asks the 42 friend’s mental health. professional, "What role can the volunteer play? " Volunteers also receive professional Compeer staff support regarding their friends, when needed. Finally, volunteers are taught about psychotropic medications, monthly reporting responsibilities to the agency and are given guidelines to assist them in various situations they might encounter. Although Compeer’s coordinators are available as consultants throughout the course of the volunteer/client relationshipIand.matches are made through.a rigorous screening process, some relationships do not flourish. Unexpected changes in life can disrupt the friendship process. Volunteers move away or no longer have the time to devote to their friendships. The mental illness may be more severe than the volunteer expected and beyond the volunteer’s confidence, or comfort, level. The client’s social skills may be too borderline, or inappropriate, for a nonprofessional volunteer offering only friendship and the friendship cannot develop. Feelings of hopelessness, abandonment and/or rejection are often prominent after failed relationships, especially for the mentally ill person who has had to deal with these feelings many times before. There can be a setback in the relationship between the therapist and client, especially regarding trust, since it was the therapist who referred the individual to the companionship program. Therefore, Compeer makes every effort that the best possible match between the volunteer and the client be made each and every time. I I I . METHODS The present study is both a case study of the Compeer companionship program and a cross—sectional comparison of three groups: current companionship program clients with volunteers, companionship program clients Ix: longer having volunteers and individuals on the companionship program waiting list who have never had volunteers. Subjects The subjects in this study exploring a volunteer companionship program, Compeer, include three groups of chronically mentally ill individuals from the Denver, Colorado and the Ottawa County, Michigan areas. The groups were selected to investigate the quality of life of mentally ill individuals before, during and after participating in Compeer. Group 1 were current Compeer participants to determine program effects. ' Group 2 were former Compeer participants to determine sustainability of positive program effects after termination. Group 3 were on Compeer’s waiting list and were the baseline group since they had not yet participated in Compeer but had been referred by their mental health professionals to do so. All subjects were referrals from local mental health professionals working in agencies like the Community Mental Health Institute of Denver and in Michigan’s Ottawa County Community Mental Health Agency. Each subject was individually contacted by Compeer (in Michigan), or by this researcher (in Colorado), and given the 43 44 opportunity to participate in this study. All consent forms contained a brief summary of the study’s objectives (see Appendix D). The three groups had the following similarities and differences: Similarities between groups: 1. Professionally diagnosed mental illness 2. Receiving ongoing professional mental health therapy 3 . Referred to Compeer by the mental health professional because of the capacity to benefit from the program 4. Referred.to Compeer from local multiple mental health agencies (Colorado) or Ottawa County Community Mental Health (Michigan) 5. Desires to be in the Compeer program 6. At least 18 years old 7. Nonviolent 8. Not receiving Compeer Calling 9. Not participating in any other companionship program Differences between groups Group 1: Mentally ill individuals currently in Compeer. Group 2: Former Compeer participants Group 3: 1. Mentally ill on Compeer's waiting list for a match 2. Not a former companionship program participant 45 Individuals receiving assistance through Compeer Calling were excluded from this study because interim companionship contact was being provided while the client/volunteer match was pending. No referred adult participant was excluded because of diagnosis, gender, employment, educational level, marital status, premorbid adjustment factors or length.of time in Compeer. However, interactive effects which might be produced by these variables were included in the analysis to determine patterns or themes in score differences. Informants Volunteers and referring mental health professionals were asked to evaluate program effectiveness qualitatively through open-ended questionnaires. This helped explain the subjects’ quantitative results. Therefore, they were classified as informants and the mentally ill participants were the subjects. Referring nental health professionals represented numerous mental health agencies and varied mental health career fields with their own philosophies, such as social workers and. psychologists. 'They' were asked. demographic questions, their opinions about.working”with volunteers, about Compeer and about any changes they'perceived in their clients. Compeer volunteers also completed questionnaires. Some of the volunteers were mental health professionals or in professional positions other than. mental health” Other volunteers were "ordinary" citizens who wished to help make someone’s life better and some were college students. None of 46 the volunteers were former Compeer participants. Volunteers were asked demographic questions, their opinions about Compeer, about Compeer’s staff and other mental health professionals with whom they worked as a Compeer volunteer and their impressions regarding any changes in their friends. Instruments Each. subject’s background. was ‘unique with the many variables that occur in social, physical and environmental contexts. There were also differences in the degree and types of mental illnesses and variabilities in the criteria and techniques therapists used to treat the illnesses. Hence, measures of global client functioning were used rather than tests that pinpointed a targeted type of mental illness. All three subject groups completed identical closed-ended self-report questionnaires and concept scales. Scale scores were analyzed by analysis of variance to determine if there were statistically significant differences at <.05 among the groups that could be attributable to the intervention of the independent variable, Compeer participation. There were three separate sections on the client questionnaire. Section 1 was the standardized and revised Community Activity Index (Appendix A) . Section 2 was a 6-item school and therapy questionnaire (Appendix B). Section 3 was a set of three adjective pairing Likert scales (Appendix C). Reliability of coefficient tests were dependable using coefficient alpha equal to .82 or greater (Appendix H). 47 Community Activity Index The original Community Activity Index was developed by the New York State Office of Mental Health (Fabisiak, Becker and Earle, 1978). It has been used successfully by mental health agencies to assess client progress as measured by the consistency of scores within groups and client improvement consistencies reported by therapists and volunteers (Seig, 1980). Portions of the questionnaire were revised by this researcher to embrace modern lifestyles in today’s society. Subjects recalled activities over the past one week period. Some questions were specific to independent living, (e.g. ongoing activities such as club or organization membership, and information such as education, level of self- supporting earnings and living arrangements). Test validity was confirmation of client progress by the clients’ mental health professionals, by the clients’ companionship volunteers (if applicable) and by findings from other studies about companionship programs for the mentally ill. Self—care (SC) and pleasant activity (PA) items were identified on the Community Activity Index and constructed into two independent scales. The scales were used for cross- sectional comparisons between the three subject groups and between subject groups and the perceptions of their mental health professionals and, if applicable, their volunteers. The "SC" and "PA” identification.markers were not shown on the subjects' questionnaires (see Appendix A). 48 The items on each scale are: Self—care wNH Work in exchange for room and board Drive a car, motorcycle, truck Work in exchange for clothing, cigarettes, or other small compensation Use public transportation Prepare a meal for yourself of a friend Pay a bill by mail Budget money for the week Write a check or money order Cash a check Purchase or pay for something costing more than $35.00 Purchase or pay for something costing $5.00 to $35.00 Purchase or pay for something under $5.00 Have a major responsibility for the physical well being and appearance of children, elderly or sick persons Launder or iron clothing Prepare a meal for a dependent or spouse Plan meals Purchase groceries for a few days Vacuum, mop, sweep or dust at home Repair a car, appliance, etc. at home Paint, hang wallpaper, mow a lawn, shovel snow, other do maintenance work at home Go to a food store Go to the bank and deposit/withdraw money Go to the post office Go to a doctor, dentist, lawyer or other professional Have a visit from a doctor, dentist, lawyer, or other professional Pleasant activities \OQQGU’IQWNH Do formal volunteer work Attend a club meeting Write a letter Read a book Read a newspaper Sit and think Knit, crochet or sew something Work on a hobby Listen to the radio or stereo Watch television/VCR Play cards, pool or other games Go to a movie, concert or theater Go to church or synagogue Go to the library Go to a tavern Go to a party at someone else’s home 49 17. Take a pleasure walk 18. Play with children 19. Visit friends 20. Have a party in your home 21. Get together with friends to do something 22. Start a conversation 23. Help someone who needed help or directions 24. Play golf, tennis, bowling or softball, etc. 25. Take a vacation Overlapping self-care and pleasant activities fromflboth.scales Make a telephone call Talk to someone who called you on the phone Write a letter Mail a letter Go to a drug store Buy a meal in a restaurant Go to a beauty parlor or barber shop Go to a department or hardware store ooqmmtbwtoI—J The Community Activity Index also contained 11 items which Seig (1980) found.were not likely to effect the subjects regardless of their Compeer participation status (e.g. "Do you own real estate?). School and therapy questionnaire The 6-item school and therapy questionnaire determined psychiatric and medical hospitalizations, enrollment in a school or training program, the amount of time spent with family and friends, and other programs in which clients were involved besides Compeer (see Appendix B). This questionnaire supplied insight into the clients’ social lives. Test reliability was demonstrated in the test-retest of outcome measures used to determine various functioning levels of Compeer participants (Seig, 1980). Reliability of the coefficients were not tested by Seig. In the present study, reliability was tested by a coding accuracy verification. A 50 random sample of the questionnaires was recoded to ensure the codes from the sample were the same as the original codes. Adjective pairing scales Adjective pairing scales, the Semantic Differential (Osgood, Suci and Tannenbaum, 1957; Snider and Osgood, 1969), measured clients’ perceptions about themselves, about their therapists and about other people inmgeneral (see Appendix C). Subjects rated each concept on 16 bipolar adjective pairs. There a range of seven selections between the positive and negative adjectives from which to choose. The number one was determined the most positive answer for each item and seven was the negative answer. The range of possible scores was between 16 and 112 in which the lowest scores were most positive. Adjectives were listed in a mixed fashion so that one side of the scale would