The extra-individual social support responses provided by adults with intellectual disabilities on a social support survey were examined. Felton and Berry (1992) argued that extra-individual social support, support from larger than individual entities, is a valid source of social support for older adults. We hypothesized that extra-individual support would also be relevant to adults with intellectual disabilities. Results showed that 43% of participants with intellectual disabilities listed extra-individual support sources in their social networks or as providers of one of several support functions. The most popular forms of extra-individual support were those from “staff” and “work.” Findings are discussed in terms of implications for social support measurement for persons with intellectual disabilities.
Editor in charge: Marsha Mailick Seltzer
As the goal of full inclusion grows near, researchers, clinicians, and policymakers have been forced to pay closer attention to issues of social support. Many adults with intellectual disabilities have difficulty establishing and maintaining social support over time. Often, these adults can be isolated, even while integrated physically in their communities (e.g., Rapley & Beyer, 1996). The presence of social support for these individuals is positively correlated with quality of life and negatively correlated with depression, as has been found for the general population (Lunsky & Benson, 2001a; Meins, 1993; Reiss & Benson, 1985). Research in the area of social support and intellectual disability is still in an early stage, however. In particular, many questions remain about how to best measure social support (Lunsky & Benson, 1999).
Recently, social support researchers have acknowledged that measuring social support from the perspective of an informant, such as a caregiver, may not be sufficient (Krauss, Seltzer, & Goodman, 1992; Lunsky & Benson, 1999; Wenz-Gross & Siperstein, 1996). Even children with mild intellectual disabilities can report “who provides what” with some accuracy (Wenz-Gross & Siperstein, 1996), and their perspective is very important. Sometimes, however, the measures adopted from the general population, along with scoring procedures, may not quite fit how individuals with intellectual disabilities conceptualize their relationships with others. Several authors have argued for social support measurement based on conceptions of people with intellectual disabilities rather than conceptions of researchers (Clegg & Standen, 1991; Kultgen & Hawkins, 1992; Lunsky & Benson, 2001b; Mest, 1988). In this paper we highlight one way in which traditional social support measures may omit an important piece of information about how adults with intellectual disabilities understand or categorize their relationships.
A current idea in community psychology regarding social support research in the general population is that the construct has become too individually biased, neglecting community aspects. Early definitions of social support included the feeling of belonging to a community. Cobb's (1976) classic definition, for example, included a feeling that one “belongs to a network of communication and mutual obligation” (p. 300). Closely tied to this definition is Sarason's (1974) definition of one's sense of community, “the feeling that one is part of a larger, dependable and stable structure” (p. 157). Community psychologists have argued that current measurement strategies selected by social support researchers focus on individual ties between people, precluding examination of larger units of support, taking away from a much needed community emphasis (Heller, 1990; Milardo & Wellman, 1992; Shinn, 1987).
The importance of categorizing support based on the respondent's conception and not the conception of the researcher is illustrated through a study on the social support of older adults in New York City (Felton & Berry, 1992). When providing information on their social networks and support functions, many of the study participants provided community or group level answers. Rather than treating such responses as invalid or as equivalent to an individual source of support, Felton and Berry chose to examine the responses in greater detail.
Felton and Berry (1992) referred to these larger support systems as extra-individual support. They argued that such support should not be dismissed as a default response in social support research. Disregarding such responses may lead to underrepresentations of support or misrepresentations of the type of support available. Furthermore, they argued that there could be advantages to extra-individual support. For example, support from a group of people can provide reassurance of worth and a sense of belonging. Extra-individual support may serve as a buffer when there are losses within a group because groups can survive losses of individual members (Felton & Shinn, 1992). It seems extra-individual support would, therefore, be especially valued by people who have difficulty recruiting and maintaining individual relationships, particularly if they are faced with frequent support losses because of old age, illness, or lack of environmental control (Maton, 1989). In Maton's research group level support accounted for individual outcome above and beyond individual variables, particularly for individuals under high stress. Such individuals benefited from being part of a supportive group/setting. Group support may have come from specific individuals within the group, but it also could have come from group members' sense of belonging, discussed by early social support researchers (e.g., Cobb, Sarason), which would not be dependent on the individuals within the group or setting.
