Preparing youth with disabilities for employment: An analysis of vocational rehabilitation case services data

Abstract

BACKGROUND: Despite national efforts to improve post-school outcomes, many students with disabilities are unprepared to enter the workforce. Coordination with Vocational Rehabilitation (VR) can offer opportunities for improved employment outcomes.

OBJECTIVE: This study examined Oregon VR data to identify predictors of positive closure status for youth with disabilities.

METHODS: Researchers used logistic regression to explore the effects of individual, in-school, post-school and contextual factors on VR case closure status among 4,443 young adults with disabilities who received and completed services from Oregon VR between 2003 and 2013.

RESULTS: Being female, having a mental illness, traumatic brain injury, multiple disabilities, interpersonal or self-care impediments to employment, receiving SSI, and closing with VR in periods of high unemployment reduce the likelihood of a positive VR closure. Participating in a collaborative transition program, earning a high school completion certificate, receiving a greater number of VR services, closing below the median number of days to closure, and closing in low unemployment periods increase the likelihood of a positive VR case closure status.

CONCLUSION: These findings highlight specific barriers to employment for vulnerable groups of young adults with disabilities, and identify service and contextual factors that can support positive employment outcomes.

Keywords

Transition youth with disabilities vocational rehabilitation employment outcomes

1 Introduction

Despite ongoing national efforts to develop effective transition services and improve post-school outcomes, young adults with disabilities still lag behind their peers without disabilities on a number of critical school and post-school indicators (Newman et al., 2011; Samuels, 2015). The transition from high school to post-school employment can be particularly challenging. In 2013, young adults with disabilities ages 16 to 19 experienced an unemployment rate of 42% compared to 26% for their same age peers without a disability (U.S. Department of Labor, 2013). In an analysis of a national sample of students with disabilities (National Longitudinal Transition Study-2), Sima and colleagues found that only 38% of students with cognitive, emotional, sensory, physical and learning disabilities were competitively employed within six-years of leaving high school (Sima, Wehman, Chan, West, & Leucking, 2014).

State vocational rehabilitation agencies are uniquely positioned to provide services for youth with disabilities preparing for employment (Honeycutt, Bardos, & McLeod, 2015). Under the Rehabilitation Act of 1973, VR agencies are required to coordinate with education officials to help facilitate the transition from in-school services to post-school services for eligible individuals with disabilities (Rehabilitation Act of 1973). VR counselors can develop customized plans for employment and provide an array of services to support the transition to the labor market. Recent reauthorization of the Workforce Innovation and Opportunities Act in 2014 strengthens the connection between education and rehabilitation systems by mandating that state VR agencies spend at least 15% of their federal VR program funds to deliver Pre-Employment Transition Services (PETS) to all potentially eligible students with disabilities in schools (U.S. Department of Education, 2014). PETS include: (a) job exploration counseling, (b) work-based learning experiences, (c) counseling on postsecondary education options, (d) workplace readiness training, and (e) instruction in self-advocacy.

The Youth Transition Program (YTP) is one example of a collaborative transition program that utilizes a partnership between state education and rehabilitation agencies to support school based transition services. Since 1990, schools in Oregon have used the YTP model to structure the coordination of in-school transition services and post-school VR services (Benz, Lindstrom, & Lata, 1999). Participating schools apply for this enhanced VR service through a competitive grant process that combine fiscal support from VR agencies and local school districts to hire an in-school transition specialist who coordinates transition services for individuals determined eligible for VR services (Benz et al., 1999; Benz, Lindstrom, & Yovanoff, 2000). Essential program characteristics of YTP include: (a) individualized student focused planning on post school goals and self-determination and help to coordinate school plans with relevant community agencies; (b) instruction in academic, vocational, independent living and personal social skills and help to stay in and complete high school; (c) career development services including goal setting, career exploration, job search skills, and self-advocacy; (d) work experience including connection with local employers, on the job assessments, placement and training; (e) support services such as individualized mentoring and support or referrals for additional specific interventions; and (f) follow-up support for one year after leaving the program to assist in maintaining positive outcomes in employment or post-secondary education (Alverson et al., 2015). These components are in alignment with recent research documenting a number of specific predictors of successful post-school employment and education outcomes including work experience and interagency collaboration (Mazzotti et al., 2016; Test et al., 2009).

Although VR agencies are mandated to provide rehabilitation services in collaboration with school systems, there is scant research on the effects of VR involvement on individual employment outcomes for young adults with disabilities in their transition to the labor market. In a comprehensive review of the literature, Poppen (2015) found that only seven studies on this topic have been published in peer-reviewed journals through 2015. Those studies identified a number of factors that contribute to either an increased or decreased likelihood that a young adult will close as rehabilitated after receiving services from VR. Overall, predictors of VR case closure outcomes have been attributed to individual characteristics such as disability type; in-school services such as work experience or vocational education; post-school services such as job placement services; and contextual factors, such as location or unemployment rates (Giesen & Cavenaugh, 2012; Gonzolez et al., 2011; Flannery, Benz, Yovanoff, McGrath Kato, & Lindstrom, 2011; McDonnal & Crudden, 2009; Migliore, Timmons, & Butterworth, 2012; Schaller, Yang, & Trainor, 2006; Sung, Sanchez, Kuo, Wang, & Leahy, 2015).

Individual characteristics identified as predictors of VR case closure outcomes include age, sex, race/ethnicity and disability category (Flannery et al., 2011; Giesen & Cavenaugh, 2012; Migliore et al., 2012). In their study investigating VR case closure outcomes among young adults with visual impairments, being female and African American significantly reduced the likelihood of positive VR case closure; being Hispanic and having more education at the time of application significantly increased the likelihood of positive VR case closure; and having a greater severity of visual impairment and additional disabilities significantly reduced the likelihood of positive VR case closure (Giesen & Cavenaugh, 2012). In contrast, Sung et al., (2015) found that for young adults with autism, there were no significant differences by sex in rates of positive VR case closures. Another study found that for young adults with visual impairments, high levels of academic competencies, self-determination, internal locus of control, and use of assistive technology were all significantly associated with positive VR case closure outcomes (McDonnall & Crudden, 2009).

