The development and validation of the workplace acceptance scale: Evidence from a sample of workers with disabilities

1.1Theoretical background

For most of us the affiliation with a group of colleagues goes without saying and satisfies our need to belong to some extent. However, there are groups of people who are seriously underemployed and lack opportunities of being associated with groups of acquaintances. Among those who have only limited access to this opportunity of belonging are people with disabilities (Vornholt et al., 2018). Not only do they have limited access to the labor market, they are also still afflicted with extremely high de-grees of employment discrimination (Bjelland et al., 2010; Stuart, 2006) such as stereotyping, harassment, marginalization, as well as discrimination aga-inst hiring, promotion, and job retention (Durand, Corbière, Coutu, Reinharz, & Albert, 2014; Scheid, 2005; Williams, Fossey, Corbière, Paluch, & Harvey, 2016).

About 14% of the European population of working age (15– 59 years) experience a moderate to severe work disability (WHO, 2004). Disability is a broad concept and varies from minor difficulties in functioning to major impacts on a person’s life. Disability is interpreted in relation to what is considered normal functioning. However, the common social norm can differ based on the context, age group, or other environmental factors (Muschalla, Vilain, Lawall, Lewerenz, & Linden, 2012). According to the International Classification of Functioning (ICF; WHO, 2001), a disability exists when the degree and type of impairment causes difficulties in a certain environment with particular activity demands, such as the work environment. In terms of this model, disability is used as an umbrella term for all possible deficits in body functions, structures, activities and participation (Linden, Baron, & Muschalla, 2010).

The workplace integration of people with disabilities has become an important new topic in work and organizational psychology and research has increased especially since the passage of the Americans with Disabilities Act (ADA, 1990) in the United States (Shore et al., 2009). The concepts “participation” and “integration” are frequently defined in terms of the prevalence of activities or contact with others rather than in terms of social outcomes, such as meaningful relationships. Social inclusion, however, also contains the personal experience of inclusion, embracing perceptions of being accepted and recognized as an individual beyond the disability (Lysaght et al., 2012). Similarly, workplace inclusion is considered to be the extent to which people with disabilities are accepted, helped, and treated as others by their cowo-rkers (Stone & Colella, 1996). Acceptance can thus be seen as an important aspect of inclusion into the social network and dynamics at work. The definition of acceptance we use for the present study is based on the Theory of Reasoned Action (Fishbein & Ajzen, 1975), which suggests that a person’s behavior is pre-dicted by the attitudes toward this behavior as well as the person’s expectations regarding reactions of others when performing the behavior in question. Acceptance can then be understood as an outcome or consequence of attitudes toward people with disa-bilities; an evaluative response on multiple dimensions (Vornholt, Uitdewilligen, & Nijhuis, 2013). Ev-aluative responses to attitudes contain a cognitive, affective, and behavioral dimension, sometimes ref-erred to as the three components of attitudes in tr-aditional social psychological research (Eagly & Chaiken, 1993). The cognitive component includes thoughts and beliefs, the affective component refers to feelings and emotions, and the behavioral component includes reactions and behaviors (see Fig. 1 for a graphical display).

Fig. 1

Relationship of the theory of reasoned action and the concept of workplace acceptance.

In this study, we operationalize the concept of acceptance from the individual perception of an employee with disability regarding displayed behavior, affect, and cognitions of the direct colleagues. We outline the concept of perceived acceptance as acknowledging oneself as an important member of a group of colleagues, experiencing support, feeling comfortable and appreciating being part of the group, having access to all common areas and events, and being integrated in professional and social activities.

It should be noted that, although we validate the present scale in a very specific sample, the scale is not particularly developed for certain types of disabilities or other minorities. It can be used for a large range of groups at risk of experiencing rejection.