The findings reported in Felton and Berry (1992) and Maton (1989) as well as the commentary offered by Felton and Shinn (1992) have direct relevance to persons with intellectual disabilities, who have difficulty recruiting and maintaining social support and who suffer support losses because of a lack of environmental control (Ashman & Suttie, 1996; Rosen & Burchard, 1990; Sands & Kozleski, 1994). Beginning in their youth, they can have a harder time making friends and may turn more often to formalized structures, such as school (teachers) and family (Wenz-Gross & Siperstein, 1996). Because adults with intellectual disabilities often lack social opportunities, places can become important sources of social support for them. Instead of arranging individual social engagements, they can simply seek out a social environment; they can go to their community center or church (McNair & Smith, 2000) or to a local store or restaurant. Even the local health center can become a reliable place of support (Aylott & Toocaram, 1996).
Support from organized groups also appears to play an important role. Although many adults with intellectual disabilities lack the skills to maintain one-to-one relationships in their spare time (Schalock & Lilley, 1986), they are frequently involved in organized groups. Many of them work in groups, such as workshop teams or group enclaves, and many live in group settings, such as group homes or larger intermediate care facilities. Several authors have discussed the topic of support that groups provide, including a choir (Bogdan, 1995), public speaking groups (Fisher, 1995), religious institutions (McNair & Smith, 2000), place of employment (Lutfiyya, 1995), or volunteer organizations (Miller & Keys, 1996). McNair and Swartz (1997) presented a framework on group sources of support versus individual supports for persons with intellectual disabilities. They suggested that groups such as work colleagues or staff members are natural supports in that they are part of a person's daily routine and ongoing in that even with changes involving individuals, the source of support remains stable.
Neely and Lunsky (1998) retrospectively examined the verbatim responses of 50 adults with intellectual disabilities given during a social support structured interview for any unsolicited community level support trends. They noted that several individuals spoke about larger entities when they were asked to identify their individual friends on a social support measure. In post-hoc analysis of responses, the authors found that 36% of individuals provided at least one extra-individual level response when asked, “Who are your friends?” For example, one unemployed older man with a mild intellectual disability who lived alone in a small town reported that he spent his days at the McDonald's down the street. When asked, “Who are your friends?” he identified “McDonald's” and then “the people at McDonald's” as his best friends, but he could not give any names of specific people when asked for clarification. McDonald's was not considered a scorable response for this survey, although the interviewer thought it was clearly an important source of support for him. Additional response examples from other participants included work, organized sports activities, and religious involvement (e.g., church). Participants may have had difficulty recalling the names of their friends because of cognitive deficits and gave the group responses as a default response. They may have found it simpler to list the group name rather than list each member. Alternatively, participants may have conceptualized their support in terms of a larger entity, different than the naming of the individuals within it. If McDonald's was the support for the gentleman just described, then he could receive it regardless of who was working there on a given day. Because their study was based on a post-hoc analysis of the data, the researchers could not go back to participants to clarify what was meant by their extra-individual level responses. Also, extra-individual responses were given in response to a question about friends, but it is not known whether similar responses would have been given to questions about support from other sources.
As the preceding review suggests, traditional measures of support adopted from the general population for use with persons who have intellectual disabilities have neglected extra-individual sources in favor of individual names. Groups and places may serve as alternative sources of support but have never been examined directly, with the exception of the post-hoc analysis by Neely and Lunsky (1998). Our purpose in the present study, therefore, was to examine the amount and type of extra-individual support reported by individuals with intellectual disabilities, unsolicited as part of a traditional social support interview. This study was exploratory; we adopted the same measures as those used by Felton and Berry (1992) to determine how adults with intellectual disabilities describe their social support and whether the concept of extra-individual support is relevant to them. Our study also serves as a methodological investigation in that past extra-individual level support responses in social support interviews of persons with intellectual disabilities were unscored and, hence, undocumented.
Data reported here were gathered as part of a larger longitudinal study on the social support of adults with mild intellectual disabilities (Lunsky, 1999).
