Vocational Expectations and Self‐Stigmatizing Views Among Collegiate Recovery Students

Rates of substance dependence for college‐age (ages 18–24) individuals are 2 to 3 times higher than those of any other age group (Substance Abuse and Mental Health Services Administration, 2013); some studies suggest that as many as 1 in 4 college students meet the diagnostic criteria for substance dependence at some point (Hunt, Eisenberg, & Kilbourne, 2010).

Substance use disorders (SUDs) are complex and chronic (Laudet & White, 2010; National Institute on Drug Abuse [NIDA], 2014). According to NIDA (2014), relapse rates are estimated to be around 60%, suggesting it is essential for researchers to identify the best methods for supporting the needs of those in recovery to reduce relapse risk. Navigating the collegiate environment poses a particular risk to individuals who are in recovery from addiction because of the high rates of substance use and the drinking culture that are prevalent on many college campuses (Cleveland, Harris, Baker, Herbert, & Dean, 2007). Exposure to substances, cues related to substance use, and stress, in general, are three of the most common factors related to relapse (NIDA, 2014); these factors are widely prevalent on college campuses. Because of these stressors, some students may be forced to choose between sustaining recovery and pursuing education. Furthermore, as the college environment poses substantial risk for individuals in recovery, returning to substance use may result in severe life consequences, including dropping out of college (Harris, Kimball, Casiraghi, & Maison, 2014; Hunt et al., 2010), that are likely to impede successful transition from college to work (Arria et al., 2013).

Collegiate Recovery Programs

In response to the unique needs of individuals in recovery from SUDs and behavioral process addictions, some universities provide academic‐ and recovery‐oriented support using collegiate recovery programs (CRPs). The primary focus of CRPs is to help students maintain recovery while providing support for academic success, which may be direct (on‐site advising, study skills workshops, etc.) or indirect (by connecting students to on‐campus resources, such as health services, counseling, etc.; Harris et al., 2014). CRPs provide access to peer support recovery groups (such as 12‐step fellowships and other mutual help groups) and psychoeducational seminars focused on substance use and mental health–related issues, and they connect students to various modes of professional support, such as counseling and local agencies (Laudet, Harris, Kimball, Winters, & Moberg, 2015). These programs provide a much‐needed form of social support and continuing care, which is a foundational aspect of successful recovery (Laudet, Magura, Vogel, & Knight, 2000). In addition to sustaining recovery from alcohol, drugs, and behavioral addiction, many students also report mental health–related issues. In some samples (Laudet et al., 2015) as many as 75% have reported receiving a mental health–related diagnosis or seeking help for mental health–related issues, emphasizing the need for specialized support to meet the complex needs of this student group.

In response to the growing need for recovery support on college campuses, there has been a substantial increase in the number of CRPs nationwide (Laudet et al., 2015; Transforming Youth Recovery, 2015). Recent data indicate there are approximately 239 operating CRPs in the United States, 40% of which are in the earliest stages of development (Transforming Youth Recovery, 2018). The range and depth of support offered by CRPs vary greatly, often depending on the stage of development and university resources. Research is limited in the area of collegiate recovery (Laudet et al., 2015; Perron et al., 2011); specifically, little empirical support exists that examines the precise needs of students in recovery. Education and career development have been stated as primary goals for college students generally, including students in recovery (Laudet & White, 2010); however, research among students in recovery is severely lacking. As more CRPs continue to develop across colleges and universities, it is essential that research efforts reflect findings intended to be useful to CRP directors and other university stakeholders as they refine their programs and work toward addressing student barriers to successful recovery during the college years and beyond.

Career Development

Regardless of the length of time in recovery, continuing education and career development are perceived as top priorities for many individuals in recovery (Laudet & White, 2010); continuing education offers new opportunities for individuals to explore career interests and strengthen skills essential for work (Arria et al., 2013). However, research has indicated that a low level of educational attainment has been perceived as one of the significant barriers to employment by individuals with SUDs; those with SUDs also report lower levels of self‐perceived basic skills (e.g., reading, writing, arithmetic) and office skills as compared with those without SUDs (Atkinson, Lee, Dayton‐Shotts, & French, 2001).

