The clubhouse as an empowering setting.
The first clubhouse, Fountain House in New York City, was founded in 1948 by a group of former patients from a nearby state hospital to provide refuge, support, and rehabilitation to adults with psychiatric disabilities (Macias, Jackson, Schroeder, & Wang, 1999). As of 2000, there were at least 350 clubhouses worldwide, in 44 U.S. states and 21 other countries (Wang, Macias, &Jackson, 1999), including Japan, Korea, Pakistan, South Africa, Germany, Sweden, and Australia, serving an estimated 25,000 psychiatric consumers ("Gold Award: The Wellspring of the Clubhouse Model," 1999; Lucca, 2000; Mastboom, 1992). Despite this proliferation and the large numbers of individuals served, surprisingly little research has been conducted on the operation of the clubhouse model. For example, Stein and colleagues (1999) noted that characteristics of consumers using these programs have not been systematically assessed. Macias and colleagues (1999) voiced concern over variations in descriptions of the clubhouse model in published literature, even in regard to services provided. Macias and colleagues (2001) raised the question of whether external constraints (for example, lack of Medicaid funding) or organizational resources might affect fidelity of a clubhouse to the intended model's principles and operations.
The lack of quantitative data on clubhouse operations has been remedied, to some extent, by a series of recent publications reporting findings from data collected through the clubhouse certification process. However, the clubhouse standards, which undergird certification, focus primarily on the model's structural characteristics (that is, administration, staffing, position descriptions, eligibility criteria, rules, space, services, and activities) rather than on its process aspects (for example, relationships between staff and members, how members are treated, hierarchies, opportunities, personal development).
LITERATURE REVIEW: WHAT IS KNOWN ABOUT CLUBHOUSES?
Characteristics of Clubhouse Users
The literature reviewed ranges from studies of members in a single clubhouse (Accordino & Herbert, 2000; Kosciulek & Merz, 2001; Stein et al., 1999; Warner, Huxley, & Berg, 1999) to surveys of directors from numerous clubhouses (that is, 40 in Mastboom, 1992; 128 in Macias et al., 2001; and 173 in Macias et al., 1999) (Table 1). The number of members covered by the studies ranges from 26 to more than 20,000. Despite the diversity of the samples, some commonalities emerge. Female members are consistently in the minority--usually around 40 percent to 45 percent. Members of ethnic minority groups account for from 10 percent to 41 percent of clubhouse populations. The typical age of clubhouse members seems to be late 30s to early 40s; members 50 years and older are less prevalent. In terms of diagnosis, schizophrenia or schizoaffective and other psychotic disorders are the largest single group in all the studies, with percentages ranging from 40 to 78. The other diagnoses mentioned include major depression and bipolar disorder. A few studies described the percentage of members who were working as ranging from 24 percent to 58 percent, which appears comparable to employment rates reported in recent multisite and experimental studies (for example, 56 percent in Bond, Dietzen, McGrew, & Miller, 1995; 53 percent in Rogers, Anthony, Toole, & Brown, 1991).
Characteristics of Clubhouse Operations
In these studies, the number of clubhouses surveyed varied from 39 to 173 (Table 1). Most of the clubhouses studied were about 10 years old. The number of active members in a clubhouse varied widely across the surveys from 25 (Mastboom, 1992) to more than 500 (Macias et al., 2001). Daily attendance was only about 40 percent of the active members. Clubhouses were typically open about 40 hours a week. The staff to member ratios ranged from .049 to. 147, or slightly under 1:10. The surveyed clubhouses reported spending about $3,500 per member per year.
Clubhouses as Empowering Settings
PSR practice has as a core value the empowerment of consumers, so that they are enabled to take control of their lives, identify their goals, voluntarily pursue goals, have available choices to attain goals, reduce reliance on professionals, and independently make decisions that are goal and value congruent (Accordino & Herbert, 2000; Dickerson, 1998). Thus, the concept of an empowering setting should be relevant to all PSR programs, including clubhouses. Zimmerman and colleagues (1992) characterized empowering organizations as those that assist individual members to feel empowered. Maton and Salem (1995) used a case study approach to identify organizational attributes of community settings that make them empowering for their members, that is, assisting members to gain resources or competencies needed to increase control over their lives and accomplish significant life goals.
Empowerment is an important component of what should go on in clubhouses, according to several authors: "Clubhouses usually have an egalitarian hierarchy that promotes consumer empowerment" (Accordino & Herbert, 2000, p. 269). Clubhouse principles seem very congruent with empowerment, for example, making members feel that the clubhouse belongs to them and that they are needed and wanted as contributors (Warner et al., 1999) and improving their sense of competence and control. Jackson's (2000) book on the clubhouse model, subtitled Empowering Applications of Theory to Generalist Practice, described how the model is designed to empower and provided examples of empowering acts that can be initiated in clubhouses (for example, including members in decisions about club operations, using member skills and aptitudes to carry out the work of the clubhouse, giving members keys to club property for which they have responsibility, and promoting expectations that members will contribute).
Using data collected from an in-depth, on-site survey of a statewide sample of clubhouses in Michigan (N = 31), supplemented by state and federal government sources, we addressed the following four research questions:
1. What are the characteristics of clubhouse users?
2. What characterizes clubhouse operations and their community context?
3. To what extent can clubhouses be characterized as empowering settings?
4. What predicts variations among clubhouses in their empowering features--constraining variables related to community context (that is, household income, crime rate, and so forth) and client characteristics (that is, demographic and impairment variables) or variables reflecting program resource levels (that is, budget, staffing ratio, and so forth) and organizational attributes, such as those describing staff, size, and age of the program?
Out of 43 clubhouses in Michigan, we selected 31 from all regions of the state, about evenly split between rural and urban-suburban locations. A research team of three trained project staff members made in-person site visits to each clubhouse. Data collection during visits included a semistructured interview of the director, which typically lasted about an hour and a half (Mdn = 99 minutes), onsite observations of members and their activities, and data collection from members.
