The Multidimensional Character of Biased Perceptions of Individuals with Disabilities.

The Americans with Disabilities Act of 1990 has sensitized many individuals in business and industry to the fact that discrimination aimed at individuals with disabilities is as unjust as discrimination against ethnic and racial minorities. The Civil Rights Act of 1964 and the moral consciousness it aroused helped eliminate many of the discriminatory employment practices directed toward minorities. Individuals with disabilities have not been as fortunate in terms of employment over the past 35 years. While African-Americans and other racial minorities have made gains toward equality in employment, a significant percentage of the American disabled population remains unemployed. The Senate Committee on Labor and Human Relations reported that 67% of individuals with disabilities between the ages of 16 and 64 are unemployed (Simpkins & Kaplan, 1991). Fifty-eight percent of all disabled men and 80% of all disabled women are unable to find gainful employment (Thornburgh, 1991). Furthermore, Harris and Associates (cited in Satcher & Hendren, 1991) found that approximately 66% of the unemployed individuals with disabilities that they surveyed wanted to work, suggesting that these unemployment figures do not simply reflect self-selection out of the workforce.

Despite such statistics, most research on discrimination in the workplace has continued to center on either race or gender-based discrimination. Little research is found on determining the current level of discrimination or on determining the possible causal and/or mediating variables that lead to the biased perceptions of an individual with a disability. An additional concern regarding much of the research on bias and discrimination is that many scales designed to measure attitudes toward individuals with disabilities rely on unidimensional measures of affect toward the "disabled." Scale developers have concentrated on unidimensional measures that seem to assume constant "bias potential" within and even across disabilities. That is, an assumption is made that individual's with a given disability are perceived similarly by a non-disabled individual. This approach also often makes the unlikely assumption that individual's with vastly different disabilities are perceived equally by the non-disabled. This approach has resulted in scales such as the Attitudes Towards Disabled Persons Scale (Yuker, Block, & Younng, 1966) and the Interaction with Disabled Persons Scale (Gething, 1994) that ask questions about "disabled" people [sic] in general terms in an effort to generate and then measure attitudes of the test-taker. Moreover, these scales typically use the computation of "total scores" on these instruments. This practice has sometimes led researchers and practitioners to erroneously assume that a single dimension can account for perceptions of an individual with a disability.

An estimated 43 million individuals were covered by the Americans with Disabilities Act when it was originally introduced (Pierce, 1991). These 43 million individuals are not a homogenous group and perceptions of them are likely to vary as well. An individual's perception of a worker with a learning disability is likely to be quite different than their perception of someone with alcoholism or epilepsy. Although some perceptions toward individuals with disabilities may have a common factor, it is not likely the entire explanation. It seems more plausible that many interrelated factors determine an individual worker or employer's attitude toward a given individual with a disability.

Cursory evidence exists in the literature to support the idea that such underlying factors are important, and that, a single generalized attitude does not provide a complete or satisfactory explanation of perceptions of the person. Bordieri and Drehmer (1986) found that if a job applicant was personally responsible for their own disability that they were likely to receive lower evaluations from raters. Thus, one might expect two individuals with disabilities to be judged differently if one suffered their injury while driving under the influence while the other was disabled at birth.

Another dimension thought to underlie perceptions is the disruptiveness of the disability. Stephens and Clark (1987) have found that people with visible disabilities are given more personal space than those with hidden disabilities. Despite the obvious conclusion that this might mean that those with visible disabilities suffer from lower evaluations, Bordieri and Drehmer (1986) found that hidden disabilities typically result in a greater negative impact in ratings. While the direction of bias remains debatable, the fact remains that the disability's concealability may help determine the attitude of the rater toward the individual.

Another dimension that has received attention is the course of the disability (Stone, Stone, & Dipboye, 1992). This dimension is perhaps best characterized as "closeness to death." There are both social and economic reasons why the course of the disability might bias ratings given by both supervisors and coworkers. Many other factors such as the unpredictability (Drehmer & Bordieri, 1985) and contagiousness of the disability might also be added to the list of potential variables influencing perceptions of individuals with disabilities. These studies support the belief that not all disabilities are perceived equally and that more than a single factor may be necessary to explain individual attitudes toward people with disabilities.