In many cases, reductions in work and disruptions to career development occur as a result of past substance use (American Psychiatric Association, 2013; Laudet, 2008). Consequences related to substance use may directly affect work adjustment, job performance, and interpersonal relationships in the workplace. Individuals with a SUD history are more likely than the general population to have a legal/criminal history, which can affect employability or licensure in specific fields. Thus, completing postsecondary education becomes critical for students in recovery as one meaningful way to offset the barriers that many face. For many people in recovery, these perceived barriers to employment may reduce motivation to work toward academic and employment goals (Lent & Brown, 2013). Self‐efficacy and outcome expectations are stable predictors of behaviors related to career development (Lent, Ireland, Penn, Morris, & Sappington, 2017). Very little is known about collegiate recovery members and their outcome expectations and efficacy beliefs with respect to career and employment.

Outcome Expectations and Self‐Efficacy

Self‐efficacy is defined as an individual's belief in his or her capacity to accomplish prospective goals. Outcome expectations (a core principle of self‐efficacy theory) are defined as a person's appraisals of the expected consequences that a particular goal or behavior will yield (Bandura, 1986, 1997). For example, the belief that one can obtain employment would be a self‐efficacy belief. The anticipation that one will succeed in one's career and receive self‐satisfaction from one's work are examples of outcome expectations associated with the efficacy belief of obtaining employment. According to Bandura's (1997) theory, outcome expectations set the foundation for self‐efficacy beliefs as they focus on likely or anticipated outcomes related to potential goals. These beliefs are central to setting goals, motivation, and overall goal attainment (Lent et al., 2017) because individuals who perceive themselves as being capable of achieving a goal that is rewarding in some capacity are more likely to persist despite challenges. Recent research findings (Lent et al., 2017) demonstrate that self‐efficacy and outcome expectations are central to intentional behavior related to career development and career decidedness. Furthermore, research findings indicate that outcome expectations may causally influence one's level of perceived self‐efficacy (Williams, 2010), and such causal relationships between outcome expectations and self‐efficacy may potentially affect individuals' motivation to pursue postgraduation employment.

Individuals' beliefs regarding their capacity to accomplish career‐related goals are influenced by perceived capabilities, past successes or failures, and previous experiences with employment (Bandura, 1997). According to social cognitive theory (Bandura, 1986), human behavior, cognitive processes, and environmental events all mutually shape one another, suggesting that behaviors are influenced by personal beliefs that are primarily influenced by individual and environmental factors. Individuals often avoid tasks and situations that they perceive to exceed their capabilities, such as taking a particular course, studying for a challenging exam, or submitting a job application. These perceptions are especially vulnerable to environmental influence, which could take the form of societal stigma (Bandura, 1986). Self‐efficacy and outcome expectations are quite vulnerable to internalized negative stereotypes and self‐stigma (Corrigan et al., 2012); these beliefs often result in self‐deprecation that discourages engagement in goal‐directed behavior.

Stigma

Stigma is a primary issue facing individuals in recovery who are pursuing postsecondary education (Dearing, Stuewig, & Tangney, 2005), yet it has been understudied in recent years (Scott, Anderson, Harper, & Alfonso, 2016). Such stigma is societally pervasive and is prevalent within the collegiate environment (Dearing et al., 2005; Perron et al., 2011). Stigma can be conceptualized as a negative set of societal beliefs, often unfounded, that result in stereotyping and social separation (Link & Phelan, 2001). A majority of the research on stigma has focused on its relationship to mental health (Parcesepe & Cabassa, 2013; Sickel, Seacat, & Nabors, 2014), but stigma is a prevailing issue encountered by individuals with both mental health disorders and SUDs (Corrigan et al., 2005; Kulesza, Larimer, & Rao, 2013; Ronzani, Higgins‐Biddle, & Furtado, 2009; Room, 2005; Schomerus et al., 2010). In fact, individuals with SUDs often experience higher levels of societal stigma when compared with individuals with mental illness, psychiatric disabilities, or other disabilities (Corrigan, Kuwabara, & O'Shaughnessy, 2009), as they are more likely to be perceived as dangerous and to blame for their current diagnosis when compared with individuals with mental illness.