Specific in-school and post-school services, and contextual factors also significantly impact the likelihood of young adults achieving positive VR case closure outcomes. For example, findings from a study investigating differences in rates of VR case closure outcomes for individuals ages 18 to 25 with ADHD suggest that predictors of positive case closure include receiving VR counseling and/or job placement services (for males only), and receiving job search assistance services (for males and females; Schaller et al., 2006). Further, an analysis of a national sample of young adults with autism ages 16 to 26 who received services from VR found that job placement services, participation in post-secondary education, closing below the median number of days to closure, receiving miscellaneous training and receiving job search services, all significantly contributed to an increased likelihood of positive VR case closure outcomes (Migliore et al., 2012). Additionally, annual state unemployment rate is an example of a contextual factor shown to significantly correlate with the proportion of youth who closed from VR with employment between 2006 and 2008 (Honeycutt, Thompkins, Bardos, & Stern, 2013).

Although previous research has identified factors that predict the likelihood of successful and unsuccessful VR case closure for young adults with disabilities, there are gaps in the existing literature. Almost all of the existing empirical studies focus on specific subpopulations of young adults with disabilities (e.g., students with specific learning disabilities, students with visual impairments, students with autism) and examine data from a single year as compared to longitudinally. In addition, there are few empirical studies examining the impact of contextual factors, such as geographic characteristics and labor market factors, as predictors of VR case closure outcomes for young adults with disabilities.

In line with the clear need to develop a better understanding of key factors that lead to positive employment outcomes for young adults with disabilities (Madaus, Dukes, & Carter, 2013), the present study examines predictors of case closure outcomes for young adults with disabilities receiving services from VR. We did so by addressing the following research question: How do individual characteristics, in-school experiences, post-school experiences, and contextual factors predict VR case closure status among young adults with disabilities?

2 Method

This study focused on identifying individual demographic characteristics, school related experiences, post-school experiences, and contextual factors that predict positive VR closure outcomes among a sample of 4,443 young adults with disabilities who received services from VR in Oregon between July 1st, 2003 and June 30th, 2013. Individual demographic characteristics included sex, race, primary disability category, impediments to employment, and receipt of Social Security Income (SSI). School related experiences included participation in a collaborative transition program, and high school completion status. Post-school experiences included number of VR services received, number of days to closure, and participation in post-secondary education. Contextual factors included community type, and federal fiscal year of closure.

2.1 Data source

Data for this study were obtained from the Oregon Rehabilitation Case Automation system (ORCA). ORCA is an integrated case management database used by Oregon Vocational Rehabilitation to record and track individual demographic characteristics, services provided, and employment outcome data for individuals who receive services. Information about each individual who applies for VR services in Oregon is entered into ORCA by a VR counselor, and reports are accessed by regional and state offices for caseload statistics, reminders of due dates and impending actions, documentation for federal and state reports, and for budgeting, planning and reporting purposes. The data used for this study were made available by the Oregon VR office in a de-identified format for research and evaluation purposes. This dataset has particular advantages over the RSA-911 public use files, including having more detailed and complete information about each individual who received VR services from Oregon during a ten-year period.

2.2 Participants

We used a de-identified data set of 9,117 young adults with disabilities (age 21 or younger at the time of application) who had applied for VR services between 2003 and 2013 and had a valid case record in the ORCA database. From the initial sample, 4,443 participants were included for this study, utilizing the following inclusion criteria: (a) applied for VR services between July 1st, 2003 and June 30th, 2013; (b) were determined eligible to receive services; (c) had a completed Individualized Plan for Employment (IPE); and (d) closed from VR services between July 1st 2003 and June 30th, 2013. Data from 4,674 (or 51%) individuals were not included because: (a) they had applied for services outside of the selected time frame (n = 111), (b) were not determined eligible for VR services (n = 681), (c) did not have a completed IPE (n = 2,819), or (d) had not yet closed with VR (n = 1,063).

2.3 Dependent variable

2.3.1 Vocational rehabilitation closure status

VR closure status is a dichotomous outcome variable with two levels: (a) closed rehabilitated, and (b) closed other than rehabilitated. A closed rehabilitated outcome indicates that an individual has entered into full-time or part-time employment in the integrated labor market, has satisfied the vocational outcome of supported employment, or has any other vocational outcome determined appropriate by the Secretary of Education including self-employment, telecommuting, or business ownership (Rehabilitation Act, 1973). If an individual does not meet the criteria to be closed rehabilitated, they are closed as other than rehabilitated. Other than rehabilitated may indicate that the individual refused services, could not be located, or was placed in a sheltered workshop or non-competitive work setting (Oregon Administrative Rules, 2016).

2.4 Independent variables

We used a multi-level approach for identifying potential factors associated with VR case closure status. Specifically, we examined the effects of individual characteristics, in-school experiences, post-school experiences, and broader contextual factors on VR closure status.

2.4.1 Individual demographic characteristics

Previous studies have found that rehabilitation outcomes can vary across individual demographic characteristics (e.g. age, gender, disability; Flannery et al., 2011; Geisen & Cavenaugh, 2012; Shaller et al., 2006). Using previous findings and data available in ORCA, we included age, sex, race, primary disability, impediments to employment, and receipt of SSI in our model and hypothesized that these characteristics would influence VR case closure status.

2.4.1.1 Age and sex. Age at application was included as a continuous variable. Sex (male/female) was included as a dummy variable with males coded as the reference category (male = 0; female = 1).

2.4.1.2 Race. Racial background was assessed using five categories: (a) White, (b) Native American, (c) Asian, (d) Black, or (e) Hawaiian. Less than 7% of the sample (n = 305) reported a race other than White. Therefore, we re-coded the responses to a dummy variable using White as the reference category (coded as 0), and non-White coded as 1.

2.4.1.3 Primary disability... Primary disability was coded using a combination of two disability specific variables that were included in the ORCA data set: (a) primary disability impairment code, and (b) primary disability cause/source code. There were 213 unique combinations of these two codes that were recoded into nine primary disability categories based on the labels defined in federal special education legislation (Individuals with Disabilities Education Improvement Act, 2004); this approach was modeled after previous research using VR data (Cimera, 2009). The nine disability categories used for this study included: (a) Attention Deficit Hyperactivity Disorder (ADHD), (b) Autism, (c) Intellectual Disability, (d) Mental Illness, (e) Other Cognitive Impairments, (f) Physical and Mobility Impairments, (g) Sensory or Communication Impairments, (h) Specific Learning Disability (SLD), and (i) Traumatic Brain Injury (TBI). The largest group of participants (1,602; 36.4%) were coded with a primary disability of SLD and were used as a reference category to compare to all other disability categories.

