Abstract: Varied research has addressed social integration at the workplace of persons with disabilities. Goffman (1963) and others propose that disability interferes with ordinary social interactions, creating a class of stigmatized interactions. Whether perceived or experienced, stigmatization leads to negative outcomes for the recipient of stigma. Allport's Intergroup Contact Hypothesis (1954), expanded by Pettigrew (1998) specified conditions to ensure positive outcomes of intergroup contact.

The Modified Labeling perspective predicts that the greater perceived stigma toward people with disabilities, the less likely that an employee with a disability engages in workplace interactions leading to friendships. Allport's perspective predicts that newly-employed people with a disability are less socially-integrated than others. Over time, differences in social integration between the two disabled and non-disabled employees decline as contact leads to social comfort and a basis for friendship.

This study brings the person with the disability into the discourse. It measured social support, in general and in the workplace, for workers with disability, compared to non-disabled coworkers. It also compared perceived stigma toward persons with a disability held by both groups. The effect of length of time employed and its interactions with disability and perceived stigma were tested. Finally, it obtained respondents' expectations of making friends in the workplace.

The study found that individuals with disabilities had higher levels of perceived stigma than those without disabilities. Disability status did not affect social support in general or at the workplace, but perceived stigma did affect social support in general. Thus, disability affects social support by increasing perceived stigma. Length of time employed moderates the effect of disability on perceived stigma. For individuals with disabilities, but not those without, perceived stigma is reduced over time in these employment settings.

The movement of people with disabilities into work groups is one of the goals of the Americans with Disabilities Act of 1990. One of the ideas behind this legislation is to try to change the attitudes of people by changing their behaviors. The combination of the Americans with Disabilities Act and the Independent Living Movement has helped to give individuals with disabilities the opportunity to participate more fully in society by exercising their right to live in society as whole people who are accepted for what they are, and their right to contribute what they can in the ways that they can (Marks, 1999). Individuals with disabilities are creating "a new sociopolitical definition" of what a disability is (Hahn, 1993).

Goffman (1963) and others propose that the possession of a disability interferes with ordinary social interactions and creates a special class of stigmatized interactions. Susman (1994) defined stigma as "any persistent trait of an individual or group which evokes negative or punitive responses" (p.16). A stigma is perceived to be a form of deviance (Goffman, 1963). As such, fault or blame is attributed to the holder of the stigma. Stigmatized persons are somehow responsible for their predicament. They are expected to act differently because of it, should apologize for it, and preferably, hide it. Holders of the stigma learn and internalize these perceptions. Based on these perceptions, they may experience a sense of shame that affects their interactions with "normals," non-stigmatized persons without disabilities. One study quoted a person with a disability who said, "The biggest single barrier I have encountered is one of attitude, society's attitude towards me, my attitude towards both society and myself. I am suddenly considered by both society and myself to be unequal, unacceptable and inferior human being who needs to be protected, controlled and segregated." (Lee and Rodda, 1994, p.229).

The "normals" respond to the stigmatized issue and their sensed interpretations of the stigmatized individual's feelings by feigning ignorance of the stigma or ignoring the person. Neither response is conducive for positive social outcomes. As Goffman puts it, "the stage is then set for the infinite regress of mutual consideration that Meadian social psychology tells us how to begin but not how to terminate" (Goffman, 1963, p.18). Susman (1994) reported on the negative interplay of stigma and deviance in mixed social interactions and the fact that societal images portray a person with a disability as being broken and in need of repair. The normal's response to stigma is grounded in our societal values of independence and physical beauty, a personal dread of becoming disabled, and our concept of a good and just world. "The research shows that disability throws a wrench into the works, so to speak, and discomforts normals as well as disabled participants in the encounters" (Susman, 1994, p.17).

It has been argued that stigma has been overused in disability research. First, to use a widely inclusive definition of what a stigma is may water down individual effects and cause researchers to miss something small but important. Second, society is ever changing, as is our concept of what is stigmatizing. Finally, a concentration on stigma as deviance encourages concentration on negative interactions. Yet, studies that looked at the positive experiences, the public kindness, shown to a person with a disability (Cahill and Eggleston, 1995) note that public kindness comes with a price. The price is to remain within a socially scripted and limited role in which the person with a disability is not seen as a whole person, but as one who is subordinate and needs assistance.

The concept of liminality has been applied to the study of disability. "Liminal people, as the word denotes, are at a threshold. They are marginal to society - poised perhaps to enter, but still outside its boundaries" (Murphy, Scheer, Murphy and Mack,1988, p.237). It is because of this nonperson or intermediate status, they argue, that interactions are befuddling. The person holding this intermediate status is perceived as potentially dangerous and the response is to isolate the person from society. Here it is not the reaction of the person with the disability to the perceptions of others but the others' preconceived ideas of how to treat those holding the stigma. Murphy et al. argue that the liminal model is more beneficial than that of deviance because it delineates disability as a symbolic system and does not omit the perspective of the subjects, persons with a disability themselves. Stigmatized, deviant, or liminal, those analysts are converging on the same concept from different perspectives, the social interactions between individuals with and without a disability, and the results thereof (Susman, 1995).