Corrigan and Watson (2002) emphasized two different types of stigma: (a) Social stigma is a public endorsement of bias often conveyed through discrimination, and (b) self‐stigma is the internalization of these public biases. Internalized stigma is significantly and positively related to reduced self‐esteem, reduced self‐efficacy, poor quality of life, and self‐disrespect (Corrigan, Bink, Schmidt, Jones, & Rusch, 2016; Corrigan, Sokol, & Rusch, 2013; Corrigan, Watson, & Barr, 2006). Corrigan et al. (2016) illustrated the relationship between public stigma and internalized stigma through the “why try” model. The “why try” model explains that individuals with mental illness exist inside a culture with widespread adverse beliefs regarding mental illness. Individuals who internalize these negative stereotypes and attitudes develop negative self‐perception, which has a profound effect on self‐esteem and self‐efficacy (Corrigan et al., 2006). These internalized beliefs can present barriers to the pursuit of life goals, including pursuing postgraduation employment. For instance, a person who has internalized negative self‐beliefs because of his or her substance use history may question whether trying to achieve academically or prepare for employment is worth the effort when little is expected to come from these efforts. It is clear how stigma is theoretically linked to a construct such as self‐efficacy when this model is used.

Purpose of the Study

CRPs are increasing in response to more students in recovery pursuing higher education. Research regarding CRPs is also growing; however, the needs of students who utilize their services remain underexamined (Laudet et al., 2015). As CRPs grow in numbers and existing CRPs develop their programming, it is essential to understand the specific barriers faced by students in recovery so that CRPs can continue to position themselves as uniquely qualified to address the barriers.

Stigma is frequently cited as a significant issue facing college students in recovery and has been connected to adverse outcomes in prior research (Corrigan et al., 2016; Dearing et al., 2005; Perron et al., 2011). Little research has been done specifically related to self‐stigma and its relationship to career development and other factors among students in recovery pursuing higher education (Laudet & White, 2010; Scott et al., 2016). Because career development is a primary goal for the majority of individuals seeking postsecondary education, and self‐stigma is a prevalent issue facing students in recovery, the purpose of this study was to examine the relationship between self‐stigma and vocational expectations. Self‐stigma has been shown to hurt self‐efficacy in previous research (Corrigan et al., 2012). Less is known about specific outcome expectations related to career goals for students in recovery pursuing higher education. We hypothesized that significant relationships would be detected between self‐stigma and vocational outcome expectations. We also investigated a related research question, seeking to determine whether other variables, such as grade point average (GPA), quality of life, and psychological health, would significantly relate to these variables in our sample and potentially aid in our determination of recovery‐based needs for this sample.

Method

Participants

Participants (N = 80) for this study were recruited from 33 university CRPs and collegiate recovery communities nationwide. Information regarding the nature of the study and informed consent was forwarded to the Association of Recovery in Higher Education and to Transforming Youth Recovery. The Association of Recovery in Higher Education and Transforming Youth Recovery were specially selected to assist with data collection as they are the only associations currently designated to support college students in recovery. Administrative staff forwarded information regarding this study to CRP and collegiate recovery community directors nationally, who then forwarded this information to their communities through email or social media (i.e., Facebook and Twitter). Information included a link to an online Qualtrics survey; participants were offered the opportunity to take part in a drawing for a $50 Visa gift card as an incentive for participating in this study by providing an email address at the end of the survey (overall four drawings took place). Of the 121 surveys that were initiated, 82 participants submitted complete surveys, resulting in a 68% response rate. During analysis, two cases that were considered extreme outliers were removed. Nationwide, there were roughly 58 programs that served 1,354 students in Fall 2014 (Transforming Youth Recovery, 2015). The current sample represented a portion of the overall population of students in recovery pursuing higher education, and the sample size was consistent with previous research (Cleveland et al., 2007; Scott et al., 2016).