Measures of Empowerment: Dependent Variables
The dependent variables reflected diverse aspects of agency operations, congruent with the multifaceted construct of empowerment. The three variables used in this analysis were member involvement in program operations and decision making, support and problem-solving assistance provided by the program, and total number of specialized services offered.
Member Involvement. The degree of consumer involvement in operations and decision making was scored using director responses to six items (Who makes up the budget? Who makes decisions about changing the budget? Who signs contracts and agreements? Who signs checks to pay bills? Who hires the director? and Who makes other decisions about hiring staff?.). Responses were coded using a five-point Likert scale measuring degree of consumer involvement in decision making: 1 = decision making outside the center, 2 = input from the center, but outside funder approves, 3 = director/ staff, 4 = board and director, 5 = board; standardized [alpha] = .67) (Mowbray, Robinson, & Holter, 2002).
Support and Problem-Solving Assistance. Our measure of support and problem-solving assistance consisted of items taken from the Reasons to Come Scar (Mowbray & Tan, 1993), a 19-item questionnaire asking directors the reasons that consumers attend the program (for example, to get support from staff). The following eight Reasons to Come items were rated on a scale ranging from 1 = no one to 4 = most people: (1) to take part in classes or discussion groups? (2) to find out about other kinds of help? (3) to get support from staff? (4) to get support from other consumers? (5) to get help finding a job? (6) solving a problem? (7) finding a place to live? (8) with their mental health issues? A mean score was calculated to represent degree of support and problem-solving assistance that the clubhouse provided (standardized [alpha] = .61).
Total Specialized Services. This variable consisted of the total number of specialized services, from a list of 14--food bank, meals, washer and dryer, showers, telephone, temporary housing or shelter, help locating housing, help finding jobs, transportation to activities or appointments, clothes, a mailing address, supported education, scheduled discussion groups, and advocacy on behalf of the consumers-that the director reported the program provided. In that empowering agencies offer services to meet individualized needs, it was thought that clubhouses with a greater diversity of services would be more empowering.
The predictor variables were chosen from organizational literature, reflecting measures that typically relate to the performance of nonprofit agencies (Daft, 1998). Predictors included variables reflecting community context, consumer characteristics, program resources, and internal organizational characteristics.
Community Context Variables. Community-level variables were gathered from several different sources. Crime rate (number of crimes committed per population served) was taken from the Uniform Crime Report (Michigan State Police, 2001). Median household income for the county was taken from the U.S. Census Bureau; information from the state funding agency was used to calculate variables such as community mental health (CMH) services cost per capita. The urban-rural designation of the area in which the clubhouse was located came from the Michigan Department of Community Health. (Rural and urban distinctions for the city or county area in which each clubhouse was located were made according to categorizations provided by the Michigan Department of Community Health.)
Consumer Characteristics. Interviewers asked directors about the characteristics of each agency's service recipients, such as race and gender. In addition, percentage of consumers with major disability (overtly symptomatic, with symptoms constituting an impairment of the individual's functioning, or being disruptive to others) was determined on the basis of field staff observations. Age of consumers came from interviews conducted with consumers by on-site field staff, as did the consumers' self-reported diagnosis.
Program Resource Levels. Information about program resources, such as staffing, budget, and expenditures, was obtained from directors and from agency documents. Information gathered was used to calculate variables such as client to staff ratios and hourly total budget per consumer. Directors were also asked about their working relationship with the local CMH administrative board, using 10 five-point Likert-scale items (for example, "How helpful was CMH in assisting your clubhouse to obtain its goals?" "How well informed are you about the specific goals and services of CMH?" with scale anchors ranging from 1 = not at all to 5--a lot; standardized 0t = .76).
Internal Organizational Characteristics. Directors reported characteristics of program operations, such as the percentage of regular attendees, number of people served in a typical day, and duration of program operation. Also, directors were asked about the length of time they had been in their position and with the program. Director education level was measured using a six-point scale: 1 = less than high school, 2 = high school or equivalent, 3 = vocational or technical training, 4 = some college, 5 = graduated college, and 6 = graduate degree.
We first generated descriptive statistics for variables that characterized clubhouse users (for example, race), clubhouse operations (for example, daily attendance), and clubhouses as empowering settings (for example, member involvement in program operations or decision making). The second part of the analysis focused on predicting variations among clubhouses in their empowering features. In preliminary analyses, Pearson product-moment correlations were performed to identify significant bivariate relationships between empowering outcome variables (for example, member involvement in program operations or decision making) and potential correlates such as community context variables (for example, median family income), client characteristics (for example, race), program resources (for example, hourly budget per consumer), and organizational attributes of the clubhouse (for example, percentage of regular attendees, director characteristics). Relationships were checked for possible curvilinearity by inspecting scatterplots using lowess smooth (a nonparametric procedure that reflects departures from linearity) (Cleveland, 1993). Diagnosis was not examined as a predictor variable because of the large amount of missing data.
Correlates identified in preliminary analyses were entered in blocks into a hierarchical regression analysis, with a planned order of entry reflecting increased mutability and amenability to intervention. The hierarchical method allowed systematic assessment of the unique and joint contributions of each block. Regression analyses were performed separately for each of the three outcomes of interest--member involvement in program operations and decision making, support and problem-solving assistance, and total specialized services provided. In each model, a community context variable (for example, rural-urban) was entered as block 1 and a characteristic of consumers (for example, percentage with cognitive disability) was entered as block 2. This allowed us to capture the variability in agency empowerment associated with relatively immutable community and consumer characteristics. In block 3, we added clubhouse resources and organizational attributes. Thus, the analyses allowed us to determine the effect of resources and other organizational factors on the empowerment variables, controlling for the characteristics of their settings and the characteristics of the individuals they served. Variables that did not make an independent and significant contribution to the variation in empowering outcomes at each block were dropped from the model reported here. With our sample size of 31, we could detect as significant (two-tailed p < .05, power = .80) a single predictor accounting for at least 20 percent of the variance in the dependent variable, after accounting for the effects of up to four other control variables that collectively accounted for 10 percent of the variance (Cohen, 1988).
