Quality of life experienced by persons with mental retardation in various residential settings.

There have been numerous calls for applying quality of life measurement concepts in assessing the rehabilitation outcome of persons with physical disabilities (e.g., Anderson, 1982; Freed, 1984; Kottke, 1982), and others have begun exploring their applicability in assessing mental retardation (Schalock, Keith, Hoffman, & Karan, 1989) and mental illness interventions (Blau, 1977; Fabian, 1990). The growing concern regarding quality of life stems from a variety of disability-related issues (Brown, 1988). However, the overriding appeal of quality of life assessment is that it is conceptually consistent with the holistic orientation of rehabilitation. Despite attractiveness of the measurement concept, there are a number of problems that must be addressed before its potential utility can be realized (Fabian, 1991). These include further clarification of the quality of life construct, development of appropriate measures to use with persons with disabilities, and gaining a better understanding of how quality of life assessment can be fitted into the rehabilitation evaluation process.

Calls for development and utilization of quality of life assessment approaches have been especially widespread within the mental retardation field. Continuing debate has followed the movement toward deinstitutionalization of persons who are mentally retarded (Alexander, Huganir, & Zigler, 1985). To a large extent, the debate centers around questions such as whether the numerous alternative living environments enhance life quality, which settings can maximize the life quality for individuals functioning at different levels, and whether traditional institutional settings can be altered to the extent that their environment is more supportive of life quality for some individuals than the alternatives (Walker, 1988).

It has been suggested that quality of life is likely to become the issue of the 1990's, superseding normalization, deinstitutionalization, and community adjustment issues (Schalock, Keith, Hoffman, & Karan, 1989). This view is consistent with that of Landesman (1986), who earlier identified quality of life and personal life satisfaction as the new buzz words in the field of mental retardation. At the same time, Landesman urged the American Association on Mental Deficiency to take leadership in developing the means by which to measure those qualities. Moreover, Walker (1988) indicated that care needed to be exercised in attempting to define quality of life from the standpoint of the mentally retarded population. She emphasized that consideration be given not only factors such as where these individuals are living and working, but their ability to make decisions and have choices within their life.

The need to develop effective, meaningful quality of life assessment tools and methods provided the impetus for implementing a multi-faceted study centered on assessing the quality of life supported by a regional service system in South Carolina for its citizens with mental retardation. A primary objective was to assess the quality of life of persons with mental retardation residing in a variety of residential settings used for placement.

The study was also designed to explore the feasibility and utility of an assessment strategy or methodology designed to address a number of the needs expressed by authorities in the field. In that respect, the stvategy followed a multidimensional quality of life assessment model similar to that described by Lawton (1991). Lawton's model rests on utilization of intra-personal and socio-normative criteria in assessing the person-environment system of individuals, with the resulting assessment considered to reflect on life quality. The model was based on three assumptions. First, there are unique, internal standards and evaluations of life generated within individuals that are not totally represented by external elements. This aspect or domain is commonly referred to as the perceived or subjective domain of quality of life. Second, a socio-normative domain exists that consists of objectively measurable or consensual evaluations of various facets of life. Finally, the person-environment system must be considered because life quality is influenced by the oft-changing interplay between the two. Thus, to be optimally meaningful, quality of life assessment must address a variety of factors, such as individuals' satisfaction with an array of significant life elements, their psychological well-being and behavioral competencies, and selected characteristics of the environment.

Method

Research Design

An ex post facto study was implemented based on two major research hypotheses. It was hypothesized that there would be a difference in the measured life quality experienced by persons who were mentally retarded according to residential settings varying in restrictiveness and overall environmental quality. Second, it was hypothesized that quality of life would vary according to level of intellectual functioning, that is, those who were higher functioning would experience a higher quality of life than those in the lower ranges.

Variables

The independent variables in the study were intellectual functioning and residential setting restrictiveness and quality. Levels of intellectual functioning were based on the American Association on Mental Deficiency's mild, moderate, severe, and profound retardation categories based on Wechsler or Stanford Binet intelligence test scores.

The six distinct residential arrangements existing in the service system administered by the South Carolina Department of Mental Retardation's (SCDMR) Pee Dee Regional Office provided varied settings for the study. Restrictiveness was based on intuitive ordering of the settings used in the resident placement process by the SCDMR. Listed according to increasing restrictiveness, the setting continuum included the family or natural home, supervised apartments, boarding homes, community training homes, community residences, and regional campus facilities.