The perception that one will be stigmatized can lead to social isolation and lack of employment. Modified labeling theory posits that negative labels lead to negative outcomes for the recipient of the label. The greater the belief people have that they are or will be stigmatized by others, the more they will withdraw from social contact (Link, Struening, Cullen, Shrout, and Dohrenwend, 1989). Link et al., studied mental disorders and found a relationship among the level of belief of rejection, coping responses adopted and social networks of the mental patients.

Modified labeling theory and the effects of perceived and experienced stigma have been supported in many studies. Stigmatized interactions were found to complicate the lives of mental patients with a dual diagnosis of substance abuse, even after an improvement in their health had been achieved (Link, Struening, Rahav, Phelan, and Nuttbrock, 1997). Stigma was found to have long-term negative effects on the self esteem and social outcomes of mental health patients (Wright, Gronfein, and Owens, 2000). Various forms of illness and its associated stigma, HIV-AIDS (Crandall and Coleman; 1992; Demi, Bakeman, Sowell, Moneyham and Seals, 1997; Pierret, 2000), a combination of HIV -AIDS and cancer (Fife and Wright, 2000), negatively impacted the social relationships of patients and their caregivers. The negative reactions included an increase in depression and an increased sense of burden for both patients and caregivers.

A study of the interactions of perceived stigma and facial pain on social relationships also found that higher levels of perceived stigma, held by the person with the pain, increased the estrangement level in their social relationships (Lennon, Link, Marbach and Dohrenwend, 1989). Research on gendered stereotypes and stigma found that "(p)eople high in stigma consciousness were more likely to perceive discrimination directed toward their group and toward them personally and were more likely to provide sound evidence for the perceptions" (Pinel, 1999, p. 126). High levels of perceived stigma have been found to lower self-esteem, increase depression and negatively affect social relationships. From the perspective of modified labeling theory, we would expect that the greater the degree of perceived stigma toward people with disabilities, the less likely an employee with a disability would be to engage in the interactions at the workplace that lead to workplace friendships.

Cognitive theory (Allport, 1954) proposes that to change the attitudes of people toward others, there needs to be contact between them. The placement of people with disabilities into the workplace is thought to create points of contact for social interactions to occur, thereby normalizing the interactions. Contact would change the attitudes held by people and attitude change would lead to behavior change. One of the intended benefits of moving persons with a disability into the mainstream workplace was to create more fulfilling lives for them, by increasing opportunities for friendship development. The underlying assumption is that people with disabilities have less of a support system in general and at the workplace compared to the social support systems of coworkers without disabilities.

Allport's Intergroup Contact Hypothesis (1954) gives four critical conditions to ensure positive outcomes of intergroup contact: equal group status within the situation, common goals, intergroup cooperation, and support of authorities, law or custom. Allport's hypothesis was extended by Pettigrew (1998) with the additional factors of time and friendship potential. A work group environment meets Allport's condition of common goals because there is a job that must be completed. Intergroup cooperation and support of authorities are factors necessary for successful job completion. Equal group status is met if the members of the work group are all at the same level, all employees or all supervisors; if not, the inconsistency must be controlled for in the analysis. Work groups also have the potential for friendship development, which Pettigrew felt was a necessity. The time factor is the length of time employed.

Research has supported the contact theory both directly and indirectly. Executives responsible for hiring decisions were found to have more positive attitudes toward people with disabilities if they had contact with the employee with a disability (Levy, Jessop, Rimmerman and Levy, 1992). Employers who had contact with employees with a disability were found to have generally more favorable attitudes toward workers with a disability than employers who had no contact (Kregel and Unger, 1993). Contact theory has been applied to the population of children with disabilities. It was found that attitudes did change favorably, more so for males than females. In general, females held more positive attitudes both before and after contact (Slininger, Sherrill and Jankowski, 2000). Adults with and without disabilities were compared in the outdoor adventure recreation arena to test the contact theory. Despite the challenges of the wilderness, an increase in relationship development and social integration occurred along with more positive attitudes toward persons with disabilities (Anderson, 1994). From this perspective, we expect that a newly employed person with a disability will be less socially integrated than a person without a disability. Over time, any differences in social integration between the two types of employees will be reduced as contact between the groups leads to social comfort and ultimately provides a basis for friendship.

The purpose of this study was to test hypotheses based on the stigma literature, modified labeling theory and contact theory by comparing the levels and interrelationships of perceived stigma and social integration over time among employees with disabilities, to the levels and interrelationships of perceived stigma and social integration among their coworkers who do not have disabilities. This addresses a significant gap in the disability literature.

Previous research on the social integration of individuals with disabilities at the workplace were often observational studies with counts of the number of work-related and social types of communication and behaviors exhibited by both the employees and their peers towards each other (Parent, Kregel, Metzler, and Twardzik, 1992; Parent, Kregel, Wehman, and Metzler, 1991). Other studies have looked at the social, vocational and communication skills of the workers with a disability (Lignugaris/Kraft, Salzberg, Rule, and Stowitschek, 1988; Storey and Provost, 1996). The assumption has been that as long as the workers with a disability speak or are being spoken to, if they greet or are greeted, the goal of social integration has been met. What the person with a disability thinks about the issue, and how this compares to their coworkers' opinions, has rarely been addressed.

