The factor structure of satisfaction ratings for selected employment concerns among people with multiple sclerosis

Abstract

OBJECTIVE:

The purpose of this study was to examine the factor structure of satisfaction ratings on a collection of 38 employment concerns items among people with Multiple sclerosis (MS).

RESULTS:

Analyses included a cross validation to establish and confirm a factor structure. The original 38 items were reduced to 17, which were found to form three distinct factors during the calibration portion of the cross validation. These three factors were Environmental/Personal Resources, Fair Treatment, and Legal Rights, which were confirmed during the validation portion of the cross validation.

CONCLUSION:

It appears that the three-factor structure of these 17 items is suitable for use with individuals with MS in the evaluation of their employment-related concerns and in the subsequent delivery of rehabilitation interventions.

Keywords

Measurement multiple sclerosis employment factor analysis dichotomous data

1. Introduction

1.1 Multiple sclerosis overview

Multiple sclerosis (MS), also known as disseminated sclerosis or encephalomyelitis, is an inflammatory disease that results in damage to the central nervous system (National Multiple Sclerosis Society [NMSS], 2018). The range of symptoms that can result from this illness has the potential to affect virtually every aspect of human activity and functioning. Its impact results in about 40% of those diagnosed with MS not reaching their 70th year of life. Two-thirds of deaths in people with MS are caused by disease-related factors (Compston & Coles, 2008). Although a variety of medical and therapeutic treatments can help to alleviate the impact of this disease, none can completely stop its progression, which often leaves individuals with multiple impediments to daily life and significant barriers, both psychological and physical, to gainful employment (Rumrill et al., 2015).

1.2 Epidemiology

As of 2017, there were an estimated 2–2.5 million people with MS globally, which translates to about 30 for every 100,000 people, although illness rates vary widely across regions of the globe (NMSS, 2018). MS-related deaths are estimated at about 18,000 per year worldwide. Race/Ethnicity is a significant factor in the occurrence of MS, as demonstrated by global regional differences. The prevalence of MS is lowest in Africa, followed by South East Asia, the Americas, and highest in Europe. For those of Northern European descent, the prevalence climbs to over 200 cases for every 100,000 people (Bishop, Rumrill, & Timblin, 2016). The worldwide number of new cases increases at an approximate annual rate of 2.5 for every 100,000 (World Health Organization, 2008). Rates of MS appear to be increasing, though this may be a result of improved detection methods. The current national prevalence of MS is estimated to be around 450,000 cases in the U.S. (NMSS, 2018). This prevalence figure suggests that approximately 1 out of every 700 Americans is dealing with MS at any given time (Fraser et al., 2002). Along with this prevalence assessment, it is estimated that 10,000 new diagnoses of MS are made every year in the U.S., an incidence rate that has steadily increased over the past 50 years (Rumrill & Roessler, 2015).

Although the initial symptoms and subsequent diagnosis of MS can occur during any time of one’s life, most cases begin in early adulthood, generally between the ages of 20 and 40 (Schapiro, 2012). More than half of all MS diagnoses occur before the age of 30, and 75% happen before the age of 40 (Fraser, Kraft, Ehde, & Johnson, 2006). Women also acquire the disease more frequently than men, at a rate of 200–300%, which is somewhat typical of autoimmune diseases in general (NMSS, 2018). Regardless of whom MS affects and when, this disease intrudes upon every important personal and social activity in one’s life including employment.

1.3 MS and employment

A variety of factors contribute to the significant and deleterious effect that MS has on a person’s employment (Rumrill & Roessler, 2015). The sporadic and unpredictable symptom cycles, which vary in intensity and duration, can be particularly disruptive. Symptoms that spread over a wide range of functional domains including psychological and physiological health are particularly problematic. Additionally, difficulties in communicating the unpredictable and multifaceted effects of MS to employers and coworkers can be challenging (Nissen & Rumrill, 2016).

With the typical onset of MS occurring in people’s 20 s and 30 s, MS often strikes during the “prime of life.” In relation to vocational growth and activity, this span of time is often the busiest and most productive period in one’s entire career. This age range has also been characterized by important career exploration, establishment, and maintenance activities (Super, 1980). Unfortunately, the onset and progression of MS makes it extremely difficult for people to continue their engagement in these career developmental activities. The vast majority, more than 90%, of people with MS have employment histories, meaning that they have worked for pay at some time in the past (Rumrill et al., 2015). Additionally, 82% of people with MS were working at the time of diagnosis (Roessler, Rumrill, Li, & Leslie, 2015). As the disease begins to progress, however, there is a dramatic reduction in labor force participation. The nature of this decline has been estimated by Fraser et al. (2002) to result in only 20 to 30% of people with MS retaining employment 15 years after their initial diagnosis. Furthermore, less than half of people with MS in the United States are currently employed (Rumrill et al., 2015). The specific reasons for these high rates of unemployment are not known, but it is likely some combination of factors such as demographics, illness-related symptoms, medical and insurance issues, and workplace factors such as discrimination and accommodation availability.