not represent all negative adjectives while the other side represented all positive adjectives. Concept scores were calculated as the sum of the scale scores for each of the three concept measures. Reliability of coefficients were dependable with coefficient alpha equal to .82 or greater (see Appendix H). Therapist and volunteer instruments Mental health. professionals’ questionnaires were qualitative and open-ended to better understand the professionals’ basic perceptions and points of View. They were asked about their experiences with, and opinions about Compeer and nonprofessional volunteers as an effective supplement towmental health.treatment and the perceived.effect 51 on clients (see Appendisz). Questions referred to areas such as criteria used.to determine which.mentally ill clients would benefit from.Compeer, how clients felt about the companionship experience, what types of evaluation processes were used when determining client mental health status and how improvement, or lack of improvement, was measured. A descriptive narrative was used to explore themes and compare subject response data with volunteer and mental health professionals’ responses. Procedures There was one procedural difference in subject data collection and between Compeer programs in Denver, Colorado and Ottawa County, Michigan. Besides geographical setting and community density differences, introductory subject contacts were made by this researcher in Colorado and by the Compeer coordinator in Michigan. The Compeer agency was a more loosely run non-profit agency in Colorado and under the auspices of the Ottawa County Community Mental Health agency in Michigan. Despite these variations, the summary of scale scores .between. the two states showed. the groups had. no significant differences (refer to Table 3). In Denver, Colorado, a list of the current and former Compeer clients and individuals on the waiting list was obtained from the Compeer office. This researcher contacted everyone on the list by telephone, gave them a brief summary of the study's objectives and asked them if they would be willing to participate. After receiving oral consents, this researcher set the appointment times and locations to meet 52 with the participants for questionnaire completion. Consent forms (see Appendix D) were signed by each subject prior to being given the questionnaire (see Appendices A, B and C). In. Ottawa County; Michigan, the Compeer* coordinator contacted each potential subject by telephone to obtain an oral consent. Written consent forms were then mailed to those agreeing to participate, signed and returned to Compeer. The listing of only those individuals who had returned signed consents to participate in this study was available to this researcher. All data collection was done by, or in the presence of, this researcher. Data were collected at sites selected by each subject. Alternative sites were: (1) at the subject’s mental health agency, (2) at the Compeer office, (3) in the subject’s home or (4) at a public place such as a restaurant. The most requested sites were at the Compeer office and at home. The quantitative, closed-ended questionnaire was presented to each subject as privately as possible. The degree of assistance required depended upon the severity of the mental illness, physical handicap, or level of literacy. If a subject could read and respond to the questions in writing, this researcher was available only to clarify questions. If a great deal of assistance was needed to complete the questionnaire, each question was read aloud and the answers written as the subject responded. Sometimes a subject requested the volunteer or the Compeer coordinator to 53 be present“ If so, that person would assist the subject. Before leaving the premises, all questionnaires were checked for completeness. This researcher contacted, by telephone, each mental health professional in the Denver, Colorado area who had one or more client(s) in the subject groups. A brief overview of the study’s objectives was given. If the professional agreed to participate, the written consent form (see Appendix G) and the therapist questionnaire (see Appendix E) were mailed. If necessary, telephone follow-up served as a reminder to complete the questionnaire. In Michigan, each Ottawa County Community Mental Health professional who had a client in one or more of the three subject groups was given a brief overview of this study and asked to participate by the Community Mental Health Program Director during a monthly staff meeting. Consent forms were signed and questionnaires were distributed during the meeting. The list of participants was obtained from the Compeer coordinator. In. Denver, Colorado, this researcher* contacted. each Compeer volunteer by telephone from the listing made available in the Compeer office. Volunteers were given a brief summary of the study’s objectives and asked to participate. After oral consents, consent forms and volunteer questionnaires were mailed. Telephone follow-up was used as needed. In. Ottawa. County; Michigan, the Compeer‘ coordinator contacted each volunteer by telephone. After oral consents, 54 the coordinator mailed consent forms to participating volunteers. When a signed consent form was returned to the Compeer office, this researcher mailed the volunteer a questionnaire. The study was a cross-sectional investigation of the three subject groups. An experimental study would have increased control over the introduction of the independent variable and the extrinsic and intrinsic variables through randomization and yielded more accurate results. However, experimental studies can be very expensive, time-consuming and can raise the concern of human subject ethics. Thus, this study compared.current Compeer clients, former Compeer clients and persons on Compeer's waiting list (eligible for service but have not yet received services). All subject groups had ongoing mental health professionals and current Compeer subjects also had their volunteer matches at the time of data collection” The subject groups' questionnaires were quantitatively analyzed. Informants’ questionnaires were qualitatively analyzed since open-ended perception questions were asked about the subjects. The triangulation. method. of incorporating' both the quantitative and qualitative methods helped depict a better picture of the study’s results. All statistical testing was by analysis of variance, using the F-test. The SPSS UNIQUE Analysis of Variance program, rather than standard analysis of variance, was used to statistically correct for unequal group sizes since the 55 subject population numbers were small. The small amounts of information in some categories would have resulted in several empty cells if computed using standard analysis of variance. Group differences can be caused by unequal group numbers, physical, mental and social factors, by environmental conditions and by varied mental health treatments. The F-test was used since three different subject groups were compared for statistical significance using the level/stage of Compeer participation status as the independent variable. Each of the three groups represented a different level of participation, from never participating in Compeer to being a current participant to being a former participant. Pre-determined quality of life mental health indicators were the dependent variables. Community Activity Index scores compared subjects’ functioning levels among the three groups (see Table 3). Scores included all self-care, pleasant activity and general items. The sum of "yes" answers were calculated. Possible scores ranged from 0-69. A score of zero indicated the subject did not answer "yes" to any item. A score of 69 indicated the subject answered "yes" to all items. The independent Self-care and Pleasant Activity scale scores each had a score range from 0-33. The Self-care and Pleasant Activity scores included eight overlapping items which appeared on both scales. The Community Activity Index also contained 11 general items that asked about the subjects’ current life situations. 56 These items remained part of the total Community Activity Index scores only. Michigan and Colorado subject groups were compared in community activities, self-care and pleasant activities for equivalence by analysis of variance according to the state in which they were living (see Table 3). The impact of Compeer participation status and seven independent variables on mental health indicators were assessed separately in a 2-way analysis of variance with P: <.05 significance. The second independent variables were: 1. Employment status 2. Educational levels 3. Levels of self-supporting earnings 4. Time spent with family over the previous six months 5. Time spent with friends over the previous six months 6. Psychiatric hospitalization over previous six months 7. Gender The variable "employment status" included (a) employed — full time (9.4%), part time (18.9%) and self-employed (1.9%), (b) ‘unemployed, - rm) job (11.3%) and. unemployed. disabled (47.2%), and (c) other - student (3.8%), homemaker (1.9%) and retired (5.7%). The variable "educational level" included (a) less than high school - grades 1-12 with.no graduation (28.4%), (b) high school -Igraduated (30.2%) and (c) college —.attended.college, technical or trade school post high school (41.4%). The variable "self—supporting earnings" included (a) a 57 job with pay - earned enough money to support self without other financial assistance (1.9%), earned some money but not enough to support self without other financial assistance (18.9%) or sheltered.workshop employment where most financial support came from outside sources (15.1%), (b) a job with no pay - in job training (3.8%) or volunteer work (13.2%) and (c) no job — did not work for training or money or served as volunteers (47.2%). A 2-way analysis of variance compared interactions using the independent variable, Compeer participation status, and each of the seven second independent variables. These variables were again analyzed one at a time by 2-way analysis of variance in order to compare score results among the three subject groups. The Self-concept, Concept of Therapist and Concept of Others scale scores were also used as mental health indicators. Concept scales were 16-item Likert scales. Score possibilities ranged from 16 to 112. If all items were marked number one (the most positive concept), the score would be 16. If all items were marked number seven (the most negative concept), the score would be 112. Hence, the lowest scoring group had the most positive outcomes. The impact of each of the seven second independent variables one at a time and.of Compeer participation status on concept scale scores were tested for statistical significance at <.05 by analysis of variance and the F-test. The 2—way analysis of variance tested the effects of each second 58 independent variable, of Compeer participation status (the baseline group, which was not yet influenced by Compeer, and the current and the former Compeer groups which had been influenced by having, or having had, volunteers) and of their interaction on the dependent variables. Two-way analysis of variance measurement is more sensitive to differences than the 1—way analysis of variance, especially in its power to detect interaction. The F-test indicates that there is a difference among groups but does not indicate which group was significantly different from the others. Therefore, in the present study, if the F-test was not significant at P: <.05, no further testing was done. IV . RESULTS The present cross-sectional study tested the effective- ness of nonprofessional volunteers in the Compeer companion- ship programs in Denver, Colorado and.Ottawa County; Michigan. All subjects were adults with a variety of mental illnesses who had been referred to Compeer by their ongoing mental health professionals. Most of the subjects were referred to help them improve their social skills and to