Sheltered Care: A Typology of Residential Facilities

Abstract

With the return of large numbers of ex-mental patients to the community, a new, privately owned system of sheltered care has developed for those individuals unable to establish an independent living situation for themselves or return to a family. Given a sample of 211 facilities in the State of California, five relatively independent dimensions were identified to characterize sheltered-care facilities: complexity, program orientation, control, support, and medical orientation. An analysis of facility score profiles generated six basic types of facilities. A typology of this nature can now be used as a descriptive model to examine placement and effectiveness issues related to the sheltered care environment.

Over the past three decades the mental health system in California has been transformed from a system dominated by the use of the mental hospital to one utilizing, whenever possible, some type of community-based sheltered care (Segal & Baumohl, 1982). This phenomenon has resulted in a return to the community of large numbers of ex-mental patients. Within this context and supported in part by federal funding, a new, privately owned system of sheltered care has developed for those individuals who are unable to establish an independent living situation or return to some type of family unit.

Sheltered care in the most general sense is a supervised residential living arrangement. The most traditional sheltered-care facilities have been the small, more intimate family care homes (McCoin, 1983) and the temporary and transitional halfway houses. During the last 20 years a large number of more permanent and structured board and care type facilities have developed. These latter facilities (licensed in numerous states) now provide residential services to increasing numbers of formerly institutionalized people or people who might now need institutionalization if this resource were unavailable.

Services included in sheltered care facilities, as well as the form of their delivery, appear to vary a good deal. Some facilities may pursue rehabilitation as a goal, others may not. Some adopt a medical orientation while others attempt to develop a large therapeutic community. These same facilities may also differ as to the restrictiveness of their environment, their size, and the composition of their resident population. While there have been various theoretical statements as to the function of these facilities, as well as to their general character (Lamb, 1976a&b), there has been little research geared to the development of a descriptive model that would permit the examination of the essential characteristics differentiating sheltered-care facilities of any type except for the work of the American Psychiatric Association (1982).

It is the purpose of this paper to develop a typology of community-based sheltered-care facilities based on empirically derived characteristics; a typology which could be used as a descriptive model in examining issues (e.g., placement, effectiveness) related to a given facility’s environment.

Method

Data collection within this study consisted of structured interviews with 211 operators of sheltered-care facilities in the State of California. Operators were interviewed as part of a larger study of both residents and operators of sheltered-care facilities (Segal & Aviram, 1978). Though the data were gathered in the 1970s, the breadth of facility characteristics in the sample appears to reflect the variance in characteristics in today’s facilities.

Sample

The sample is a self-weighting (Kish, 1985) representative sample of all sheltered-care facilities (i.e., family care, halfway house, board and care) servicing a nonretarded, mentally ill population between 18 and 65 years of age. At the time of the study (1973) there were approximately 1155 such facilities in the State of California. Of these facilities, 28% were family care homes that served 14% of the residents in sheltered care. Halfway houses constituted 2% of the facilities and served only 3% of the sheltered-care population. Board and care homes which developed in an unplanned, ad hoc manner, serviced 82% of California’s sheltered-care population and comprised 72% of the state’s facilities (Segal & Aviram, 1978). Current estimates based on an 85% growth rate found to be characteristic of facilities serving the developmentally disabled (Hauber, Bruininks, Hill, Latkin, & White, 1984), indicate that in 1984 there were approximately 3597 such facilities in California.

In order to obtain the sample, the state was divided into three master strata: Los Angeles County, the San Francisco Bay Area (and adjacent counties), and all other counties of the state. Facilities were stratified by size in both Los Angeles and the Bay Area and a sample was drawn probability proportionate to size. The operator of each selected facility was then interviewed. In the third stratum made up of “all other counties,” a cluster sample was designed using counties as primary selection units. Two counties from the northern and two from the southern part of the state were selected probability proportionate to size. From each pair, samples of facilities were selected also probability proportionate to size. Operators from these facilities were then interviewed. Of the 234 operators contacted 10% refused to participate in the study (Segal & Aviram, 1977).