Studies that did question the person with a disability were usually interested in their assessment of the job performance of those who provided services to them (Gilson, Bricout, and Baskind, 1998; Pless, 1997). A face-to-face interview intent on discovering the work and disability experience of the participants was done from an occupational social work perspective with employees who had a history of mental illness (Akabas and Gates, 2000). The HELIOS Programme (which stands for "handicapped people in Europe living independently in an open society") analyzed the ways that persons with disabilities actively participated in the program or not. The authors felt that:

(T)here is an urgent need to ensure greater participation on the part of disabled people in these programmes as one small but achievable step towards their more active involvement in society....The sense of involvement and participation, a more informed understanding and awareness of issues, and the new knowledge and confidence acquired as a result, are all important steps towards achieving empowerment, (Powell and Lovelock, 1997, p.578).

The initial assumption in the literature implies there exists a difference in level of social integration based on disability status. The difference suggests a problem in need of repair. The question arises, Is there a difference? Research to date appears to have left persons with disabilities out of the discourse. To help address this gap, the following specific hypotheses were tested:

1. Employees with disabilities perceive more stigma toward people with disabilities than do employees without disabilities. The level of perceived stigma held toward people with disabilities by both groups will be measured. The assumption is that employees with disabilities will perceive more stigma toward persons with disabilities than do employees without disabilities. Persons with the disability are the recipients of the negative attitudes; they have the experiences. Unless a non-disabled person -- a TAB ("temporarily able bodied," a term used by one of the study participants with a disability) -- knows a person with a disability fairly intimately, he or she does not fully understand the import, nor the frequency encountered, of stigma on daily life. Even the most common comment of workers without disabilities heard during the interviews, "I can't believe that people will speak to me instead of him/her when we are out together," only skims the surface.

2. The length of time employed reduces perceived stigma toward people with disabilities for employees with disabilities more than for employees without disabilities. If Allport's Contact Theory holds true, then the length of time employed will reduce perceived stigma toward people with disabilities more for employees with, than for those without, disabilities. For employees without disabilities, in fact, contact with coworkers who experience stigma due to their disability may actually increase perceived stigma overtime.

3. Employees with disabilities have lower levels of social support in general than do employees without disabilities. The assumption is that a person with a disability is unable to have a normal social life due to physical constraints and societal attitudes. They are liminal people, as Susman suggests, and are placed outside the normal boundaries of social interactions.

4. The length of time employed increases social support in general more for employees with, than for those without, disabilities. Over time and armed with new experiences of acceptance and successful social interactions at the work place, the person with a disability will have new social skills and higher levels of self esteem which will translate into higher levels of social support. No change is expected in those without disabilities.

5. Among employees with (but not those without) disabilities, perceived stigma toward people with disabilities reduces social support in general. The expectation is that perceived stigma will reduce the amount of social support for the possessor of the stigma. Link et al.'s 1989 study implies that employees with disabilities have lower levels of social support overall than do employees without disabilities because of the higher level of perceived stigma held by persons with disability, and its negative effect on their social interactions.

6. Employees with disabilities have lower levels of social support from coworkers than do employees without disabilities. This is a carry-over from the personal to the public life, extending the second hypothesis.

7. The length of time employed increases social support from coworkers more for employees with, than without, disabilities. Again, over time, the person with the disability will experience an increase in social support from coworkers because of contact that is conducive to friendship development and a reduction in perceived stigma. No change is expected in those without disabilities.

8. Among employees with (but not those without) disabilities, perceived stigma toward people with disabilities reduces social support from coworkers. The negative effect that perceived stigma has on the level of social support in general and between coworkers will be greater for those with disabilities than those without.

In addition to these hypotheses, the study addresses the research question of whether or not there is a difference in the value placed on work relationships between individuals with a disability and those without.

The research design used a survey questionnaire containing 39 questions that took from 20 minutes to one hour to complete; most participants finished in less than 30 minutes. The questionnaire includes queries about perceived stigma, available social support, personal opinions of social support from the workplace and demographic characteristics. The definition of disability used for inclusion in the study is that of a physical disability or impairment. A disability of mental illness or mental retardation was excluded from the sample. The study participants' disabilities included, but were not limited to, visual and hearing impairments, a loss of one or more limbs, cerebral palsy, and spinal cord injuries.

The sample for this study was one of convenience. Participants (persons with disabilities who were employed part- or full-time and their associated coworkers) were recruited from supported employment agencies, self-help agencies, service providers and independent businesses. No misleading cover story was used. The place of employment was contacted to confirm that proper consent was given to allow all interested participants from each work site to complete the survey questionnaire at the place of business. Approval was received from all but one site, whose employees offered to meet during their lunch hour. All participants signed the required consent forms. Participants with a visual impairment and their associated coworkers had the questionnaires read to them and the interviews were taped. All other participants self-administered the questionnaire.