2.4.1.4 Impediments to employment. Impediments to employment are determined by a VR counselor as a part of the eligibility determination process and are used to guide service delivery. An impediment to employment refers to a physical or mental impairment that prevents the individual from accessing employment. Without an impediment to employment, an individual cannot receive services from VR. Impediments to employment are classified within seven domains: (a) communication, (b) interpersonal, (c) mobility, (d) self-care, (e) self-direction, (f) work skills, and (g) work tolerance. Individuals may have an impediment to employment in more than one of the seven domains, and for that reason the effect of each of the seven domains on VR closure status were tested independently using a unique dummy variable for each of the seven impediment domains (no = 0; yes = 1).

2.4.1.5 Receipt of social security income at application (SSI). SSI is a nationwide disability benefit program for persons with limited income administered by the Social Security Administration. Individuals with disabilities are eligible for SSI if they have a monthly income that does not exceed the relevant SSI rate determined by the state. The relevant SSI rate for a single individual living in Oregon was $721/month in the last year of data analyzed for this study (U.S. Social Security Administration, 2014). For our study, the effect of being a recipient of SSI on VR closure status was tested using a dummy variable (no = 0; yes = 1)

2.4.2 School-related experiences

Previous studies have also found that school related experiences can have an effect on post-secondary outcomes for young adults with disabilities (Benz, et al., 2000; Sanford et al., 2011). Based upon these findings, and the available data in ORCA, we included participation in a transition program during high school, and earning a high school completion certificate by the time of closure as school-related experiences hypothesized to influence VR case closure status.

2.4.2.1 Youth transition program. (YTP) Participation in a transition program during high school is another factor that has been shown to positively influence post-school employment outcomes for young adults with disabilities (Benz et al., 2000; Repetto, Webb, Garvan, & Washington, 2002). We used participating in YTP as a school-level predictor variables in our model, coded 1 if it was reported that the individual participated in this program and 0 if not.

2.4.2.2 High school completion. High school completion (HSC) is a school-related experience that has been shown to influence post-secondary employment outcomes for young adults with disabilities (Mazzotti et al., 2016; Sanford et al., 2011, Test et al., 2009; Wagner, Newman, & Javitz, 2014). In the ORCA database, level of educational attainment is reported for each individual at the time of closure, with response categories including: (a) no formal education; (b) elementary education (grades 1-8); (c) secondary education, no HS diploma (grades 9–12); (d) special education completion certificate; (e) high school graduate or equivalency certificate; (f) post-secondary education, no degree or certificate; (g) AA degree or Vocational-Technical school certificate; (h) bachelor’s degree; and, (i) master’s degree or higher. We used this information to create a dummy variable for HSC, coding 0 for individuals who were reported as not having earned at least a special education completion certificate (categories a-c), and 1 for those who reported having earned a special education completion certificate or greater (categories d-i).

2.4.3 Post-school experiences

2.4.3.1 Number of VR services received The number of services an individual received from VR was computed from 27 separate variables that indicated if an individual received one of 27 specific VR services offered by Oregon VR between 2003 and 2013. The services received by young adults in the sample included: (1) assessment (64.8%), (2) other services (60.2%), (3) transportation services (49.3%), (4) VR counseling and guidance (36.5%); (5) job placement assistance (32.1%), (6) job search assistance (26.7%), (7) job readiness training (17.7%), (8) occupational/vocational training (17.6%), (9) on-the-job supports (15.4%), (10) information and referral services (15.1%), (11) college or university training (11.1%), (12) diagnoses and treatment of impairments (9.9%), (13) job development (9.9%), (14) job preparation (9.4%), (15) on-the-job training (7.8%), (16) job placement (7.6%), (17) job retention (6.5%), (18) rehabilitation technology (5.8%), (19) basic remedial or literacy training (4.8%), (20) job coaching (4.7%), (21) disability related augmentative skills training (4.1%), (22) maintenance (2.9%), (23) miscellaneous training (2.9%), (24) interpreter services (2.4%), (25) technical assistance services (<1%), (26) personal attendant services (<1%), (27) reader services (<1%). We calculated the number of services that an individual had received and treated it as a continuous variable in analyses that ranged from 0 to 16, where 0 indicates no record of services received. The variable ranged from 0 to 16, because 16 was the maximum number of services any one individual received.

2.4.3.2 Below the median number of days to closure. The total number of days between an application date and closure date were calculated for all participants. A variable was then computed to test the effect of closing below the median number of days on VR closure status (i.e., 490 days, or 1.36 years). Individuals who had closed with VR at or above the median number of days were coded as a 0 and those who had closed below the median number of days were coded as a 1. The median was used as the measure of central tendency for this variable to account for individuals who may have extreme values on either end of the spectrum.

2.4.3.3 Participation in post-secondary education. Coded similarly to HSC at closure, some post-secondary education at closure was computed using the “Closure Education Level” field in ORCA. We used this information to create a dummy variable for participation in at least some post-secondary education at closure, coding 0 for individuals who reported yes to categories 1–5 (“no formal education” through “high school graduate or equivalency certificate”), and 1 for those who reported educational attainment of 6 (“post-secondary education, no degree or certificate” through “master’s degree or higher”) or greater.

2.4.4 Contextual factors.

We considered contextual factors as variables that would broadly impact the likelihood of an individual finding employment (i.e., outside of the scope of individual, in-school, and post-school factors). Contextual factors included both community type where services were provided and Federal Fiscal Year (FY).

2.4.4.1 Community type. Following the Metropolitan Statistical Area (MSA) definition set by the U.S. Office of Management and Budget and used by the U.S. Census Bureau (U.S. Census Bureau, 2014), we coded the county where an individual received services into one of three types: metropolitan (population, ≥50,000), micropolitan (10,001–49,999), or rural (≤10,000). Individuals who received services in a metropolitan area were coded as 0, micropolitan as 1, and rural as 2; with metropolitan as the reference variable for analysis.

2.4.4.2 Federal fiscal year of closure. The federal fiscal year represents the period between July 1 and June 30 of the following calendar year (for example, FY13 represents the period July 1, 2012- June 30, 2013). Deviation from the means coding was used to compare the effect of the FY when an individual closed to the grand mean. This is a particularly relevant variable given the fluctuations in the labor market between December 2007 and June 2009 (U.S. Bureau of Labor Statistics, 2012). The results of this coding scheme reports the average effect of FY of closure on VR case closure status, and allows for a comparison of the effect of each individual FY with the overall effect of FY. Any significant deviation identifies FY’s in which the odds of obtaining a particular VR case closure status were more or less than the overall effect of FY. This method of coding was achieved by setting all the design variables equal to -1 for FY13, and then using 0,1 coding for FY’s 2003 through 2012 (Hosmer et al., 2013).