In the family or natural home, clients were receiving services such as case management and day programming. In the supervised apartments, clients typically resided with a roommate in an apartment, rented house, or room, and received periodic monitoring, supervision, and training consistent with individual needs provided by agency personnel. The boarding home program provided room, board, and limited personal assistance from the boarding home operators with minimal supervision and no training activities. In the community training home settings, the clients resided with up to two peers in trained caregivers' personal homes. The community residences consisted of licensed, community based group homes that accommodated six to ten clients and provided 24-hour supervision and ongoing training. Finally, the regional campus facilities constituted an ICF/MR program, and clients resided in dormitories ranging in size from 16 to 32 beds. The dormitories were segregated on the basis of gender, and clients attended an on-campus workshop and participated in community based activities and outings that were scheduled and supervised by staff members.

To augment the restrictiveness rankings of the SCDMR, the overall quality of the six residential settings was established using the Program Analysis of System Implementation of Normalization Goals (PASSING) measure, which was developed to address the structure of settings as it relates to normalization (Wolfensberger & Thomas, 1983). The use of this assessment instrument related to the need to consider person-environment factors, as delineated in the Lawton (1991) quality of life model.

The dependent variables reflecting quality of life represented both the subjective and objective domains. Measures were selected that also addressed the intra-personal domain through self-rating of life situation (life satisfaction), together with others in the socio-normative domain that were focused on decision making opportunity and relevant behavioral competencies. Thus, the selected measures provided for both self-ratings by persons who were mentally retarded and others by professionals who were in situations that enhanced their familiarity with person-environment elements.

Life satisfaction was ascertained using the Life Situation Survey (Chubon, 1990). Autonomy and decision making opportunities were determined with the Resident Choice Assessment Scale (Durant, Kearney, & Mindell, 1987), and behavioral competencies essential to social comfort and fit were determined with the American Association on Mental Deficiency Adaptive Behavior Scale, Parts I and II.

Sample

A stratified random sample of adult clients on the service rolls of the South Carolina Department of Mental Retardation's Pee Dee Region Office, which contains approximately 2500 listings, was used in the study. The sample was stratified to equally represent the four levels of intellectual retardation, i.e., mild, moderate, severe and profound. Intelligence test scores utilized were those contained in the clients' case files, which were obtained during earlier routine diagnostic activities. The sample was also stratified to equally represent clients residing in the six residential service setting alternatives comprising the system. Thus, 20 individuals were selected to represent each of the six settings, and five of the persons residing in a given setting category were classified as functioning at each of the four levels of retardation.

Consent for participation was solicited from parents or guardians in accordance with agency policy and human subjects committee approved procedures. No client whose participation was invited in the selection process was denied parental-guardian consent and excluded for this or any other reason. Thus, the sampling procedure was robust. The 120-member sample was comprised of 62 males and 58 females. On the basis of information abstracted from the participants' case files, the mean age of the sample was 42.48, with ages ranging from 22 to 83 years.

Instruments

The Program Analysis of Service System Implementation of Normalization Goals (PASSING) has been described as a program evaluation tool, but has been used for teaching normalization concepts and other purposes (Wolfensberger & Thomas, 1983). The PASSING has reportedly been developed as an objective measure of a variety of human service programs and delivery systems. It has undergone several revisions since the original 1969 version and has an extensive record of usage. It consists of a 42-item scale representing seven measurement domains. Four of the domains are concerned with social image enhancement, and the remaining three focus on personal competency enhancement. Items are rated on a 1 to 5 negative-positive scale.

The Life Situation Survey (LSS) is a 20-item Likert-type rating scale developed by Chubon (1990) for use in assessing perceived life quality by persons with disabilities and chronic illness, as well as the general population. The items are clear and concise, with readability estimated to be in the fourth grade to fifth grade range. Although designed to be self-administered, provisions are included for oral administration to persons who have reading or other limitations that prohibit them from completing the written self-report form. Item scores range from one to seven.

To ascertain reliability of the LSS, inter-item consistency (Cronbach's alpha) has been determined for a variety of samples, such as persons with a variety of disabilities, prison inmates, and college students, and has ranged from .74 to .95 (Chubon, 1987, 1990). Validation derives in part from the construction process. It was developed using item areas considered by social scientists and other researchers to be widely accepted as reflecting quality of life together with item areas obtained from samples representing target populations having a variety of disabilities and chronic illnesses through a critical incidents procedure. Studies were also conducted that supported its capacity to appropriately discriminate between groups commonly accepted as differing in life quality. Sensitivity to changes stemming from intervention, i.e., pre- post-treatment was also supported by developmental studies.