The requisite criteria for beneficial contact outlined by Allport were met in the settings in which the research was conducted. Equal group status was confirmed by using the job classification of the respondents and their income level. Any hierarchical differences within the groups surveyed were controlled for in the analysis. At five of the sites, the person with the disability was the manager, director, company owner or in some way a lead worker. The reasons for this included workers with adult onset of a disability and who had remained in their existing managerial position; and business owners who had disabilities. The agencies that had a requirement that a large percent of their staff be persons with disabilities appeared to have their hierarchical structure fairly mixed, because they were fairly small work sites.

The frequency of interactions among group members was examined by questioning how often the coworkers converse during a typical workday. The exact type of conversation, whether work-related or non-work related, was not defined. Intergroup cooperation appeared (to the investigator) to have been within the normal range found at any work site. Cooperation is required of any successful work group in order to get the job done. Getting the job done is the common goal of a work group, whether they have a service to provide or a product to produce.

Bosses, the authorities, need successful completion of the work, therefore they will encourage positive intergroup cooperation. Support of authorities has also been ensured by the American with Disabilities Act (ADA) and the many lawsuits that have occurred since its enactment. All of the workers I met were very conscious of ADA, aware of possible repercussions if it is not adhered to, and its overall effects were reported as positive.

The length of time employed is the time factor described by Pettigrew and is a major variable in the analysis.

(Note: Tables presenting the data described below, and the survey questionnaire, are available on request from the author.)

Fifty-one percent of the sample were persons with disabilities. The groups, employees with disabilities and those without, were found to be similar in terms of most of the characteristics measured. Results of paired samples t-tests showed that within both groups, the average respondent was in his/her forties, and married or living with a partner. He/she had a post secondary education, was employed full-time in a professional or managerial position, had a monthly salary of $2600, had been employed at the current job for more than six years, and at some job for more than 20 years. Over half of the respondents were female. Thirty-three percent of employees with disabilities and 22% of those without, reported being Protestant. The other religion options were: Catholic (33% of persons with disabilities, 44% percent of non-disabled), Judaic (six percent of persons with disabilities, none of non-disabled), other (24% of persons with disabilities, 27% of non-disabled), or none.

The only demographic variable that was found to differ significantly between the groups was the self-reported health rating. Those without a disability rated their overall health more favorably (mean, 4.06) than those with a disability (mean, 3.39).

Disability. The disability status of the respondent was determined by simply asking respondents if they have a disability or have ever been told that they have one. Respondents with disabilities were asked to describe their disability in several subsequent questions. Respondents with a disability were asked to answer "yes" or "no" as to whether or not they considered their disability to be obvious to others. Sixty percent responded positively. I found it interesting that those who reported their disability as being not obvious included persons who were deaf and those who were blind. Twenty-seven percent reported onset of the disability occurring prior to age 12. The severity of the disability, measured by a self-report scale of perceived severity, slight (1) to very severe (4), revealed a mean of 2.13 with a standard deviation of 1.04. The respondents with a disability appeared, to me, to have quite optimistic attitudes toward their disabilities. Both paraplegic and totally blind respondents referred to the severity of their own disability as being slight.

Response categories for a measure of limitations in "Activities of Daily Living (ADL)" were: quite a bit (2), a little bit (1) or none (0). Scores were summed, with the maximum possible score of 20. The average response was 3.69 with a standard deviation of 4.08.

Stigma. All items, except trustworthiness, from the stigma scale created by Link, et al. (1989) for their study of mental patients were adapted to persons with disabilities. Four items were added to the scale - whether or not most people believe a person with a disability is as independent, motivated or hard working as the average person; whether or not the possession of a disability makes personal failure more likely; whether or not people feel sad when meeting a person with a disability; and whether or not people feel nervous and/or awkward when meeting a person with a disability. The responses to the six-category Likert type scale were coded or were recoded so that (1) represents the lowest level of perceived stigma and (6) the highest. The observed scores averaged 51.52 with a standard deviation of 11.59. The alpha coefficient for the adapted scale is .88, compared to Link's alpha of .71.

Social Support. A measure of the level of social support received in general by the respondents was adapted from Fishers' social support scale (1982). The queries included whether or not the respondent had someone: who would help with household tasks, housesit during vacation, baby-sit their children, to talk to about personal issues, to socialize with, and/or borrow money if needed. The response categories were yes, no, or not applicable. Because of the large number of respondents without children (31), the babysitting item was omitted from the measure. All other items were recoded to yes (1) or no (0) because all were potentially applicable to all respondents. The summed value of (0) is no social support and (7) is the maximum support possible. The mean is 6.22 with a standard deviation of 1.62. The adapted scale has a reliability alpha of .86. The measure was again adapted to reflect the amount of social support received from coworkers, a mean of 4.69, a standard deviation of 2.16 and a reliability alpha of .82.

Paired sample t-tests showed that both for individuals with disabilities and those without, the scores were significantly higher on the measure of social support in general than on the social support at work (individuals with disabilities, t = 4.70, p<.001; non-disabled, t = 3.39, p<.01). A repeated measures ANOVA showed, further, that the amount of difference between the two measures did not differ significantly for the two groups (F = 0.75, p = .39).