Only by fully understanding the various inter-related dimensions that lead to the formation of such biased attitudes, can unfair evaluations be eradicated. The current study was designed to test the hypothesis that attitudes toward persons with disabilities are not unidimensional. While some multidimensional scales exist, it is important to note that simply factor analyzing existing scales would provide only information concerning the dimensionality of those particular scales, not necessarily the dimensionality of the construct in question. This study attempts to generate a more appropriate potential list of variables that might underlie perceptions of individuals with disabilities.

In order to accurately measure a person's propensity for bias it is important to include all of the relevant dimensions on which such bias may be based. More importantly, to understand how biased perceptions are formed we must go beyond the assumption that the perceptions of all individuals with disabilities are the same. The primary goal of the current research was thus to determine the number and characteristics of the dimensions underlying perceptions of workers with disabilities. A secondary goal was to determine if those underlying dimensions could then be used to predict criterion measures independent of disability-specific connotations (i.e. affect).

Experiment 1

Method

Participants. Eighty-seven undergraduate students at a major southern university completed the pilot study materials. All participants were recently employed and worked at least 20 hours a week. Approximately 57% of the sample was male and the majority of the sample was Caucasian (86%). The subjects were recruited through the university subject pool and received extra credit in psychology classes for their participation.

Instruments. The materials described were developed specifically for the current pilot study and were administered to 50 of the 87 participants. These materials simply asked the subjects to list "all possible concerns" that they would have about certain individuals with disabilities in specific situations. Twelve scenarios were created based on manipulating two separate dimensions. The first dimension was whether the rater was a subordinate, coworker, first-line supervisor, or an employer. The second dimension concerned whether the disability itself was a physical disability, a mental disability, or a communicable disease. Again, by manipulating these two dimensions a total of twelve scenarios resulted.

Procedures. The pilot study materials were designed to elicit possible variables that underlie biased perceptions related to individuals with disabilities. The three various disability types were used in order to prevent participants from basing all of their answers on one particular disability exemplar. The participants were asked to take on different perspectives to provide the participant with an opportunity to express a wider range of possible concerns for a given disability type.

The remaining thirty-seven students (in groups ranging from 2 to 4) were asked open-ended questions in interview format. The questions were asked from a less professional/more personal perspective similar to a focus group. Subjects were asked questions such as "what do you think about an individual with a disability on your intramural team or as a roommate." Once ideas were introduced by participants, the discussion was allowed to continue with minimal interference from the researcher. These discussions were later transformed into statements similar to those provided on the paper and pencil materials.

Results and Discussion

A rigorous quantitative analysis of the data, from both of these exercises was not possible due to its open-ended nature. First, all responses reflecting concepts other than those relevant to the current research were discarded (this would include such statements as" I do not treat the disabled differently," etc.). Next, where possible, each statement was reduced to a one-word concern ("could I catch it" was reduced to contagiousness, etc.). Similar statements were grouped together under these one-word categories. While some judgment calls were required, the statements typically grouped quite naturally into categories. Thus, for each participant, a list of one word "concerns" resulted.

As a check of the researcher's conceptualization, the participants who had engaged in the focus group activities (N-37) were also asked to take the original statements and group them under appropriate headings. While a few statements were inconsistently classified into categories, the majority of the statements (70%) were classified identically across all thirty-seven participants. Of the remaining 30% of the statements, all were associated with a particular category by at least half of the 37 participants. In general, the categorizing frameworks that the participants produced were remarkably similar to the researchers'. A total of sixteen concerns resulted from this exercise. The sixteen concerns and a brief description of each are presented in Table 1.

Table 1

Sixteen Concerns Regarding Workers with Disabilities

Aesthetics (Degree to which disability is physically unappealing)

Causality (Degree to which disability was internally caused)

Concealability (Degree to which disability can be hidden)

Contagiousness (Degree to which disease/disability can be spread
and if so the extent to which it can)

Controllability (Through medication/therapy can disability be
controlled)

Curability (Degree to which disability is curable)

Distractiveness (Degree to which disability is distracting to
either coworkers, employers or clients in a way other than aesthetic)

Preferential treatment (Degree to which if hired disabled person
would receive preferential treatment/special privileges from
workers/boss that might interfere with effective workflow)

Need for assistance (Degree to which disability would necessitate
structural accommodations)

Novelty (How much is known about the disability?)