Measures

Previous research devoted to developing an empirical taxonomy of environments has included work by Blake (1987), Price and Moos (1975), Moos (1975), and Barker (1968). Barker attempted to characterize naturally occurring community environments as “behavioral settings.” Salient variables that were considered by Barker to be descriptive of such settings included predominant behaviors in the setting, characteristics of setting inhabitants, setting duration and frequency, system characteristics, and population size. Price and Moos (1975) examined treatment environments and described them in terms of an assessment of interactions between setting personnel and clients and between clients themselves. Blake (1987), using Moos’ Multiphasic Environmental Assessment Procedure (MEAP), assessed the social environment of boarding homes describing their physical and architectural features, policy and program features, aggregate resident and staff characteristics, and social climate. From our point of view, a true descriptive taxonomy of sheltered care facilities should focus on the practical goal of optimizing the fit between individuals and their environments and therefore should include variables that reflect the full range of contextual factors influencing the outcome of individual behavior.

In this study operators responded to an extensive interview and assessment schedule. Included in the assessment were variables descriptive of the social psychological climate of the facility reflected in resident/operator and resident/resident interaction, the facility’s physical and organizational structure, its operating procedures and policies, its orientation to services, and its resident population.

The social-psychological climate of the facility was assesssed by operators’ responses to the Community Oriented Programs Environment Scale (COPES). The COPES consists of 102 forced-choice items appropriate to the description of community-based, mental-health programs (Moos, 1974). The items sample three general areas that describe aspects of the internal climate of a facility or program. The first area includes items that describe personal relationships within a facility, both resident to resident and staff to resident. The second area includes items that relate to the characteristics of the treatment program itself and the third, to the facility system maintenance efforts (e.g., its emphasis on order and organization with respect to resident behavior and facility operation).

Indices relating to the physical and organizational structure of the facility included information on staffing patterns, size of facility and its overall group structure (e.g., family vs. non-family oriented). Operating and policy variables relate, for instance, to whether or not curfews are exercised for some or all residents, whether or not there are procedures for the supervision of medications, and whether facilities employ special criteria for selecting and accepting prospective residents.

Information regarding a facility’s service program included whether or not the facility provided in-house services (for rehabilitation or therapy) or whether they relied on outside services (or did not make any provisions) and, in addition, their evaluation as to whether or not outside services of all kinds (and which ones) were useful. Finally, characteristics of the resident population referred to the age and sex composition of the resident group, the types of disabilities they possessed (since mixing of different categories of disability within facilities is often practiced), and residents’ history of hospitalization and length of stay in the present facility.

Analysis

Following preliminary analyses and utilizing traditional factor-analytic techniques, a selected sample of 36 characteristics of sheltered-care facilities were reduced to simple-sum factor scores on five factor dimensions. Each facility could then be described by a profile of factor scores over these five dimensions. On the basis of similarities and differences among facility profiles, a typological analysis was performed (Tryon & Bailey, 1970).

In this case, a typological analysis consists simply of the identification of the fewest, most distinct score patterns that emerge over the five factor dimensions. Each facility is “typed” by virtue of its similarity to one or another of the most frequently occurring patterns. The analysis then attempts to condense the resulting most-frequent-types into the fewest possible types that are still distinctly different to the analyst.

In terms of specific method, the profile of scores for each facility is reduced to a point in five-dimensional factor space. The resulting projection contains as many points as there are facilities. The typological analysis proceeds to partition this space into arbitrarily defined sectors and assigns each point to a single sector based on its profile of scores. The sectors that then contain significant numbers of points define what is referred to as core types and profiles not originally falling into core-type sectors are then assigned to the sector nearest to them in the factor space. (Here the five dimensional space was partitioned into sectors defined in terms of standard deviation units. Each dimension was divided into three parts: + s.d., −1 s.d., and between +1 and −1 s.d. The result is 3⁵ possible sectors. The criterion used for a sector to be considered as a core type was that it should contain at least 2% of the facilities. Distance scores were used to assign profiles to their nearest sector.)