Workplace as a source of friendship. The final question was whether or not participants consider the workplace to be a source for new friendship development, or they prefer not to socialize with coworkers. The question was open-ended and their comments were encouraged. The opinions expressed here are used throughout the findings section to help interpret the quantitative results.

Control variables. Because of the non-random nature of the sample, it is important to control for all variables that might possibly cause a spurious relationship between disability status and any of the three dependent variables regardless of the significance levels. Therefore, demographic variables with a Pearson correlation of .15, or greater, with disability status and at least one of the dependent variables (stigma, social support in general, or social support from coworkers) were used as control variables, regardless of the significance of the correlation. Variables with lower correlations would not be expected to cause spurious relationships between disability status and other variables. Based on this criterion, gender [male (0), female (1)], education (in years), and type of job [professional or managerial (1), other (0)] were chosen as the control variables.

The analysis used the Statistical Package for the Social Sciences (SPSS) 10.1. Multiple linear regressions were run to test the effect of the independent variables and the demographic control variables on the dependent variables: stigma (Equation 1), social support in general (Equation 3) and social support at the workplace (Equation 8). Interaction terms were created by multiplying disability status, first with the length of time employed, and then, the level of perceived stigma. The interaction terms were added to the regression runs (Equations 2, 4 & 9).

Hypothesis 1: Equation #1 Persons with a disability had higher levels of perceived stigma than those without disabilities, (B = 7.41, SE = 2.85, p<.05). This supports the Modified Labeling Theory that persons who are the object of the stigma are more aware of it and its repercussions. Responses of workers with a disability to the opinion question about socialization at work noted that they were well aware that they have something negative to overcome:

We have to work twice as hard to prove to ourselves and others that we are just like anyone else on the job and we are sometimes expected to do more work than others. Sometimes an able bodied person does not realize how hard it is to work as a normal person can. Case #1200

They also felt a responsibility to help to reduce stigma:

I feel that as a disabled person it is important to project a positive attitude because people will, more than likely, return a positive attitude. If a disabled person projects a positive attitude, then someone who hasn't ever had any contact with someone with a disability will be more accepting to them as opposed to a disabled person who is bitter, angry and negative. Case #1504

Hypothesis #2: Equation #2. Analysis related to hypothesis #2 shows that the length of time employed moderates the difference between employees with and without disabilities in terms of the perception that individuals with disabilities are likely to be stigmatized by others. The interaction of disability status and length of time employed was negative and relatively large, (B = -1.14, SE .39, p<.01). The length of time employed reduces perceived stigma more for employees with, than those without, disabilities. This provides clear support for Allport's theory in that while employees with disabilities perceive more stigma than employees without disabilities, this difference is reduced through workplace contact. Subsequent tests in which the length of time employed and background variables were regressed on stigma, separately for each group, showed that perceived stigma decreased with the length of time employed for employees with disabilities (B = -.80, SE = .33, p<.05). For employees without disabilities, perceived stigma increased with the length of time employed in these settings where at least one coworker has a disability (B = .43, SE = .24, p= .09). Thus, it appears that employees with disabilities learned to trust employees without disabilities and employees without disabilities developed a more empathetic view of the way those with a disability are treated by others.

Hypothesis 3: Equation #3. Results related to hypothesis #3 indicate there was no significant impact of disability group membership on the level of social support in general, B = -.19, SE = .45.

Hypothesis 4: Equation #4. The length of time employed also did not increase social support in general more for either group (B = -.01, SE = .04) and disability status did not moderate the relationship between length of time employed and social support as hypothesized. The interaction term for disability X length of time employed was small and insignificant.

Hypothesis 5: Equation #4. The level of perceived stigma did not significantly affect the level of social support in general for employees with or without disabilities, (B = -.02, SE = .03).

In summary to this point, the findings indicate that neither disability status nor stigma affect social support in general. This seems odd since there is a bivariate correlation between stigma and social support from this source (r = -.28, p < .05). It is possible that including both disability status and stigma in the same analysis has masked the impact of stigma on social support due to the correlation between disability status and stigma reported above. Therefore, the regression analyses were rerun with disability and stigma in separate equations.

The data show that disability status still fails to affect social support in general even when stigma is not in the equation (B = .67, SE = 1.22, p = .59). Furthermore, the interaction term for disability and length of employment is not significant (B = .18, SE = .17, p = .31). Persons with a disability, like those without, appear to have an existing social support network and while the participants therein may fluctuate, the overall amount does not increase over time because of workplace interactions.

I have always developed workplace friendships and sometimes they move into out-of-work friendships...Most times, workplace friendships remain so because outside of work we have little in common due to age and financial status. Case #1104

However, when stigma is included without disability, its impact on social support in general is significant (B = -.04, SE = .02, p < .05). The greater the amount of stigma toward people with disabilities an employee perceives, the lower the level of social support in general. Since disability status and stigma cannot be entered together, the only way to test for their interaction was to repeat the analysis for the two groups separately. While the regression coefficients for stigma were not significant in either group, because of the reduced sample size, they remained negative both for employees with disabilities (B = -.04, SE = .03) and those without (B = -.02, SE = .03). It appears that people who are pessimistic about how others treat individuals with disabilities have fewer friends in general, whether or not they themselves have a disability. Conversely, it may be that those with few friends perceive more stigma.