2.5 Data analysis

First, we examined the descriptive statistics and bivariate associations between the study variables. Subsequently, missing data and collinearity were assessed. Multicollinearity between variables was determined to pose no problem for analyses using zero-order correlations, which ranged from 0.001 and 0.400. Lastly, logistic regression model building procedures were used to develop a statistical model predicting VR case closure status. Our procedures for handling missing data and building a logistic region model are detailed below.

2.5.1 Missing data

Patterns of missing data were explored prior to model building to determine any significant group differences between those records with missing values and those without. Missing data was present for the variable race (n = 1;<0.1%) and primary disability (n = 40; 0.9%). Chi-square tests revealed that there were no statistically significant differences between individuals with and without missing data for any of our variables. It is not clear why these records were missing; however, because these missing values represent such a small percentage of the sample (n = 41; 0.9%) and no patterns of missingness were identified, these records were excluded from model building analysis using listwise deletion (Allison, 2002).

2.5.2 Model building procedures

A statistical model that predicted a positive VR case closure status of rehabilitated was built using a seven-step logistic regression model building procedure (Hosmer et al, 2013). In the first step we conducted a single-variable examination of each independent variable using a standard contingency table. Variables that were statistically significant at the bivariate level (p≤0.20) were identified as candidates for the first multivariable model. The initial significance level was set at p≤0.20 to ensure that variables providing a meaningful contribution to the model were not omitted from the model. In the second step we fit a multivariable logistic regression model and removed variables that were not statistically significant at our pre-established level for this study (p≤0.01). In the third step, we compared the results from the first model with those from the larger model. During this step we added back into our model, variables that were excluded during the second step, but determined to cause a change in the estimated coefficients of greater than 20% in other variables. Fourth, we added each variable that was omitted from either step one or step two back into our model, one at a time, to check if it was significant in the presence of other variables. During the fifth step, we tested and confirmed the assumption of linearity for the continuous variable (average number of VR services received). In the sixth step, we tested all plausible interaction effects by examining their bivariate relationship with the dependent variable, and adding the interaction effect to the main effects model if it was statistically significant at the bivariate level (p≤0.01 for this study). Subsequently, we added each of the independent variables removed in previous steps back into our model and examined these variables to see if any of them were statistically significant in the presence of the recently added interaction effect. Lastly, we determined the fit of the logistic regression model using Hosmer & Lemeshow goodness of fit test and Nagelkerke’s pseudo R².

3 Results

Of the 4,443 individuals included in this study, 60.8% were closed as rehabilitated (e.g., employed) and 39.2% were closed other than rehabilitated. Our sample consisted of more males than females (61.0% vs. 39.0%) and was predominately white (93.1% vs. 6.9% non-white). The four most common primary disabilities for individuals in our sample were Specific Learning Disability (36.4%), Mental Illness (11.4%), Intellectual Disability (10.5%) and Autism (9.6%). The most frequently reported impediments to employment included work skills (85.4%), interpersonal (51.5%), self-direction (51.2%), and communication (51.5%). Almost one fifth (18.6%) of the sample were receiving SSI at the time of application. Just over two-thirds of our sample participated in YTP (68.4%), and 78.0% earned some type of high school completion certificate by the time that closed with VR. On average, participants received just over four services from VR (M = 4.22; SD = 2.47; Min = 0; Max = 16), 35.3% closed below the median number of days to closure (490 days; or, 1.36 years), and 12.3% had participated in at least some post-secondary education. Furthermore, 74.2% of participants received services from a VR branch office in a metropolitan area, 20.2% received services from a branch in a micropolitan area, and 5.6% were served in a rural area.

3.1 Bivariate findings

Table 1 provides descriptive statistics for participants with a VR closure status of rehabilitated compared with those who had a VR case closure of other than rehabilitated. A significant bivariate relationship with VR case closure status was present for the following variables: sex, primary disability, multiple disabilities, impediments to employment, receipt of SSI, participation in YTP, HSC, average number of VR services, closing below the medium number of days to closure, and federal fiscal year.

3.2 Logistic regression analyses

The first step of the seven step model building procedure (described in detail in the methods above) revealed that at the bivariate level, all variables were statistically significant predictors of VR closure status at a probability level of p≤0.20, and were thus retained for subsequent multivariable model building procedures (see Table 2). The second, third, fourth and fifth steps indicated that in the multivariable model, age, race and five of the seven impediments to employment were not statistically significant predictors of VR closure status at a probability level of p≤ 0.01, and were removed from the multivariable model. In the sixth step, all plausible interaction effects were tested, and one interaction effect, YTPxHSC, was statistically significant at the bivariate level (p = 0.01); suggesting that the interaction between participating in YTP and having earned a high school completion certificate (HSC) may provide a meaningful contribution to the model and that the effect of one of those two variables was not constant across the levels of the other variables (Hosmer et al., 2013). Once added to the main effects model, however, the effect of YTPxHSC was non-significant (p = 0.012); following the model building procedures this interaction remained in the final model (Hosmer et al., 2013). Lastly, the Hosmer and Lemeshow test failed to reject the null hypothesis, suggesting an adequate model fit ready for interpretation, (χ² [8, 4,401] = 7.888, p = 0.444). The final multivariable model is presented in Table 3.

Our final multivariable model indicates that while holding the effects of all other variables constant, the individual characteristics of being female, having a primary disability of mental illness or traumatic brain injury, multiple disabilities, having an interpersonal or self-care impediment to employment, and being a recipient of SSI were associated with significantly lower odds of closing VR with a case closure status of rehabilitated. Specifically, the odds of females closing VR rehabilitated are 72% as likely as the odds of males. (OR = 0.72, 95% CI [0.63, 0.82], p < 0.001). Additionally, the odds of individuals with mental illness and individuals with traumatic brain injury closing VR rehabilitated are 62% and 37% (respectively) as likely as the odds of individuals with SLD (OR = 0.62, 95% CI [0.49, 0.79], p < 0.001; OR = 0.37, 95% CI [0.19, 0.71], p < 0.01), and the odds of individuals with multiple primary disabilities closing VR rehabilitated are 81% as likely as the odds of individuals with one primary disability (OR = 0.81, 95% CI [0.71, 0.93], p < 0.01).