Concerning characteristics of clubhouse members, their average age (self-reported) was in the mid-40s, with a rather narrow range, from about 35 to 50 (Table 2). Clubhouse directors estimated, on average, that the centers served more men than women (55 percent compared with 45 percent) and that the vast majority of center members were white (83 percent), although this percentage varied markedly from less than 0.5 percent to 100 percent. According to field staff observations, fewer than one-third (29 percent) of consumers were characterized as having a major disability. From self-reported diagnosis data, the largest number of respondents had a schizophrenia diagnosis (nearly one-third), whereas about one-fourth had an affective disorder diagnosis (depression or bipolar disorder). About another one-fourth did not report their diagnosis, and the remainder mentioned another diagnostic category, a medical disorder, mental or emotional disorders in general, or some other (nonpsychiatric or health-related) reason for their problems.
The clubhouses in the sample were in urban (61.3 percent) as well as rural (38.7 percent) locations. In describing the economic context of these locations, state and census data indicated wide variations. On average, the crime rate of the city or county location for the clubs was .20 (20 crimes per 100 population), ranging from .03 to .83; the average for the state of Michigan was .11. The average median household income for the counties where clubs were located was about $43,000, which is just slightly above the median for the state ($41,963). Again, however, there was wide variation in this range across clubs, from about $27,000 to more than $65,000. Similarly, the CMH funding level per capita for each county averaged about $66 but varied from less than half of that average ($26.40) to more than double it ($148.20). The average across all CMH county agencies in Michigan was $86.17.
Data on clubhouse operations were obtained from the clubhouse director's survey. On average, these clubhouses were open a few hours more than the standard work week, but ranged from 25 to 65 hours per week. They served an average of about 34 members each day, varying from 12 to 90 across the sample. The mean staffing level was seven members per paid full-time-equivalent (FTE) staff position (ranging from about 3:1 to 15:1). The budget per consumer served per hour open, averaged across this sample of clubs, was $6.26, but ranged from about a third of that amount ($2.12) to about twice that amount ($12.36). On average, the clubs had been in operation nearly nine years (ranging from two to 24 years). The working relationship with the local CMH board (the funding authority) averaged 3.76 on a five-point scale, indicating a moderate level of involvement with the board, ranging from % little" (1.7) to "a lot" (4.8) across the clubs. As reported by directors, on average more than two-thirds of clubhouse members were regular attendees (ranging from 25 percent to 95 percent across clubhouses). Directors had been in their positions, on average, 4.22 years and had worked for the program an average of 6.1 years. The directors reported, on average, an educational level between college graduate and graduate degree. All the directors had at least some college education.
Examining descriptive statistics on the empowerment variables, we see that the Member Involvement Scale score was actually rather low (M = 1.61), indicating that most decision making about operations was made outside the clubhouse, but with club input on some items. On the other hand, the average score on the Reasons to Come Scale item Support/Problem-Solving Assistance was nearly three on a four-point scale, indicating that at least some of the members came specifically for this type of assistance. Of the eight items, those most highly scored were "to get support from staff," "to get support from other consumers" and "to get help with their mental health issues." For the final empowerment outcome variable, the count of total specialized services, the mean score indicates that the typical club offered slightly more than 11 of the 14 services; the minimum number offered was eight and the maximum was 14. Nearly all the clubhouses provided the following services: meals, telephone, transportation to activities and appointments, and help finding jobs. In addition, 80 percent to 90 percent of clubhouses provided mailing addresses for consumers and help locating housing.
As described in the analysis plan, the predictor variables with relationships (at p < .10) to each outcome variable (Table 3) were entered into hierarchical regression analyses.
The final model for member involvement had an adjusted [R.sup.2] of. 17 (Table 4). No variables from the community context block were significant. In terms of consumer characteristics, the field staff's estimate of the proportion of members with a major disability was a significant predictor at trend level (having more members with major disability related to less member involvement) as was, in the program resource block, the consumer to staff ratio (more consumers per staff, less member involvement).
The adjusted [R.sup.2] for the model predicting total specialized services was .48, and there were significant predictors from all three levels (Table 5). At the level of community context, the higher the crime rate, the greater the number of specialized services. At the level of consumer characteristics, controlling for crime rate, the lower the percentage of white consumers, the higher the number of specialized services. Finally, controlling for crime rate and percentage white, resource variables still made a highly significant contribution to the prediction of total specialized services: The greater the number of people served in a day and the greater number of years the director had been in his or her position, the higher the number of total specialized services.
From the community context block, being in an urban rather than rural environment predicted coming for support and problem-solving assistance when entered into the regression alone; however, once the block for program resources was added, urban-rural was no longer significant. None of the consumer characteristics were significant predictors (Table 6). From the resource category, years the director had been at the program was a significant predictor (the greater number of years the director had worked at his or her program, the more members who reportedly attended the program to get support and assistance with problem solving). Finally, the number of FTE paid staff was a trend-level predictor. Having more FTE staff was related to having more members who reportedly attended the program to get support and assistance with problem solving. This effect remained even when average daily attendance was added as a control for clubhouse size (not shown in Table 6).
What Are the Characteristics of Clubhouse Members?
The average age of members in our sample of clubhouses was mid-40s and generally consistent with other studies, although perhaps somewhat older. Also generally consistent with other research were the data on diagnoses, with schizophrenia being the most prevalent category of self-reported diagnosis in this sample (32.3 percent). In many other studies, this percentage is much higher (up to 78 percent). However, basing the percentage on only those participants who reported a psychiatric diagnosis, the percentage of members with schizophrenia in our sample increases to 55.4 percent--closer to other study results. Also, some literature suggests that reporting no diagnosis or some other (nonpsychiatric) basis for problems (which is rather high in this sample, approximately 35 percent) is more prevalent with schizophrenia diagnoses (Selten, Gernaat, Nolen, Wiesma, & van den Bosch, 1998). Less than a third of members appeared to have a major disability; this moderate level of disability should be helpful for clubhouse operations or could perhaps reflect the positive result of clubhouse membership; it is also congruent with current notions of recovery.