The Resident Choice Assessment Scale (RCAS) is a rating instrument designed for administration by persons familiar with the individual whose quality of life is being assessed. It was selected because it reportedly measured autonomy and freedom of choice in one's living environment, which are considered to be key elements of life quality in-so-far as the mentally retarded population is concerned (Durant, Kearney, & Mindell, 1987). These socio-environmental elements seem essential to empowerment. Choices available to persons are rated according to estimated frequencies on a 25-item, 1-7 point Likert-type scale. Durant, Kearney, and Mindell (1987) reported that test-retest reliability determined over a 2-week span was .74. Inter-rater reliability calculated from administrations every three months during the course of a year was .52. In their study of inter-rater reliability, ratings were done by facility staff in intermediate care facilities or a state developmental center and involved clients classified as profoundly retarded. Validity rests on item content.

The American Association on Mental Deficiency Adaptive Behavior Scale (ABS), which is comprised of two parts, was used in the study. It was selected to address the issue of assessing behavioral competencies integral to attaining a good quality of life in accordance with the Lawton (1991) model. Part I of the ABS focuses on 10 different behavior domains including independent functioning, physical development, economic activity, language development, numbers and time, domestic activity, vocational activity, self-direction, responsibility, and socialization, and reflects level of adaptive functioning. Part II addresses 14 domains of maladaptive behavior related to personality and behavior disorders, ranging from violent and destructive behavior to withdrawal. The ABS was designed for administration by persons familiar with the individuals to be evaluated.

Isett and Spreat (1979) reported two-week test-retest reliability coefficients ranging from .62 to .98 for Part I of the ABS. Inter-rater reliability was reported to be .86 (King, Soucar, & Isett, 1980). For Part II, test-retest reliability, coefficients ranged from .60 to .97, while inter-rater reliabilities ranging from .32 to .84 were reported (Isett & Spreat, 1979). Validation began with item development based on critical incident studies (Nihira, 1978), and factor analyses (Nihira, 1969).

Procedure

Administration of the instruments occurred during a three-week period in June, 1991. The Life Situation Survey was administered to all subjects orally by a single SCDMR staff psychologist. Administration was restricted to one person to minimize reliability problems. That consideration was necessary because no studies were reported regarding inter-oral administrant variation or reliability by Chubon (1990). The reliability of the Life Situation Survey results was ascertained by calculating Cronbach's alpha, and found to be .74 for the High IQ group and .61 for the Low IQ group. These were lower than those reported for most samples having non-cognitive disabilities, but are at levels sufficient to render the results usable in interpreting group data. Given the fact that half the participants had been determined to be functioning at the severe and profound retardation levels, it was anticipated that a lower reliability would be manifest.

The RCAS and ABS were completed by the staff member who was determined to have the most contact with the client being assessed. In the home setting, the client's case manager was the administrant, and in the other settings, the staff member who was assigned as the primary program implementor for the client assumed responsibility for administration. A total of 51 staff persons provided ratings. Prior to the study, a South Carolina Department of Mental Retardation staff psychologist instructed those who were selected in how to complete the rating scales.

The PASSING instrument, which was used to evaluate the various settings, was completed by a staff psychologist during visits to each of the residences occupied by participants in the study.

Statistical Analyses

Administrative considerations restricted the number of participants in the study. Therefore, the four levels of intellectual retardation, i.e., mild, moderate, profound, and severe, were collapsed into two groups to provide a more powerful statistical design. Preserving high and low IQ groups was essential to TABULAR DATA OMITTED examination of the interaction between level of retardation and quality of life.

Because of the number of variables examined in the study, multiple comparison procedures were utilized when appropriate. A preliminary analysis used an ANOVA to determine whether the six residential settings varied in overall quality, using mean PASSING scores. Because the participants' intellectual functioning was compromised, concern for their capacity to appropriately respond to the Life Situation Survey necessitated that the reliability of the obtained scores be evaluated. Chronbach's alpha, a measure of inter-item consistency, was used for that purpose. A two-factor, between subjects MANOVA was then used to test the hypotheses that the dependent (quality of life) measure scores would vary according to residential setting, and that the scores would differ between the high and low functioning groups. Follow-up ANOVA's were used to ascertain differences according to each dependent variable. Supplemental post hoc correlational analyses were carried out using the scores from the five instruments to reveal any interrelationships.