Even though the jobless rate for people with MS is very high in general, it seems disproportionately so for women and non-Caucasians (Roessler et al., 2015). People with MS who experience cognitive impairments are four to six times more likely to be unemployed than people with MS who do not experience cognitive impairments, and people with MS who have a progressive disease course are more likely to disengage from the workforce than those with the relapsing-remitting form of MS (Bishop et al., 2016). Job satisfaction has been found to be a key component of both job tenure and employment status for both the general populous and for people with MS (Dawis & Lofquist, 1984; Hartung, 2013). Li, Roessler, Rumrill, and Ahmed (2017) found that many people with MS who were employed experienced high levels of dissatisfaction with their jobs. The key factors accounting for this dissatisfaction were the excessive amount of work they were being expected to do, not receiving enough pay, too few opportunities for advancement, not enough on-the-job training, and low job recognition. Some of these dissatisfaction factors were results of people with MS feeling that they were being treated differently than their co-workers in the workplace.

To identify high-priority employment concerns from the perspectives of people with MS, Rumrill et al. (2015) conducted a national survey containing 38 employment concerns items. The survey instrument was created with input from 13 members of the NMSS. Survey respondents assigned the highest priority to concerns related to access to information on disability benefits programs, discrimination in hiring retention practices or employer behavior, employment protections under major legislation, the quality of rehabilitation and employment services, access to assistive technology, support for returning to work and planning for the future, and reactions of employers and coworkers after the disclosure of one’s disability status.

1.4 Purpose of this study and research questions

Although the 38 employment concerns items from the Rumrill et al. (2015) survey have been used in previous research to determine groups of items in which participants find the most and least satisfaction, no analysis prior to the present study had been conducted to explore the nature of their interrelations. For the present study, an Exploratory Factor Analysis (EFA) was first conducted to examine the factorial structure of the items; then, the EFA was followed by a Confirmatory Factor Analysis (CFA) to validate any structure that emerged in the EFA. The combination of EFA and CFA in this format is referred to as a cross validation. Using the cross validation approach, this study aimed at identifying the composing factors of the MS employment concerns rating scale. To be specific, this study sought to answer the following questions:

How many factors will emerge from the data when exploratory factor analysis is conducted?

How are the factors interrelated and are the intercorrelations displayed among the factors substantively meaningful?

How well do these factors fit the data, in other words, are the factors identified throughout the modeling process sufficient for the data?

2. Method

2.1 Participants

Data analyzed in this study were extracted from a national survey investigating the employment concerns of people with MS from nine NMSS chapters, representing 21 states and Washington, DC in the United States (Rumrill et al., 2015). Participants in this study were 1,844 people with MS who submitted complete data for the employment concerns items in the survey. Descriptive statistics for the sample in this study are presented in (Table 1). The sample included 759 men (41%) and 1,080 women (59%; 5 participants chose not to identify their gender). Twenty-five percent of respondents indicated that they lived in urban areas, 53% resided in suburban areas, and 22% resided in rural areas. The sample was predominantly White (76%) with an average age of 54 (SD = 12.13). The remaining participants were of minority backgrounds with African Americans and Hispanic groups being over-sampled (12% were African American; 11% were Hispanic; and 1% were Asian, Pacific Islanders, Native American, or Alaskan Natives). Forty one percent of the participants reported that they were employed at the time of the survey. Educational status reports indicated a high rate of both high school (98%) and college degree completion (46%) among the participants.

Table 1
Descriptive statistics of the survey participants (N = 1,844)

Characteristic	n	%
Employment Status
Unemployed	1,036	56.2
Employed	808	39.3
Gender
Female	1,443	78.3
Male	401	21.1
Race/Ethnicity
White	1,381	74.9
Non-white	463	25.2

Illness-related symptoms reported by the group ranged from mild to severe and suggested a variety of MS conditions, including physiological, sensory, and psychological effects. Commonly reported symptoms were fatigue (indicated by 79% of respondents), balance/coordination problems (65%), diminished physical capacity (63%), gait/mobility impairment (56%), numbness (50%), bowel or bladder dysfunction (50%), cognitive impairment (49%), spasticity (44%), pain (43%), depression (35%), vision problems (33%), sexual dysfunction (27%), anxiety (28%), and speech problems (16%).

2.2 Procedure

A national random sample of individuals with MS was drawn from 9 NMSS chapters across the United States (Rumrill et al., 2015). The aim of this sampling was to ensure that participants were representative of the total population of people with MS with regard to geographic location, gender, race/ethnicity, income, and education. For a complete description of sampling and data collection procedures that were used in the original survey, readers are referred to Rumrill et al. (2015).

Of the 7,369 people with MS whom Rumrill et al. (2015) included in the target sample, 1,932 people returned questionnaires, resulting in a response rate of 26%. A total of 1,844 members of the original respondent sample provided complete information on the variables of employment concerns and therefore formed the sample for the present investigation.