A majority (63.8%) of participants were female, 35.0% were male, and one participant reported transgender. Participants' ages ranged from 18 to 52 (M = 26.96, SD = 7.05). Seventy‐eight percent of participants identified as straight/heterosexual, 17.5% reported bisexual identity, and 5.0% reported lesbian or gay identity. Most (67.5%) were single/never married, 17.5% reported living with a partner, 13.8% reported being married in a domestic partnership or civil union, and one participant reported being divorced. The majority (74.1%) of participants reported they were White, 14.8% reported Hispanic or Latina/o, 4.9% reported Asian American or Pacific Islander, 1.2% reported Native American, 1.2% reported Black or African American, and 3.7% reported other. Most (63.9%) were undergraduate students: freshman (6.3%), sophomore (6.3%), junior (20.0%), and senior (31.3%). A little over a third (36.3%) reported working on a graduate degree. A majority of students (63.8%) were employed, and a majority (71.3%) lived off campus. Participants reported being in recovery for drugs (65.0%), alcohol (70.0%), and behavior or process addiction (32.5%). Over two thirds (67.1%) self‐reported as receiving a mental health–related diagnosis such as depression (45.0%), anxiety (31.3%), bipolar disorder (8.8%), posttraumatic stress disorder (6.3%), or attention‐deficit disorder (6.3%); 31.3% reported receiving more than one mental health–related diagnosis; and 55.0% reported receiving specialty treatment outside the CRP for a SUD. Percentages in this section do not all total 100 because of rounding.

This study was approved by the first author's institutional review board. Inclusion criteria required that all participants be at least 18 years of age, be of any gender, and be currently enrolled in college and involved in a CRP or collegiate recovery community). Participants were informed that involvement in this survey was voluntary, and no personally identifying information would be collected.

Variables

Self‐stigma. We used the Self‐Stigma Mental Illness Scale–Short Form (Corrigan et al., 2012) to evaluate self‐stigma. The Self‐Stigma Mental Illness Scale measures four domains of self‐stigma related to the following: (a) awareness: knowledge of common stereotypes surrounding mental illness; (b) agreement: agreement that stereotypes surrounding mental illness are accurate and truthful; (c) application: the degree to which an individual applies negative stereotypes surrounding mental illness to himself or herself; and (d) hurts self: the degree to which an individual experiences decreased worth or decreased perceived ability as a result of applying negative stereotypes related to mental illness to himself or herself. The individual domains and the total scale can be used depending on how the subscales relate to each other in a given sample (Corrigan et al., 2012). Only the Hurts Self subscale was used for the current study as this subscale best captures the operational definition of how self‐stigmatizing beliefs can sabotage a person's goals, which is of interest in the current study. Corrigan et al. (2006) established discriminant validity for the Hurts Self subscale as it is significantly and negatively correlated with the Rosenberg Self‐Esteem Scale (Rosenberg, 1965; r = –.57) and the Sherer and Adams Self‐Efficacy Scale (Sherer & Adams, 1983; r = –.44). Minor adaptations were applied to this subscale to include recovery from addiction. Sample items include “Because I am in recovery and have a mental illness, I cannot be trusted” and “I currently respect myself less because I am to blame for my problems.” Items are scored on a 9‐point Likert scale (1 = strongly disagree, 9 = strongly agree), with higher scores indicating higher levels of self‐stigmatizing tendencies. This 10‐item subscale demonstrated an acceptable degree of internal consistency reliability for this sample (α = .79).