Additionally, the odds of individuals with an interpersonal impediment to employment closing VR as rehabilitated are 80% as likely as the odds of those without an interpersonal impediment to employment (OR = 0.80, 95% CI [0.69, 0.93], p < 0.01), and the odds of individuals with a self-care impediment to employment closing VR as rehabilitated are 76% as likely as the odds of those without a self-care impediment to employment (OR = 0.76, 95% CI [0.65, 0.89], p = 0.001). Furthermore, the odds of individuals who were receiving SSI at application closing VR as rehabilitated were 71% as likely as the odds of those who were not receiving SSI at application (OR = 0.71, 95% CI [0.59, 0.85], p < 0.001). There were no individual characteristics included in our multivariable analyses that were associated with significantly greater odds of closing VR with a case closure status of rehabilitated.

The school related experiences of participating in YTP and earning a high school completion certificate by the time of VR closure were associated with significantly greater odds of closing VR as rehabilitated. Specifically, young adults who participated in YTP were 2.34 times more likely to close VR rehabilitated than young adults who did not participate in this enhanced transition service (OR = 2.34, 95% CI [1.69, 3.23], p < 0.001). Those who had received a high school completion certificate or higher at the time of VR closure were 2.03 times more likely close VR rehabilitated than those young adults who had not yet earned a high school completion certificate (OR = 2.03, 95% CI [1.49, 2.75], p < 0.001).

Furthermore, the post-school related experiences of receiving a greater number of VR services and closing below the median number of days to closure were also associated with greater odds of positive VR closure. Specifically, for each VR service that an individual received, they were 1.29 times more likely to close VR rehabilitated (OR = 1.29, 95% CI [1.25, 1.34], p < 0.001). Young adults who closed from VR below the median number of days to closure were 2.14 times more likely to close rehabilitated (OR = 2.14, 95% CI [1.85, 2.49], p < 0.001).

Finally, compared to individuals who had closed in any other year, individuals who closed in FY04 had a greater likelihood of closing as rehabilitated (OR = 1.46, 95% CI [1.12, 1.91], p < 0.01; and the odds of individuals who closed in FY08 or FY09 closing as rehabilitated ate 74% and 62% as likely as those who had closed in any other year (OR = 0.74, 95% CI [0.61, 0.90], p < 0.01, and OR = 0.62, 95% CI [0.50, 0.77], p < 0.001).

4 Discussion

This study examined the effects of individual characteristics, in-school experiences, post school experiences and contextual factors on VR case closure status among 4,443 young adults with disabilities who received services from Oregon Vocational Rehabilitation between FY2003 and FY2013. Overall, the findings from this study confirm and extend a sparse body of research examining factors that influence VR case closure status for young adults with disabilities in transition from high school to the workforce. By understanding the implications of individual characteristics, in-school experiences, post-school experiences and contextual factors on VR case closure status, rehabilitation professionals can be better prepared to enhance career development opportunities for youth with disabilities – especially for individuals who face significant barriers to employment (e.g., young women; those with a mental illness, multiple disabilities or traumatic brain injury, and those who are trying to find work in a difficult labor market). Further, this study lays the groundwork for additional investigation of the influence of collaborative transition programs on post-school outcomes and provides an important reminder about the critical influence of labor market factors on post-school outcomes for students with disabilities.

4.1 Individual characteristics

Previous studies investigating VR case closure status for young adults with disabilities have focused on specific subgroups of young adults with disabilities (e.g. young adults with visual impairments or autism; Giesen & Cavenaugh, 2012; Gonzolez et al., 2011; McDonnal & Crudden, 2009; Migliore et al., 2012; Schaller et al., 2006; Sung et al., 2015). Our study extends the literature by examining outcomes for young adults across nine different disability categories and offers important new information about groups of young adults who may need additional services and supports to successfully enter employment. The results from our study suggest that young women across all disability groups are significantly less likely than their male counterparts to be closed as rehabilitated and thus less likely to enter the labor market. This finding is consistent with previous research documenting sex differences in employment outcomes for young adults with disabilities (Carter, Austin, & Trainor, 2012; Newman et al., 2011) and supports the contention that young women with disabilities may need targeted efforts to increase career options even within the existing school transition and VR systems (Doren, Gau, & Lindstrom, 2011; Doren, Lombardi, Clark, & Lindstrom, 2013).

Additionally, while previous research has documented poor employment outcomes for young adults with more severe disabilities (i.e., intellectual disabilities, autism, etc.; Boeltzig, Timmons, & Butterworth, 2009; Newman et al., 2011; Metzel et al., 2007; Roux et al., 2013), our findings indicate – after controlling for the effects of other individual characteristics, in-school experiences, post-school experiences and contextual factors – that young adults with intellectual disabilities or autism were not significantly less likely to close VR rehabilitated than young adults with specific learning disabilities. One possible explanation for these findings is that the in-school experiences and services that are offered to these students (i.e., high school completion, collaborative transition programs) may address many of the structural and institutional barriers that have prevented these young adults from entering the workforce in the past. However, our findings also indicate that young adults with a mental illness, multiple disabilities or traumatic brain injury are less likely to close rehabilitated than their peers with a specific learning disability. Our findings may suggest that individuals with mental illness, traumatic brain injuries, and multiple disabilities face unique sets of circumstances that are more difficult to address within current school and VR services, or were not accounted for in this model (i.e., public stigma, impulsivity, etc.). These differences across disability groups reinforce the need to individualize career development and transition planning services offered to young adults with disabilities while they are still in school.

Although our descriptive data showed that white youth had a higher percentage of VR case closures of rehabilitated, multivariate analysis did not find race (white/non-white) to be a significant predictor of VR closure outcomes. This finding is inconsistent with earlier studies and multiple reports documenting disparities in post school employment outcomes between Hispanic and African American young adults with disabilities and their white peers (Giesen & Cavenaugh, 2012; Newman et al., 2011; Newman, Wagner, Cameto, & Knokey, 2009). We wondered if this finding might be related to our method of combining all individuals who were non-white into a single grouping, which could be covering up important differences found in more marginalized racial sub-groups. Furthermore, this finding may also indicate that like sex, the inclusion of other factors in our model captured variance that would have otherwise been attributed to race. However, it is also important to note that our data had a disproportionately low rate of young adults who were non-white. When compared to the proportion of non-white students with disabilities who exit the Oregon public school system each year (25% during the 2011-2012 school year; Oregon Annual Performance Report, 2013), we noted that only 6.9% of young adults who received and completed VR services were non-white.