2.3 Measures

The 38 employment concerns items that were analyzed in this research included a two-part response set that elicited participants’ importance and satisfaction ratings for each item. For example, in responding to the item regarding health insurance coverage, respondents were asked, “Is it important that people with MS have adequate health insurance coverage?” and “Are you satisfied that people with MS have adequate health insurance coverage?” The 38 items cover a wide range of topics including access to information on disability benefits programs (e.g., “have adequate information about benefits such as health and disability insurance”), discrimination in employment such as hiring and retention practices (e.g., “are treated fairly in termination or demotion decisions made by employers”), employment protections under major legislation (e.g., “understand the employment protections of Title 1 in the Americans with Disabilities Act as Amended”), the quality of rehabilitation and employment services (e.g., “have access to service providers who understand the needs of people with MS”), access to assistive technology (e.g., “have access to assistive technology resources needed for work”), and support for returning to work and planning for the future (e.g., “are encouraged to take control of their lives”). Respondents answered either “yes” or “no” with regard to whether they thought each concern was important and if they were satisfied that the concern had been addressed in their communities. In this study, only the responses related to satisfaction with the employment concerns were analyzed, and the 38 items were dichotomously coded with 1 indicating “satisfied” and 0 indicating “dissatisfied”.

2.4 Data analysis

The goal of this study was to establish and validate a factor structure for the satisfaction ratings on the employment concerns of people with MS. A cross validation procedure was adopted, and the sample of participants was randomly split into two groups of roughly equal size. The first half was the calibration sample used for an exploratory factor analysis (EFA) to assess the psychometric properties of the items and to explore the possibilities of any overarching factor structure. After a factor structure emerged in the EFA, the factor structure was tested and validated with the second half of the sample (i.e., the validation sample) using confirmatory factor analysis (CFA).

As there is currently no single-model goodness-of-fit index that can directly confirm the fit of the model, five model goodness-of-fit indices were used to evaluate the fitted model: (a) the χ² test, (b) the root mean square error of approximation (RMSEA), (c) the comparative fit index (CFI), (d) the Tucker-Lewis Index, and (e) the weighted root-mean-square residual (WRMR). Data management and descriptive statistics in this study were performed in the Statistical Package for the Social Sciences software (SPSS 21). The factor analysis was conducted in Mplus 7, a latent variable modeling program with a wide variety of analytic capabilities (Muth $\overset{´}{e}$ n & Muth $\overset{´}{e}$ n, 2012). A sampling weight variable was also included in the analyses to account for the over sampling of some ethnic groups in the original survey conducted by Rumrill et al. (2015).

3. Results

3.1 EFA

An EFA was conducted on the calibration subsample (N = 921) to explore the possibility of various factor structures within the data. Due to the dichotomous nature of the items in the survey, the weighted least squares mean and variance (WLSMV) estimator was used with CF-Parsimax oblique rotation to produce more interpretable factors. Mplus offers the option to run multiple EFA models, with the ability to specify how many factors are produced in each. Eigenvalues were generated and examined first to determine which range of factor solutions might be worth further testing. Among the 38 eigenvalues, the first three were greater than one (e.g., the largest six eigenvalues are λ₁ = 26.41, λ₂ = 1.84, λ₃ = 1.67, λ₄ = .98, λ₅ = .94 and λ₆= .71). Given the consideration of sampling errors, eigenvalues appeared to suggest the possibility of a three, four, and possibly five factor solution if based on the eigenvalue greater than 1 rule. Because the employment concern topics were taken from a variety of source themes during the creation of these 38 items, a minimum of a two factor solution was expected.

To find a model that makes the most intuitive sense, forced two-, three-, four-, and five-factor solutions were analyzed. The model fit statistics from each factor solution in the series of EFA models are presented in (Table 2). Generally speaking, as the models increased in number of factors, each index of fit also improved, indicating better model fit. The three factor model shown in (Table 2) is the first of the four models tested that exhibited good fit on all of the listed goodness-of-fit statistics (i.e., χ² = 1274.86, df = 592, p < 0.001; RMSEA = 0.036; CFI = 0.990; TLI = 0.998; WRMR = 0.039). Scree plot of eigenvalues was also examined (Fig. 1). The values associated with λ₁, ₂, and ₃ appear to function similarly before dropping slightly to λ₄ through λ₃₈. Based on the analysis of the eigenvalues, it appeared that a three factor solution was the best choice, and the details of its functioning were explored.

Table 2
Summary of model goodness-of-fit indices for all EFA models

Model	χ ²	df	RMSEA	CFI	TLI	WRMR
EFA model comparisons
Two factor solution	2086.31 ***	628	0.051	0.978	0.957	0.056
Three factor solution	1274.85 ***	592	0.036	0.990	0.998	0.039
Four factor solution	949.22 ***	557	0.028	0.994	0.992	0.032
Five factor solution	701.13 ***	523	0.019	0.997	0.996	0.025

Notes.^***p < 0.001. χ² = Chi-Square Test (i.e., Minimum Fit Function); RMSEA = Root-Mean-Square-Error of Approximation; CFI = Comparative Fit Index; TLI = Tucker Lewis Index; WRMR = Weighted Root-Mean-Square Residual index.