The next step in analysis is to combine core types into even fewer types based on pattern similarities. The method used is the condensation method by which core types are combined beginning with the two most similar (again using distance measures as the similarity criterion) and proceeding two at a time (including previous combinations) until all types have been reduced to a single type—that is, until a flat, patternless profile is produced. The more condensation that is accepted, the more similar the resulting profiles become. The trade-off is parsimony for differentiation (and perhaps meaning). Condensation is totally under the control of the analyst.

Thus a typological analysis seeks to reduce the number of types to the fewest possible while still maintaining distinct, interpretable differences among the final emergent types. In this analysis 14 core types condensed to six final types.

Results

Dimensions

Using Tryon and Bailey’s (1970) method of factor analysis, five relatively independent dimensions were identified that characterize community-based, sheltered-care facilities regardless of the formal name they choose. These five dimensions are:

1. The degree of complexity or structure of the facility.

2. The program orientation of the facility.

3. The extent of control over residents exercised by the facility.

4. The degree of support residents are seen to offer each other.

5. The medical orientation of the facility.

Tables 1, 2, 3, 4, and 5 present the items that make up the factors, their factor loadings, the percent of communality (h²) accounted for by the factor and its reliability. Intercorrelations between factor scores are also included.

Table 1.

FACTOR I Facility Complexity Synonyms: Structure, Formalization, Scope

Item/scoring	Factor Coefficient	Factor Score Interpretation: (High Factor Scores =)
SERVICE USE Do residents use outside services? 5 level variable, High use = Low score	.71	Low use of Services
SIZE The sum of the number of males + females reported to be in the facility. Continuous variable, Large size = High score	−.63	Small sized facility
POLICE PICKUP How many residents were picked up by the police during previous year? 6 level variable, High pick up = Low scores	.61	Low number picked up by police
SERVES FEMALES Do you serve females (only)? Dichot variable. Yes = Low score	.58	Does not serve only females
STAFF Employ a staff? Dichot variable. Yes = Low score	.55	Does not employ a staff
SERVES BOTH SEXES Do you serve both …? Dichot variable. Yes = Low score	−51	Serves both males and females
MESSY Some residents look messy. Dichot variable, Yes = High score	−.51	Are not messy
RARELY ARGUE Residents here rarely argue. Dichot variable, Yes (rarely) = High score	.50	Rarely argue
FAMILY ORIENTATION Based on statements of purpose made by operators. Dichot variable, Yes (a family orientation) = Low score	−.43	Facility has family orientation
RARELY ANGRY Residents here rarely become angry. Dichot variable, Yes (rarely) = High score	.41	Rarely become angry

Factor statistics:

%h² 37% Alpha Reliability .82

Inter r factor scores = I/II .53 I/III −.21 I/IV .27 I/V .03

Table 2.

FACTOR II The Program Orientation of the Facility Synonyms: Outward/Future Directed, Youth Orientation, Active

Item/scoring	Factor Coefficient	Factor Score Interpretation: (High Factor Scores =)
PROGRAM Does the facility have (or make arrangements for) a rehab/therapy program? 4 level variable, Yes, it does = High score	−.65	Does not have program
AGE The reported age range of the residents divided by 2 (i.e. mean age in facility). Continuous variable, Older age = High score	65	Older residents
JOB TRAINING This program emphasizes training for new kinds of jobs. Dichot variable, Yes = High score	−.63	No, it does not emphasize job training
ACUTE Based on the number of residents reported to have been treated in a psych hospital during previous year. 3 level variable, High number treated = High score	−.63	Low number of “acutes”
SOCIAL ACTIVITIES This program has very few social activities. Dichot variable, Yes, it has few = High score	.54	Few social activities
FUTURE PLANS There is relatively little discussion about exactly what residents will be doing after they leave the facility. Dichot variable, Yes, little discussion = High score	.53	Little discussion concerning future outside
DISCUSS SEX Residents hardly ever discuss their sexual lives. Dichot variable, Yes, little discussion = High score	.51	Little discussion re own sex life
CHANGEABLE People are always changing their minds here. Dichot variable, Yes, changing = High score	−.39	They do not change their minds here