4.2 In school experiences

Another important finding from this study is the correlation between a collaborative transition program (YTP) and VR closure outcomes for young adults with disabilities. After controlling for the effects of other factors included in our model (see Table 3), our model suggest that students who participated in YTP were 2.47 times more likely to be closed as rehabilitated. Previous research investigating program characteristics of YTP has linked paid work experience and transition goal completion to positive in-school and post-secondary outcomes (Benz et al., 2000). These findings are also similar to a recent study conducted using statewide VR data in Oklahoma, that reported evidence that school transition programs improve both hourly wages and work hours (Jun, Kortering, Osmanir, & Zhang, 2016). Interestingly, Jun et al., report that the effects of school transition programs were larger for students with non-cognitive disabilities, and that a much smaller percentage of participants in their data base participated in a school based transition program than did the participants in the Oregon study (46% compared to 68%). One possible explanation for the effects of this program on VR closure outcomes may be the capacity that it creates for expanded and enhanced transition service delivery in the schools. The findings from our study suggest that the enhanced services provided to VR applicants through this collaborative transition program are correlated with improved access to employment for young adults with disabilities, and that additional efforts to explain this correlation are needed.

Furthermore, our findings also indicate that individuals who had earned at least a high school completion certificate were 2.14 times more likely to close rehabilitated than those who had not (in this study high school completion certificate was defined as special education completion certificate, high school diploma, or equivalency certificate). This finding is consistent with previous research reinforcing the importance of high school completion in developing skills and knowledge that support pathways to employment (Kortering, 2012; McDaniel & Kuehn, 2012; Rabren, Hall, & Brown, 2003).

4.3 Post-school experiences

Previous research has demonstrated that the types of services provided by VR are related to positive closure outcomes among young adults with disabilities (Flannery et al., 2011; Geisen & Cavenaugh, 2012; McDonnall & Crudden, 2009; Migliore et al., 2012; Schaller et al., 2006). Specifically, these services include job placement, general or vocational supports, college services, job search assistance, miscellaneous training, VR counseling, financial support, and assistive technology. The results of our study support existing literature, and also provide a different, and meaningful, insight into the relationship between number of VR services and rehabilitation outcomes. Our findings indicate that for each additional service that an individual received, they were 1.29 times more likely to close VR rehabilitated. This means that an individual who received the average number of VR services (about 4) was about 3.87 times as likely to close rehabilitated as someone who had only received a single service from VR. In addition to exploring the number of VR services received, our findings also suggest that young adults with disabilities who closed services with VR on or below the median number of days (490 days or 1.34 years) were 2.29 times more likely to close VR as rehabilitated. These findings are also consistent with previous research from Migliore et al. (2012) who found that young adults with autism who closed below the median number days for their sample (820 days or 2.25 years) were 1.54 times more likely to be competitively employed. While these findings provide new evidence about the relationships between VR services and closure status, they do not necessarily indicate that a greater number of VR services, or, closing below the median number of days to closure, leads to a greater probability of closing VR as rehabilitated. Both of these factors are dynamic and are a result of careful efforts to address the unique needs of each individual who applies for services through VR. Instead, these findings may serve as a foundation for future research investigating the relationships between VR services and closure outcomes, and serve as a discussion point as state and local teams seek to provide effective services to young adults with disabilities.

4.4 Contextual factors

Although previous research has demonstrated that VR closure outcomes rise and decline along with the performance of the general economy (Gilmore, Schuster, & Butterworth, 2001; Chan et al., 2013), and that job rates for individuals with disabilities follow the same basic trends as for those without disabilities (U.S. Bureau of Labor Statistics, n.d.), labor market factors are often omitted when exploring post-secondary outcomes for young adults with disabilities. Between December 2007 and June 2009, the number of job openings, or unfilled jobs, in the U.S. decreased 44% from its pre-recession peak in March 2007 (U.S. Bureau of Labor Statistics, 2012). The unemployment rates in Oregon followed the same trend seen at the federal level and rose sharply from 5.1% in June 2007 to 11.6% in June of 2009 (U.S. Bureau of Labor Statistics, n.d.). Furthermore, during the four years between July 2003 and June 2007, the Oregon unemployment rate dropped more than three full percentage points, from 8.6% to 5.1%, (U.S. Bureau of Labor Statistics, n.d.). About one third of this drop, 1.1%, was experienced during FY 2004 (U.S. Bureau of Labor Statistics, n.d.). Findings from our study support previous trends on the effects of labor market factors on VR outcomes for young adults with disabilities and suggest that: (a) young adults with disabilities who close from VR during a period with high rates of unemployment (i.e., FY 2008, FY2009) are less likely to close VR as rehabilitated; and (b) those who close during periods of low unemployment rates (i.e., FY 2004) are significantly more likely to close rehabilitated.

4.5 Limitations

This study had several limitations. First the extant data set analyzed was developed for service delivery and not research purposes. All data were entered into ORCA by local VR counselors in the course of providing services. This limitation may raise some concerns about the reliability of the data that should be considered when interpreting the results. Additionally, this limitation prevented us from measuring the quality or degree of implementation of school and VR services provided to individuals included in this study, including the differentiation of services that at face value appear similar (i.e., job placement assistance vs. job placement). Second, because this study utilized an extant dataset, we did not have control of variables available for analysis and were unable to include other predictor variables previously shown to influence VR case closure status among young adults with disabilities (i.e., self-determination, vocational skills self-efficacy, course of study, previous work experience, family socio-economic status, etc.). Third, the results are correlational, not causal, and the population included in this study was not a nationally representative sample and only included individuals from a single state in the northwest; thus, readers are cautioned to not over generalize the findings from this study.

4.6 Implications for practice

The results of our study identified a number of vulnerable groups that continue to face barriers to entering the workforce, despite receiving individualized services from schools and VR. Examples of these vulnerable groups include young women, individuals with mental illness, multiple disabilities, and traumatic brain injury. Our findings indicate a need for an increase in targeted career development opportunities for these particularly vulnerable groups of individuals, including additional opportunities for career exploration, school and community work experiences, and individualized instruction in work skills and behaviors (Doren et al., 2013). Our findings also point to the importance of increasing access to collaborative transition programs that build capacity for transition services, promote high school completion and create linkages between school and rehabilitation systems.