Hypothesis 6: Equation #8. Disability status did not significantly affect the level of social support from the workplace, (B = -.03, SE = .02).

Hypothesis 7: Equation #9. The length of time employed was not shown to increase the social support from coworkers more for employees with than for those without disabilities; again the results were not significant, (B = .05, SE = .09).

Hypothesis 8: Equation #9. The level of perceived stigma did not significantly affect the level of social support from coworkers for employees with disabilities, (B = -.05, SE = .04).

As in the case of social support in general, to test for multi-collinearity effects, these regression analyses were rerun with disability status and stigma entered separately. The results show that disability status does not affect social support from co-workers, (B = -2.67, SE = 1.87, p = .16).

Also, unlike applied to social support in general, social support from co-workers is unrelated to perceived stigma (B = -.04, SE = .03, p = .12). Subsequent tests show virtually identical results for employees with disabilities (B = -.04, SE = .04, p = .33) and those without (B = -.04, SE = .04, p = .32). Thus, regardless of length of employment, perceived stigma does not affect social support from co-workers.

Perhaps factors other than stigma affect relationships between coworkers with disabilities and those without. The lack of a complete knowledge of, or the unwillingness of those without a disability, to overcome the hardships of the built environment for the person with the disability may play a role. In this post-modern world, we all have busy lives and it is hard to go the extra mile to create a new friendship whether or not that person has a disability. Something has to trigger the effort:

There are some people at work I would never care to have as friends and others with common interests that have become friends as you do spend eight hours a day with them. Case #3200.

I feel that a person's place of employment should be enjoyable and comfortable to them. If that includes making friendships with other co-workers, I think that is terrific! It just depends on the individual person. Case #1204.

One factor that could have affected the relationships among disability, stigma and social support is the type of employment settings in which the study was conducted. The self-help agencies, six sites, were the most willing to participate and recommend referrals. The agencies also had more persons with a disability on staff. This was found to be a requirement of the by-laws of two sites. Government jobs, including schools and regulatory agencies, counted for another six sites. Here it was the person with the disability who held a position at the site who functioned as the gatekeeper. Four businesses participated, again, because my first contact was with the person with the disability. In order to test for the potential impact of employment settings on the results related to social support and stigma, the employment settings were coded, agencies (1), government and business (0), and the analysis reported above were replicated with type of employment as an additional control variable. The findings indicate no substantive changes. Setting, then, does not appear to have affected the outcome.

Attitude about workplace friendships. Responses to the open-ended question of whether or not work is a good place to find friends were coded as a dichotomous variable: workplace a good place to find friends was coded (1), the opposite was coded (0). Responses that were not clearly positive or negative about this point offer were coded as "missing." An independent sample t-test found that respondents with a disability (N=29), had a mean of .86, standard deviation of .35. The mean for the non disabled group (N=27) was .85 with a standard deviation of .36. There was no significant difference, t (54) = -.11, p = .92. Having a disability does not affect expectations of workplace socialization in general. The responses were then analyzed for themes.

Some respondents do not seek work friendships because their social lives are already full:

I have chosen not to socialize outside of the workplace. Occasionally, but not much. I have a full active social life without adding to it. I have a husband. Not really enough time - I have a large family, too. I haven't noticed a lot of relationships to develop in the workplace. I feel as if I am friends with the people at work, but don't carry it over to personal. Case #1301.

Others place a high value on workplace friendships:

I feel that friendships with fellow workers is an important part of life, it not only helps with daily activities at work, but it helps both people see each other in a different light. Case #2302

Still others say it depends on the circumstances and the attitude of co-workers:

I feel you should be friendly as far as the other person allows it. Everyone should be "business friendly." One should follow the natural flow of what is allowed. Case #2200.

Whether disabled or not, people tend to add to their sphere of friends the people whom they meet at work - if there are common interests outside of the work relationship:

If there is an opportunity to build a friendship I would facilitate it. However, I do not seek out employment with friendships in mind. Case #1503

The person with the disability was aware of the need to reach out to coworkers however:

A person is responsible for themselves on a social level. Each person, regardless of disability is personally responsible to initiate conversations and establish friendships and working relationships with those around them. People with disabilities have to work on this even harder than someone without a disability. The person with a disability needs to overcome the initial hesitancy of others to know what to say. Case #1404

The findings are not generalizable to the general population because of the convenience sample and its small size. The broad spectrum of disabilities used in this study may have affected the levels of perceived stigma and social support, as may the high numbers of those with adult onset of a disability. Saranson's developmental model of perceived social support in Ross, Lutz, Lakey (1999) found that it is in early childhood that the basic perceptions of social support are formed. Those with adult onset of a disability, age 12 and over in this study, 12.3%, would not have had the issue of disability negatively affect their early perceptions of social support as those with an early onset of a disability would because they experienced a non disabled childhood.