Oversensitivity (Degree to which disabled person will be overly
sensitive about condition)

Peril (Degree to which death is likely in the near future)

Response of disabled to stressful situation (Degree to which time
dead-line crunch would effect person)

Severity (Degree to which disability interferes with day to
day activities)

Stability (Degree to which disability is static in nature and course)

Work Longevity (Degree to which disability will render the person
unable to continue working in the near future)

One way to determine the dimensions that affect people's attitudes or perceptions of the disabled would be to simply ask people what factors affect their feelings toward individuals with disabilities at work. This straight-forward approach is relatively common in the field of rehabilitation psychology. Unfortunately, such an approach suffers from several shortcomings. First, social desirability could be problematic in such an easily transparent study. Second, even if participants were willing to give honest answers, they may be unable to do so. Individuals may have little knowledge of how to combine information about disabilities in forming attitudes toward a given person. A more sophisticated approach would be to have participants make actual decisions, and then try to capture how they used the independent variables in making those decisions. Methodologically, this approach is known as policy capturing. Madden (1981) and Rossi and Madden (1985) have shown that this methodology was capable of detecting prejudicial attitudes where other more direct measures could not due to socially desirable responding.

Experiment 2

Method

Participants. One hundred and seventy-one introductory psychology students at a major southern university completed the study. Approximately 55% of the sample was male and the majority of the sample was Caucasian (82%). While this lack of diversity may be somewhat troublesome to some, in all likelihood, it mirrors the demographic make-up of the decision-makers at many large firms potentially employing individuals with disabilities. The participants in the current study were recruited through the university subject pool and received extra credit in psychology classes for their participation. As in the first experiment, a necessary condition of participation was current employment of at least 20 hours a week.

Instruments. Two sets of materials were required for the policy capture. Set 1 materials asked participants to indicate the degree of association they felt existed between a person with one of 15 common disabilities and the 16 concerns that resulted from the pilot study. Participants were asked to make this rating on a Likert-type scale ranging from 1 (not associated) to 5 (highly associated). The 15 disabilities were chosen based on two general criteria. First, common disabilities that would be well-known by most participants were preferred over uncommon ones. Second, an effort was made to select a diversity of disabilities (i.e. physical, mental, and communicable diseases). The resulting disabilities that were used as stems in the current study were persons with alcoholism, amputee (arm), AIDS, blindness, cancer, deafness, diabetes, epilepsy, heart disease, learning disorder, leprosy, manic/depressiveness, paraplegia, schizophrenia, and tuberculosis.

The Set 2 materials presented participants with various scenarios and asked subjects how each of 15 disabilities would affect (not likely to very likely) their perception of the individual with the disability. The criterion measures included many variables of interest to organizations. These measures were: (a) perceived effect on working relationship of the workgroup, (b) hiring decision, (c) promotability, (d) tardiness/absenteeism problem, (e) willingness to work with, (f) willingness to work for, (g) trainability, (h) work motivation, and (i) predicted productivity. Examples of both Set 1 and Set 2 materials are presented in Tables 2 and 3, respectively.

Table 2

Dimensions Potentially Influencing Perceptions (Set 1 Materials)

Please indicate the degree to which each dimension is commonly
thought to be associated with this particular
disability:

INDIVIDUAL WITH ALCOHOLISM

Not                                       Highly
Associated                              Associated
1                2       3        4         5

-- Aesthetics (Disability is physically unappealing)

-- Causality (Disability is thought to be person's own fault)

-- Concealability (Disability can be hidden)

-- Contagiousness (Disease/disability can be spread and if so
the extent to which it can)

-- Controllability (Disability can be controlled through
medication and/or therapy)

-- Curability (Disability is curable)

-- Distractiveness (Disability is distracting to either
coworkers, employers or clients in a way other than aesthetic)

-- Preferential Treatment (If hired person would likely receive
preferential treatment/special privileges from coworkers/boss)

-- Need for Assistance (Disability would necessitate structural
accommodations)

-- Novelty of Disability (Disability's nature, causes, cures are
not well known)

-- Oversensitivity (Disability likely to lead individual to be
overly sensitive about own condition)

-- Peril (Disability is likely to result in death in the near future)

-- Response of Disabled to Stressful Situation (Time dead-line crunch
would affect him/her)