Fig. 1

Scree plot for exploratory factor analysis.

The rotated factor loadings for the three-factor model were examined (Table 3) and loadings less than the cutoff value of 0.32 were suppressed to improve readability of the factor structure (Tabachnick & Fidell, 2001). The factor loadings of the 38 items were all significant (p < 0.05); however, many of the items appeared to load onto more than one factor in the model. In order to achieve a simple solution in which each item mainly contributed to the content of one latent construct, loadings were examined to identify items that cross-loaded onto more than one factor (Millsap & Yun-Tein, 2004). Twenty-one of the original 38 employment concerns items had a factor loading greater than 0.32 and cross-loaded onto multiple factors. According to the standards of Tabachnick and Fidell (2001), these were determined to be cross-loading items and were removed from the scale, which resulted in 17 items remaining in the model.

Table 3

Rotated factor loadings for three factor EFA model based on the original 38 items

Item	Loadings
	Three Factor
	Factor 1	Factor 2	Factor 3
A1	0.329	0.431
A2	0.645
A3	0.508
A4	0.651
A5	0.536	0.458
A6	0.535	0.530
A7	0.689	0.486
A8	0.666	0.388
A9	0.633		0.421
A10	0.691		0.408
A11	0.664		0.493
A12	0.711
A13	0.585
A14	0.615
A15	0.671		0.331
A16	0.581
A17	0.379	0.518
A18	0.401	0.586
A19		0.330	0.599
A20			0.706
A21		0.739
A22		0.667
A23		0.744
A24		0.331	0.629
A25			0.803
A26		0.421	0.391
A27			0.804
A28		0.669
A29		0.607	0.324
A30		0.559	0.336
A31			0.702
A32		0.580	0.369
A33			0.647
A34		0.535	0.369
A35		0.564	0.374
A36		0.603	0.363
A37		0.636	0.330
A38			0.761

Note. Loadings less than 0.32 were suppressed to improve readability.

The remaining 17 items were subjected to another EFA to confirm if the three factor structure would still be produced after such a large change in the number of items. The results of this three-factor 17-item model is exhibited below in (Table 4), which indicates that this three factor model with fewer items retained the simple structure suggested in the previous analysis. This model also had a close fit to the data (i.e., χ² = 203.96, df = 592, p < 0.001; RMSEA = 0.039; CFI = 0.990; TLI = 0.998; WRMR = 0.039).

Table 4

Manifest content of three factor 17-item model and factor loadings

Item		Loading
Environmental/Personal Resources
A2	Have opportunities for job training or retraining.	0.58 *
A3	Are encouraged to take control of their lives.	0.49 *
A4	Have access to service providers who understand the needs of people with MS.	0.63 *
A12	Can expect employers to respond to their accommodation needs in a timely manner.	0.69 *
A13	Have physical access to workplace facilities.	0.68 *
A14	Are considered for other jobs in the same company if their MS prevents them from returning to their former jobs.	0.66 *
A15	Have access to assistive technology resources needed for work.	0.73 *
Fair Treatment
A21	Are provided the same retirement benefits as other workers.	0.85 *
A22	Receive the same on-the-job training opportunities as other workers.	0.79 *
A23	Have their seniority honored in the same way as other employees.	0.86 *
A28	Are evaluated no more frequently than other workers.	0.65 *
Legal Rights
A20	Have adequate information about benefits such as health and disability (short and long-term) insurance.	0.64 *
A25	Have access to adequate information about Social Security programs.	0.79 *
A27	Understand the employment protections of Title I in the Americans with Disabilities Act as Amended (ADA).	0.76 *
A31	Know what to do if they encounter discrimination at work.	0.66 *
A33	Understand the risks and benefits of disclosing disability status to employers.	0.59 *
A38	Understand the health insurance provisions and protections of the Affordable Care Act (i.e., Obamacare).	0.72 *

Note.^*p < 0.05.

The three factors were named Environmental/Personal Resources, Fair Treatment, and Legal Rights. The Environmental/Personal Resources factor was composed of seven items that were related to external supports helping an individual with MS in her or his maintenance of, or return to, employment (i.e., Items A2, A3, A4, A12, A13, A14, and A15). The resources reflected in these items included access to service providers, the physical workplace, assistive technology, and reasonable accommodation. One item (i.e., Item A3) also addressed support that may come from one’s social circle in the form of encouragement. The Fair Treatment factor contained four items examining training, benefits, evaluation, and the honoring of seniority that were primarily related to the expectation that an individual with MS should be treated in the same way as her or his coworkers who may or may not have disabilities (i.e., Items A21, A22, A23, and A28). The Legal Rights factor included six items that were related to the intersection of disability-related legislation, discrimination, and government benefits (i.e., Items A20, A25, A27, A31, A33, and A38).

The three factors were moderately inter-correlated with correlation coefficients ranging between 0.531 (between Legal Rights and Fair Treatment) and 0.568 (between Fair Treatment and Environmental/Personal Resources) (see Table 5). The fact that the correlations were not extremely high (r > 0.85) suggested that there was not significant overlap or multicollinearity among the factors (Mansfield & Helms, 1982). At the same time, the moderate factor inter-correlations provided evidence for discriminant validity among the latent constructs.