3.1Development of the scale

The items of the questionnaire were generated using a deductive item-generation approach. The theoretical foundation of the item generation is based on relevant studies identified in a literature review (Vornholt et al., 2013) and according to the theory of acceptance described above. We thus generated items that are in accordance with the definition of acceptance and specific to the workplace setting. Consist-ent with recommendations by Hinkin (1998), we developed items that were concise, short and easy to understand. In addition, we made sure that the items were clearly differentiating between the three dimensions of cognitive, affective, and behavioral acceptance. All items were positively worded and were created in simple language, making the scale accessible for people with mental and intellectual disorders. We supplemented these items with items from several well-established questionnaires that measure (a part of) acceptance (e.g., Burge, Ouellette-Kuntz, & Lysaght, 2007; Chadsey, Shelden, Horn, & Cimera, 1999; Cramm, Tebra, & Finkenflügel, 2008). Although these scales were not meant to measure acceptance per se, the scales contained items that were consistent with the definition of acceptance developed earlier (Vornholt et al., 2013). Using these two techniques, we obtained 14 items. More specifically, in this initial version of the questionnaire we had four items for the cognitive dimension (e.g. I think that I am an important member of my team at work), five items for the behavior dimension (e.g. My colleagues treat me no different than other colleagues), and five items for the affect dimension (e.g. I like to work with my colleagues). We consider the three dimensions as latent factors, which are not directly observable, but measurable by specific items.

3.2Participants and procedure

In two separate studies, the items developed were presented to employees with disabilities for further refinement and examination of the underlying factor structure. We chose this specific group of employees for validating the scale because this is a population that in particular experiences stigmatization, discrimination, and low levels of acceptance (Bos, Kanner, Muris, Janssen, & Mayer, 2009). In study 1 we conducted an exploratory factor analysis using sample 1, and in study 2 we applied a confirmatory factor analysis using sample 2.

3.3Statistical analysis

The underlying structure (cognitive, affective, and behavioral acceptance) of the Workplace Acceptance Scale was explored using both exploratory factor analysis and confirmatory factor analysis. Expl-oratory factor analysis was performed on the first sample. The number of participants exceeds the recommended ratio of 10 to 1 (e.g. ten cases for each item; Field, 2009). A first inspection confirmed that the data was suitable for factor analysis. Principal component analysis (PCA) was conducted with oblique rotation (oblimin). The number of factors to be retained was guided by Kaiser’s criterion (eigenvalues above 1), inspection of the Scree plot, and by using Horn’s parallel analysis (Horn, 1965). We only retained factors with eigenvalues obtained in PCA that exceeded the values attained from a randomly generated data set of the same size using the software MonteCarlo (Watkins, 2006). In the next step we conducted confirmatory factor analysis (CFA) on the second sample using Mplus version 7 to evaluate model fit and to replicate the factorial structure.

3.4Results

3.4.2Confirmatory factor analysis

We conducted confirmatory factor analysis (CFA) using maximum likelihood estimation on the second independent sample with 159 cases. As identified in the exploratory factor analysis, we investigated a one-factor model with twelve items and evaluated model fit by using several fit indices, including Chi-square, Comparative Fit index (CFI), Tucker-Lewis Fit Index (TLI), Root Mean Square Error of Approximation (RMSEA), and Standardized Root Mean square Residual (SRMR).

In a first analysis, we found a moderate model fit, with Chi-square test χ² = 156.33, p = 0.000, incremental indices of fit CFI = 0.915 and TLI = 0.896, and absolute indices of fit RMSEA (90% CI)=0.109 (0.089 – 0.129) and SRMR = 0.051. Model modification indices indicated that item 5 (i.e. My colleagues offer their help when I need it) prevented a better model fit, due to covariance with other items without clear conceptual rational. We decided to remove the item. Modification indices furthermore recomme-nded to allow two residual covariance’s to the model (item 8 and item 6, item 13 and item 2), which we accepted based on their conceptual relationship. Item 8 (i.e. I feel comfortable around my colleagues at work) and item 6 (i.e. I like to work with my colleagues”) both originate from the items developed for the affect dimension and describe a person’s feelings about the group of colleagues. Item 13 (i.e. My colleagues involve me in their daily conversations) and item 2 (i.e. My colleagues regularly talk to me about work-related topics) both originate from the items developed for the behavior dimension and thus investigate how a worker with disability experiences the interaction with colleagues. The model fit indi-ces improved considerably and presented a good fit (χ²= 74.930, p = 0.001, CFI = 0.968, TLI = 0.958, RMSEA [90% CI] = 0.070 [0.043 – 0.096] and SRMR = 0.040). Figure 2 visualizes the modified me-asurement model with parameter estimates. It is not-iceable that item 1, item 2, and item 13 show rather low standardized loadings (>0.60). We decided to keep these items regardless of low magnitudes, as they may be based on response patterns unique to the sample of respondents. The eleven-item scale ex-hibited a high internal consistency reliability (Cron-bach’s α= 0.91).