-- Severity (Disability significantly interferes with day-to-day
activities)

-- Stability (Disability is static/stable in nature)

-- Work Longevity (Disability will likely make individual unable
to continue working in the near future)
Table 3
List of disabilities used as examples (Set 2 Materials)

How willing would you be to work with as a co-worker a person
with each of the following disabilities?   (1-5 scale)
      Not                                   Very
     Likely                                Likely
       1       2     3                 4     5
              --   Alcoholism
              --   Amputee
              --   AIDS
              --   Blindness
              --   Cancer
              --   Deafness
              --   Diabetes
              --   Epilepsy
              --   Heart Disease
              --   Learning disorder
              --   Leprosy
              --   Manic/Depressive
              --   Paraplegia
              --   Schizophrenia
              --   Tuberculosis

Procedures. A slight deviation from the traditional policy capturing approach was necessary. Traditionally, vignettes are presented to participants. In these vignettes the independent variables of interest are manipulated across the vignettes. As an example, one might manipulate the gender, race, and education level across the vignettes. Let us assume for now that we have two genders, four races, and that we use four levels of education. By manipulating those three variables we could produce 32 different vignettes (2 X 4 X 4). The participants are then asked to rate the individuals in each of the vignettes on some dependent variable. Then, by simply conducting a regression analysis it is possible to determine how the participants used the different independent variables in forming a rating on the dependent variable. Thus, the participants are never explicitly asked about the independent variables.

In the current study, however, forming vignettes that manipulated the 16 relevant concerns simply in terms of high versus low levels of each concern would have resulted in 65,536 vignettes ([2.sup.16]). Based on the impracticality of this task, a different approach was utilized. In the current study, vignettes were essentially replaced by the 15 specific disability exemplars. The participants' inherent disability-dimension associations tapped by the Set 1 materials thus served as the independent variable manipulation. The dependent variable was ascertained by having the participant complete the Set 2 materials. These materials captured the effect that each disability had on the various criterion measures. Thus, as with a traditional policy capture, the participants were never asked explicitly about the effect each of the dimensions had on any of the dependent variables. The two sets of materials were completed approximately one week apart to control for any memory effects.

Results and Discussion

One hundred and seventy-one students completed both sets of materials. The correlations among the 16 concerns are presented in Table 4. These correlations reflect the associations between the different concerns across all subjects (note that the correlations among the 16 concerns were appropriately pooled across the 15 disabilities). Thus, Table 4 presents the pooled within-disability correlations. As the table illustrates these correlations were high in some cases. In order to simplify the analysis and to provide insight into the underlying relationships among the 16 concerns, a factor analysis was conducted.

Table 4
Intercorrelations Between Sixteen Concerns Regarding Disabled Workers
(Decimals Omitted)

                             1    2    3     4     5      6     7

 1. Aesthetics               --   04   -41    22    -28   -20   -39
 2. Causality                     --    16    24     09    21    23
 3. Concealability                      --   -03     28    18   -18
 4. Contagiousness                            --   -.03    03    21
 5. Controllability                                  --    50   -12
 6. Curability                                             --    03
 7. Distractiveness                                              --
 8. Preferential Treatment
 9. Need for Assistance
10. Novelty
11. Oversensitivity
12. Peril
13. Stress Response
14. Severity
15. Stability
16. Longevity

                             8     9     10    11    12    13    14

 1. Aesthetics                31    38    02    28    11   -15     22
 2. Causality                -15   -22    22    22    45    33     14
 3. Concealability           -38   -45    09   -14    16    02    -31
 4. Contagiousness            03   -18    31    15    59   -12     00
 5. Controllability          -35   -49    01   -24    12    21    -21
 6. Curability               -25   -35   -08   -09    09    26   -.05
 7. Distractiveness           39    24    22    45    05    32     54
 8. Preferential Treatment    --    65    05    45   -15    01     45
 9. Need for Assistance             --   -11    25   -26   -11     45
10. Novelty                               --    23    44    13     04
11. Oversensitivity                             --    15    27     42
12. Peril                                             --    12    -02
13. Stress Response                                         --     38
14. Severity                                                       --
15. Stability
16. Longevity