Table 5

Factor Correlation Coefficients for Three Factor Solution

	Environmental/Personal Resources	Fair Treatment	Legal Rights
Environmental/Personal Resources	1
Fair Treatment	0.568 *	1
Legal Rights	0.532 *	0.531 *	1

Note. ^*p < 0.05.

3.2 CFA

Because three factors had emerged from the data, factor loadings less than 0.32 in the previous three-factor 17-item exploratory factor model were then fixed to zero. In other words, a three-factor 17-item confirmatory factor model, in which each item was only allowed to load onto one factor with factors inter-correlated and measurement errors uncorrelated, was subjected to model fitting and testing. Figure 2 displays the graphical representation of the three-factor 17-item confirmatory model. The CFA model was first fitted to the calibration subsample (N = 921). Based on the model fit indices of RMSEA, CFI, and TLI, this analysis indicated a close model fit (i.e., χ² = 282.99, df = 116, p < 0.001; RMSEA = 0.040; CFI = 0.992; TLI = 0.990). Examination of the Lagrange Multiplier (LM) modification indices did not suggest that any changes to this model would dramatically improve the model fit.

Fig. 2

The three-factor 17-item factor solution.

Next, this confirmatory factor model was fitted to the validation subsample (N = 923) for the purpose of cross-validation. The model in the validation subsample also produced a close model fit, indicating that the factor structure was the same when data were different (i.e., χ² = 316.38, df = 116, p < 0.001; RMSEA = 0.045; CFI = 0.989; TLI = 0.987). The assessment of LM modification indices for the confirmatory factor model in the validation subsample did not indicate the need for model re-specifications that would result in dramatic improvement to the overall model fit. For easy comparison, Table 6 displays the model fit statistics for the three-factor 17-item EFA solution in the calibration subsample, the three-factor 17-item CFA solution in the calibration subsample, and the three-factor 17-item CFA solution in the validation subsample. To conclude the analysis, the internal consistency as measured by Cronbach’s Alpha coefficient was calculated for each factor. The coefficients were 0.96, 0.98, and 0.98 for Environmental/Personal Resources, Fair Treatment, and Legal Rights respectively, indicating excellent or very high internal reliability of all three factors.

Table 6

Summary of model goodness-of-fit for three factor models

Model	χ ²	df	RMSEA	CFI	IFI	WRMR
Calibration Subsample
Three factor solution (EFA, 17 items)	203.96 ***	88	0.039	0.994	0.991	0.033
Three factor solution (CFA, 17 items)	282.99 ***	116	0.040	0.992	0.990	…
Validation Subsample
Three factor solution (CFA, 17 items)	316.38 ***	116	0.045	0.989	0.987	..

4. Discussion

Existing research clearly indicates that MS has a unique and usually negative impact on a person’s prospects for ongoing employment, but it remains unclear why this is so. Certainly, the impact of acquiring a chronic, usually episodic, and unpredictable disease such as MS during the peak-performance years of career development requires significant personal, social, environmental, lifestyle, and financial adjustments. However, a unifying theory or explanation for the employment-related difficulties that await people with MS following diagnosis is not evident in current medical or rehabilitation literature. It is well established that the vast majority of people with MS have employment histories (98% in the present sample) and most (82% in the present sample) were still working at the time of diagnosis, yet only 20 to 30 % of people with MS are still employed 15 years after diagnosis (Fraser et al., 2002). Even those people with MS who are employed report significant dissatisfaction with many aspects of their jobs (Li et al., 2017; Roessler & Rumrill, 1995).

The overall understanding of the employment experiences and outcomes of people with MS could be enhanced by close examination of the highest-priority employment-related concerns of this population. This study examined the factor structure of 38 key employment concerns that formed the basis for a large, national survey of Americans with MS. Specifically, this research attempted to identify patterns in respondents’ satisfaction ratings on the 38 items.

The first question addressed in the present study sought to determine whether the collection of 38 employment concerns items contained separate individual factors. Results indicated that there were three distinct factors in the scale. Once items that did not contribute to a single factor were removed, the three themes that emerged were Environmental/Personal Resources, which was related to opportunities and services available to people with MS (e.g., “Have opportunities for job training or retraining,” “Have access to service providers who understand the needs of people with MS,” “Can expect employers to respond to their accommodation needs in a timely manner”); Fair Treatment, which related to the way people with MS are treated in comparison to other workers (e.g., “Are provided the same retirement benefits as other workers,” “Have their seniority honored in the same way as other employees,” “Are evaluated no more frequently than other workers”); and Legal Rights (e.g., “Understand the risks and benefits of disclosing disability status to employers,” “Know what to do if they encounter discrimination at work,” “Understand the employment protections of Title I in the Americans with Disabilities Act as Amended [ADA]”).