Fig. 2

Modified model and parameter estimates for the Workplace Acceptance Scale.

4.1Individual-level variables

Acceptance is thought to be an essential component of social integration at work and refers to positive relationships with others (Vornholt et al., 2013). While long-term exposure to negative interpersonal reactions causes poor psychological and physical health, positive reactions of others stimulate psychological and physical well-being (Baumeister & Leary, 1995; Richman & Leary, 2009). We assume that being accepted in a group of colleagues has a positive influence on several aspects related to the work context, such as health, motivation, and satisfaction. Being accepted by colleagues is furthermore likely to be important for the sustainability of employment, which implies that employees are able to function on the long term considering achievement and maintenance of participation and health (van der Klink et al., 2011). More specifically, we hypothesize that acceptance is related to health and productivity, in a way that it reduces absence from work and counteracts ‘presenteeism’, which is the condition of being present at work without being productive (Koopman et al., 2002). Furthermore, we expect that acceptance is positively related to the worker’s organization-based self-esteem (Pierce, Gardner, Cummings, & Dunham, 1989), in that being accepted positively influences the self-perceived value within one’s organizational context.

Motivation and commitment are essential components of sustainable employability and are closely related to well-being, health and productivity (Ka-nste, 2011; Meyer, Allen, & Smith, 1993; Warr, Cook, & Wall, 1979). We therefore hypothesize that both motivation and organizational commitment are positively linked with workplace acceptance. Ultimately, we expect a positive relationship between the feeling of being accepted and job satisfaction; and a negative link between workplace acceptance and the intent to leave. Both factors depend on positive relationships with colleagues (Bentein, Vandenberg, Vandenberghe, & Stinglhamber, 2005; Judge, Locke, Durham, & Kluger, 1998; Wolgast & Fischer, 2017) and we expect that the positive effects can only be achieved by being accepted (Vornholt et al., 2013). Also, acceptance can potentially be considered being a part of the concept of job satisfaction. A closer look at scales measuring job satisfaction (e.g., The Job Satisfaction of Persons with Disabilities Scale; Smedema et al., 2016) reveals compliances in some items. To summarize, we hypothesize that acceptance is positively related to health and productivity, self-esteem, motivation, organizational commitment, and job satisfaction, and negatively related to the intent to leave the organization.

4.2Team-level variables

Besides the person with a disability, also a number of colleagues filled in a questionnaire, representing the work team as a whole. Since many European countries adopted a disability quota system that forces companies to employ people with disabilities (Wuellrich, 2010), diversity is more and more turning into the lived reality in organizations. Moreover, teamwork is one of the core elements of new organizations (Eurofound, 2007) and it can be assumed that future work increasingly asks for cooperation between workers with and without disabilities. As mentioned above, acceptance is intimately bound to the need to belong to a group of acquaintances (Baron et al., 2009; Baumeister & Leary, 1995). A group of colleagues is an important social group most people in some way or another belong to. Therefore, we argue that workplace acceptance is also a team-level variable, capturing part of the social dynamics in groups.

In order to test the nomological network of acceptance on the team-level, we included a number of relevant team-level variables. Primarily, we expect a positive relationship between workplace acceptance and pro-social team values. Pro-social values are part of organizational citizenship behavior and describe motives which are strongly associated with behavior directed at individuals, such as being helpful or the desire to build positive relationships with others (Rioux & Penner, 2001). Similar to pro-social values, also the concept of workplace acceptance incorporates interactions between individuals and behavior toward others. The concept cohesion includes interpersonal and task cohesion and describes the commitment to a common task as well as the forces that act on a group to remain together (Carless & De Paola, 2000). For that reason we hypothesize that in a cohesive group, members tend to form stronger ties and will therefore be more likely to include and accept other group members. Similarly, psychological safety refers to a shared belief that the team is safe for interpersonal risk taking, which means that mutual respect and trust exist among team members (Edmondson, 1999). We argue that in an environment where people feel safe to take interpersonal risks, they are less likely to exclude others. Finally, the attitudes of colleagues toward people with disabilities play a major role for the perceived acceptance of the employee with a disability (Vornholt et al., 2013). We expect a positive relationship between both variables. Summing up, we hypothesize that acceptance is positively related to pro-social values, cohesion, psychological safety, and the attitudes of colleagues.