provide them consistent companionship via one—on-one "friendships" with their Compeer volunteers. Subjects were appropriately' placed into the current Compeer client group, the former Compeer client group or the waiting list group, which was also the baseline group. There were no significant differences between the Michigan and Colorado groups. Subjects completed a 3—section, closed-ended question- naire which included a Community Activity Index with Self-care and Pleasant Activity subscales, a school and therapy questionnaire and Likert Self-concept, Concept of Therapist and Concept of Others scales. All scales were quantitatively analyzed by measuring objective and subjective quality of life indicators and second independent variables of predetermined mental health indicators/demographic variables. Compeer volunteers and the subjects’ mental health professionals were informants. They completed open-ended questionnaires which were qualitatively analyzed. Informants’ 59 60 results were compared to the subjects’ results to help explain similarities or discrepancies between subjects and informants. Subjects in the present study included 26 of the 36 current Compeer clients, 10 of the 28 former Compeer clients and 17 of the 21 individuals on Compeer’s waiting list. Sixteen therapists and one psychologist of the 25 mental health professionals participated (only four therapists were from Colorado). Thirty-three of the 55 volunteers participated. Of this number, 25 volunteers were active and eight were past volunteers or part of a husband/wife team. Table 1 presents a general description of the subjects’ characteristics. Characteristic categories were primary occupations, marital status, living arrangements, enrollment in a school or training program and mental illness diagnoses. Overall characteristics of the groups showed most subjects were not currently working or unskilled laborers (54.7%, 26.4%), single (52.8%), living in an apartment, group home or family care home (30.2%, 26.4%, 22.6%, respectively), not enrolled in a school or training program (86.8%) and schizophrenic (67.7%). Ages ranged from 18 to 75 years old with a mean of 49 years. There were 19 males and 34 females. In the mental illness diagnosis category, schizophrenia types were combinations of undifferentiated (n=24), schizoaffective disorder (p=6), residual (p=1) and paranoid (p=5). Diagnoses information was not matched to particular subjects or groups. Rather, it was a categorical listing obtained through Compeer coordinator interviews. 61 Table 1 Characteristics Within All Subject Groups p=53 PRIMARY OCCUPATION p Percentage Skilled trades 3 5.7 Business/sales 2 3.8 Clerical 3 5.7 Unskilled Laborer 14 26.4 Temporary 2 3.8 Not working 29 54.7 MARITAL STATUS Single, never married 28 52.8 Married 7 13.2 Legally separated 1 1.9 Divorced 14 26.4 Widowed 3 5.7 CURRENTLY LIVING IN Own house/condominium 6 11.3 Apartment 16 30.2 Boarding house 1 1.9 Group adult home 14 26.4 Family care home 12 22.6 Health-related facility 4 7.6 ENROLLED IN SCHOOL/TRAINING Yes 7 13.20 NO 46 86.80 DIAGNOSES Schizophrenia (all types) 36 67.7 Organic Personality Disorder 1 1.9 Psychotic Disorder with Delusions 2 3.8 Depression with Adjustment Disorder 3 5.7 Bipolar Disorder 3 5.7 62 Table 1 (cont’d). Borderline Personality Disorder 5 9.5 Post Traumatic Stress Disorder 1 1.9 Anorexia Nervosa 1 1.9 Avoidant Personality 1 1.9 All subjects were required to be in ongoing therapy to qualify as participants in this study. The type of mental health therapy each subject was receiving is shown in Table 2. Compeer clients were involved in mental health therapy in addition to their Compeer participation. Twenty-five Compeer clients were receiving individual psychotherapy, four of the 26 clients were attending day treatment, one was attending group therapy and one was in family therapy. Table 2 Mental Health Treatments of Current. Former and Waiting List Groups p=53 Percent Receiving p the Treatment Individual psychotherapy 46 86.8 Compeer 26 49.1 Day treatment/rehabilitation 13 24.5 Group psychotherapy 2 3.8 Couple/family psychotherapy 1 1.9 63 Table 3 presents a summary of the Community Activity Index and Self-care and Pleasant Activity scale scores by state by analysis of variance and the F-test. The states were Colorado and Michigan. Despite slight procedural and program differences between states, groups were found similar and therefore could be combined. 64 Table 3 Summary of Community Activitv Index. Self—care and Pleasant Activitv Scale Scoresfibv State by ANOVA p=53 Sig p Mean S.D F of F COMMUNITY ACTIVITY INDEX Entire population 53 25.45 11.24 .94 .34 Colorado 22 27.23 12.09 Michigan 31 24.19 10.63 SELF-CARE Entire population 53 12.57 6.42 1.67 .20 Colorado 22 13.91 6.67 Michigan 31 11.61 6.17 PLEASANT ACTIVITIES Entire population 53 13.83 6.11 .39 .54 Colorado 22 14.45 6.77 Michigan 31 13.39 5.68 65 A summary of the Community Activity Index and Self—care and Pleasant Activity scales on Table 4 verifies equivalence among the subject groups. The hypothesis presented in this study was that there would be a significant difference among the groups. The Compeer group that currently had volunteers was expected to do better than the former Compeer group that no longer had volunteers and the waiting list group that never had volunteers. However, as a result of the summary of the Community Activity Index, Self-care and Pleasant Activity scale scores, the extent of compeer participation did not impact significantly on these indicators of the dependent variable, quality of life. Hence, mentally ill clients did not have significant impacts in their quality of life regardless of their Compeer participation status. 66 Table 4 ANOVA Summary of Community Activity Index, Self-care and Pleasant Activity Scores by Compeer Participation Status Group p=53 Sig Q MESH S-D- E Q£_E COMMUNITY ACTIVITY INDEX Entire population 53 25.45 11.24 .06 .94 Currently has volunteer 25 25.96 11.22 No longer has volunteer 10 25.40 14.40 Therapy only-never volunteer 17 24.70 9.83 SELF-CARE Entire population 53 12.57 6.42 .03 .97 Currently has volunteer 26 12.69 6.40 No longer has volunteer 10 12.80 8.30 Therapy only-never volunteer 17 12.24 5.57 PLEASANT ACTIVITIES Entire population 53 13.83 6.11 .20 .82 Currently has volunteer 26 14.35 5.90 No longer has volunteer 10 13.00 7.09 Therapy only-never volunteer 17 13.53 6.16 67 Table 5 is a summary of overall concept scores using analysis of variance and the F-test at the <.05 level. The lowest mean scores represented the most positive outcomes. There were no significant differences found among the groups. However, the group who had never had volunteers scored somewhat (but not significantly) more positively’ on the Concept of Others scale than the other two groups. This may be due to the waiting list group having more frequent contact with friends than the other two groups had. This will be detailed in the discussion for Table 12. One current client did not complete the Self-concept scale, three current clients and one former client did not complete the Concept of Therapist scale and one current client did not complete the Concept of Others scale. 68 Table 5 ANOVA Summary of Concept Scales Status Group p=53 SELF—CONCEPT Entire population Currently has volunteer No longer has volunteer Therapy only-never volunteer CONCEPT OF THERAPIST Entire population Currently has volunteer No longer has volunteer Therapy only-never volunteer CONCEPT OF OTHERS Entire population Currently has volunteer No longer has volunteer Therapy only—never volunteer b 52 25 10 17 49 23 17 52 25 10 17 for Compeer' Participation Sig Mean S.D. E of F 40.60 16.20 .32 .73 41.24 16.05 43.10 17.97 38.18 16.04 27.27 13.45 .54 .59 27.43 12.83 30.89 19.34 25.12 10.80 46.90 19.68 2.38 .10 51.04 20.63 50.70 16.75 38.59 18.13 69 According to the mental health professionals, the length of time their current Compeer participants had been in the program ranged from three months to 10 years with a mean of 4.89 years and median of four years. Seven of the mental health professionals did not know how long some of their clients had been Compeer participants because those clients had been referred to Compeer by someone else. Comparisons among current Compeer clients who have volunteers, former Compeer clients who no longer have volunteers and the individuals on Compeer’s waiting list who never had volunteers were made to ascertain if the current client group had better results than the other two groups on measures of the dependent variables: Community Activity Index, Self-care, Pleasant Activity, Self-concept, Concept of Therapist and Concept of Others. Measurements were compared using 2-way analysis of variance and the F-test with significance levels at‘:JM5. Groups were compared controlling for each of the second set of independent variables: Employment status, educational level, level of self-supporting earnings, time spent with family over the previous six months, time spent with friends over the previous six months, psychiatric hospitalizations over the previous six months and gender. After controlling for the second set of independent variables one at a time, no significant differences were found 70 at the <.05 level among current, former, and waiting list Compeer groups on any of the Self-concept scales. Neither were significant differences found among current, former, and waiting list groups on any of the scales regarding employment status, level of self-supporting earnings or time spent with family over the previous six months. A 2-way analysis of variance compared interactions among the baseline waiting list group and the current and former Compeer groups using the independent variable, Compeer participation status, and each of the seven second independent variables. Concept scale score were tested for statistical significance at <.05 by analysis of variance and the F-test for each of the second independent variables (mental health indicators and demographic variables) separately between the three groups. .A second 2—way analysis of variance was used to compare the interaction effects of the baseline waiting list group, which had not been influenced by Compeer, and the current and former Compeer groups which had.been influenced.by having volunteers. As shown on Table 6, there was a significant 2-way interaction (P: <.04) between the Community Activity Index scale scores and gender. There were no significant main effects. Current Compeer males scored lower than males in the other two groups and the former client males scored highest. 71 Current Compeer females scored higher than females in the other two groups and waiting list females scored the lowest. Schizophrenia is the mental illness most often seen in Compeer clients. According to Flics and Herron (1991), schizophrenic females have a better prognosis than males with schizophrenia and have a lower relapse rate because females are more social, more help-seeking and have a greater ability for intimacy and verbal expression. 72 Table 6 Communitv Activity Index by Gender and Compeer Participation Status in 2-Wav ANOVA p=53 In Compeer/Not in Compeer Sum of Mean Source of variation Sgpares Sguare Main effects 196.67 98.34 Gender 195.04 195.04 Compeer participation status 1.76 1.76 2-way interactions 562.67 562.67 Group Scores p Mean MALES (all) 19 26.63 Currently has volunteer 10 21.80 No longer has volunteer 2 34.00 Therapy only-never volunteer 7 31.43 FEMALES (all) 34 24.79 Currently has volunteer 16 28.56 No longer has volunteer 8 23.25 Therapy only-never volunteer 10 20.00 .61 .01 .63 10.01 11.97 12.14 15.16 Sig .45 .21 .91 .04 73 Table 7 shows a significant 2—way interaction (P: <.05) on Self-care scales when the second independent variable was gender. The main effects were not significant. Overall, males scored higher than females. However, current Compeer males scored lowest and.the waiting list group scored.highest. Conversely, current Compeer females scored highest and waiting list females scored lowest. To speculate, the majority of Compeer clients are women and the most prevalent mental illness is schizophrenia. Schizophrenia is a more serious disease for men than woman, with poorer long-term adjustment (Torrey, 1983). Both Table 6 and Table 7 reinforce Flics and Herron’s (1991) analysis of predictors of premorbid adjustment between male and female schizophrenics. They found females had a higher premorbid adjustment and increased premorbid competence than males as well as a better prognosis because of greater socialization capacities. Women seem to have been able to make effective use of the socialization opportunities offered through Compeer, while 'men. receiving’ 'volunteer socialization opportunities did not do well in socialization or in self— care . 