These three factors are informative, but they are not surprising given the fact that the identified themes appear quite often in existing vocational rehabilitation literature. For example, Environmental/Personal resource-related topics are often cited as factors that tie into the availability and effects of support within one’s immediate social system, but also into the availability of accommodations and amenities at work (Strauser, 2014; Wehman, 2013). Fair Treatment, or the absence thereof, is a commonly cited issue in the employment experiences of people with disabilities in general and people with MS in particular (Leslie et al., 2015; Smart, 2016). Similarly, Legal Rights are known to be a key factor in the employment experiences of people with MS and other disabilities (McMahon & McMahon, 2016), a factor that includes the navigation of federal protections and disability benefits programs.

To address the second set of questions related to the validity of the three-factor model, a cross-validation was conducted on the validation subsample of a split data file. When this model was tested on this specific subsample, it continued to exhibit excellent fit. Additionally, all factor loadings, factor intercorrelations, and error variances remained constant. This step of the analysis confirmed that this factor structure including Environmental/Personal Resources, Fair Treatment, and Legal Rights themes that were previously identified held their structure. The relevance and implications of these specific themes extend beyond the immediate realm of these analyses. Environmental/Personal Resources, Fair Treatment, and Legal Rights exist as active themes in defining what it means to work while coping with MS, and some of the most prominent employment issues facing Americans with MS cluster into those themes (Bishop et al., 2016; Li et al., 2017).

4.1 Environmental and personal resources

A variety of environmental and personal resources have been shown to influence the employment outcomes of people with MS. They include considerations such as the availability of on-the-job accommodations (Leslie et al., 2015), support from coworkers (Vickers, 2012), and the availability of vocational rehabilitation services (Fraser et al., 2006). Tailored interventions are needed to address the specific concerns contained within the environmental/Personal resources factor (i.e., “Have access to service providers who understand the needs of people with MS”; “Can expect employers to respond to their accommodation needs in a timely manner”; “Have physical access to workplace facilities”; “Are considered for other jobs in the same company if their MS prevents them from returning to their former jobs”; “Have access to assistive technology resources needed for work”).

To further underscore the importance of workplace issues and barriers, another study of unemployed people with MS indicated that, although 75% left their previous employment voluntarily, the majority of them cited worksite barriers as the primary reason (Allaire & LaValley, 2003). Somewhat conversely, in the study by Rumrill et al. (2015) that generated the data for the present analyses, people with MS were generally satisfied with their access to insurance, employment supports, equitable personnel policies, and the accessibility of accommodations. They were generally dissatisfied with limited information about funding for assistive devices, retraining, and the availability of employers to discuss their accommodation needs.

4.2 Fair treatment

The perception of unfair treatment or discrimination by coworkers and employers is not uncommon among people with MS. Nissen and Rumrill (2016) described workplace discrimination as the unifying feature of the employment experience for Americans with MS, suggesting that it is the number one explanation for the high rate of workforce attrition that follows the onset of the illness. Indeed, in a national survey conducted by Roessler, Rumrill, and Hennessey (2002), no fewer than six items related to implementation of the ADA and the Family and Medical Leave Act (FMLA) were reported among the 12 most prominent employment-related problems identified by Americans with MS. Specifically, the majority of respondents reported having been treated unfairly in the hiring process by employers (73%), having been denied reasonable accommodations (58%), having received lower pay compared to their peers who are without disabilities (53%), being refused schedule modifications that would have enabled them to continue working (59%), having received inadequate health insurance coverage (73%), and having received little or no information about their legal rights from employers (69%). Vickers (2012) reported similar findings highlighting workplace discrimination as a major employment obstacle for people with MS.

4.3 Legal rights

It is not surprising that legal rights emerged as a distinct factor in respondents’ satisfaction ratings on the employment concerns items. Kalb (2016) emphasized how important it is for people with MS and their significant others to stay abreast of the legal protections that are accorded them in the Affordable Care Act, the Americans with Disabilities Act Amendments Act, the Family and Medical Leave Act, and other Federal statutes. The Job Accommodation Network at West Virginia University (askjan.org) and the National ADA Centers Network (adata.org) are excellent sources of information on disability-related legal rights, as is the NMSS’ (2018) interactive website. Nissen and Rumrill (2016) reported that a lack of awareness of legal protections is one of the top five employment barriers facing Americans with MS. To address this problem, Strauser (2014) encouraged the deployment of disability navigators who can assist people with various disabilities in understanding disability benefits and Federal legal protections. He also exhorted vocational rehabilitation professionals to serve as liaisons to health care providers to ensure that people with MS and other significant disabilities receive much-needed legal information and advocacy in an early intervention context.

4.4 Support for the three factor model

The occurrence of the three factors identified in this series of analyses is supported not only by the rigor of the statistical techniques used, but also in previous research findings. Following the development of the original 38 employment concerns items, one report of their functioning explored those items that demonstrated the greatest overall satisfaction (i.e., employment strengths) and overall dissatisfaction (i.e., employment weaknesses; Rumrill et al., 2015). Rumrill and his colleagues emphasized the potential importance of multifactor models that could be used to understand the impact of MS on various personal issues and their impact on quality of life. One of the key components of this notion of quality of life was job satisfaction, which has been identified as a strong predictor of continued employment in several studies of people with MS (Li et al., 2017; Rumrill, Roessler, & Fitzgerald, 2004).