Other issues that may well have contributed to the lack of differences in the groups on the social support measures include the possibility that persons with disabilities who have very high levels of perceived stigma are often not to be found in the workplace. Farina, Gliha, Boudreau, Allen and Sherman (1971) found that mental health patients who believed that coworkers were aware of their psychiatric history felt less valued, performed tasks less adequately and reacted so negatively to their perceptions that a neutral observer rated them as being "more tense, anxious, and more poorly adjusted" (Farina et al., p.4). Have they withdrawn from the possibility? Link et al.'s withdrawal step is where social interactions are limited to only those "who know and accept one's stigmatized condition" (1989, p.403). One could argue that those already in the workplace are a special subset of persons with a disability with strong wills.

A larger sample size with more general businesses participating would give a truer picture of the work-a-day world. A larger sample would also allow the differing types of disabilities to be analyzed separately. The role played by the severity or the obviousness of the disability should be analyzed. The effects of the age of onset of the disability could be teased out also. Despite these limitations, however, findings presented here make several important contributions to what we know about the impact of disability, stigma and workplace contacts on the lives of people with disabilities.

First, neither group considered the workplace to be a major source of new friends, unless the coworkers held common interests and were the type of individuals who would be chosen as friends in any situation. This does not mean that a workplace is not a place from where individuals augment their support group. Support systems can be fluid and include family, friends, church groups, or work groups. Support groups are important because they function as buffers between the individual and the larger world helping maintain one's mental health. (Caplan, 1974). They are a place to go for friendship, solace, or when one needs someone to talk to. The important thing is that everyone have an adequate supply to draw from for a personal need that may arise (Fischer, 1982).

Second, the workers with disabilities did perceive more stigma than those without disabilities. This finding supports the Modified Labeling Theory in that the perceptions shared by all are actually intensified for the bearer of the stigma. Disability status affects the level of social support in general through the effect of perceived stigma. One of the most interesting findings was that for those without disabilities the higher their level of perceived stigma, the lower their level of social support in general. It would be intriguing to explore whether or not this holds true for other types of stigmatized classifications.

Third, the length of time employed did not affect the level of social support in general or at the workplace, for either group, but it did reduce the level of perceived stigma for those with a disability. Since perceived stigma is inversely related to social support, this supports Allport's Contact theory in that those with a disability learned to trust their non-disabled coworkers. This should lead to unencumbered interactions and eventually to higher levels of social support. Most people require a source of income and work is the place where we sell our labor for income. Allport stated that "only the type of contact that leads people to do things together is likely to result in changed attitudes." (1954, p.276) A work group is a setting of forced contact that meets Allport's criteria in that it forces people to interact, do things together, toward the successful completion of a common goal. Thus, the workplace offers "forced" opportunities for people with and without disabilities to interact with one another which, over time, reduced the expectation of the workers with disabilities that they will be stigmatized.

Fourth, another point of interest is that, over time, the persons without disabilities experienced an increase in their level of perceived stigma. During the interviews the fact was shared, more than once, that the coworker with the disability was the first individual with a disability met by the coworker without a disability. This had given them the opportunity to learn about life with a disability and it had opened their eyes to the amount of, and effect of, stigma that an individual with a disability actually deals with in a normal day. During the course of the interviews, there were side comments made by non-disabled workers as to how enlightening the experience of working with a person with a disability had been. Putting a face on the term allowed them to empathize. This finding argues against the casual order implied by Modified Labeling Theory (Link et al., 1989) in that it shows that stigmatizing labels can be shed. If it could be shown that stigmatizing labels do not have to solidify into traits, it might encourage the enactment of more policies such as the ADA that encourage contact among stigmatized and non-stigmatized groups, regardless of the stigmatizing factor.

Akabas, Sheila H. and Lauren B. Gates (2000). "A Social Work Role: Promoting Employment Equity for People with Serious and Persistent Mental Illness," Administration in Social Work, 23(3/4):163-184.

Allport, Gordon W. (1954). The Nature of Prejudice, Addison-Wesley Publishing Company, Inc., Cambridge, MA.

Anderson, Lynn S (1994). "Outdoor Adventure Recreation: A Social-Psychological Perspective," Dissertation, University of Minnesota.

Cahill, Spencer E. and Robin Eggleston (1995). "Reconsidering the Stigma of Physical Disability: Wheelchair Use and Public Kindness," The Sociological Quarterly, 36:681-698.

Caplan, Gerald, M. D.(1974). Support Systems and Community Mental Health: Lectures on Concept Development, Behavioral Publications, New York.

Crandall, Christian S. and Robert Coleman (1992). "Aids-Related Stigmatization and the Disruption of Social Relationships," Journal of Social and Personal Relationships, 9:163-177.

Demi, Alice, Roger Bakeman, Richard Sowell, Linda Moneyham, and Brenda Seals (1997). "Effects of Resources and Stressors on Burden and Depression of Family Members Who Provide Care to an HIV-Infected Woman," Journal of Family Psychology, 11(1):35-48.

Farina, Amerigo, Donald Gliha, Louis A. Boudreau. Jon G. Allen, and Mark Sherman (1971). "Mental Illness and the Impact of Believing Others Know About It," Journal of Abnormal Psychology, 77(1):1-5.