74 Table 7 Self-care by Gender and Compeer Participation Status in 2—Way ANOVA p=53 In Compeer/Not in Compeer Sum of Mean Sig Source of variation Sgpares Sguare E of F Main effects 27.58 13.79 .34 .71 Gender 26.04 26.04 .65 .43 Compeer participation status 1.58 1.58 .04 .84 2-way interactions 163.78 163.78 4.06 .05 Group Scores p Mean S.D. MALES (all) 19 12.74 5.93 Currently has volunteer 10 9.90 4.68 No longer has volunteer 2 17.50 7.78 Therapy only-never volunteer 7 15.43 5.77 FEMALES (all) 34 12.47 6.76 Currently has volunteer 16 14.44 6.83 No longer has volunteer 8 11.63 8.48 Therapy only-never volunteer 10 10.00 4.42 75 Self-care scale scores shown on Table 8 were significant when the second independent variable was time spent with friends over the previous six months (2-way interaction, P: <.04). However, significance was not present in the nein effects. The waiting list group had better self-care (more independence) when they spent time with friends less than once a week and less self-care (less independence) when they spent time with friends more than once a week. Inversely, the former client group displayed more independence when they spent time with friends more than once a week than when they spent time with friends less frequently. The current client group fell mid—range between the subject groups whether they spent time with their friends more or less than once a week although they did slightly better when they spent time with their friends more than once a week. These scores were similar to time spent with family over the previous six months in the waiting list group. This group displayed more self-care when they spent less time with family and less self-care when they spent more time with family. Sometimes close family ties can discourage independent behavior (Clausen and.Huffine, 1975). The former client group had higher levels of self-care the more frequently they spent time with family. The current client group scores were mid- range in all three variable measurements. 76 Table 8 Self-care bv Compeer Participation Status and Time Spent with Friends in 2—Wav ANOVA p=53 In Compeer/Not in Compeer Sum of Source of variation Sguares Main effects 38.25 Time with friends 24.70 Compeer participation status 13.85 2—way interactions 277.32 Group Scores LESS THAN ONCE A MONTH (all) Currently has volunteer No longer has volunteer Therapy only—never volunteer LESS THAN ONCE A WEEK (all) Currently has volunteer No longer has volunteer Therapy only-never volunteer MORE THAN ONCE A WEEK (all) Currently has volunteer No longer has volunteer Therapy only-never volunteer Q 12 29 14 Mean Sguare 12.75 12.35 13.85 138.66 Mean 12.58 12.25 16.00 12.33 11.37 14.25 12.66 13.57 15.14 .32 .31 .35 .50 5.45 3.40 5.67 8.47 Sig .81 .73 .56 .04 77 Pleasant Activity scale scores on Table 9 show that psychiatric hospitalizations luui a. significant effect on pleasant activities (P: <.04). No significance was found due to Compeer participation status. Subjects with one hospitalization during the past six months had higher mean Pleasant Activities scores than those not hospitalized. The overall mean score for those with one hospitalization was 18. The overall mean score for those not hospitalized was 13. This finding might have been related to hospital aftercare programs. IHospital aftercare was not included in this study. Four out of the 26 current clients, two out of the 10 former clients and three out of the 17 waiting list individuals were ixiaa psychiatric hospital once during the previous six months. The subjects’ hospitalization histories were unknown if the hospitalizations occurred prior to the past six months. 78 Table 9 Pleasant Activitv Scales by Compeer Participation Status and Psychiatric Hospitalizations Durinq Previous 6 Months in 2-Wav ANOVA Q=53 In Compeer/Not in Compeer Sum of Mean Sig Source of variation Sguares Sguare E of F Main effects 162.52 81.26 2.24 .12 Psychiatric hosp. 161.29 161.29 4.44 .04 Compeer participation status 2.61 2.61 .07 .79 2-way interactions 19.22 19.22 .53 .47 Group Scores - Not Hospitalized p=44 p Mean S.D. Currently has volunteer 22 14.50 6.22 No longer has volunteer 8 10.50 4.93 Therapy only-never volunteer 14 12.36 5.68 Groups Scores - Hospitalized Once p=9 Currently has volunteer 4 13.50 4.20 No longer has volunteer 2 23.00 5.66 Therapy only-never volunteer 3 19.00 6.25 79 Table 10 indicates subjects' Concept of Therapist scale scores were significantly affected by educational levels (P: <.03). Compeer participation did not affect the concept of therapist. (The lower the mean scale score, the more positive the concept.) Overall, subjects with less than a high school education had a more positive concept of their therapists, those who were high school graduates the least positive and those with college educations in the middle (20.00, 27.00, 34.93, respectively). There were no 2-way interactions. 80 Table 10 Concept of Therapist by Compeer Participation Status and Education in 2;Way ANOVA p=49 In Compeer/Not in Compeer Sum of Mean Sig Source of variation Sguares Sguare E of F Main effects 1214.03 404.68 2.58 .07 Educational level 1146.44 573.22 3.65 .03 Compeer participation status 103.37 103.37 .66 .42 2-way interactions 183.15 91.57 .58 .56 Group Scores p Mean S.D. Entire population 49 27.27 13.45 LESS THAN HIGH SCHOOL 14 20.00 5.74 Currently has volunteer 7 20.14 6.49 No longer has volunteer 3 17.00 1.73 Therapy only-never volunteer 4 22.00 6.48 HIGH SCHOOL 14 34.93 14.24 Currently has volunteer 6 36.67 15.33 No longer has volunteer 3 37.00 13.11 Therapy only-never volunteer 5 31.60 16.04 COLLEGE 21 27.00 14.23 Currently has volunteer 10 27.00 12.06 No longer has volunteer 3 38.67 29.74 Therapy only-never volunteer 8 22.63 7.67 81 Table 11 reflects a very mixed finding when measuring the impact of Compeer participation status on Concept of Others scale scores, with educational level controlled as the second independent variable. Compeer participation status affects Concept of Others scores with P: <.03. There were no 2—way interactions. 82 Table 11 Concept of Others by Compeer Participation Status and Educational Level inyg-WayfANOVA p=52 In Compeer/Not in Compeer Sum of Mean Sig Source of variation Sguares Sguare E of F Main effects 1882.64 627.55 1.69 .18 Educational level 300.67 150.34 .41 .67 Compeer participation status 1762.95 1762.95 4.75 .03 2-way interactions 623.92 311.96 .84 .44 Group Scores p Mean S.D. LESS THAN HIGH SCHOOL 15 46.27 21.93 Currently has volunteer 8 55.88 18.11 No longer has volunteer 3 42.33 23.71 Therapy only-never volunteer 4 30.00 22.14 HIGH SCHOOL 15 44.20 20.43 Currently has volunteer 7 39.57 23.12 No longer has volunteer 3 57.67 16.56 Therapy only-never volunteer 5 42.60 18.58 COLLEGE ‘ 22 49.18 18.18 Currently has volunteer 10 55.20 19.53 No longer has volunteer 4 51.75 12.82 Therapy only-never volunteer 8 40.38 16.93 83 As Table 12 shows, Compeer participation status was significantly related to Concept of Others scale scores when time spent with friends during the previous six months was controlled and the second independent variable (main effect was P: <.03 and Compeer participation status was P: <.03). There were no 2-way interactions. The waiting list group was most positive in its concept of others in all three variable outcomes. The current client group had the least positive concept of others if they spent less than once a week with their friends and the former client group had the least positive concept of others if they spent more than once a week with their non-Compeer friends. Compeer volunteers were not to be considered friends for this variable. Nevertheless, there was a possibility that some subjects did.not follow this instruction. One current client did not complete the scale. In contrast, the Concept of Others scale scores were not significantly related to Compeer participation when time spent with family during the previous six months was the second independent variable. There were no significant interactions. 171'“, . I . . .l . a . C 2 ..J v. r“ z... ..1 Cu .1. 1.. St .a a... .... 1.; II: .C r u c. .3 .J ..1 W ct Cc “R T; Q» C N T Cu Na d n o C __ C a . w u at Cu O a . n o C. L Co H3. 2 E w L To HR. 84 Table 12 Concept of Others by Compeer Participation Status and Time Spent with Friends During Previous 6 Months in 2-Way ANOVA p=53 In Compeer/Not in Compeer Sum of Mean Sig Source of variation Sgpares Sguare E of F Main effects 3408.14 1136.05 3.22 .03 Time with friends 1239.63 619.82 1.76 .18 Compeer participation status 1897.18 1897.18 5.38 .03 2-way interactions 99.24 49.62 .14 .87 Group Scores p Mean S.D. Entire population 52 46.90 19.68 LESS THAN ONCE A MONTH 12 44.50 24.61 Currently has volunteer 4 52.25 29.60 No longer has volunteer 3 49.00 30.12 Therapy only-never volunteer 5 35.60 19.50 LESS THAN ONCE A WEEK 12 57.00 14.67 Currently has volunteer 8 62.38 9.98 No longer has volunteer 0 - - Therapy only-never volunteer 4 46.25 18.03 MORE THAN ONCE A WEEK 28 43.61 18.38 Currently has volunteer 13 43.69 20.73 No longer has volunteer 7 51.43 10.80 Therapy only-never volunteer 8 36.63 18.72 .w. r . S. s , , . . . . .. 1 n c «L. a-» rt. L. :3 r: a V). I“ n“ Co ”W mm a C no .3 Po Au no no by Wu C... rd. n o n C ..nl. I. l .. . 85 Table 13 shows significance in the Concept of Others scale scores when the second independent variable was psychiatric hospitalizations during the previous six months. Compeer participation levels were significantly related to Concept of Others scores. The waiting list group with one hospitalization during the past six months had a significantly lower mean score than the current and the former groups (the lowest scores were most positive). There were no 2-way interactions. II . P: Q. .. . .. .. .. . .. ... I. . i. a . C C 3 I K .3 3.. C C L. L C L. ..C r u n. 7n 1 .l w C N T E C N T. .a no .t _ no .3 . pd. P}. ,. fl C «H. Al. A: ”It.“ a]. Pin Aiv 86 Table 13 Concept of Others by, Compeer Participation Status and Psychiatric Hospitalizations During the Previous 6 Months in g-way ANOVA Q=53 In Compeer/Not in Compeer Sum of Mean Sig Source of variation Sguares Sguare E of F Main effects 1934.99 967.50 2.66 .08 Psychiatric hosp. 357.78 357.78 .98 .33 Compeer participation status 1771.77 1771.77 4.87 .03 2-way interactions 300.31 300.31 .83 .37 Group Scores g Mean S.D. NO HOSPITALIZATIONS 44 47.89 19.77 Currently has volunteer 22 51.00 21.15 No longer has volunteer 8 51.13 17.19 Therapy only—never volunteer 14 41.14 18.41 ONE HOSPITALIZATION 8 41.50 19.51 Currently has volunteer 3 51.33 20.23 No longer has volunteer 2 49.00 21.21 Therapy only—never volunteer 3 26.67 12.90 87 In summary, at P: <.05 significance, there were no significant differences between groups by state (Table 3), on the overall Community Activity Index, Pleasant Activity, Self— care scale summary (Table 4), nor on Self-concept, Concept of Therapist and Concept of Others scale summary. There was, however, a significance on Concept of Others scale scores at the P: <.10 level. The waiting list group had a considerably more positive concept of others than the other two groups (Table 5). On Table 6, the Community Activity Index scores and the second independent 'variable, gender, showed. there was a significant 2-way interaction of P: <.04. Scores were higher for males in the former client group than for males in the waiting list group (34.00, 31.43) and males in the current client group scored lowest (21.80). For females, the higher scores were in the current client group than in the former client group (28.56, 23.35) and the waiting list group scored lowest (20.00). On Table 7, Self-care scale scores with gender as the second independent variable showed a significant 2—way interaction of P: <.05. Male former clients again scored higher than.males in the waiting list group (17.50, 12.47) and males in the current client group scored lowest (9.90). Conversely for females, the current client group scored higher than the former client group (14.44, 11.63) and females in the waiting list group scored lowest (10.00). A tally of the highest scores on the Community Activity ‘ :fififl' $LAJC- .. . C. S no pf. C. c 5.; x. A \. 