In another study, a large international group of people with MS was surveyed regarding their employment status (Uccelli, Specchia, Battaglia, & Miller, 2009). Several factors were identified as correlates of employment status, including MS-related symptoms, workplace environment, one’s attitude toward work, others’ attitudes toward people with MS, financial considerations, and personal issues. In particular, workplace environment, attitudes of others, financial considerations, and personal issues had indicators similar to those that composed the final three-factor model in the present study.

Further support for the potential relevance of the present findings can be found in a qualitative analysis of employment themes among people with MS (Johnson, Yorkston, Klasner, Kuehn, Johnson, & Amtmann, 2004). Following a series of interviews, the researchers identified four overall themes that appeared to impact employment. They were the cost-benefit economy of working, fatigue and cognitive changes, stress in the workplace, and accommodations to address barriers. Specifically within the theme of accommodations to address barriers, participants expressed the importance of coworker education regarding MS, accommodation through personal action, changing of roles and activities, and employer participation in the accommodation process. The importance of accommodations was frequently stressed as an element of both fair treatment in the workplace and environmental/personal supports.

4.5 Limitations

One limitation of this study is that data were collected in a purely self-report format. Self-report information is subjective, i.e., grounded in the unique experiences of each respondent and often biased in favor of responses that cast the respondent in a favorable light (Heppner et al., 2015). That said, because the purpose of this study was to analyze data regarding the specific attitudinal concerns of respondents, a self-report survey was the only way to elicit the requisite information regarding respondents’ employment concerns.

The low response rate (26%) in the original survey from which data for this study were drawn is another limitation. Even though the respondent sample was generally representative of known population parameters of people with MS, it is unknown whether the 74% of the target sample who did not return surveys are characteristically different from those who did respond. It is certainly possible that non-responders have different employment concerns, are employed at different rates, and have different experiences with MS symptoms than the respondent sample.

This sample was also limited to participants currently living in the U.S, although MS has been shown to exist throughout the world. International and intercultural differences in the incidence and prevalence of MS, treatment approaches and availability of health care services, societal attitudes toward people with disabilities, and laws regarding labor practices and disability benefits would almost certainly yield different results if a study like the present one were conducted in another country.

5. Conclusion

Findings from this factor analytic study provide new insights into how people with MS conceptualize their employment related concerns. The 38 items that formed the basis for the original national survey were reduced to 17 items that loaded into three distinct factors: environmental/personal resources, fair treatment, and legal rights. The stability of the underlying factor structure of these 17 employment concerns items lends to their validity and reliability. Additionally, the cross-validation tests that were conducted on these data were done so using dichotomous indicators reflecting a strong model fit. Future researchers may use these confirmed employment concern factors as both predictor and criterion variables to obtain a parsimonious and nuanced understanding of the themes inherent in the satisfaction that people with MS ascribe to their overall employment situation.

Most importantly, the three factors identified in this study illuminate the dynamics of employment acquisition, retention, and advancement among adults with MS. If individuals with MS, or any other severe disability or chronic illness for that matter, are to succeed in employment, critical long-term supports are needed with respect to Environmental and Personal Resources, Fair Treatment, and Legal Rights. For rehabilitation counselors, diagnostic consideration of the extent to which their clients are satisfied with each of these factors should guide the development and delivery of interventions needed to improve employment outcomes.

Conflict of interest

None to report.

Footnotes

Acknowledgments

This research was funded through a Health Care Delivery and Policy Research grant from the National Multiple Sclerosis Society, New York, NY. The authors wish to thank the National Multiple Sclerosis Society, its participating chapters, and the study participants for their support and assistance with this research.

References

Allaire, S. H., Li, W., & LaValley, M. P. (2003). Reduction of job loss in persons with rheumatic diseases receiving vocational rehabilitation: a randomized controlled trial.Arthritis & Rheumatism, 48(11), 3212–3218.

Bishop, M., Rumrill, P., & Timblin, R. (2016). Medical, psychosocial, and vocational aspects of multiple sclerosis: Implications for rehabilitation professionals.Journal of Rehabilitation,82(2), 6–13.

Compston, A., & Coles

(2008). Multiple sclerosis.Lancet,372 (9648), 1502–1517.

Dawis, R., & Lofquist, L. (1984). A psychological theory of work adjustment. Minneapolis, MN: University of Minnesota Press.

Fraser, R., Clemens, D., & Bennett, F., (2002). Multiple Sclerosis: Psychosocial and Vocational Interventions. New York, NY: Demos.

Fraser, R., Kraft, G., Ehde, D., & Johnson, K. (2006). The MS workbook: Living fully with multiple sclerosis.Oakland, CA: New Harbinger Publications, Inc.

Hartung, P. (2013). The life-span, life-space theory of careers. In

Brown and

Lent (Eds),? Career development and counseling (2nd ed., pp. 83–114). Hoboken, NJ: John Wiley.