Concerning race and gender, congruent with the other studies reviewed, in this sample of clubhouses, women were in the minority and white consumers were in the majority, the latter percentage being similar to that reported for the white population of the state of Michigan (80.2 percent; U.S. Census Bureau, 2000). The underrepresentation of women in clubhouses is an area of concern, and so far it has not been explained. Epidemiological studies of mental illness prevalence in middle-aged adult populations do not reflect a disproportionate number of women compared with men. Perhaps their underrepresentation reflects the vocational focus of clubhouses, which may be seen as more relevant to men.
Although there are fewer studies on clubhouse operations available for comparison, we can examine some of our data in this area in reference to published studies. Years in operation of our clubhouses (average 8.68 years) is generally in the range of what has been reported in other studies (eight to 12 years). However, the clubhouses in our sample had fewer participants than those studied by Macias and colleagues (1999, 2001) or by Mastboom (1992), with an average daily attendance of 34.2. However, those studies reported number of active members, which could be inflated compared with the number that come on a daily basis. When numbers are adjusted by the percentage that come on a daily basis, the average number of members for clubhouses in our sample is about 50--still smaller, but somewhat closer to the numbers reported by others. Another explanation for the size difference may be that the Macias and colleagues (1999, 2001) sample consisted of certified clubhouses; clubhouses that can afford certification may necessarily be larger than those in our statewide sample, due to the cost of the surveying and training involved in certification.
The average staffing level in our club sample is similar to that reported by Mastboom (1992) at about one to seven (staff to consumers), but substantially different from averages reported by Macias and associates (1999, 2001) for certified U.S. clubs (between 1:14 and 1:20) and by Brekke and Test (1992) for a Wisconsin clubhouse (1:15). This may reflect the fact that larger clubs are more likely to be certified or to participate in research studies. In terms of other resources, the average annual budget for our clubs is, at $442,443, within the range reported in the literature. Adjusted for a typical work week of 40 hours, the average annual cost per member in our clubhouse sample is about $13,000, more than double the annual cost per member reported by Macias and associates (1999). Even adjusted for inflation over a two- to three-year period, the costs in our clubhouses are much higher, probably reflecting their enriched staffing ratio as well as the lowered efficiency of smaller organizations.
This study was unique in examining the community context of this clubhouse sample. The clubhouse locations appear to generally reflect what is typical of the state of Michigan in terms of median household income. The clubs are located in areas that have higher crime rates than the average for the state as a whole, perhaps due to the greater percentage of urban locations represented. They are also located in counties where the average per capita CMH funding (approximately $66) is lower than the state average (approximately $86). Examining correlations, we noted with surprise that the context variables (urban-rural, crime rate, median county income, CMH funding per capita) had little relationship to each other; only two trend-level correlations were found (median household income with CMH cost per capita and with urban-rural). However, urban-rural has significant relationships with other consumer and program variables: Being in an urban location relates to a lower percentage of white consumers and to a higher number of consumers being served daily.
Clubhouses as Empowering Settings?
Analyses of sample data produced a mixed response to this question. In terms of offering members opportunities and assistance in making their own decisions, the clubhouses, on average, seem to be doing well. That is, most clubhouses offered a wide variety of specialized services, beyond employment experiences and job search. In most clubhouses, the majority of members were seen as coming to the club because they get help and support in solving their problems (for example, finding a job or a place to live) rather than being placed or told what to do. This is similar to the "Opportunity Role Structure" of empowering settings, as described by Maton and Salem (1995), wherein programs offer, but do not dictate, activities that are useful to and requested by their members.
On the other hand, in terms of shared decision making in operating the program and responsibility for its governance, these clubhouses seem to offer much less. The survey responses indicate that, for the most part, governance activities, like making rules, hiring and firing staff, deciding on budget issues, allocating funds to meet members' needs, and so on, are not allocated to members in many clubhouses. Rather, this is a function of an outside board, an administrator outside of the club, the club director, staff, or all of these.
Although the statistical models (regression analyses) for predicting all three of the empowerment-related variables were significant, that for member involvement predicted the least amount of variance. This may be because of the rather restricted range of member involvement scores in this sample. The predictors in the model included proportion of consumers with major disability and staffing resources; there was less member involvement in governance and decision making in the clubhouse when there was a higher proportion with major disability and when there was a higher number of consumers per staff. The latter finding is counter to the club-house philosophy, which indicates that having too many staff available detracts from member initiatives and opportunities to practice important role-playing behaviors. In this sample of clubhouses, when there was a more intensive staffing level, members tended to be more involved in governance and in making decisions about their club.
Predictions for total specialized services and for the Reasons to Come Scale items have much higher proportions of variance accounted for than does member involvement. For total specialized services, significant predictors include a higher crime rate, lower percentage of white people, and a higher percentage of people served in a day--all indicative of more urban areas. Urban areas may allow clubs to offer more services because of service density (access to referral networks) or service availability, as well as having more jobs, more variety in living arrangements, and so forth. Another significant predictor for this variable is the director's experience (number of years in the position), which may reflect his or her ability to access these services.
The regression model predicting Reasons to Come is somewhat similar to that for total specialized services, except that the urban-rural variable is not significant. The director's experience is, once again, significant. At a trend level, the total number of FTE staff is also a significant predictor. When the clubhouse has more total staffing, more members are likely to come for problem solving and support. Again, this runs counter to clubhouse philosophy.
The limitations of this study should be acknowledged. First, this was a statewide, not a national, clubhouse sample, and it therefore may reflect factors particular to the mental health system in Michigan. Second, the sample size was rather small (N = 31). Although we had adequate power to detect rather large size effects, the power to detect small effects was limited; that is, to detect as significant (at a two-tailed p < .05 and power = .80, after accounting for the effects of up to four other control variables) a single predictor accounting for 10 percent of the variance, a sample size of 65 would be needed; a sample size of 136 would be needed for a predictor accounting for 5 percent (Cohen, 1988). Finally, much of the data for this study came from reports of the clubhouse directors. Although the questions were fairly straightforward and minimized judgments, directors' biases could have affected their answers, for example, in terms of exaggerating the services and benefits of their clubs.