Factor statistics:

%h² 17% Alpha Reliability .80

Inter r factor scores = II/I .53 II/III −.22 II/IV .34 II/V −.05

Table 3.

FACTOR III The Controlling Aspect of the Facility Synonyms: Authoritarian/Permissive, Management vs Client-Centered

Item/scoring	Factor Coefficient	Factor Score Interpretation: (High Factor Scores =)
SCHEDULE Once a schedule is arranged for a resident, the resident must follow it. Dichot variable, Yes = High score	.70	Must follow the schedule
REGULARITY Residents here follow a regular schedule every day. Dichot variable, Yes = High score	.48	A regular schedule here
CURFEW Do your residents have a curfew? 3 level variable. Number for whom there is a curfew = Low scores	−.46	Has a curfew
PERMISSIVE Residents can generally do whatever they feel like. Dichot variable, Yes, do whatever = High score	−.42	Cannot do whatever they like

Factor statistics:

%h² 12% Alpha Reliability .60

Inter r factor scores = III/I −.21 III/II −.22 III/IV .06 III/V −.27

Table 4.

FACTOR IV Mutual Support and Assistance among Residents in the Facility

Item/scoring	Factor Coefficient	Factor Score Interpretation: (High Factor Scores =)
HELPING Members seldom help each other. Dichot variable, Yes, seldom help = High score	.63	Seldom help each other
CONCILIATORY It’s hard to get people to argue around here. Dichot variable, Yes, its hard = High score	.57	Hard to get an argument
ASSISTING The healthier residents here help take care of the less healthy ones. Dichot variable, Yes, they assist = High score	−.54	They do not assist
SHARING There is relatively little sharing among the residents. Dichot variable, Yes, little sharing = High scores	.48	Little sharing here
RARELY ARGUE Residents here rarely argue. Dichot variable, Yes, rarely = High scores	.52	Rarely argue
UNEXPRESSIVE When members disagree with each other, they keep it to themselves. Dichot variable, Yes, keep it to themselves = High score	.42	Keep it to themselves

Factor statistics:

%h² 10% Alpha Reliability .66

Inter r factor scores = IV/I .27 IV/II .34 IV/III .06 IV/V −.24

Table 5.

FACTOR V Medical Orientation Within the Facility

Item/scoring	Factor Coefficient	Factor Score Interpretation: (High Factor Scores =)
MEDICATIONS Do you or staff supervise the dispensing of medications? 3 level variable, High supervision (in terms of numbers of residents supervised) = Low scores	.63	Few if any are supervised for medication
HOUSE DOCTOR Have you arranged for a house doctor? Dichot variable, Yes = Low score	.54	No house doctor
CONCILIATORY It’s hard to get people to argue around here. Dichot variable, Yes, it’s hard = High score	−.43	Arguments are easy to come by
DAILY PLANNING Residents’ activities are carefully planned Dichot variable, Yes, planned = High scores	−.43	Activities not planned
LIFE FUNCTIONS How many (with a history of mental illness) need help to carry out basic life functions? 6 level variable, High frequency (number needing help) = Low scores	.33	Few if any need help with basic functions
UNEXPRESSIVE When members disagree with each other, they keep it to themselves. Dichot variable, Yes, keep it to themselves = High score	−.29	Disagreements are expressed

Factor statistics:

%h² 12% Alpha Reliability .64

Inter r factor scores = V/I .08 V/II −.05 V/III −.27 V/IV −.24

The most general dimension in sheltered-care facilities is complexity (I). The complex facility is likely to be larger, employ a staff, not to follow a family model for its structure, have a group composition that is likely to accommodate more troublesome behavior, and have a greater orientation toward using or accepting outside services. The second factor that emerged, and one closely related to the complexity of the facility, is the facility’s program orientation (II). The program-oriented facility is likely to be characterized by having a program catering particularly to younger residents. It is also likely to be future oriented, to emphasize job training and rehabilitation, to have many social activities and to generally serve an acute, recently hospitalized population. While these first two dimensions are moderately related (the correlation between factor scores is r = .53), the three remaining dimensions vary more or less independently of complexity and program orientation (see Tables 1 & 2, inter r factor scores). An orientation toward controlling residents, an emphasis on mutual support offered each other by residents, and a medical orientation (see Tables 3–5) are facility characteristics found either in simply or complexly organized facilities and also in facilities irrespective of the age of their residents or program orientation. Medical orientation refers to a facility organized to meet medical needs of residents, with a resident group that has such needs.

Typology

Six basic types of facilities were delineated by the analysis of score profiles generated by the five factor-analytically derived dimensions. The profile of facility type is given by its standardized, mean factor score across all five dimensions (see Figure 1).

Figure 1.

The six types can be described as follows:

The family care home.

This facility is not complex. It is small (average bed capacity for seven residents) and family oriented (I). It is oriented toward a more stable, older resident (the average age of residents is 46) and does not offer a formal program geared to moving residents out of the facility (II). Family care homes of this type are neither controlling nor directing of their residents (III) and tend to offer only a slightly supportive environment (IV). They are not medically oriented facilities (V). The family care home as defined here constitutes 17% of California’s sheltered-care facilities.

The family care home, medical.

The medically oriented family care home is also a small (average bed capacity for eight residents), unstructured, family-oriented facility without a rehabilitation-focused program. The average age of the residents here is 56 years, the oldest average of the six facility types. The medically oriented family care home differs from the family care home in two respects. It offers a low level of social support (IV) and has a strong medical orientation (V) (i.e., it caters to the medical needs of residents and has residents with such needs). This type of facility constitutes a very small proportion of all facilities in California, only six percent.

The group home.

This facility is not family oriented. It is above average in size (average bed capacity for 30 residents) and has a complex structure. It employs a staff and is involved with other community social and psychological services. It is similar to the two previous types in that it does not have a formal outreach program and serves an older population (mean age 49 years) (II). It also does not emphasize to any significant degree control over its resident population (III). It is average in terms of the support it offers to its residents (IV) and average as to medical orientation (V). The Group Home has a profile that closely approximates the average on all five dimensions. Twenty percent of all sheltered-care facilities in California are of this type.

The group home, programmatic.

The programmatic group home is like the group home, although somewhat larger (bed capacity for 77 residents), more structured and differentiated. It differs, as the name indicates, in having a more youth-oriented, programmatic orientation (II). The average age of the resident is 40 years. It also differs from the group home in that it is much more controlling of its residents (III). It offers a slightly more supportive environment than does the group home (IV), but like the group home, it is average in its medical orientation (V). The program-oriented group home is the most prevalent facility in the state. It constitutes 24% of all sheltered-care facilities.

The medically oriented facility.

This facility is average in its size (average bed capacity for 17 residents), level of complexity (I), and in its program orientation (II). The average age of residents is forty-seven. It is distinguished, however, by its highly controlling and directing policies (III) and by its low scores on the amount of support its social environment offers (IV). Also, as the label implies, it is very high on medical orientation (V). Facilities of this type make up 19% of all California facilities.

The therapeutic community.