4.3Participants and procedure

4.4Measures

We included several measures in this study of which some were filled in by all participants (wor-kplace acceptance, intent to leave, intrinsic job mo-tivation, affective organizational commitment, job satisfaction), some only by the participants with dis-abilities (health status and employee productivity, organization-based self-esteem), and others were filled in by colleagues only (attitudes, pro-social values, cohesion, psychological safety).

4.5Results

The correlations and partial correlations between acceptance and the variables of interest are presented in Table 2. With all of the variables examined, it is evident that acceptance as perceived by the employee with a disability is more strongly related to variables measured on the individual-level than to those measures on the team-level. Our first hypothesis predicted that acceptance is positively related to health and productivity, self-esteem, motivation, organizational commitment, and job satisfaction, and negatively related to the intent to leave the organization. This hypothesis was fully supported. All constructs correlated with acceptance as predicted: health and productivity (r = 0.48, p < 0.01), self-esteem (r = 0.60, p < 0.01), motivation (r = 0.45, p < 0.01), organizational commitment (r = 0.30, p < 0.05), job satisfaction (r = 0.52, p < 0.01), and the intent to leave (r = – 0.42, p < 0.01).

The second hypothesis concerns the team-level variables. Because the sample consisted of individuals embedded in teams, we applied multilevel analyses, using Random Coefficient Modeling in R (Pinheiro & Bates, 1996), with individuals nested in teams to assess the relationship between acceptance and pro-social values, cohesion, psychological safety, and the attitudes of colleagues. We found a significant positive relation between attitudes of colleagues and acceptance (r = 0.36, p < 0.05) and a marginally significant (two-tailed) positive relation between psychological safety and acceptance (r = 0.26, p < 0.08). No other concept measured predicted perceived acceptance by people with disabilities. Additional to the above-mentioned constructs, we measured acceptance as indicated by the colleagues. Important to mention is that this scale is not validated, but closely follows the structure and content of the items developed earlier. We found acceptance measured on the individual-level and on the team-level significantly correlated (r = 0.37, p < 0.05).

5.1Limitations and future research directions

Despite the promising results, our study is not without limitations and those should be addressed in future research and validity testing. Although we consider acceptance as a common factor relevant for all types of impairments, it still remains to be tested whether our results are universally valid among all possible samples. As mentioned above, we purposely did not differentiate between types of disabilities in this study, but we realize that variation might exist especially between mental and physical disorders. In future research, the study population should be selected carefully so that more nuanced conclusions can be drawn about differences in the level of acceptance considering types of disabilities.

Another limitation concerns particularly study three. The number of teams in this study is rather small, making strong inferences about acceptance as a team variable difficult. More research is needed that assesses acceptance in larger and more diverse samples. Besides that, the items of some scales appeared to be difficult to understand for the target group. More instruments in easy language are required for high quality research in samples of people with mental disorders.

Future research should work out further details of the positive dynamics that arise when work teams become more diverse. While for employers organizational diversity still often takes on a negative connotation (Erickson, von Schrader, Bruyère, & VanLooy, 2014), it is the challenge for psychologists and others involved to emphasize the attractive potential that diversity actually offers. This includes the basic discussion on the role of employer attitudes, how these attitudes translate into the actual hiring of people with disabilities and what role workplace acceptance plays in this dynamic decision-making process. Besides the focus on the group of employers, also people with disabilities themselves should actively claim their entitlement on sustainable work (United Nations, 2006) on the primary labor market as possible. Research in work and organizational psychology needs to deliver convincing arguments for both, the employer to hire a person with a disability, and for the person with a disability to have the courage to compete on the primary labor market. Although work is no panacea for societal inclusion, it represents a major part of people’s lives (Jahoda, 1981) and its influence on inclusion should not be underestimated. If (the access to) work becomes barrier-free, then soon other parts of social life will follow. Research can do its bit by making relevant findings accessible for practice.