                             15    16

 1. Aesthetics                03    15
 2. Causality                -22    39
 3. Concealability           -14    08
 4. Contagiousness           -22    44
 5. Controllability          -22    04
 6. Curability               -25    15
 7. Distractiveness          -10    29
 8. Preferential Treatment    15    00
 9. Need for Assistance       31   -19
10. Novelty                  -21    32
11. Oversensitivity           01    26
12. Peril                    -32    52
13. Stress Response          -26    31
14. Severity                  07    21
15. Stability                 --   -35
16. Longevity                       --

Principal axes factoring with varimax rotation was utilized. Although the Kaiser-Guttman criterion (eigenvalues greater than unity) suggested a four-factor solution, an analysis of the scree plot and the parallel analysis criterion (Humphreys & Montanelli, 1975; Lautenschlager, 1989) both indicated a three-factor solution was more appropriate. The rotated three-factor solution also resulted in a more easily interpretable solution.

This solution is presented in Table 5. Variables loading on the first factor (using a saliency level of .40) included need for assistance, preferential treatment, controllability, aesthetics, curability, and concealability. This factor was labeled overtness and was thought to capture the visibility of a disability and how that disability might require accommodations. Variables loading on the second factor were peril, contagiousness, work longevity, causality, and novelty. Factor 2 was labeled the risk factor and seems to represent the degree of risk or uncertainty associated with a disability. Factor 3 consisted of distractiveness, severity, response to stressful situation, oversensitivity, and preferential treatment. This factor was termed the response factor as it seemed to tap the person's response toward the environment. The three-factor solution accounted for 45.9% of the variance in the data set.

Table 5
Rotated Factor Loadings for the 16 Concerns of Individuals with
Disabilities

                         (Overtness)   (Risk)   (Response)

Need for Assistance         .75         -.29        .30
Preferential Treatment      .62         -.07        .43
Controllability            -.62          .05       -.06
Aesthetics                  .61          .25        .14
Curability                 -.51         -.09        .11
Concealability             -.50          .10       -.17
Peril                      -.09          .83       -.05
Contagiousness              .11          .68       -.06
Work Longevity              .08          .67        .26
Causality                  -.21          .48        .24
Novelty                    -.01          .47        .13
Stability                   .30         -.39       -.08
Distractiveness            -.03          .13        .76
Severity                    .32         -.04        .68
Response to Stress         -.31          .15        .59
Oversensitivity             .31          .27        .53

The next question addressed by the current study was whether or not these dimensions could be used to predict the nine criterion responses. Particularly important was the question of whether these dimensions could predict the criterion responses in lieu of the specific disability in question. Based on the above factor analysis, factor scores were computed for each of the 171 subjects on the overtness, risk and response factors. These variables were then entered simultaneously as the first step of a hierarchical regression for each of the nine dependent variables. Next, the disability-specific components were forced into the hierarchical regressions (this was accomplished via dummy coding each disability as a within-subjects factor). This second step essentially reflects the value of knowing what specific disability was being rated above and beyond knowing a disability's perceived status on the three underlying dimensions (i.e. the disability-specific connotation). Finally, the 45 possible interaction effects (3 factors X 15 disabilities) were entered into the hierarchical regression simultaneously. This step would indicate if affect moderated the relationship between the three dimensions and the criterion measure. Table 6 presents the results for this analysis for all nine dependent variables. [Note, as nine regressions were conducted, Bonferoni's correction was utilized to maintain a .05 family-wise error rate. Thus, incremental variance was significant only if it exceeded a significance level of .0057.]

Table 6
Incremental Variance Accounted for in Hierarchical Regressions by
Dimensions, Disability, and Interactions

                        Dimension   Disability   Dimension X Disability

Working Relationship     .34(*)      .17(*)              .01
Hire                     .23(*)      .14(*)              .03(*)
Promotability            .25(*)      .20(*)              .01
Tardiness/Absenteeism    .32(*)      .17(*)              .01
As Coworker              .32(*)      .19(*)              .01
As Boss                  .34(*)      .22(*)              .01
Trainability             .36(*)      .17(*)              .01
Work Motivation          .35(*)      .18(*)              .01
Productivity             .34(*)      .15(*)              .02

Note. (*) 12 [is less than] .006; Bonferoni correction used to
maintain a family-wise alpha of .05.