4.7 Implications for research

There are also a number of implications for research that stem from this study. First, it would be useful to replicate this study using extant data from other state VR agencies, as well as the nationally available RSA-911 database. Second, additional research is needed to understand and address the barriers to employment for young women, and individuals with mental illness, multiple disabilities, and traumatic brain injuries. Third, we need to carefully examine, operationalize and replicate essential features of collaborative school based transition programs that build capacity for transition services, and promote high school completion. Fourth, researchers should be reminded to consider labor market factors when conducting longitudinal research on employment outcomes. Fifth, based upon our discovery that nearly 51% of VR applicants age 21 or younger do not complete VR service plans the demographic characteristics and outcomes of these young adults needs to be explored and better understood. Lastly, future research should include measures beyond employment status such as measures related to the quality and type of employment attained (e.g., setting, wages, hours, benefits received, etc.).

Conflict of interest

The authors have no conflict of interest to report.

References

Allison

(2002). Missing data. Sage University Papers Series on Quantitative Applications in the Social Sciences. Thousand Oaks, CA: Sage.

Alverson

, Burr

, FitzGerald

, Dickinson

, Johnson

, Ozols

, & Simich

(2014). Oregon Youth Transition Program Procedures Manual (2nd Ed.). Eugene, OR: University of Oregon, College of Education. Retrieved from: http://www.ytporegon.org/sites/default/files/fileattachments/Master%20Sep%209%20v5.pdf

Benz

, Lindstrom

, & Latta

(1999). Improving collaboration between schools and vocational rehabilitation: The Youth Transition Program model. Journal of Vocational Rehabilitation, 13, 55–63.

Benz

, Lindstrom

, & Yovanoff

(2000). Improving graduation and employment outcomes of students with disabilities: Predictive factors and student perspectives. Exceptional Children, 66(4), 509–529.

Boeltzig

, Timmons

, & Butterworth

(2008). Entering work: Employment outcomes of people with developmental disabilities. International Journal of Rehabilitation Research, 31, 217–223. doi: 10.1097/MRR.0b013e3282fb7ce5

Butterworth

, Migliore

, Sulewski

J. S.

, & Zalewska

(2014). Trends in employment outcomes of young adults with intellectual and developmental disabilities 2004-2012. Boston, MA: University of Massachusetts Boston, Institute for Community Inclusion.

Carter

, Austin

, & Trainor

(2012). Predictors of postschool employment outcomes for young adults with severe disabilities. Journal of Disability Policy, 23, 50–63.

Chan

, Wang

, Ditchman

, Kim

, Joseph

, Chan

, & Dries

(2014). State unemployment rates and Vocational Rehabilitation outcomes: A multilevel analysis. Rehabilitation Counseling Bulletin, 57, 209–218. doi:10.1177/0034355213509855

Cimera

(2009). The national costs of supported employment to Vocational Rehabilitation: 2002 to 2006. Journal of Vocational Rehabilitation, 30, 1–9. doi:10.3233/JVR-2009-0448

10.

Doren

, Gau

J. M.

, & Lindstrom

L. E.

(2012). The relationship between parent expectations and postschool outcomes of adolescents with disabilities. Exceptional Children, 79, 7–23.

11.

Doren

, Lombardi

, Clark

, & Lindstrom

(2013). Addressing career barriers for high risk adolescent girls: The PATHS curriculum intervention. Journal of Adolescence, 36, 1083–1092.

12.

Flannery

, Benz

, Yovanoff

, McGrath Kato

, & Lindstrom

(2011). Predicting employment outcomes for consumers in community college short-term training programs. Rehabilitation Counseling Bulletin, 54, 106–117.

13.

Giesen

, & Cavenaugh

(2012). Transition-age youths with visual impairments in Vocational Rehabilitation: A new look at competitive outcomes and services. Blindness, 106, 475–487.

14.

Gilmore

, Schuster

, & Butterworth

(2001). Vocational Rehabilitation outcomes for people with mental retardation, cerebral palsy, and epilepsy: An analysis of trends from 1985 to 1998. Boston, MA: Institute for Community Inclusion. Retrieved from https://www.communityinclusion.org/pdf/mon29.pdf

15.

Gonzalez

, Rosenthal

, & Kim

(2011). Predicting vocational rehabilitation outcomes of young adults with specific learning disabilities: Transitioning from school to work. Journal of Vocational Rehabilitation, 34, 163–172. doi:10.3233/JVR-2011-0544

16.

Honeycutt

, Bardos

, & McLeod

(2015). Bridging the gap: A comparative assessment of Vocational Rehabilitation agency practices with transition-age youth. Journal of Vocational Rehabilitation, 43, 229–247. doi:10.3233/JVR-150772

17.

Honeycutt

, Thompkins

, Bardos

, & Stern

(2013). State differences in the vocational rehabilitation experiences of transition-age youth with disabilities. Mathematica Policy Research. Retrieved from https://www.mathematica-mpr.com/our-publications-and-findings/publications/state-differences-in-the-vocational-rehabilitation-experiences-of-transitionage-youth-with-disabilities

18.

Hosmer

, Lemeshow

, & Sturdivant

(2013). Applied logistic regression (3rd ed.). Hoboken, NJ: Wiley.

19.

Individuals with Disabilities Education Improvement Act of 2004 (IDEIA). Pub.L.No.108-446, 118 Stat. 2647 (2004)

20.

Jun

, Kortering

, Osmanir

, & Shang

(2015). Vocational rehabilitation transition outcomes: A look at one state’s evidence. Journal of Rehabilitation, 81(2), 47–53.

21.

Kortering

(2012). School completion and adolescent transition education. In Wehmeyer

& Webb

(Eds.), Handbook of Adolescent Transition Education for Youth with Disabilities (pp. 208–222). New York, NY: Routledge.

22.

Madaus

, Dukes

, & Carter

(2013). Introduction to the special issue: Reflecting on the past, present and future of CDTEI. Career Development and Transition for Exceptional Individuals, 36, 4–5. doi:10.1177/2165143413485225

23.

Mazzotti

, Rowe

, Sinclair

, Poppen

, Woods

, & Shearer

(2016). Predictors of post-school success: A systematic review of NLTS2 secondary analyses. Career Development and Transition for Exceptional Individuals, 39(4), 196–215. doi:10.1177/2165143415588047

24.