Results

Setting Differences

The mean scores obtained with the various measures are contained in Table 1. Using an ANOVA, a significant main effects difference between residential settings was found with the PASSING scores (F = 1771.94, df = 5/5, p = .0001). Employing Tukey's studentized range test set at the .05 level for follow-up analysis, all six settings were determined to differ from one another with regard to quality or adherence to the principles of normalization.

Hypothesis Testing

In testing the hypotheses, the MANOVA revealed an overall setting effect, with the Wilks' lambda criterion test producing an L of. 1611, and an F approximation of 11.35 (df = (20/309) and p .0001. An overall IQ effect was also revealed, with L = .4572, and the F approximation, 7.07 (df = 12/246) and p .0001. The loci of the differences revealed by the MANOVA procedure were then established through post hoc ANOVA's.

With regard to Life Situation Survey scores, significant main effects differences were found between the settings (F = 4.44, df = 5/5, p .001). Tukey's studentized range test (p = .05) was then performed to discern the specific differences. The combined high-low IQ group scores obtained for the community training homes were significantly higher than those for the regional campus and boarding homes, and supervised apartment scores were higher than those from the boarding homes. In addition, comparing the mean score of the high IQ groups from all settings (M = 86.37) with that of the low IQ groups (M = 82.88) revealed that they were significantly different (F = 6.92, df = 1/119, p .001).

The results of the ANOVA conducted with the Resident Choice Assessment Scale (RCAS) scores revealed a significant main effects difference between the settings (F = 42.85, df = 5/5, p .0001). Using Tukey's studentized range test, scores from the supervised apartments and community training home settings were found to be higher than those from the other four settings. The scores for the community residence and home settings were found to be higher than those from the regional campus and boarding home settings. The high IQ group's mean rating of 124.20 was significantly higher than that of the low IQ group, which was 99.07 (F = 14.95, df = 1/119, p .0005.

The Adaptive Behavior Scale-Part I scores were subjected to an ANOVA and a significant settings difference was revealed (F = 12.84, df = 5/5, p .0001). The results of the Tukey procedure indicated that the community training home scores were higher than those from the home, community residence, and boarding home settings. Scores from all other settings were significantly higher than the boarding home group scores. The mean from the high IQ groups (152.95) was significantly higher than that of the low IQ groups (128.30) (F = 15.33, df = 1/119, p .0005).

An ANOVA of scores from Part II of the ABS resulted in a significant setting effect, with F = 4.92, df = 5/5, and p .0005. Both the community residence and regional campus group scores were significantly higher than those of the community training home and home scores. It is noted that in contrast to the other measures used in this study, higher scores on the ABS-II are more negative. The respective high and low IQ group means of 22.8 and 18.48 were not significantly different (F = 1.1, df = 1/119, p = .30).

Table 2

Correlations of individual quality of life measures

Instrument       LSS       RCAS       ABS-I       ABS-II

LSS                        .35(*)     .35(*)        .04
RCAS                                  .58(*)       -.04
ABS-I                                               .08

* p [is less than].01

Supplemental Analyses

The scores obtained with the measures for the combined high and low IQ groups in each setting category are plotted in Figure 1. The resultant graph is suggestive of correlation between the various measures, with the exception of the ABS-II. However, the disordered pattern of mean scores between the settings ordered on restrictiveness continuum established by the South Carolina Department of Mental Retardation do not reflect a hierarchy.

To further examine the dependent variable interrelationships, correlational analyses were carried out on the mean scores. A correlation matrix of the individual instrument scores is contained in Table 2. Last, the PASSING scores from the six different settings were correlated with the individual instrument mean scores for each of the settings. Both the Life Situation Survey (r = .91, p .01) and the Resident Choice Assessment Scale (r = .98, p .001) correlated significantly with the PASSING scores, while the ABS-I (r = .78) and ABS-II (r = -.32) did not.

Discussion

The findings from the use of the PASSING to evaluate the various settings indicated the existence of significant differences between all of them. Ranked by diminishing score, they were supervised apartments, community training homes, family homes, community residences, regional campus, and boarding homes. This ranking by PASSING scores is depicted in Figure 2, which also contains a plotting of the mean scores of the other instruments. This hierarchy does not coincide with the pattern in the SCDMR's ordering according to restrictiveness, which has been plotted in Figure 1. From least to most restrictive, the Department's intuitively derived ranking was family homes, supervised apartments, boarding homes, community training homes, community residences, and regional campus. The differences between the Department' s ordering and the PAS SING hierarchy are notable when comparing Figures 1 and 2.