031'). pro, i 88 Index, the Self-care and Pleasant Activity scales and the most positive scores on the concept scales that had a significance of P: <.05 showed that the current client group was least represented on scales with significant variables and the waiting list group was the most represented, especially on the concept scales. The impact of significant variables were directly related to Compeer'participation status. Some of the scales had interactive effects with the various mental health indicators and demographic variables as second independent variables. For a list of mental health and quality of life indicators with the most positive scores in each Compeer participation status group (P: <.05 significance) see Appendix J. Informant qualitative results Up to this point, quantitative research methods were used to measure and analyze subject data. However, quantitative research restricts the scope of inquiry. The hypotheses guided the present study. They predicted what this researcher believed the study’s results would most likely be and the measurements were built upon that premise“ .Scales, along with other questions, measured each variable. Responses were processed by the computer-driven statistical package. The present study used.the SPSS package. IResults were analyzed.by analysis of variance and by 2-way analysis of variance with a second set of independent variables. ‘Two-way interactions and significance of P: <.05 were reported. Qualitative methods do not depend upon the hypotheses to v u 'Y‘J“ Oll'u‘ l l u ,- .F' ;_ “8““ c ‘ I Mifif "--A‘ ~— . i (4 r1, NC“! «p. 89 guide the study. Methods are much less confining and the researcher is open to new information that may redirect the study. The subjects’ mental health professionals and Compeer volunteers were informants in the present study. Qualitative research was used to analyze how informants perceived the effectiveness of volunteers as a supplement to professional mental health treatment for the mentally ill. Structured, open-ended questions elicited viewpoints about themselves and their relationships with the each other and with the subjects. Indicators of patterns and themes found in the narrative helped to address issues raised by the quantitative data analysis. Mental health professionals The:17 mental health professionals in this study included 13 from Michigan and four from Colorado. These professionals in this study represented four different categories: Social workers (50%), mental health nurses (12.5%), psychologists (12.5%) and other mental health clinicians such as licensed counselors (25%) . They completed individual questionnaires on 29 of the 53 clients in the subject groups. Some had more than one client participating. One social worker had as many as eight clients participating. To protect client anonymity, there was no identification of the client being described. Thus client-by-client comparisons of data from subjects and from informants are not possible. The range of time mental health professionals had been in their professions was between two and.40 years. The mean time 90 was 13.66 years and the median time was seven years. The amount of time at their current agencies was between three months and 26 years with a mean of 8.11 years and a median of five years. Mental health professionals’ observations of client functioning Most of the professionals evaluated their clients’ ongoing mental health status by continued assessments (75 . 1%) . Other methods included observation (6.3%), referral to a psychiatrist (6.3%), documentation (6.3%) and psychological testing (6.3%). Social integration (58%) was the major reason why professionals referred cflients tx> Compeer. Companionship (17%) was a distant second reason for referrals and client request (11.8%) was third. Other referrals totaled 5.9%. Mental health professionals perceived their clients’ attitudes as very positive (69.2%), positive (23.1%) and somewhat positive (7.7%) when discussing Compeer during therapy. The professionals reported that none of their clients felt negatively toward Compeer or its volunteers. Client improvement which the therapists attributed to Compeer participation was reported in 65.4% of the cases. In 15.4% of the cases, mental health professionals thought their clients remained. about the same ‘whether' in” or‘ not in, Compeer. Other professionals had clients on the Compeer waiting list and did not respond. Positive client changes were reported in the combination LA; 'th l . :81 S Q.» .hu Lt» 38.: 91 of independent living and mental health (52.6%) since Compeer. If either independent living or mental health improvements occurred, the percentages were 10.5% and 20.7%, respectively. Only one professional noted no apparent change in either area. Table 14 Types of Client Improvement since Compeer Participation as Perceived by Mental Health Professionals g=2l (coded from open-ended questions) Percentage n of Clients More social comfort 11 52.4 More independent 3 14.3 More trusting 2 9.5 Substance abuse decreased 1 4.8 Less need for therapist I 4.8 No change 3 14.3 Seven out of the 17 mental health professionals had 11 clients whO‘were former Compeer participants. The majority of the professionals asserted that positive changes they noted while their clients were active in Compeer had been sustainable (sustainable - 83.3%; not sustainable ~ 16.7%). The likelihood of positive improvements continuing to be sustainable were seen as very likely (28.6%), likely (57.1%) and somewhat likely (14.3%). Four professionals had clients on Compeer’ s waiting list. These professionals completed only the demographic portion of the therapist questionnaire since their clients were in the baseline group and still not influenced by the independent 92 variable, participation in Compeer. Two-thirds of the mental health professionals had worked with nonprofessional volunteers before Compeer involvement. One-third had not. Professionals’ ratings of Compeer volunteers regarding their clients’ mental health conditions were very' helpful (79.2%), helpful (16.7%) and somewhat helpful (4.2%). None of the volunteers were rated.negatively. Mental health professionals considered their clients to be very satisfied (79.2%), satisfied (12.5%) and somewhat satisfied (8.3%) with Compeer. There were no dissatisfaction responses. They rated Compeer volunteers as very helpful, (57.7%), helpful (19.2%) and somewhat helpful (23.1%). They also believed nonprofessional volunteers, in general, were a positive and vital part of the mental health system - very positive (82.8%), positive (13.8%) and. somewhat positive (3.4%). All of the mental health professionals in this study said that they would recommend Compeer to other clients and to colleagues. There were no negative responses. Most of the therapists have had minimal, if any, contact with the volunteers. Eighty-five percent of the volunteers said they had never had contact with their friends’ mental health professionals after the initial screening process to become volunteers. Mental health professionals seemed to be forming their conclusions about the effectiveness of Compeer volunteers on the fact that their clients have not had negative experiences with the volunteers and that clients spoke positively about them to their therapists. 93 Compeer volunteers Compeer volunteers represented many walks of life. The 33 volunteers in the present study listed 28 different professions. Among these volunteers, the profession with the most volunteers was school teacher, retired.(n=3). .Almost all of the volunteers were white (32 out of 33). One was African- American. Volunteers’ religious persuasions were Protestant (63.6%), Catholic (24.2%) and other (3%). Three volunteers did not respond to the question on religion. Twenty-nine percent of the volunteers were male and 71% were female. Volunteers’ ages ranged from 23 to 78 years old. Their mean age was 49 years, which was identical to the mean ages of the subjects. The median age was 48 years old. The length of time volunteers had been in Compeer ranged from six months to 12 years. The mean was 6.6 years with a median of five years. The majority of volunteers had been recruited through either newspaper advertisements or by word of mouth. Seventy-nine percent had been volunteers elsewhere. Twenty-four percent were still at those volunteer positions besides being volunteers at Compeer. Table 15 depicts the most important reasons why volunteers said they joined Compeer. The reasons were to help someone and because they had an interest in mental illness. 94 Table 15 Reasons Why Volunteers Joined Compeer (n=33) (Coded from open-ended questionnaire) n Percentage To help someone 15 45.5 Interest in mental illness 9 27.3 Compeer’s philosophy 4 12.1 Religious reasons 2 6.1 Flexible hours 2 6.1 Influenced by someone I 3.0 Compeer volunteer observations of the friend’s functioning Compeer volunteers’ average monthly contacts with their friends ranged from one to 10 times with a mean of 4.28 times and a median of four times. Volunteers rated their friends’ benefits due to the one-on—one associations as - very high (50%), high (21.9%), medium (25%) and low (3.1%). Based on the coding from open-ended questionnaires, volunteers felt their friends’ mental health status was improving. The friends were becoming less isolated and more trusting of relationships and of social settings. Table 16 shows how volunteers perceived their friends benefitted from the Compeer program. Two of the volunteers’ questioned if their friends benefitted at all. One volunteer stated that "the friend. always tried. to borrow' money". Realistic expectations must be discussed with the Compeer friend toward the beginning of the relationship for a better 'understanding of that relationship. Another volunteer felt 95 that "it was impossible to get close to the friend". The Compeer friend might not have been ready for socialization, especially if the friend was male and schizophrenic (Torrey, 1983). This writer believes volunteers need to consult with the Compeer coordinator or the friend’s therapist to learn more about the mental illness and ways to facilitate the friendship. Volunteers perceived their friends’ satisfaction wdth Compeer as very satisfied (66.7%), satisfied (15.2%), somewhat satisfied (15.2%) and. somewhat dissatisfied. (3%). This supports the volunteers’ view'of the benefits of their one-on- one associations with their friends. Table 16 The Most Important Ways Compeer Friends Benefitted from their Friendships with Volunteers as Reported by Volunteers n=33 (Coded from open-ended questionnaire) Percentage n of Friends Less isolation 9 30.0 More trust 6 20.0 Improved mental health 5 16.7 Provides advocacy 3 10.0 Improved personal appearance 2 6.7 Temporary benefits only 2 6.7 More independence 1 3.3 Always tried to borrow money 1 3.3 Impossible to get close 1 3.3 No response 2 6.7 96 In the present study, volunteers were asked to make recommendations how Compeer could be improved. TWO of the major