McDaniel

, & Kuehn

(2012). What does a high school diploma get you? Employment, race and the transition to adulthood. Review of Black Political Economy, 40, 371–399. doi:10.1007/s12114-012-9147-1

25.

McDonnall

, & Crudden

(2009). Factors affecting the successful employment of transition-age youth with visual impairments. Journal of Visual Impairment and Blindness, 103, 329–341.

26.

McDonough

, & Revell

(2010). Accessing employment supports in the adult system for transition youth with autism spectrum disorders. Journal of Vocational Rehabilitation, 32, 89–100. doi:10.3233/JVR-2010-0498

27.

Metzel

, Boeltzig

, Butterworth

, Sulewski

, & Gilmore

(2007). Achieving community membership through community rehabilitation provider services: Are we there yet?Intellectual Developmental Disabilities, 45, 149–160.

28.

Migliore

, Timmons

, Butterworth

, & Lugas

(2012). Predictors of employment and postsecondary education of youth with autism. Rehabilitation Counseling Bulletin, 55, 176–184.

29.

Newman

, Wagner

, Cameto

, & Knokey

(2009). The post-high school outcomes of youth with disabilities up to 4 years after high school. A report from the National Longitudinal Transition Study-2 (NLTS2). Menlo Park, CA: SRI International. Retrieved from www.nlts2.org/reports/

30.

Newman

, Wagner

, Knokey

A.-M.

, Marder

, Nagle

, Shaver

, Wei

, with Cameto

, Contreras

, Ferguson

, Greene

, & Schwarting

(2011). The Post-High School Outcomes of Young Adults With Disabilities up to 8 Years After High School. A Report From the National Longitudinal Transition Study-2 (NLTS2) (NCSER 2011-3005). Menlo Park, CA: SRI International. Retrieved from: www.nlts2.org/reports/

31.

Oregon Administrative Rules, 9 O.A.R. § 582-060 (2016). Oregon Secretary of the State. Retrieved from: http://arcweb.sos.state.or.us/pages/rules/oars_500/oar_582/582_060.html

32.

Oregon Annual Performance Report (2013). Post school outcomes 2013 follow up. Retrieved from: http://triwou.org/projects/tcn/topicspage/639/post-school-outcomes-2013-follow-up

33.

Poppen

(2014). Vocational Rehabilitation: Predicting employment outcomes for young adults with disabilities (Doctoral dissertation). University of Oregon, Eugene, OR. Retrieved from: https://scholarsbank.uoregon.edu/xmlui/bitstream/handle/1794/18734/Poppen_oregon_0171A_11175.pdf?sequence=1&isAllowed=y

34.

Rabren

, Hall

, & Brown

(2003). Employment of transition age rehabilitation consumers: Demographic and programmatic factors. Journal of Vocational Rehabilitation, 18, 145–152.

35.

Rehabilitation Act of 1973. (Pub. L 93-112). 29 U.S.C. § 701 (1973).

36.

Repetto

, Webb

, Garvan

, & Washington

(2002). Connecting student outcomes with transition practices in Florida. Career Development for Exceptional Individuals, 25, 123–139.

37.

Roux

, Shattuck

, Cooper

, Anderson

, Wagner

, & Narendord

(2013). Postesecondary employment experiences among young adults with an Autism Spectrum Disorder. Journal of the American Academy of Child & Adolescent Psychiatry, 52, 897–899.

38.

Sanford

, Newman

, Wagner

, Cameto

, Knokey

, & Shaver

(2011). The post-high school outcomes of young adults with disabilities up to 6 years after high school. A report from the National Longitudinal Transition Study-2 (NLTS2). Menlo Park, CA: SRI International. Retrieved from www.nlts2.org/reports/

39.

Samuels

(2015). Students face uncertain paths after special education. Education Week, Diplomas Count 2015: Report and Graduation Rates, 34(33), 2–4.

40.

Schaller

, Yang

, & Trainor

(2006). Transition-age adults with ADHD: Gender and predictors of vocational rehabilitation outcomes. Journal of Applied Rehabilitation Counseling, 37, 3–12.

41.

Sima

, Wehman

, Chan

, West

, & Leucking

(2014). An evaluation of risk factors related to employment outcomes for youth with disabilities. Career Development and Transition for Exceptional Individuals, 38, 89–11. doi:10/1177/2165143414534887

42.

Sung

, Sanchez

, Kuo

H. J.

, Wang

C. C.

, & Leahy

(2015). Gender differences in Vocational Rehabilitation service predictors of successful competitive employment for transition-aged individuals with autism. Journal of Autism Developmental Disorder, 45, 3204–3218. doi:10.1007/s10803-015-2480-z

43.

Test

D. W.

, Mazzotti

V. L.

, Mustian

A. L.

, Fowler

C. H.

, Kortering

L. J.

, & Kohler

P. H.

(2009). Evidence-based secondary transition predictors for improving post-school outcomes for students with disabilities. Career Development for Exceptional Individuals, 32, 160–181.

44.

U.S. Bureau of Labor Statistics (n.d.). Data retrieval for Current Population Survey: State seasonally adjusted unemployment rates. Retrieved from: http://tinyurl.com/nee463q

45.

U.S. Bureau of Labor Statistics (2012). BLS spotlight on statistics of the recession of 2007-2009. Retrieved from: http://www.bls.gov/spotlight/2012/recession/pdf/recession_bls_spotlight.pdf

46.

U.S. Census Bureau (2014). Urban and rural classification. Retrieved from: https://www.census.gov/geo/reference/urban-rural.html

47.

U.S. Department of Education (2014). The Workforce Innovation and Opportunity Act overview of Title IV: Amendments to the Rehabilitation Act of 1973. Retrieved from: https://www2.ed.gov/about/offices/list/osers/rsa/publications/wioa-changes-to-rehab-act.pdf

48.

U.S. Department of Labor (2013). Persons with Disabilities: Labor force characteristics summary. Retrieved from: http://www.bls.gov/news.release/pdf/disabl.pdf U.S.

49.

U.S. Social Security Administration (2014). Disability Benefits. Retrieved from: http://www.socialsecurity.gov/pubs/EN-05-10029.pdf

50.

Wagner

M.M.

, Newman

L.A.

, & Javitz

H.S.

(2014). The influence of family socioeconomic status on the post-high school outcomes of youth with disabilities. Career Development and Transition for Exceptional Individuals, 37(1), 5–17. doi:10.1177/2165143414523980