The results from this analysis are fairly consistent across the nine dependent variables. Knowledge of the disability-dimension associations typically accounted for around 30% (ranging from 23% to 36%) of the variance and was significant in all cases. This is interpreted as evidence that these dimensions can be used to predict criterion responses effectively; however, column two in Table 6 shows that in all cases the disability-specific components added significant predictive power (on average accounting for an additional 18% of variance). Thus, there was a benefit in knowing the connotation associated with a specific disability above and beyond what would be expected based only on the overtness, risk and response of that given disability. Finally, with only one exception, the interactions failed to significantly increase incremental variance. That is, for eight of the nine dependent variables it appears as though the dimensions were used similarly across disabilities (i.e. disability does not moderate the dimension-dependent variable relationships). The standardized beta weights are thus presented for the nine dependent variables in Table 7. Obviously, those presented for hiring decision should be interpreted with caution since in that case the dimensions were moderated by the disability.

General Discussion

The primary purpose of this study was to determine the character of the dimensions underlying perceptions of individuals with disabilities. Both the Attitudes Towards Disabled Persons Scale (Yuker, Block, & Younng, 1966) and the Interaction with Disabled Persons Scale (Gething, 1994) compute and interpret summated rating scores. These scales seem to imply that there is some degree of general affect toward those labeled as disabled and that by generating and measuring that affect one can capture the essence of "attitudes toward individuals with disabilities." A modified policy capturing technique was used in order to minimize the effects of social desirability. The resulting factor analysis found that at least three distinct dimensions underlie perceptions of individuals with disabilities, thus rejecting the unidimensionality hypothesis.

Another important outcome of the current study involves the composition of the three dimensions found to underlie perceptions of individuals with disabilities. In particular, rehabilitation professionals may be interested in knowing what aspects of an individual with a disability are used by raters in forming attitudes. First, a dimension relating to the overtness of the disability was found to be important in predicting eight of the nine outcomes (see Table 7). Factor 1 seems particularly important in any criterion measure involving personal contact with the individual with the disability. These dependent variables would include working relationships, eventual promotability, and the ability to benefit from training. Individuals with a highly visible disability should be advised accordingly. They may want to emphasize their social skills and provide evidence of their ability to get along with colleagues and supervisors.

Table 7
Standardized Regression Weights for Predicting Various Outcome
Measures Using Underlying Dimensions

                        Overtness    Risk      Response

Working Relationship     .24(**)     .03        .40(**)
Hire                     .13(**)     .03        .40(**)
Promotability           -.11(**)    -.13(**)   -.49(**)
Tardiness/Absenteeism   -.19(**)     .20(**)    .37(**)
As Coworker             -.14(**)    -.12(**)   -.40(**)
As Boss                  .00        -.16(**)   -.42(**)
Trainability            -.13(**)     .22(**)   -.43(**)
Work Motivation          .15(**)    -.21(**)   -.41(**)
Productivity             .08(**)    -.03       -.54(**)

Note. (**) p [is less than] .01; (*) p [is less than] .05

The second dimension found to underlie perceptions of the individuals with disabilities was the risk associated with the disability. Risk was found to be predictive of all criteria except working relationship, hiring decision, and productivity. It is important for rehabilitation professionals and individuals with disabilities to realize that the level of risk associated with their given disability is an important consideration for raters. This fact may imply the need for better education of non-disabled colleagues on the contagiousness, stability, and course of a given individual with a disability. As an example, an HIV-positive worker may be well served to educate his/her colleagues on the remote chances of contracting HIV through casual contact. To the degree that individuals with disabilities can minimize the assumed risk associated with their disability, one would expect the negative consequences of this factor to be reduced.

The third dimension found to underlie perceptions of individuals with disabilities was a response dimension. This dimension appears to tap the disabled person's response to the environment. Unlike the other two dimensions, the response dimension was not prominent in previous literature. It is therefore somewhat surprising that this dimension appears to be the most important predictor of all nine dependent variables (note standardized beta weights in Table 7). This indicates that a major determinant in the formation of attitudes toward the individuals with disabilities concerns the disabled individual's response to their environment. Individuals with a disability may be well advised to refrain from displaying an overly large "sense of entitlement" during job interviews, etc.