recommendations were more professional involvement (30.8%) and more volunteers (30.8%). Fifty-eight percent believed more seminars about mental illness would greatly improve volunteer effectiveness. Contact between the volunteer and the Compeer friend’s mental health professional Eighty-five percent of the volunteers said they had.never contacted their friends"mental health professionals after the initial screening process to become a volunteer. Confirming this percentage, 86.2% of the mental health professionals said they did not make contact with the volunteers. Despite these numbers, half of the volunteers said their friends’ mental health professionals seemed very interested (34.6%) or interested (19.2%). Twenty-seven percent were somewhat interested and 19% were not interested in the Compeer association. Eighty—eight percent of the volunteers thought mental health professionals gave quality advice when asked (although few of them ever asked). This researcher discovered there was some confusion by the volunteers between the friends’ mental health professionals and the mental health professionals at Compeer. Toward the end of the data collection, two volunteers asked this writer to which professional the question about the "interaction between volunteer and the friend’s mental health professional" was referring, the Compeer coordinator or the 97 friend’s therapist. This writer was alerted again when only 9.7% of the volunteers said they had no contact with the friend’s therapist and 3.2% said the friend’s therapist was helpful but needed to show more interest. On the other hand, 35% of Michigan’s volunteers said the Compeer staff made them feel appreciated and 31.6% said they had positive contact with the Compeer coordinator. In Colorado, volunteers did not ask the writer about the differentiation between the friends’ mental health professionals and the Compeer coordinator. V. DISCUSSION Data collection from the subjects being treated for mental illness was done by, or in the presence of, this researcher. Questionnaires were checked for completeness immediately following the data collection meeting with each subject. All data were coded by this writer and entered into the computer for analysis by SPSS statistical programs. Codes were independently double checked for errors to verify accuracy. The present study's outcome offered no personal gain to any subject, therapist or volunteer. The only obvious reward for participating was the opportunity for thoughtful interaction.and.the opportunity'to‘assist Compeer to’determine future program directions and.to assess the place of volunteer programs offering friendshiptin.the'mental health field.and in the community. Self-report scales Data for the quantitative analysis were collected directly from the subjects using closed-ended self-report scales. This made up the backbone of the quantitative study. The Community Activity Index scale, including the Self-care and Pleasant Activity subscales, were self-reported about specific types of activities which the subject either did or did not do over the previous week. Answers were either "yes" or "no" or left blank. Total scores were calculated for each subject on the number of "yes" answers. The scales also 98 99 included closed—ended demographic questions about the subject’s current status and life ewents (e.g. education, employment, financial status, and living arrangements). All of the subjects completed this part of the questionnaire, using this researcher for assistance as needed.(e.gu blind or illiterate clients required total assistance). The School and Therapy portion of the questionnaire asked for concise responses about health and socialization factors (e.gu psychiatric hospitalizations, time spent.with family and friends). All possible responses were listed on the questionnaire. It was also completed by all of the subjects, with assistance as needed. The Self-concept, Concept of Therapist and Concept of Others measures were 16-item Likert scales. On each item, subjects chose one of the possible seven adjective variations about their perceptions of each adjective as it applied to them. There was a mixture of from positive to negative and from negative to positive answers. The scales called for the subjects’ introspection about feelings and personal opinions. Hence, although the measures were reliable, the responses must remain somewhat more suspect on the concept scales than those obtained from the other measures. One person in the current client group did not complete any of the concept measures. Three additional subjects from the current client group and one from the former client group did not complete the Concept of Therapist measure. Self-report scales can be underreported or overreported. 100 Answers relying on memory and/or judgement leave considerable opportunity for error or distortions because of possible limited recall abilities of some of the subjects. Additionally, subjects sometimes try to give what they think are socially acceptable responses or try to present a more favorable picture of themselves, especially if they think the researcher is a representative of the agency. Each subject wrote responses privately, using closed- ended questionnaires. The researcher was available to answer respondent questions and to assist in unusual circumstances when a subject responded. orally (e.g. an illiterate or severely handicapped respondent). After comparing responses with. volunteers and. therapists ‘via informant qualitative questionnaires, no remarkable inconsistencies were identified by this researcher. Research results and the hypotheses In hypotheses 1 and 2, the current client group with companionship volunteers would improve their degree of objective and subjective quality of life and the former client group no longer with companionship volunteers would display sustainability in any positive changes made while they had volunteers. Both of these groups were compared to mentally ill adults receiving only professional therapy and never having companionship volunteers. Quantitative testing employed an analysis of variance, using the F-test with a significance level of P: <.05. The instruments measuring quality of life were the Community Activity Index, Self-care, 101 Pleasant Activity, Self—concept, Concept of Therapist and Concept of Others scales. There were no significant differences among groups with or without volunteers on any of these scales. Thus there was no statistical evidence of program impact. Potential effects on mental health indicators of seven second independent variables each taken one at a time and the main independent variable, Compeer participation status, and interactions in a 2-way analysis of variance was also tested. These second independent variables were predetermined mental health indicators and demographic variables. Significant relationships at P: <.05 were found on the Pleasant Activity scale (higher Pleasant Activity scores if subjects had one psychiatric hospitalization during the previous six months) and on the Concept of Therapist scale (more positive Concept of Therapist scores if the subjects had less than a high school education). These results controlled for Compeer participation levels. Concept of Others scores (controlling for educational level, for time spent with friends during the previous six months and for psychiatric hospitalizations) were significantly related to Compeer participation status. Two-way interaction effects of Compeer participation and a second independent variable were significant at P: <.05 on the Community'Activity Index and on the Self-care scale due to Compeer participation status when the second independent variable was gender and on the Self-care scale due to Compeer participation status when the second independent variable was 102 time spent with friends during the previous six months. When gender was the second independent variable, Compeer participation status and gender showed significant 2-way interactions at P: <.04 on the Community Activity Index and significant 2-way interactions at P: <.05 on self-care. When time spent with friends was the second independent variable with Compeer participation status, there were significant 2— way interactions at P: <.04 on self-care. ' When psychiatric hospitalizations during the previous six months was the second independent variable, its impact on pleasant activity was significant at P: <.04. This was possibly due to something about the psychiatric hospital, which can. be seen. when Compeer‘ participation status is controlled. When educational level was the second independent variable, it impact on the concept of therapist was significant at P: <.01, when Compeer participation status was controlled. Concept of others controlling for educational level, time spent with friends during the previous six months and psychiatric hospitalizations during the previous six months as second independent variables each showed a significant impact P: <.03 due to Compeer participation status, but not in the hoped-for direction. The waiting list group scored most positive on all of these Concept of Others scales. This was possibly due to this group seeing their friends more often than the current and former groups did (e.g. friends at the "clubhouse" at Michigan’s Compeer). 103 Hypotheses 3 and 4 were supported in the qualitative measurements. Mental health professionals reported positive changes in 65.4% of their clients’ degree of objective and subjective quality of life when they currently had volunteers and positive changes were sustained in 83.3% of their clients who formerly had volunteers. 'These changes were found in both the clients’ independent living and mental health status (52.6%). As shown in Table 14, clients improved in areas of social comfort, trust, independence, substance:abuse recovery, and had less need for a therapist. Hypothesis 5 ‘was supported. in the qualitative measurements. Volunteers perceived that their one-on—one associations with their friends were helpful in their friends’ mental health.progressiont Volunteers reported their friends experienced less isolation, more trust, improved personal appearance, more independence and improved mental health than when the volunteer friendships began. Hypothesis 6 was not supported. Contact was not maintained between the friends’ mental health professional and the volunteer. Instead, volunteers contacted the Compeer coordinator if advice was needed. Approximately 86% of the volunteers and the friends’ mental health professionals did not maintain contact at all. The majority of the professionals expected volunteers to contact them if they needed advice but expressed.no wish for other contact with the volunteers. On the other hand, 31% of the volunteers expressed.a«desire for more contact with the professionals and 104 58% wished to learn more about mental illness by professionals. There was more communication between volunteers and the Compeer coordinators. Compeer was the common ground. between the ‘volunteer and. client and the coordinator was usually accessible. Nevertheless, 95% of the mental health professionals in this study rated volunteers as helpful to their clients because of the positive changes they observed in their clients and all of the professionals rated them as a positive part of the mental health system. Mental health. jprofessionals and. 'volunteers were responsive to questions about client changes. Professionals had firsthand knowledge of changes in clients’ lives and circumstances that would allow them to competently be able to associate specific areas of change with the companionship program intervention. The questionnaire response rate was much higher for mental health professionals working for Ottawa County Community Mental Health, of which Compeer is a part, than for professionals affiliated with Compeer only through clients they had referred to the agency (13 out of 17 were from.Michigan). Volunteers also had firsthand knowledge into the lives of the subjects from a friendship standpoint. Information from volunteers, however, might tend to be more biased because of their personal involvement in the program. Since volunteers do not work for wages and are more praiseworthy, feeling they are beneficial to their Compeer friends gives a sense of accomplishment and service. 