Implications for Social Work Practice
Social work is the major profession providing services in public mental health systems, which serve adults with PSR needs and therefore should be concerned with the extent of empowerment in these service settings. This study suggests that in the clubhouse the concept of an empowering setting has several dimensions. On the organizational level, clubhouses are clearly not empowerment settings in terms of governance or client control. At the level of having services, opportunities, and supports, though, the clubhouses seem to be operating more according to expectations. However, there are wide variations in the empowerment variables as well as in their potential predictors. Variations in clubhouse empowerment indicate independent contributions from factors reflecting their locations and their agency resources. Only for member involvement did we see perceived disability as a significant negative predictor.
The results suggest that clubhouses could do more to enhance member involvement in governance and decision making, such as rules and their administration, making decisions about resource allocations, and so forth. They also suggest that clubs being located in proximity to opportunities, services, and other resources may be beneficial to their members. Those staff working in rural areas may need to exercise more creativity in finding appropriate resources for members. Staffing also emerged as a significant predictor of empowerment in two ways. Directors with more experience appeared to be more able to access resources for members and to ensure a supportive and problem-solving-oriented environment. The results concerning staffing levels appeared to contradict the clubhouse philosophy. If this philosophy is to be actualized, perhaps staff in these clubhouses need more training in how to restructure operations so members make more decisions independently. The data suggest that consumer control is taking place only when there are sufficient staff to direct it.
More research is needed on the operation of the clubhouse model and its many variations and adaptations in different settings and with differing member characteristics. This should be a first step before examining the extent to which operational and member characteristic variations relate to differences in process and outcome variables.
Original manuscript received April 20, 2004
Final revision received September 27, 2004
Accepted November 22, 2004
Accordino, M. P., & Herbert, J.T. (2000). Treatment outcome of four rehabilitation interventions for persons with serious mental illness. Journal of Mental Health Counseling, 22(3), 268-282.
Bond, G. R., Dietzen, L. L., McGrew, J. H., & Miller, L. D. (1995). Accelerating entry into supported employment for persons with severe psychiatric disabilities. Rehabilitation Psychology, 40, 75-94.
Brekke, J. S., & Test, M. A. (1992). A model for measuring the implementation of community support programs: Results from three sites. Community Mental Health Journal, 28(3), 227-247.
Cleveland, W. S. (1993). Visualizing data. Summit, NJ: Hobart Press.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.
Daft, R. L. (1998). Organization theory and design (6th ed.). Cincinnati: South-Western College Publishing.
Original manuscript received April 20, 2004 Final revision received September 27, 2004 Accepted November 22, 2004
Dickerson, F. B. (1998). Strategies that foster empowerment. Cognitive and Behavioral Practice, 5, 255-275.
Gold Award: The wellspring of the clubhouse model for social and vocational adjustment of persons with serious mental illness. (1999). Psychiatric Services, 50, 1473-1476.
Jackson, R. L. (2000). The clubhouse model: Empowering applications of theory to generalist practice. Florence, KY: Wadsworth.
Kosciulek, J. F., & Merz, M. (2001). Structural analysis of the consumer-directed theory of empowerment. Rehabilitation Counseling Bulletin, 44, 209-216.
Lucca, A. M. (2000). A Clubhouse Fidelity Index: Preliminary reliability and validity results. Mental Health Services Research, 2, 89-94.
Macias, C., Barreira, P., Alden, M., & Boyd, J. (2001). The ICCD Benchmarks for Clubhouses: A practical approach to quality improvement in psychiatric rehabilitation. Psychiatric Services, 52, 207-213.
Macias, C., Jackson, R., Schroeder, C., & Wang, Q. (1999). What is a clubhouse? Report on the ICCD 1996 survey of USA clubhouses. Community Mental Health Journal, 35, 181-190.
Mastboom, J. (1992). Forty clubhouses: Model and practices. Psychosocial Rehabilitation Journal, 16, 9-23.
Maton, K. I., & Salem, D. A. (1995). Organizational characteristics of empowering community settings: A multiple case study approach. American Journal of Community Psychology, 23, 631-656.
Michigan State Police. (2001). Uniform crime report (43rd ed.). East Lansing: Michigan State Police Criminal Justice Information Center.
Mowbray, C.T., Robinson, E.A.R., & Holter, M. C. (2002). Consumer drop-in centers: Operations, services, and member involvement. Health & Social Work, 27, 248-261.
Mowbray, C.T., & Tan, C. (1993). Consumer-operated drop-in centers: Evaluation of operations and impact. Journal of Mental Health Administration, 20, 8-19.
Oliver, J, Huxley, P., Bridges, K., & Mohamad, H. (1996). Quality of life and mental health services. New York: Routledge.
Rogers, E. S., Anthony, W., Toole, J., & Brown, M. A. (1991). Vocational outcomes following psychosocial rehabilitation: A longitudinal study of three programs. Journal of Vocational Rehabilitation, 1, 21-29.
Rosenfield, S., & Neese-Todd, S. (1993). Elements of psychosocial clubhouse program associated with a satisfying quality of life. Hospital and Community Psychiatry, 44, 76-78.
Selten, J. P., Gernaat, H. B., Nolen, W. A., Wiesma, D., & van den Bosch, R.J. (1998). Experience of negative symptoms: Comparison of schizophrenic patients to patients with a depressive disorder and to normal subjects. American Journal of Psychiatry, 155, 350-354.
Stein, L. I., Barry, K. L., Van Dien, G., Hollingsworth, E.J., & Sweeney, J. (1999). Work and social support: A comparison of consumers who have achieved stability in ACT and clubhouse programs. Community Mental Health Journal, 35, 193-204.
U.S. Census Bureau. (2000). Statistical abstract of the United States. Washington, DC: U.S. Government Printing Office.
U.S. Department of Health & Human Services. (1999). Mental health: A report of the Surgeon General (DHHS Pub No. [MHEALTH] 18779). Washington, DC: Author.
Wang, Q., Macias, C., & Jackson, R. (1999). First step in the development of a clubhouse fidelity instrument: Content analysis of clubhouse certification reports. Psychiatric Rehabilitation Journal, 22, 294-301.