Participants were 84 adults (41 men, 43 women) with mild intellectual disabilities. They ranged in age from 20 to 65 years (M = 38, standard deviation [SD] =10.4). Age limits were set such that participants had completed school and were not regarded as senior citizens. These individuals were nominated by 11 residential agencies in central Ohio and 3 residential agencies in Toronto, Ontario, that provide community housing and support to persons with intellectual disabilities. Agencies were asked to nominate individuals with mild intellectual disabilities who could participate in two interviews on relationships, each 1 to 2 hours in duration, and who expressed an interest in participating. The majority of respondents were White (85%), with 14% African American and 1% Asian American participants. Forty-five percent of participants worked in community settings, 43% worked in sheltered employment settings, and 12% were unemployed. Twenty-eight percent of the sample lived alone, 30% had one housemate, and the remainder had two or more housemates. Only 7 individuals in the sample were married, 2 women were divorced, and the remainder were single. Comparable marital status ratios were reported in a study of 300 adults of similar age range in semi-independent living programs from four Western states (Halpern, Close, & Nelson, 1986).
Measures and Procedure
Prior to assessing social support networks, we asked participants to provide both oral and written informed consent. The consent form was read out loud, and they had to respond to four oral questions about the consent correctly in order to proceed. Next, the Peabody Picture Vocabulary Test, Third Edition (Dunn, Dunn, & Williams, 1997) was administered as an assessment of single-word receptive vocabulary. Participants were required to score at the 5-year age equivalent or above in order to proceed. No one was excluded based on these criteria. In order to obtain data that could be compared with previous investigations of extra-individual social support, we made an effort to use a methodology consistent with Felton and Berry's (1992) research with older adults. We used Antonucci and Akiyama's circle technique (1987) to measure social support structure through respondents' own reports of people in their lives. Respondents were shown a page with a circle labeled “you” surrounded by three concentric circles, then told to think of these circles as representing “all the people who are important in your life right now.” First, they were asked to identify those people “who you feel so close to that it is hard to imagine life without them.” After these people's names were recorded in the innermost circle, individuals were asked to identify the people “who you don't feel as close to but who are still very important to you.” Finally, they were asked to identify the people “who you didn't mention yet but who are close enough and important enough to be in your circle.” After listing each name, participants were asked how they knew the person. This information along with the name was then depicted in the appropriate circle. Before going to the next circle, participants were asked, “Is there anyone else?”
A social provisions measure, developed by Felton and Berry (1992), was also administered. We asked six questions, one for each social provision/function, to determine whether or not respondents had anyone who made these provisions. The six questions addressed whether there was “someone you can count on for sympathy and understanding (attachment), for pleasant companionship (social integration), for help with household or personal tasks (reliable alliance), or for advice if needed (guidance)”; if there is “someone you know who recognizes special abilities you have (reassurance of worth)”; and if there is “someone who relies on you for their care or part of their care (opportunity for nurturance)?” Respondents could list up to three providers for each provision. If they did not list three persons, they were asked, “Is there anyone else?” before advancing to the next question. Additional explanations were provided beyond those used in the Felton and Berry study to ensure comprehension by participants. For example, following, “Is there someone you can count on for sympathy and understanding?” interviewers added, “Like if something bad happened to you and you needed a friend who would understand and help you feel better?”
On both of these measures, respondents were asked about individual support. That is, the questions posed in both the structural and functional support measures were phrased to elicit individual names (e.g., List the people. . . Is there anyone else. . . Is there someone who ), as is typical in social support measurement (Felton & Shinn, 1992). When a participant offered an extra-individual support response, the interviewer asked for clarification, by saying “Who do you mean?” Participants had to maintain their extra-individual response upon query for it to be scored as such. This scoring procedure is based on a procedure described by Felton and Berry (1992) with two modifications. They checked each response two times through asking specific questions. For the purpose of this study, only one question was asked repeatedly (“Who do you mean?”). As well, in the present study, once the individual listed at least one extra-individual support source maintained with follow-up questioning, subsequent extra-individual responses were not always queried. Findings reported here are based on the number of people who reported and maintained extra-individual responses as opposed to the number of extra-individual responses provided.
Preliminary analyses were conducted to determine whether there were any differences between the Canadian and United States samples. No significant differences were found on gender, age, type of employment, or number of extra-individual responses. Participants living in Toronto had significantly more housemates than those living in central Ohio (Ms = 3.35 and 2.19, respectively). Participants in central Ohio had higher receptive language scores than those living in Toronto (Ms = 124 and 108, respectively). Due to the similarities between the two groups on amount of extra-individual responding, the groups were combined for the primary analyses.