The findings lend support for both hypotheses, i.e., the quality of life does vary with type of setting and that persons functioning at the higher intellectual levels were experiencing a better life quality than those at the lower end of the continuum. Scores from the four measures of life quality generally support the conclusion that differences do exist between settings. However, the differences do not neatly vary between all six, as can be seen in Figures 1 and 2. Significantly lower scores were found on the Life Situation Survey for persons residing in the regional campus and boarding home settings than those calculated for the community training home and supervised apartment settings. The community residences and family home seemed to be very close to the community training homes and supervised apartments but differences from the campus and boarding homes were not supported by the statistical analysis.

As can be seen in Figures 1 and 2, the ordering by RCAS mean scores was very similar to that of the LSS scores. However, the statistical analysis revealed three distinct groupings, consisting of the (1) supervised apartments and community training homes, (2) community residences and natural home, and (3) regional campus and boarding home. With the scores for the settings ranging from approximately 75 to 151, there appeared to be major differences in freedom of choice between the bipartite groupings.

The findings from Parts I and II of the Adaptive Behavior scale are less definitive than those from the LSS and RCAS. The ABS-I scores are relatively consistent with those two measures, as reflected in the r's of .35 and .58. There was no apparent relationship with Part II, although there did appear to be a between setting difference. The more negative scores (higher) were from the groups in the community residences and regional campus, and differed significantly from those residing in the community training home and family home settings.

In retrospect, expectation that the ABS-II scores would correlate with the ABS-I scores and those of the other measures was based on the assumption the maladaptive behaviors are the converse of adaptive behaviors, and that is not the case. Both types of behavior may be manifest in individuals, with opportunity, training emphasis, and other factors affecting type and frequency. On the other hand, the higher levels of maladaptive behavior found in the regional campus institutions was consistent with the finding of Borthwick-Duffy, Eyman, & White, 1987. It may be that individuals with high levels of maladaptive behavior are concentrated there because they are not considered ready for community placement.

Additionally, the high levels of maladaptive behavior in the community residence setting category is explainable. Researchers (e.g., Hemming, Lavender, & Pill, 1981) have reported temporary increases in maladaptive behavior following deinstitutionalization. This explanation is consistent with the fact that the SCDMR routinely utilized the community residence setting as the first step in making the transition to the community.

The correlation between the Life Situation Survey and Resident Choice Assessment Scale, and ABS-I, and in general, the apparent consistency of the results from the three instruments provide an accumulation of evidence that supports the conclusion that the quality of life being experienced by the clients residing is, in fact, different between settings. Moreover, the three individual instrument derived scores were consistent with those from the PASSING scores, although the ABS-I r was not statistically significant. In considering this data as a whole, the settings appear to fall in pairs with regard to the life quality experienced within. Highest are the supervised apartment and community training home settings, followed by the family or natural home and community residences. The regional campus and boarding home settings comprise the low end of the continuum.

Finally, across settings, the high IQ group appeared to be faring better than the low IQ group. That is consistent with the common perception that severity of disability is associated with life quality, which has yet to be substantiated by research. Moreover, it must be noted that the range of individual scores indicates that there are persons in both the high and low IQ categories who are faring better and worse than counter-group peers in all settings.

Conclusions

With the possible exception of the findings from the Adaptive Behavior Scale Part II, the multifaceted model-based approach to quality of life assessment used in the study appears to be viable and productive. It was especially effective with regard to revelation of erroneous rankings in the SCDMR's intuition-based ordering of its various system settings. It underscores the risks inherent in relying on intuition in formulating both classification of service system components policy formulation. In fact, the extremely low life quality ratings for boarding home residents has already led to reexamination of their use by the SCDMR. The additional revelation of differences in quality of life between settings and intellectual groupings provide foci for further system examination. Together, these findings provide evidence of the potential for practical application of quality of life assessment approaches to disability issues.

The findings reflect consistency between self-ratings and those of the South Carolina Department of Mental Retardation staff. That consistency, together with the reliability levels determined for the Life Situation Survey, supported the notion that this population was capable of making reliable and valid judgments about their condition. These findings underscore the feasibility and importance of securing client input in such assessment activities. Such inclusion serves as a means of empowering persons with disabilities.