Warner, R., Huxley, P., & Berg, T. (1999). An evaluation of the impact of clubhouse membership on quality of life and treatment utilization. International Journal of Social Psychiatry, 45, 310-320.
Zimmerman, M. A., Israel, B. A., Schulz, A., & Checkoway, B. (1992). Further explorations in empowerment theory: An empirical analysis of psychological empowerment. American Journal of Community Psychology, 20, 707-727.
Carol T. Mowbray, PhD (deceased), was professor, Lisa Lewandowski, PhD, is research associate; and Mark Holter, PhD, is assistant professor, School of Social Work, University of Michigan, Ann Arbor. Deborah Bybee, PhD, is professor, Department of Psychology, Michigan State University, East Lansing. Address correspondence concerning this article to Dr. Lewandoski, Institute of Social Research, Survey Research Center, 3310 ISR, P.O. Box 1248, 426 Thompson Street, Ann Arbor, MI 48106-1248; e-mail: firstname.lastname@example.org. This research was supported through National Institute of Mental Health grant R24-MH51363 to the University of Michigan School of Social Work, Center for Research on Poverty, Risk, and Mental Health.
Table 1: Characteristics of Clubhouse Members and of Clubhouse Operations, by Study Member Characteristics % % Author N Female White Accordino & 26 35 -- Herbert (2000) Kosciulek & 159 36 59 Merz (2001) Macias et al. 20,178 45 76 (1999) Macias et al. 3,692 44 79 (2001) Mastboom (1992) 4,373 42 -- Oliver, Huxley, 38 24 84 Bridges, & Mohamad (1996) Rosenfield & 157 40 90 Neese-Todd (1993) Stein et al. 38 46 88 (1999) Warner et al. 38 26 84 (1999) Member Characteristics % Not Age Author Diagnosis Married (Years) Accordino & 40%, Sz/psychosis; -- -- Herbert (2000) 32%, MD; 16%, Bip; 12%, other Kosciulek & -- 96 M = 40.0 Merz (2001) Macias et al. -- -- <35, 28%; 35-50, (1999) 53%; >50, 19% Macias et al. 52% Sz -- -- (2001) 65%, Sz/psychosis; -- <30, 28%; 31-40, Mastboom (1992) 12%, Bip; 9%, 38%; 41-50, MD; 13%, other 20%; >50,14% Oliver, Huxley, -- 86 M = 40.2 Bridges, & Mohamad (1996) Rosenfield & Most Sz 94 Most Neese-Todd (1993) late 30s Stein et al. 78% Sz/Szaff 92 M = 42 (1999) (SD = 9.3) Warner et al. -- 66 M = 40.2 (1999) (SD = 6.5) Member Characteristics Education % Author (Years) Working Accordino & -- 58 Herbert (2000) Kosciulek & M = 12.56 33 Merz (2001) Macias et al. -- -- (1999) Macias et al. -- 37 (2001) Mastboom (1992) -- -- Oliver, Huxley, -- 47 Bridges, & Mohamad (1996) Rosenfield & Most 24 Neese-Todd (1993) completed high school Stein et al. M = 14.2 -- (1999) (SD = 2.5) Warner et al. -- 47 (1999) Member Characteristics Hours Wage Author Working per Hour Accordino & -- -- Herbert (2000) Kosciulek & Merz (2001) Macias et al. -- -- (1999) Macias et al. -- -- (2001) Mastboom (1992) -- -- Oliver, Huxley, M = 14 M = $4.98 Bridges, & Mohamad (1996) Rosenfield & -- -- Neese-Todd (1993) Stein et al. -- -- (1999) Warner et al. M = 19.8 M = $5.28 (1999) Clubhouse characteristics Age No. Active Author No. Clubs (years) Members Macias et al. (1999) 173 certified M = 8.5 M = 118 (SD = 5.8) (SD = 70) Macias et al. (2001) 71 certified M = 11.8 M = 144 (SD = 4.6) (SD = 80) Mastboom (1992) 39 M = 9.7 M = 68 Clubhouse characteristics Daily No. Weekly Staffing Author Attendance Hours Ratio Macias et al. (1999) 41% M = 47.6 .078 (a) (SD = 11.4) (SD = .037) Macias et al. (2001) 36% M = 39.3 1:20.23 (b) (SD = 19.27) Mastboom (1992) -- -- 1:6.8 (b) Clubhouse characteristics BA-level Annual Annual cast Author Staff Budget per Member Macias et al. (1999) 50% M = $408,327 M = $3,66 (SD = $288,483) (SD = $1,852) Macias et al. (2001) -- M = $484,375 M = $3,364 (SD = $279,966) Mastboom (1992) -- -- -- Notes: Dashes indicate that the information was not reported by the author(s). Sz = schizophrenia, Bip = bipolar disorder; MD = major depression, Szaff = schizoaffective disorder. (a) Member to staff ratio (b) Staff to member ratio Table 2: Descriptive Statistics for Empowerment and Program Descriptor Variables in Clubhouses (N = 31) Continuous Variables Variable M SD Empowerment variables (dependent variables) Member involvement 1.61 0.42 Support and 2.97 0.36 problem-solving assistance Total specialized 11.16 1.85 services Program descriptors (predictor variables) 2001 crime rate (a) 0.20 0.19 Median household income $42,970 $10,028 (county) CMH funding per capita $66.16 $31.74 (county population) % consumers with major 29.01 15.14 disability Mean age of consumers 44.37 2.68 interviewed % white consumers 83.37 16.57 % female consumers 45.23 12.70 Hourly total budget $6.26 $2.54 per consumer Total weekly operating 42.10 10.23 hours No. consumers served 34.23 17.27 daily Total no. FTE staff 5.74 3.83 No. consumers per staff 7.08 3.02 CMH involvement (b) 3.76 0.68 Years program in 8.68 4.01 operation % regular attendees 67.94 19.22 Years director in 4.22 3.47 position Years director with 6.10 4.36 program Director's highest 5.52 0.63 grade in school Variable Min Max Empowerment variables (dependent variables) Member involvement 1.00 3.17 Support and 2.00 3.63 problem-solving assistance Total specialized 8.00 14.00 services Program descriptors (predictor variables) 2001 crime rate (a) 0.03 0.83 Median household income $26,737 $64,705 (county) CMH funding per capita $26.40 $148.20 (county population) % consumers with major 0.00 58.00 disability Mean age of consumers 35.57 49.96 interviewed % white consumers 49.00 100.00 % female consumers 0.00 60.00 