Further preliminary analyses were conducted to determine whether there were important demographic differences in overall network size between those who did and did not provide extra-individual responses. Chi-square analyses revealed no significant differences between the two groups on gender, ethnicity, or presence of physical disability. T tests revealed no significant differences between the two groups on age, number of roommates, or total number of support sources mentioned on either the circle technique or the social provisions measure. Thus, the two groups did not differ on size of overall network reported. Participants working in workshop settings were more likely to provide extra-individual responses than were those who were unemployed or working in the community, χ2(2, N = 81) = 8.48, p < .05. There was also a significant difference between the two groups on receptive language ability. Participants who provided no extra-individual response had lower receptive language ability scores than did those who did provide an extra-individual response (Ms = 114.82 [SD = 24.35] and 127.43 [27.14], respectively), t(78) =−2.19, p < .05. These findings serve as evidence against the explanation that extra-individual responses were a function of poorer verbal comprehension.
Social Network Analysis
Using the circle technique, the 84 respondents in this study provided a total of 932 network members. Individual totals ranged from 0 to 38 network members (M = 11.10, SD = 6.2). Overall, 25% of respondents provided at least one extra-individual response to the network measure. Table 1 provides the percentage of participants providing at least one extra-individual response for each of the three concentric network circles (inner, middle, and outer) and each of the six social provisions. There was no significant difference between the number of people providing extra-individual responses to each of the three concentric circles following the query.
Social Provisions Analysis
Extra-individual level social support providers were mentioned in response to all six social provisions. Overall, 23% of participants provided an extra-individual response. Table 1 provides the percentages. There was no significant difference between the number of individuals providing extra-individual responses to the six social provisions. On each of them, at least 4% of the sample maintained one or more extra-individual responses when asked for clarification.
Interestingly, the individuals who gave extra-individual responses for the social provisions measure were not the same as those who gave these types of responses on the network nomination measure. In total, 42% of the participants provided at least one extra-individual level response on either the social provisions or social network task. Only 6% of respondents provided an extra-individual response to both measures. Nineteen percent provided an extra-individual response to the network measure only and 17% to the provisions measure only.
Extra-individual responses were categorized into three response types described by Felton and Berry (1992): (a) groups of kin, (b) undifferentiated general responses (e.g., everyone, my friends, or neighbors), and (c) more formally organized groups (e.g., staff, work, or church). Formal groups such as work and staff accounted for the largest percentage of extra-individual responses (61%). Undifferentiated groups such as “everyone” and “neighbors” accounted for 26%, and groups of kin such as “cousins” accounted for the smallest percentage (13%). Formal group support consisted primarily of “staff” and “work,” suggesting these two sources as single entities.
In addition to the extra-individual support just described, participants mentioned several support sources that could not be categorized as extra-individual or typical individual responses. Eighteen participants listed deceased individuals as support sources, 2 individuals listed “God,” 3 listed pets, and 1 individual listed several characters from the World Wrestling Federation. Seven individuals considered themselves to be an important support source.
In the present study we examined the frequency with which adults with intellectual disabilities identified support sources other than specific individuals. Using the same methodology as Felton and Berry (1992), we demonstrated the importance of community as a potential source of social support for a group of individuals confronted with social support barriers. Our results bolster preliminary findings reported by Neely and Lunsky (1998). In addition to listing individual names, nearly half of the participants provided, unsolicited, at least one group or place as a support provider on one of two measures and maintained this response when queried.
Individuals who provided extra-individual responses were more likely to come from sheltered work settings, but they did not differ on any other demographic variables from those who did not provide such responses. It may be that adults from sheltered settings are more likely to consider the support they receive in an extra-individual fashion because they work in groups and are supervised by groups of staff. Individuals who provided extra-individual responses also had higher single-word receptive language ability than those who did not provide extra-individual support sources. This suggests that these responses were not due to a lack of comprehension. Our findings taken together suggest the presence of a community or extra-individual level support schema for many adults with mild intellectual disabilities living in community settings.
An important question raised by the present findings is, How meaningful are extra-individual supports? No differences were found in the total network size or total perceived social support of individuals with and without extra-individual support (i.e., overall level of support was not influenced by extra-individual support). Therefore, it does not appear that persons with more limited individual supports provide more extra-individual responses when asked about their social lives. These findings do not rule out more complex relationships or the possibility of individual differences in the need for and use of extra-individual support. For some individuals in this study, listing extra-individual support was a way to tally support despite their cognitive limitations (e.g., 10 individuals commented when asked “Who do you mean?” that they could not remember the names). For other individuals, listing extra-individual support may have been a way to describe a sense of belonging to a large group (e.g., when asked “Who do you mean?” 14 individuals made comments such as “There are too many to list” or “I can't separate them”).