Fife, Betsy L. and Eric R. Wright (2000). "The Dimensionality of Stigma: A Comparison of its Impact on the Self of Persons with HIV/AIDS and Cancer," Journal of Health and Social Behavior, 41:50-67.

Fischer, Claude S. (1982). To Dwell Among Friends: Personal Networks in Town and City, The University of Chicago Press, Chicago & London.

Gilson, Stephen French, John C. Bricout, and Frank R. Baskind (1998). "Listening to the Voices of Individuals with Disabilities," Families in Society, 79(2):188-196.

Goffman, Erving (1963). Stigma: Notes on the Management of Spoiled Identity, Simon & Schuster, Inc., New York.

Hahn, Harlan (1993). "The Potential Impact of Disability Studies on Political Science (as well as vice versa)," Policy Studies Journal, 21(4):740-751.

Kregel, John, Darlene Unger (1993). "Employer Perceptions of the Work Potential of Individuals with Disabilities," Journal of Vocational Rehabilitation, 3(4):17-25.

Lee, Tatia and Michael Rodda (1994). "Modification of Attitudes Toward People with Disabilities," Canadian Journal of Rehabilitation, 7(4):229-238.

Lennon, Mary Clare, Bruce G. Link, Joseph J. Marbach and Bruce P. Dohrenwend (1989). "The Stigma of Chronic Facial Pain and its Impact on Social Relationships," Social Problems, 36(2):117-134.

Levy, Joel M., Dorothy Jones Jessop, Arie Rimmerman and Philip H. Levy (1992). "Attitudes of Fortune 500 Corporate Executives Toward the Employability of Persons With Severe Disabilities: A National Study," Mental Retardation, 30(2):67-75.

Lignugaris/Kraft, Benjamin, Charles L. Salzburg, Sarah Rule, Joseph J. Stowitschek (1998). "Social-Vocational Skills of Workers With and Without Mental Retardation in Two Community Employment Sites," Mental Retardation, 26(5):297-305.

Link, Bruce G., Elmer Struening, Francis T. Cullen, Patrick E. Shrout, and Bruce P. Dohrenwend (1989). "A Modified Labeling Theory Approach to Mental Disorders: An Empirical Assessment," American Sociological Review, 54:400-423.

Link, Bruce G., Elmer I. Struening, Michael Rahav, Jo C. Phelan, and Larry Nuttbrock (1997)."Stigma and Its Consequences: Evidence from a Longitudinal Study of Men with Dual Diagnosis of Mental Illness and Substance Abuse," Journal of Health and Social Behavior, 38:177-190.

Marks, Deborah (1999). Disability: Controversial Debates and Psychological Perspectives, London: Routledge.

Murphy, Robert F., Jessica Scheer, Yolanda Murphy and Richard Mack (1988). "Physical Disability and Social Liminality: A Study in the Rituals of Adversity," Social Science and Medicine, 26(2):235-242.

Parent, Wendy S., John Kregel, Paul Wehman, and Helen M. D. Metzler (1991, January). "Measuring the Social Integration of Supported Employment Workers," Vocational Rehabilitation, 35-49.

Parent, Wendy, John Kregel, Helen Metzler, and Gary Twardzik (1992, March) "Social Integration in the Workplace: An Analysis of the Interaction Activities of Workers with Mental Retardation and Their Co-Workers," Education and Training in Mental Retardation, 28-38.

Pettigrew, Thomas, F. (1998). "Intergroup Contact Theory," Annual Review of Psychology, 49:65-85.

Pierret, Janine (2000). "Everyday Life with AIDS/HIV: Surveys in the Social Sciences," Social Science and Medicine, 50:1589-1598.

Pinel, Elizabeth C. (1999). "Stigma Consciousness: The Psychological Legacy of Social Stereotypes," Journal of Personality and Social Psychology, 76(1):114-128.

Pless, Michael E. (1997, April). "The Experiences of Visually Impaired People with Their Vocational Rehabilitation Counselors," Unpublished Masters Thesis, University of South Florida.

Powell, Jackie and Robin Lovelock (1997). "Participants or Recipients - Disabled People's Involvement in a European Programme," British Journal of Social Work, 27:565-583.

Ross, Lisa Thomson, Catherine J. Lutz and Brian Lakey (1999). "Perceived Social Support and Attributions for Failed Support," Personality and Social Psychology Bulletin, 25 (7):896-909.

Slininger, David, Claudine Sherrill and Catherine M. Jankowski (2000). "Children's Attitudes Toward Peers With Severe Disabilities: Revisiting Contact Theory," Adapted Physical Activity Quarterly, 17:176-196.

Storey, Keith and O'Neil Provost (1996). "The Effect of Communication Skills Instruction on the Integration of Workers with Severe Disabilities In Supported Employment Settings," Education and Training in Mental Retardation and Development Disabilities, 123-141.

Susman, Joan (1994). "Disability, Stigma and Deviance," Social Science and Medicine, 38 (1):15-22.

Wright, Eric R., William P. Gronfein, and Timothy J. Owens (2000). "Deinstitutionalization, Social Rejection, and the Self-Esteem of Former Mental Patients," Journal of Health and Social Behavior, 41:68-90.