Hourly total budget $2.12 $12.36 per consumer Total weekly operating 25.00 65.00 hours No. consumers served 12.00 90.00 daily Total no. FTE staff 1.27 16.00 No. consumers per staff 2.86 14.67 CMH involvement (b) 1.70 4.80 Years program in 2.00 24.00 operation % regular attendees 25.00 95.00 Years director in 0.25 12.00 position Years director with 0.58 15.00 program Director's highest 4.00 6.00 grade in school Categorical Variable n. % Program descriptors Urban-rural Urban 19 61.3 Rural 12 38.7 Diagnosis (c) Schizophrenia or schizoaffective disorder 308 32.3 Bipolar disorder 138 14.4 Depression 91 9.5 Anxiety-related disorders 19 2.0 Mental and emotional disorder 69 7.2 Medical disorder 56 5.9 Other disorder or problem 42 4.4 Missing, none, not reported 234 24.5 Note: CMH = community mental health; FTE = full-time equivalent. (a) From the Uniform Crime Report, number of crimes reported per population base. (b) Based on 10 items, scale ranging from t = "not at all" to 5 = "a lot." (c) Diagnosis was not used as a predictor of outcome variables, nor was it included in the correlation matrix, owing to the large amount of missing data on this variable. Table 3: Intercorrelation Matrix of Empowerment Variables and Potential Predictors for Clubhouses (N = 31) Variable 1 2 3 1. Member involvement -- 2. Support and .12 -- problem-solving 3. Total specialized .12 .51 ** -- services 4. 2001 crime rate -.08 .15 .34 ([dagger]) 5. Median household -.02 .34 .22 income 6. CMH cost per capita .05 .08 .23 7. Urban-rural -.03 .33 .29 (5 = urban/l = rural) 8. % consumers with major -.34 -.06 .03 disability 9. Mean age of consumers -.00 .09 -.01 10. % of white consumers .16 -.31 ([dagger]) -.41 * 11. % of female consumers .04 -.15 -.30 12. Hourly total budget -.16 -.03 .17 per consumer 13. Total weekly .18 .38 * .14 operating hours 14. Total no. FTE staff .03 .49 ** .51 15. No. consumers served -.08 .29 .42 * daily 16. No. consumers per -.40 * -.30 ([dagger]) -.23 staff 17. CMH involvement .10 -.16 .13 18. Years program in .10 .27 .53 ** operation 19. % of regular -.14 -.11 -.12 attendees 20. Years director in .24 .38 * .52 ** position 21. Years director with -.04 .50 ** .36 * program 22. Director highest -.16 .13 .42 * grade in school Variable 4 5 6 1. Member involvement 2. Support and problem-solving 3. Total specialized services 4. 2001 crime rate -- 5. Median household -.16 -- income 6. CMH cost per capita -.04 -.34 ([dagger]) -- 7. Urban-rural -.05 .31 ([dagger]) .23 (5 = urban/l = rural) 8. % consumers with major .25 .19 -.05 disability 9. Mean age of consumers .07 -.08 .12 10. % of white consumers -.02 -.15 -.23 11. % of female consumers -.15 -.21 -.08 12. Hourly total budget .01 .07 -.17 per consumer 13. Total weekly -.00 .14 .04 operating hours 14. Total no. FTE staff .22 .09 .11 15. No. consumers served .02 .16 .22 daily 16. No. consumers per .06 -.18 .07 staff 17. CMH involvement .08 .13 .02 18. Years program in .10 -.03 .17 operation 19. % of regular .12 .08 .26 attendees 20. Years director in .01 -.05 .36 * position 21. Years director with -.13 .02 .26 program 22. Director highest .13 .35 .14 grade in school Variable 7 8 9 1. Member involvement 2. Support and problem-solving 3. Total specialized services 4. 2001 crime rate 5. Median household income 6. CMH cost per capita 7. Urban-rural -- (5 = urban/1 = rural) 8. % consumers with major -.06 -- disability 9. Mean age of consumers -.12 -.02 -- 10. % of white consumers -.59 *** .16 -.17 11. % of female consumers -.12 -.22 -.10 12. Hourly total budget -.03 -.05 .01 per consumer 13. Total weekly .27 -.22 -.06 operating hours 14. Total no. FTE staff .45 ** -.04 -.01 15. No. consumers served .52 ** .21 .01 daily 16. No. consumers per -.20 .41 * .02 staff 17. CMH involvement -.24 .14 -.33 18. Years program in .14 .07 -.01 operation 19. % of regular .18 .14 .18 attendees 20. Years director in -.00 -.01 -.09 position 21. Years director with .21 .05 -.11 program 22. Director highest .13 .50 ** -.21 grade in school Variable 10 11 12 1. Member involvement 2. Support and problem-solving 3. Total specialized services 4. 2001 crime rate 5. Median household income 6. CMH cost per capita 7. Urban-rural (5 = urban/l = rural) 8. % consumers with major disability 9. Mean age of consumers 10. % of white consumers -- 11. % of female consumers .12 -- 12. Hourly total budget -.13 .22 -- per consumer 13. Total weekly -.27 -.10 -.48 ** operating hours 14. Total no. FTE staff -.36 * .08 .03 15. No. consumers served -.31 -.07 -.21 daily 16. No. consumers per .18 -.22 -.30 staff 17. CMH involvement .28 -.14 -.07 18. Years program in -.36 * .01 .16 operation 19. % of regular -.15 -.21 -.04 attendees 20. Years director in -.19 -.17 -.14 position 21. Years director with -.27 -.08 .12 program 22. Director highest .09 -.04 .15 grade in school Variable 13 14 15 1. Member involvement 2. Support and problem-solving 3. Total specialized services 4. 2001 crime rate 5. Median household income 6. CMH cost per capita 7. Urban-rural (5 = urban/l = rural) 8. % consumers with major disability 9. Mean age of consumers 10. % of white consumers 11. % of female consumers 12. Hourly total budget per consumer 13. Total weekly -- operating hours 14. Total no. FTE staff .41 * -- 15. No. consumers served .28 .72 *** -- daily 16. No. consumers per -.39 * -.58 *** -.04 staff 17. CMH involvement .15 -.30 -.25 18. Years program in .20 .37 * .31 operation 19. % of regular -.23 -.21 -.20 attendees 20. Years director in .31 .22 .15 position 21. Years director with .21 .29 .10 program 22. Director highest .06 .26 .39 * grade in school Variable 16 17 18 1. Member involvement 2. Support and problem-solving 3. Total specialized services 4. 