In attempting to maintain the structure of questionnaires developed for individuals without intellectual disabilities, we were met with a methodological challenge. Asking for clarification after each extra-individual response proved to be tiring for participants and even stressful or frustrating for those individuals who seemed to fear that they kept giving the wrong answer. The compromise agreed to by investigators was to follow-up on at least the first extra-individual support source listed, a less precise protocol than Felton and Berry used in their study. Failure to clarify all extra-individual responses likely led to an underrepresentation of such responses because it is probable that some percentage of missed responses would have been maintained responses had they been queried. This difficulty was likely a consequence of using a protocol not designed for this population, a problem that has been addressed with regard to the measurement of other constructs, such as fear experienced by persons with intellectual disabilities (Ramirez, Morgan, Manns, & Welsh, 2000). Questions could be designed based on responses provided in this research, as well as from ethnographic research on relationships of persons with intellectual disabilities (e.g., Day & Harry), in order to assess extra-individual support most adequately in the future (e.g., Day & Harry, 1999).
The participants in this study demonstrated that some relationships might be conceptualized in larger units than typically considered in social support research with this population. One limitation of the present study is that we did not provide the evidence required to determine whether individuals are conceptualizing their support at a group level. More research is needed to learn how extra-individual support is related to the traditional conceptualization of individual sources of support. Are the effects of two types of support additive, or do they function differently? Is one form of support better than the other, or are the two sources interchangeable? Although on the basis of the present study we cannot answer these important questions, the results do indicate that broadening the definition of social support is an issue worthy of further attention. Valuable information about the structure of an individual's network may be lost if traditional methodology is used without consideration of extra-individual units of support.
Another limitation of the present study is that only adults with mild intellectual disabilities living in the community and not in the parental home were interviewed. Persons with moderate to severe intellectual disabilities, who have increased difficulties developing their own individual networks, may list extra-individual support more frequently. It would be worthwhile to examine whether there is an increase in the amount of extra-individual support sources for those with moderate and severe intellectual disabilities and also to determine whether individuals living with families report different experiences than do those who have moved from their family home or who do not live with family. Because even those with family are in the community, it may be that persons with more restricted individual networks have an even greater reliance on extra-individual support sources. Finally, the current findings do not provide enough information to suggest whether extra-individual sources of support should be a point of intervention among adults with intellectual disabilities. Future researchers should examine whether there are benefits to increasing the available group sources of support, such as organized recreational activities or support groups.
Social support is important for good mental health (Barrera, 1986; Cohen & Wills, 1985). Understanding the construct and its measurement is a critical component of the intervention process. The present study suggests one way that social support measurement for persons with intellectual disabilities may require modification to better fit the way such individuals conceptualize their relationships. In future research on social support and intellectual disability, investigators should consider other forms of support in addition to individual support and determine how important other sources may be. Both direct questioning about such support and the recording of unsolicited responses should be analyzed so that a construct meaningful to individuals with intellectual disabilities can be developed. Then, specific hypotheses about the meaning of extra-individual support can be tested with implications for policy and intervention in terms of the development and maintenance of healthy relationships.
Note: The authors express their appreciation to Betsey Benson, Brian Lakey, and three anonymous reviewers for their helpful comments on earlier drafts of this manuscript. Thanks are also extended to Melissa Fiatal for her assistance in data collection and to the agencies and individuals from central Ohio and southern Ontario who agreed to participate in the study. The research described in this article was supported by a dissertation fellowship and alumni research grant from The Ohio State University and a fellowship awarded by the Social Sciences and Humanities Research Council to the first author.
Authors:Yona Lunsky, PhD, Assistant Professor, Dual Diagnosis Program, Centre for Addiction and Mental Health, 1001 Queen Street West, Unit 4–4, Toronto, Ontario, Canada M6J 1H4 ( email@example.com). Lynn C. Neely, MA, Doctoral Candidate, Wayne State University, Department of Psychology, 71 W. Warren St., Detroit, MI 48202