Heppner, P., Wampold, B., Owen, J., Thompson, M., & Wang, K. (2015). Research design in counseling (4th ed.) Independence, KY: Cengage Learning.

Johnson, K., Yorkston, K., Klasner, E., Kuehn, C., Johnson, E., & Amtmann, D. (2004). The cost and benefits of employment: a qualitative study of experiences of persons with multiple sclerosis.Archives of Physical Medicine and Rehabilitation, 85(2), 201–209.

10.

Kalb, R. (2016). Living with multiple sclerosis: The psychosocial challenges for patients and their families. In

Giesser (Ed.),? Primer on multiple sclerosis (2nd ed., pp. 483–498). New York, NY: Oxford University Press.

11.

Leslie, M., Kinjanui, B., Bishop, M., Rumrill, P., & Roessler, R. (2015). Patterns in workplace accommodations for people with multiple sclerosis to overcome cognitive and other disease-related limitations.Neurorehabilitation,37, 425–436.

12.

Li, J., Roessler, R., Rumrill, P., & Ahmed, E. (2017). A four-stage prediction model of job satisfaction for employed adults with multiple sclerosis.Rehabilitation Counseling Bulletin,61(1), 28–40.

13.

Mansfield, E., & Helms, B. (1982). Detecting multicollinearity.The American Statistician, 36(3a), 158–160.

14.

McMahon, M. C., & McMahon, B. T. (2016). The National EEOC ADA research project: History, available data, and basic findings.Journal of Vocational Rehabilitation, 44(3), 333–342. doi:10.3233/JVR-160803

15.

Millsap, R., & Yun-Tein, J. (2004). Assessing factorial invariance in ordered-categorical measures.Multivariate Behavioral Research, 39(3), 479–515.

16.

Muthán, L., & Muthán, B. (2012). Mplus User's Guide (Sixth Edition). Los Angeles, CA: Muthán & Muthán.

17.

National Multiple Sclerosis Society (2018). What is MS. National Multiple Sclerosis Society. Retrieved from https://www.nationalmssociety.org/What-is-MS

18.

Nissen, S., & Rumrill, P. (2016). Employment and career development considerations. In

Giesser (Ed.),? Primer on multiple sclerosis (2nd ed., pp. 483–498). New York, NY: Oxford University Press.

19.

Roessler, R., & Rumrill, P. (1995). The relationship of perceived worksite barriers to job mastery and job satisfaction for employed people with multiple sclerosis.Rehabilitation Counseling Bulletin,39(1), 214.

20.

Roessler, R., Rumrill, P., & Hennessey, M. (2002). Employment concerns of people with multiple sclerosis: Building a national employment agenda. Kent, OH: Kent State University Center for Disability Studies, Report Submitted to the National Multiple Sclerosis Society.

21.

Roessler, R., Rumrill, P., Li, J., & Leslie, M. (2015). Predictors of differential employment statuses of adults with multiple sclerosis.Journal of Vocational Rehabilitation,42(2), 141–152.

22.

Rumrill, P., & Roessler, R. (2015). An overview of multiple sclerosis: Medical, psychosocial, and vocational aspects of a chronic and unpredictable neurological disorder.Rehabilitation Research, Policy, and Education, 29(2), 122–134.

23.

Rumrill, R., Roessler, R., Li, J., Daly, K., & Leslie, M. (2015). The employment concerns of Americans with Multiple Sclerosis from a national sample.Work, 52(4), 735–748.

24.

Rumrill, P., Roessler, R., & Fitzgerald, S. (2004). Vocational rehabilitation-related predictors of quality of life among people with multiple sclerosis.Journal of Vocational Rehabilitation,20(3), 155–163.

25.

Schapiro, R. (2012). Managing the symptoms of multiple sclerosis (4th ed.). New York, NY: Demos Medical Publishing.

26.

Smart, J. (2016). Disability, society, and the individual (3rd ed.). Austin, TX: Pro-Ed.

27.

Strauser, D. (2014). Career development, employment, and disability in rehabilitation. New York, NY: Springer Publishing Company.

28.

Super, D. (1980). A life-span, life-space approach to career development.Journal of Vocational Behavior,16, 282–298.

29.

Tabachnick, B., & Fidell, L. (2001). Using multivariate statistics. Boston, MA: Pearson.

30.

Uccelli, M., Specchia, C., Battaglia, M., & Miller, D. (2009). Factors that influence the employment status of people with multiple sclerosis: A multi-national study.Journal of Neurology, 256(12), 1989–1996.

31.

Vickers, M. (2012). Antenarratives to inform health care research: Exploring workplace illness disclosure for people with multiple sclerosis (MS).Journal of Health and Human Services Administration,35, 170–206.

32.

Wehman, P. (2013). Life beyond the classroom: Transition strategies for young people with disabilities (5th ed.). Baltimore, MD: Brookes Publishing Company.

33.

World Health Organization (2008). Atlas: Multiple Sclerosis resources in the world 2008. Geneva: World Health Organization. pp. 15–16.