2001 crime rate 5. Median household income 6. CMH cost per capita 7. Urban-rural (5 = urban/l = rural) 8. % consumers with major disability 9. Mean age of consumers 10. % of white consumers 11. % of female consumers 12. Hourly total budget per consumer 13. Total weekly operating hours 14. Total no. FTE staff 15. No. consumers served daily 16. No. consumers per -- staff 17. CMH involvement .00 -- 18. Years program in -.22 .13 -- operation 19. % of regular .06 .05 -.12 attendees 20. Years director in -.20 .30 .60 * position 21. Years director with -.34 -.09 .53 ** program 22. Director highest -.04 .45 * .43 * grade in school Variable 19 20 21 1. Member involvement 2. Support and problem-solving 3. Total specialized services 4. 2001 crime rate 5. Median household income 6. CMH cost per capita 7. Urban-rural (5 = urban/l = rural) 8. % consumers with major disability 9. Mean age of consumers 10. % of white consumers 11. % of female consumers 12. Hourly total budget per consumer 13. Total weekly operating hours 14. Total no. FTE staff 15. No. consumers served daily 16. No. consumers per staff 17. CMH involvement 18. Years program in operation 19. % of regular -- attendees 20. Years director in .13 -- position 21. Years director with .05 .64 *** -- program 22. Director highest .08 .23 .12 grade in school Variable 22 1. Member involvement 2. Support and problem-solving 3. Total specialized services 4. 2001 crime rate 5. Median household income 6. CMH cost per capita 7. Urban-rural (5 = urban/l = rural) 8. % consumers with major disability 9. Mean age of consumers 10. % of white consumers 11. % of female consumers 12. Hourly total budget per consumer 13. Total weekly operating hours 14. Total no. FTE staff 15. No. consumers served daily 16. No. consumers per staff 17. CMH involvement 18. Years program in operation 19. % of regular attendees 20. Years director in position 21. Years director with program 22. Director highest -- grade in school Note: FTE = full-time equivalent; CMH = community mental health. ([dagger]) p [less than or equal to] .10. * p [less than or equal to] .05. ** p [less than or equal to] .01. *** p [less than or equal to] .001. Table 4: Hierarchical Regression: Final Model Predicting Clubhouse Member Involvement Using Consumer and Program Resource Variables (N = 31) Final Model Block [beta] t 1. Consumer characteristic: -0.19 -1.04 % major disability 2. Program resources: -0.36 -1.95 ([dagger]) No. consumers per staff Adjusted Adjusted Block [R.sup.2] [DELTA][R.sup.2] 1. Consumer characteristic: 0.08 ([dagger]) 0.08 ([dagger]) % major disability 2. Program resources: 0.17 * 0.09 ([dagger]) No. consumers per staff Note: Beta coefficients presented are standardized. ([dagger]) p [less than or equal to] .10. * p [less than or equal to] .05. Table 5: Hierarchical Regression-Final Model Predicting Total Number of Specialized Clubhouse Services Using Community Context, Consumer, Program Resource, and Internal Organizational Variables (N = 31) Final Model Block [beta] t 1. Community context characteristic: 0.33 2.49 * Crime rate 2001 2. Consumer characteristic: % white -0.24 -1.74 3. Program resources and internal organizational characteristics: No. people served in a day 0.28 1.99 ([dagger]) Years director in position 0.42 3.13 ** Adjusted Adjusted Block [R.sup.2] [DELTA] [R.sup.2] 1. Community context characteristic: 0.09 ([dagger]) 0.09 ([dagger]) Crime rate 2001 2. Consumer characteristic: % white 0.23 ** 0.14 * 3. Program resources and internal organizational characteristics: No. people served in a day 0.48 *** 0.25 ** Years director in position Note: Beta coefficients presented are standardized. ([dagger]) p [less than or equal to] .10. * p [less than or equal to] .05. ** p. [less than or equal to] .01. *** p [less than or equal to] .001. Table 6: Hierarchical Regression: Final Model Predicting Clubhouse Support/Problem-Solving Assistance Using Community Context, Program Resource, and Internal Organizational Variables (N = 31) Final Model Block [beta] t 1. Community context -0.09 -0.53 characteristic: Urban--rural (rural = 1, urban = 5) 2. Program resources and internal 0.34 1.95 ([dagger]) organizational attributes: No. FTE staff Years director at program 0.38 2.42 * Adjusted Adjusted Block [R.sup.2] [DELTA][R.sup.2] 1. Community context 0.08 ([dagger]) 0.08 ([dagger]) characteristic: Urban--rural (rural = 1, urban = 5) 2. Program resources and internal 0.32 ** 0.24 ** organizational attributes: No. FTE staff Years director at program Notes: Beta coefficients presented are standardized. FTE = full-time equivalent. ([dagger]) p [less than or equal to] .10. * p [less than or equal to] .05. ** p [less than or equal to] .01.
|Printer friendly Cite/link Email Feedback|
|Author:||Mowbray, Carol T.; Lewandowski, Lisa; Holter, Mark; Bybee, Deborah|
|Publication:||Health and Social Work|
|Date:||Aug 1, 2006|
|Previous Article:||Prevention and women's reproductive health: a matter of concern.|
|Next Article:||Significance of gender and age in African American children's response to parental victimization.|
|If you build it, they will come: creating a new clubhouse was an inclusive process.|