105 Limitations of the study Although.Compeer is nationwide, this study was limited to the Colorado and Michigan programs. Neither state had more than one active Compeer program and these programs were relatively small. If a pre-test was given at the orientation of each program, participant numbers would be too small to show significance of any kind unless the study was conducted over a long period of time. Time constraints and funding prevented this ideal procedure. The relationship between study results and mental illness diagnoses could not be used in the present research. During data collection when subjects were asked what type of mental illness they had, most of them did not want to reveal that information to this researcher. Compeer coordinators provided this writer with only the number of people they had with each particular diagnosis. Thus, there was no information on individual diagnosis. There are limits to the validity and unreliability of data in this study. This cross-sectional study could not control for many secondary variables. There was no comparative data.with other time frames in the subjects’ lives which might have been noticeable using pre-posttests. Spitzer, Endicott and Robins (1978) advise researchers studying the mentally ill to be cognizant that (1) subject variance can occur when patients have different conditions at different times, (2) occasion variance can.occur when patients are in different stages of the same condition at different 106 times, (3) information ‘variance is ‘when clinicians have different sources of information about their clients, (4) observation variance is when clinicians presented with the same stimuli differ in what they notice, and (5) criterion variance can arise when there are differences in the formal inclusion and exclusion criteria that clinicians use to summarize patient data into psychiatric diagnoses (also see Ihilevich and Gleser, 1982, p. 6). There was some confusion.about questions on the volunteer questionnaire that asked about the volunteer’s interactions with their Compeer friend’s therapist and perceptions of the therapist’s interest in the Compeer relationship. Volunteers contacted the Compeer coordinators more frequently for guidance, not their friends’ mental health professionals. Hence, results about the amount of interest shown by mental health professionals might be misleading. Finally, mental health professionals working for a public agency might be fulfilling requirements in order to go into private practice. INone of the referring professionals in this study were already in private practice. A study of employee commitment and turnover showed that "stayers" maintained a relatively constant commitment during a 15-month period but "leavers" started with a lower commitment which declined steadily as they got closer to the point of quitting (Porter, Crampon and. Smith, 1976). In. the ‘present study, all professional mental health informants worked for public agencies. VI . CONCLUS I ONS In tertiary prevention programs, such as the Compeer companionship jprogram, 'volunteers only' interact with the mentally ill clients and rarely, if ever, are in contact with the clients’ professionals. Tertiary prevention programs are not really prevention programs but services attempting to reintegrate persons suffering from mental illness into the community; The preventive function is t1) reduce relapses (Leiter & Webb, 1983). Nevertheless, in the present study, despite the absence of communication with volunteers, all of the mental health professionals felt that volunteers were indeed a viable supplement to professional therapy for the mental ill. This dissertation investigated the effectiveness of volunteers as a supplement to professional therapy for the mentally ill. The Compeer companionship program was used as the case study, which included cross—sectional comparison of three groups (Compeer clients who currently have volunteers, former Compeer clients who no longer have volunteers and individuals on Compeer's waiting list who have not yet had volunteers). Quantitatively, this study has shown that there were no significant differences in.the«quality of life among the three subject groups regardless of ‘their' Compeer' participation status as measured by mental health indicators, the dependent variables (the Community Activity Index with Self—care and 107 108 Pleasant Activity subscales, by the School and Therapy Questionnaire and by Self-concept, Concept of Therapist and Concept of Others Likert scales). Score results for the three Compeer participation groups were measured and the impact of Compeer status was tested by analysis of variance, using the F-test at P: <.05 significance. Subject groups were also compared using 2-way analysis of variance and the F-test, controlling for each of the second set of independent variables: employment status, educational level, level of self—supporting earnings, time spent with family over the previous six months, time spent with friends over the pmevious six months, psychiatric hospitalizations over the previous six months and gender. The impact of each of this second set of independent variables on mental health indicators was assessed, controlling and testing for the impact of the independent variable, Compeer participation status, and for interactions among the variables. Overall, subjects among the three groups that had one psychiatric hospitalization over the past six months had pleasant activity mean scores significantly higher than those not hospitalized. The overall mean score for those with one hospitalization was 18. The overall mean score for those not hospitalized was 13. No significance was found among the groups due to Compeer participation status. IHowever, the mean scores of the subjects in the current Compeer group were lower than the mean score in the other two groups. This might have been due to the Compeer volunteers’ decreased interactions 109 with clients during periods of hospitalizations. Four out of the 26 current clients, two out of the 10 former clients and three out of the 17 waiting list individuals were in a psychiatric hospital once during the previous six months. The subjects’ hospitalization histories were unknown if the hospitalizations occurred prior to the past six months. Also, hospital aftercare program participation was unknown. Subjects’ educational levels controlling for Compeer participation status were associated with the concept of therapists were significant at P: <.03. However, Compeer participation status did not influence these concepts among groups. Although there were no significant differences among groups, overall, subjects who had less than a high school education had the most positive concepts of their therapists and. those who ‘were high. school graduates had. the least positive concepts of their therapists. There were no 2-way interactions. The reason for this outcome was unclear. Concept of others controlling for educational levels was affected by Compeer participation status. This finding was very mixed among the groups. There were no significant 2—way interactions. The reason for this outcome was unclear. Concept of others controlling for psychiatric hospitalizations was affected by Compeer participation status . There were no significant 2-way interactions. The waiting list group had the most positive concept of others whether they had been hospitalized or not during the previous six 110 months. The current Compeer client group had the least positive mean scores. This may be due to the Compeer volunteer’s lack of interaction with the client during psychiatric hospitalizations, but this finding is not positive for the Compeer program. Controlling for time spent with friends during the previous six months and concept of others showed a significant main effect due to Compeer participation status (P: <.03). There were no significant 2-way interactions. It was interesting that the waiting list group had the most positive mean scores in their concept of others when they spent time with friends more than once a week. To speculate, this group might have social networks sufficient to create feelings of well being such as friends at work, friends at the home in which they live or in clubs to which they belong. On the other hand, the current client group had the least positive concept of others if they spent less than once a week with their non-Compeer friends and.the former Compeer group had the least positive concept of others if they spent more than once a week with their non-Compeer friends. The 2-way interaction of Compeer status and.gender on the Community Activity Index showed that females who were current Compeer clients had.higher mean scores compared to females who were not Compeer clients. They had higher scores than males who were also current Compeer clients. 'This finding supported Flics and Herron’s (1991) finding that females were more likely to participate and do well in groups that require 111 socialization because females are more help—seeking, more social and have a greater ability for intimacy and verbal expression. The 2-way interactitmlon.the Self-care scales showed that females, again, had the highest mean scores if they were current Compeer participants. ifimnrscored.higher'than females who were not Compeer participants. Males, on the other hand, had the lowest mean scores if they were current Compeer participants. The implications for Compeer programming based on these gender findings is for the agency to look at affects of programming'by'gendern Recounting Flics and Herron (1991) and Torrey (1983), males have more debilitating effects from schizophrenia (the mental illness most often seen by Compeer). There was a 2-way interaction on the Self-care scales controlling for time spent with friends. Here, self-care is synonymous with independence. The waiting list group had better self-care when they spent time with friends less than once a week and less self-care when they spent time with friends more than once a week. Inversely, the former client group displayed more self-care when they spent time with friends more than once a week than when they spent time with friends less frequently. The reason for the difference between the waiting list and the former Compeer group differences is not clear. The current client group’s mean scores were in the middle of the other two groups. There was a striking contrast between the quantitative 112 outcomes and the qualitative outcomes in this study. Quantitative results as shown above were not as glowing as qualitative results. Nevertheless, they identified various areas in which the Compeer program could investigate to possibly make the program better (e.g. more gender-related programming). Qualitatively’ speakingy 'mental health professional anui volunteer informants’ impressions cflf the effectiveness of volunteers as a supplement to professional mental health treatment showed that the majority of the professionals and volunteers believed that volunteers substantially increased the quality of life for mentally ill individuals. Mental health.professionals reported positive changes in 65.4% of their clients due to current Compeer participation. About 83% of the mental health professionals whose clients were former Compeer participants believed that improvements made while their clients were in Compeer had been sustained. Most felt that sustainability would most likely continue. Compeer volunteers perceived their one-on-one associations were beneficial for their Compeer friends. Benefits reported by volunteers included less isolation, more trust, improved personal appearance, more