Finally, Table 6 demonstrates that the overtness, risk, and response dimensions do not tell the entire story. Some disabilities appear to have specific affect associated with them above and beyond the overtness, risk, and response associated with that disability. It is important to note that the affect measured in this study is associated with specific disabilities, not the "disabled" in general. It is also important to realize that, in most cases, the interactions between disability and the three dimensions were not significant. This indicates that even in cases where disability specific connotations exist, the dimensions underlying the disabilities are still used by the rater in the same fashion. That is, the dimensions are still predictive of attitudes without being moderated by the specific disability in question.

Conclusions

In summary, the current research has shown that several factors helped determine whether an individual was likely to suffer from biased perceptions. Attitudes toward individuals with disabilities were shown to be multidimensional in basis, and could not be adequately classified as unidimensional. The current research did not entertain lofty visions of changing people's social cognitions or magically transforming the world into one free from all forms of bigotry. Instead, it was undertaken in an attempt to discover the basis of biased perceptions so that directed efforts could be undertaken to reduce such bias. The results of this study provide some guidance, in that, perceptions of individuals with disabilities seem to be dependent on the overtness, risk, and response associated with the disability. Before advising a client entering the workforce, a counselor should take into account the level of overtness, risk, and response associated with their client's specific disability. They should also carefully consider any disability-specific connotations associated with that particular disability. Professionals approaching the task of employing or advising workers with disabilities with a "one size fits all" approach may be providing a disservice to their clients.

References

Bordieri, J., & Drehmer, D. (1986). Hiring decisions for disabled workers: Looking at the cause. Journal of Applied Social Psychology. 16, 199-208.

Bose, J. L., Geist, G. O., Lam, C. S., Slaby, M. & Arens, M. (1998). Factors affecting job placement success in proprietary rehabilitation. Journal of Applied Rehabilitation Counseling, 29, 19-24.

Drehmer, D.E. & Bordieri, J.E. (1985). Hiring decisions for disabled workers: The hidden bias. Rehabilitation Psychology, 30, 157-164.

Gething, L. (1994). The interaction with disabled persons scale. Journal of Social Behavior and Personality, 9, 23-42.

Hampton, N. Z. & Crystal, R. (1999). Gender differences in acceptance of disabilities among vocational rehabilitation consumers. Journal of Applied Rehabilitation Counseling, 16-21.

Humphreys, L. G. & Montanelli, R. G. (1975). An investigation of the parallel analysis criterion for determining the number of common factors. Multivariate Behavioral Research, 10, 193-205.

Lautenschlager, G.J. (1989). A comparison of alternatives to conducting monte carlo analyses for determining parallel analysis criteria. Multivariate Behavioral Research, 24, 365-395.

Madden, J.M. (1981). Using policy-capturing to measure attitudes in organizational diagnosis. Personnel Psychology, 34, 341350.

Pierce, P.A. (1991). The Americans with Disabilities Act of 1990: A symposium. The Journal of Intergroup Relations, 17, 34.

Rossi, A.L. & Madden, J.M. (1985). A note on policy capturing as a method for studying social desirability. Bulletin of the Psychonomic Society, 23, 465-466.

Satcher, J. & Hendren, G.R. (1991). Acceptance of the Americans with Disabilities Act of 1990 by persons preparing to enter the business field. Journal of Applied Rehabilitation Counseling, 22, 15-18.

Simpkins, K.L., & Kaplan, R.K. (1991). Fair play for disabled persons: Our responsibilities under the new ADA. Journal of Career Planning and Employment, 51, 40-46.

Stephens, K.K. & Clark, D.W. (1987). A pilot study on the effect of visible physical stigma on personal space. Journal of Applied Rehabilitation Counseling, 18, 52-54.

Stone, E.F., Stone, D.L., & Dipboye, R.L. (1992). Stigmas in Organizations: Race, Handicaps, and Physical Unattractiveness. In K. Kelley (Ed.), Organizational Psychology. Amsterdam: Elsevier.

Thornburgh, R. (1991). The Americans with Disabilities Act: What it means to all Americans. The Journal of Intergroup Relations, 17, 35-41.

Yuker, H. E., Block, J. R., & Younng, J. H. (1966). The measurement of attitudes toward disabled persons. Albertson, NY: Human Resources Center.

Adrian Thomas, Ph.D., Department of Psychology, Western Kentucky University, 1 Big Red Way, Bowling Green, KY 42101-3576. Email: [e-mail]