Fabian (1991) indicated that there is considerable agreement that multiple assessments should be used in examining quality of life. Frameworks such as that set forth in Lawton's (1991) model provide a logical basis for selecting measures. Such model-based approaches are likely to be more revealing that a conglomeration of measures chosen at random or intuitively. This study should give encouragement to others to utilize quality of life assessment approaches in rehabilitation or disability service activities. In addition to overall program evaluation, research efforts directed at examining their use as progress measures and in other clinical applications seem warranted. Further work on development of measures that can be used with a variety of populations, such as the Life Situation Survey, should also be pursued to enable comparative studies. Empirically-based quality of life rankings of different disability groups have potential to provide a more sound basis for establishing intervention priorities and resource allocation.

References

Alexander, K., Huganir, L. S., & Zigler, E. (1985). Effects of different living settings on the performance of mentally retarded individuals. American Journal of Mental Deficiency, 90, 9-17.

Anderson, T. P. (1982). Quality-of-life and the individual with a disability. Archives of Physical Medicine and Rehabilitation, 63, 65.

Borthwick-Duffy, S. A., Eyman, R. K., & White, J. F. (1987). Client characteristics and residential placement patterns. American Journal of Mental Deficiency, 92, 24-35.

Brown, R. I. (ed.). (1988). Quality of life for handicapped people. New York: Croom Helm.

Chubon, R. A. (1987). Development of a quality-of-life rating scale for use in health care intervention. Evaluation and the Health Professions. 10, 186-200.

Chubon, R. A. (1990). Manual for the Life Situation Survey. Columbia, SC: University of South Carolina, Rehabilitation Counseling Program.

Durant, B., Kearney, C., & Mindell, J. (1987, November). Measuring choice in residential settings: The Residential Choice Assessment Scale (RCAS). A paper presented at the annual meeting of the Association for the Advancement of Behavior Therapy, Boston, MA.

Fabian, E. S. (1990). Quality of life: A review of theory and practice implications for individuals with long-term mental illness. Rehabilitation Psychology, 35, 161-170.

Fabian, E. S. (1991). Using quality-of-life indicators in rehabilitation program evaluation. Rehabilitation Counseling Bulletin, 34, 344-356.

Freed, M. M. (1984). Quality-of-life: The physician's dilemma. Archives of Physical Medicine and Rehabilitation, 65, 109-111.

Hemming, H., Lavender, T., & Pill, R. (1981). Quality of life of mentally retarded adults transferred from large institutions to new small units. American Journal of Mental Deficiency, 86, 157-169.

Isett, R. D. & Spreat, S. (1979). Test-retest and interrater reliability of the AAMD adaptive behavior scale. American Journal of Mental Deficiency, 84, 93-95.

King, R., Soucar, E., & Isett, R. (1980). An attempt to assess and predict adaptive behavior of institutionalized mentally retarded clients. American Journal of Mental Deficiency, 84, 406-410.

Kottke, F.J. (1982). Philosophic considerations of quality of life of the disabled. Archives of Physical Medicine and Rehabilitation, 63, 60-62.

Landesman, S. (1986). Quality of life and personal life satisfaction: Definitional and measurement issues. Mental Retardation, 24, 141-143.

Lawton, M. P. (1991). A multidimensional view of quality of life in frail elders. In J. E. Birren, J. E. Lubben, J. Chicowlas Rowe, & D. E. Deutchman (Eds.), The concept and measurement of quality of life in the frail elderly (pp. 3-27). San Diego: Academic Press.

Nihira, K. (1969). Factorial dimensions of adaptive behavior in mentally retarded adults. American Journal of Mental Deficiency, 74, 130-141.

Nihira, K. (1978). Factorial descriptions of the AAMD adaptive behavior scale. In W. A. Coultor & H. A. Marrow (Eds.), Adaptive behavior (pp. 45-58). New York: Grune & Stratton.

Schalock,R. L., Keith, K. D., Hoffman, K., & Karan, O. C. (1989).Quality of life: Its measurement and use. Mental Retardation, 27, 25-31.

Walker, V. S. (1988). Presidential Address, 1988: Conquest of mental retardation. Mental Retardation, 26, 251-255.

Wolfensberger, W., & Thomas, S. (1983). PASSING (Program analysis of service systems implementation of normalization goals). Normalization criteria and ratings manual (second edition). Toronto: National Institute on Mental Retardation.

David V. Vandergriff, Ph.D., Executive Director, Marion-Dillon County Mental Retardation Board, P.O. Box 1082, Marion, SC 29571