This facility is the most complex facility. It has an average bed capacity for 56 residents. It is staffed and makes use of outside facilities and is not like a family (I). It places most emphasis on its program orientation and caters to the youngest population (average age 39). It focuses its efforts in fostering resident involvement in outgoing rehabilitative roles (II). The therapeutic community is neither controlling nor directing, but not particularly permissive either. It is slightly supportive and slightly nonmedical in orientation. It represents 16% of all facilities in the sample.

Discussion and Conclusion

With the evolution of community care has come an increasing awareness of the consumer rights of mental patients. This has included the recognition of a responsibility to place patients in the facility most appropriate to their present condition as well as to their future therapeutic well-being. Unfortunately, a characteristic of the sheltered-care movement has been the lack of commonly accepted criteria for matching residents to facilities. In addition, mechanisms have not been established to introduce change so that facilities might better meet the needs of the ex-mental hospital patient.

In order for a rational placement system to become accepted it first becomes necessary to identify the dimensions along which matching should take place. What kinds of patients should be placed in what kinds of facilities. Or, more realistically, what kinds of facilities presently exist and what kinds of patients might be best served by each? This research examined only the first of these questions. What types of facilities exist and what are their major distinguishing features?

The study found six types of sheltered-care facilities in California. Two types of family-oriented facilities offer to the older resident a more stable environment in a more simply constructed setting leading to stability of life style and not necessarily stressing ongoing rehabilitation. One of these family-oriented environments offers specific strengths in the area of medical supports. We have also delineated two types of group homes. One can be characterized as being more like a boarding house operation and the other as being more program oriented and concerned with rehabilitating residents and seeing them move into a larger community. A fourth sheltered-care facility type, the medically oriented facility, presents strengths for the person in need of medical care, though this strength is costly in terms of lack of mutual support available to residents. The fifth type, the therapeutic community, offers a strong emphasis on the programming of a social environment geared to a younger population.

Given our findings, the needs of former hospital patients may now be assessed relative to the five dimensions of sheltered-care facilities we have noted. Can residents cope with the complex facility environment? How can they best benefit by an outwardly oriented rehabilitation program focused primarily on a youthful population? Do they need the structure of a controlling environment? Do they need a good deal of mutual support from their fellow residents, and what are their needs with respect to the medical orientation of a facility?

Our findings indicate that it will be more difficult for the worker seeking a less complex facility with a definite outwardly reaching program to make a placement since these two dimensions of care seem to vary together. Similarly, it will be difficult to find medical facilities that offer a great deal of social support. On the other hand, workers who seek to make choices for clients with respect to control, support, and medical orientation may be more fortunate in terms of finding these characteristics associated with both complex and simple, and program and nonprogram-oriented facilities.

Aside from the placement theme this work contributes to our knowledge of community-based sheltered care by raising questions about two current emphases in the field: 1. the emphasis on rehabilitation in a youthful population, and 2. the emphasis on the conflict between the therapeutic community and the medical model of care.

The finding which indicates that a programmatic orientation toward rehabilitation is primarily directed at a youthful population seems to leave older residents in sheltered care out in the cold. This finding is certainly reinforced by Segal’s (1979) observation that while the probability of a young resident in sheltered care receiving therapeutic aid is two to one against, it is five to one against for the older resident. The focus on youth in therapeutic communities, to the exclusion of the older resident, is further illustrated by the findings of a review of 26 halfway-house studies indicating that the average age of the residents was 30 (Rog & Raush, 1975). The average age of the sheltered-care resident population between 18 and 65 years old in California is forty-eight. Future efforts to consider the development of an appropriate programmatic orientation in facilities serving older sheltered-care residents should be a priority for the field.

The emphasis on the conflict between the medical model and the therapeutic community in our field, given this study’s findings, seems rather limited. These two facility types accounted for only a third of the available facility environments in California. Our perspective on facility environments should be broadened to accommodate new models and to improve existing models.

Looking ahead, we hope to be able to delineate a typology of community environments which in conjunction with the facility types specified in this paper will enable us to determine the best fit between individual need and the total social context.