independence and improved mental health. Some volunteers, however, thought the benefits of Compeer were only temporary. The majority of volunteers said their friends were very satisfied with Compeer. Likewise, the majority of mental health professionals said their clients were very satisfied 113 with the Compeer program and with its volunteers. None of the mental health professionals rated nonprofessional Compeer volunteers negatively: 'Volunteers were rated as helpful to the clients (95.9%) and a positive part of the mental health system (100%) by the mental health professionals in this study. There was no regular contact between the mental health professional and the Compeer volunteer. Eighty-five percent of the volunteers said they had never contacted the professional since the initial screening visit to become a volunteer. Confirming this, 86.2% of the mental health professionals said they did not make contact with volunteers. However, the majority said they expected volunteers to call them if they needed advice. Most of the volunteers contacted the Compeer coordinator, instead.of the friends’ mental health professionals. An implication of the qualitative aspect of this study is the amount of involvement mental health professionals have with volunteers entrusted as a supplement to professional therapy. In the past, the professionals would have insisted on being in control of all aspects of therapy, especially where volunteers were concerned” IDespite the lack of contact, most mental health professionals were confident that Compeer volunteers were making the quality of life better for their clients. This raises questions about the differences between qualitative and quantitative studies. Qualitatively, this 114 study's results were much the same as results from the literature review. Volunteer programs usually got very high ratings from the clients, volunteers and mental health professionals. Shipley (1976), however, found that studies about volunteers as companions for the mentally ill had glowing qualitative results but did not do very well when analyzed quantitatively. Quantitatively speaking, the results of the present study did not support the perceptions of the informants. This study has raised questions for future research” .Are volunteers more effective with individuals with one type of mental illness over other types of mental illness? Is there a difference in the qualitative outcome because few mental health professionals participated, or took an interest in, this study or in the work of the volunteer? In what way does formal education play a role in a mentally ill person’s concept of therapist or concept of others? Why’ would individuals on Compeer’s waiting list feel they are less independent when they are with friends more than once a week while the former Compeer group feel more independent? What differences would there have been in this study's outcome if intelligence test scores of all subjects were included? Could the volunteers have been more successful in the Compeer friends’ functional and behavioral improvements if they interacted with the friends at least three times per week rather than once a week as suggested by Oei and Tan (1981)? 115 Implications for practice In the treatment of individuals with mental illness, social work cannot be viewed in isolation from the social conditions which foster the problems the social worker is attempting to treat. A systems approach allows the social worker to operate in whatever fashion is appropriate to the client’s person-in-situation needs. It attempts to decrease emotional disorders by making community agencies more responsive to human.needs. WResource cycling" (Leiter & Webb, 1983) considers the full range of human needs, ways that can meet these needs and potential consequences should these needs not be met. Informal and personal social networks create a nonthreatening environment for the mentally ill that a professional setting' cannot offeiu Good social support networks that provide empowerment and a sense of mastery over one’s own life lead to increased self-control and a more positive outcome in.psychosis treatment (Tobias, 1990; Benton & Schroeder, 1990; Kiev, 1979). Compeer provides the informal and personal social networks to assist mentally ill persons by offering volunteers for one—on—one role—modeling, advocacy, socialization, and friendship and should be considered by mental health professionals as part of the systems approach for services to the mentally ill. However, therapists and volunteers should maintain a working relationship throughout a client's therapy to ensure 116 each volunteer’s effectiveness is perpetually optimal for the client. Since there is a difference between mentally ill males and females, especially in schizophrenia (Flics & Herron, 1991; Torrey, 1983), it is important to identify these differences and train the volunteers towmeet the special needs of each gender. APPENDICES APPENDIX A GENERAL CLIENT INFORMATION Employment Status: vvvvvvvvv Primary Occupation: AAAA vvvv vvvvv \OCDx‘IONU'I Name of Occupation: \DCDQO'NUID‘PUJNH “>wa Employed Full Time Employed Part Time Homemaker College Student High School Student Retired Unemployed/Disabled Unemployed Self-employed Professional Paraprofessional Trainee/Journeyman Craftsman, Building/ Skilled Trades or Skilled Laborer Business/Sales Clerical Laborer Temporary Help Not Working Now Other Highest Degree Earned: vvvvvvv \lmU'InhbJNl-J k0 AAAA vvvv 118 10. ll. 12. 13. Grade One or Less Some Grade School Grade School Some High School High School Some College Associate Degree (Arts or Science) Bachelor Degree Technical or Trade School Some Graduate School Masters Degree Ph.D., M.D., J.D. Other (Specify) 5. 6. 7. Gender 119 Male ( ) Female ( ) I am currently living in (check one)... vvvvvvvv (I)\lONU'lvpquvl-fl| A house or condominium which I own An apartment or house which I rent A boarding home or hotel An adult home A family care home A health related facility A nursing home Other I am currently (check one)... During vvvvvv vvvvvvv mU‘Ish-UJNH Single, was never married Married for the first time Married for the second, or later, time Legally separated Divorced Widowed the past two years, I (check all that apply)... \IONU'IhUJNH Married Divorced Had a close friend or family member die Changed living arrangements Changed employment or training program Sustained a serious injury from an accident Had other stressful life event 9. dollars working as a 10. (Please specify) During the past week I earned approximately .00 Leave blank if you did not work I live in... vvvv fiWNl-J A very large city of 500,000 + The suburbs Small town close to a city Rural farm/ranch ll. 12. 120 My status at the home in which I live is (check al that apply)... ( ) 1. Living with husband/wife ( ) 2. Married, not living with spouse ( ) 3. Living with boyfriend/girlfriend ( ) 4. Living alone independently ( ) 5. Living alone with assisted living visits ( ) 6. Living with roommate(s) independently ( ) 7. Living with roommate(s) and assisted living visits ( ) 8. Living as a dependent with relative/guardian ( ) 9. Living independently with relative/guardian ( ) 10. Living in a group home ( ) 11. Living in a nursing home or extended care facility ( ) 12. Other Of the following seven (7) choices, check the one (1) that is most true for you. Without other financial help, I earned enough money on my job to support myself and at least one other person (even if there is no other person). Without other financial help, I earned enough money on my job to support myself. Worked in a paying job and earned some money but not enough to completely support myself. Worked in a sheltered workshop or vocational training program that pays some salary. Received job training but no salary. Worked as a volunteer with no salary. Did not work for training or money. 13. Over 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 121 Please answer yes for each of the following statements that is true for you at least once over the last seven (7) days. the last seven (7) days did you... Work in exchange for room and board Work in exchange for clothing, etc. Do formal volunteer work Drive a car, motorcycle, truck, etc. Use public transportation (e.g. bus) Prepare a meal for yourself or a friend Attend a high school or college class Attend a club meeting Make a telephone call Talk to someone who called you Write a letter Pay a bill by mail Mail a letter Read a book Read a newspaper Sit and think Knit, crochet or sew something Work on a hobby Listen to the radio or stereo Watch television/VCR Play cards, pool or other games Budget money for the week Write a check or money order Cash a check Yes ( No Over the last 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 122 seven (7) days did you... Yes No Purchase or pay for something costing more than $35.00 (e.g. a bag of groceries) ( ) )SC Purchase or pay for something costing $5.00 to $35.00 ( ) )SC Purchase or pay for something under $5.00 ( ) )SC Have a major responsibility for the physical well being and appearance of children, elderly, or sick persons living in your home ( ) )SC Launder or iron clothing ( ) )SC Prepare a meal for a dependent or spouse ( ) )SC Plan meals ( ) )SC Purchase groceries for a few days ( ) )SC Vacuum. mop, sweep, or dust at home ( ) )SC Repair a car, appliance, etc. at home ( ) )SC Paint, hang wallpaper, mow a lawn, shovel snow, other maintenance work at home ( ) )SC Go to a food store (bakery, deli) ( ) )SC Go to a movie, concert, or theater ( ) )PA Go to church or synagogue ( ) )PA Go to the bank and deposit/withdraw money ( ) )SC Go to the post office ( ) )SC Go to the library ( ) )PA Go to a drug store ( ) )SC PA Buy a meal in a restaurant ( ) )SC PA Go to a tavern ( ) )PA Go to a beauty parlor or barber shop ( ) )SC PA Go to a department or hardware store ( ) )SC PA Go to a doctor, dentist, lawyer, or other professional ( ) )SC 123 Over the last seven (7) days did you... 48. Have a visit from a doctor, dentist, lawyer or other professional ( 49. Go to a party at someone else’s home ( 50. Take a pleasure walk ( 51. Play with children ( 52. Visit friends ( 53. Have a party in your home ( 54. Get together with friends ( 55. Start a conversation ( 56. Help someone who needed help or directions ( 57. Play golf, tennis, bowling, softball, go skiing, jogging, etc. ( 58. Take a vacation ( Currently... 59. Do you have a very good friend? ( 60. Do you have friends who are not close, but with whom you get together and do things? ( 61. If unmarried, do you have an intimate friend with whom you have a sexual relationship? ( 62. Are you a member of a club/organization? ( 63. Are you a member of a church or synagogue? ( 64. Do you have apy credit cards? ( 65. Do you have a major loan (e.g. car loan) or mortgage? ( 66. Are you the owner or co-owner of property (real estate)? 67. Do you own or lease a car? ( 68. Do you have a check cashing card at a supermarket? APPENDIX B CLIENT SCHOOL AND THERAPY QUESTIONNAIRE Are you currently enrolled in school or a training program? Yes Please specify. No During the last 6 months, how often were you in a medical hospital? During the last 6 months, what was the total length of time you were in a medical hospital? days weeks months During the last 6 months, how often were you in a psychiatric hospital? During the last 6 months, what was the total length of time you were in a psychiatric hospital? days weeks months During the last 6 months, I have spent time with some members of my family (check the answer which applies most): Less than once a month. Once or more a month. Once or more every 2 weeks. Once or more a week. Every day. 124 125 7. Please check all items that apply: I am currently involved in: Individual psychotherapy or counseling. Couple or family psychotherapy or counseling. Group psychotherapy or counseling. A day treatment center or rehabilitation program. A companionship program such as COMPEER 8. ‘When answering this item,