Vocational outcome expectations. We used a modified version of the Vocational Outcome Expectancy Scale (VOES; Iwanaga et al., 2017) to measure vocational outcome expectations. The full VOES was constructed using social cognitive career theory, an extension of Bandura's (1997) self‐efficacy theory, for item development (Lent et al., 2008). The scale contains 11 items across two factors: the six‐item Positive Vocational Expectancy subscale and the five‐item Negative Vocational Expectancy subscale. Strong psychometric properties were established as evidenced by results from the exploratory factor analysis and confirmatory factor analysis (Iwanaga et al., 2017). Specifically, results from the exploratory factor analysis and confirmatory factor analysis support the two‐factor scale, with each subscale demonstrating sufficient reliability (α = .79). The results of the positive subscale significantly correlated with measures of other constructs as expected, including the motivation for work (r = .39) and job performance self‐efficacy (r = .47). Only the six‐item Positive Vocational Expectancy subscale was used for the current study because the results of the confirmatory factor analysis provided stronger psychometric properties for this subscale (Iwanaga et al., 2017). The full two‐factor scale was initially developed for individuals with disabilities receiving vocational rehabilitation services. As such, the anchor for this scale was changed from “Completing my VR [vocational rehabilitation] program will likely allow me to …” to “Completing my education will likely allow me to ….” Items are measured on a 5‐point Likert scale (1 = strongly disagree, 5 = strongly agree) with higher scores indicating higher expectancy beliefs. Sample items include “If I decided to pursue my career, I could work for an employer who would be supportive of people in recovery from substances or behavioral/process addictions” and “If I decided to pursue my career, it would likely lead me to a job that is good for my lifestyle or have a job with good pay and benefits.” The six‐item scale demonstrated an acceptable degree of internal consistency reliability for this sample (α = .71).

Control variables. Participants were asked to self‐report several points of information as a way to control for potential factors that may explain negative self‐views or positive vocational expectations. Participants indicated their current GPA and indicated their overall quality of life and psychological health. All participants (100%) reported data for these demographic variables, with GPAs ranging from 2.00 to 4.00 (M = 3.41, SD = 0.51). Participants indicated their quality of life by designating a number from 1 to 10, with higher scores representing higher overall quality of life; participants' scores ranged from 3 to 10 (M = 8.20, SD = 1.55). Participants also completed the Patient Health Questionnaire for Depression and Anxiety (Kroenke, Spitzer, Williams, & Lowe, 2009), a brief four‐item screener for anxiety and depression. The Patient Health Questionnaire for Depression and Anxiety has sound psychometric properties and is significantly and positively correlated (r = .80) with the Medical Outcomes Study Short‐Form General Health Survey Mental Health Subscale (Stewart, Hays, & Ware, 1988). Participants respond on a 4‐point Likert‐type scale (0 = not at all, 3 = nearly every day) to indicate how often they have been bothered by anxious or depressive symptoms over the last 2 weeks. Sample items include “Feeling nervous, anxious, or on edge” and “Little interest or pleasure in doing things.” Internal consistency reliability was good for this sample (Cronbach's α = .84).

Data Analysis

Before conducting analyses, we conducted two preliminary tests to compare students' scores for our independent and dependent variables on the basis of employment status and graduate or undergraduate student status. Independent samples t tests demonstrated no significant differences between employed participants and unemployed participants on self‐stigma, t(80) = .201, p = .84, or positive vocational outcome expectations, t(80) = 0.457, p = .65. Similarly, no significant difference was found between undergraduate or graduate students on self‐stigma, t(80) = –.226, p = .82, or positive vocational outcome expectations, t(80) = 0.418, p = .68, justifying merging the full sample in the analyses designed to test our hypothesis.

To test our hypothesis, we used a two‐tailed Pearson correlation and regression analysis. Assumptions for ordinary least squares estimation regression analysis were verified before analysis (Cohen, Cohen, West, & Aiken, 2003); two extreme outliers were identified and removed from the analysis. Boxplots, scatterplots, and normal P–P plots were examined to determine (a) a linear relationship between variables, (b) normal distribution, and (c) independence and reasonable distribution of residuals. All assumptions for regression analyses were met before we conducted analyses.

A post hoc power analysis for R2 deviation from zero fixed linear model regression was conducted using G*Power 3.1 (Faul, Erdfelder, Buchner, & Lang, 2009) and the following parameters for three independent variables: n = 80, α = .05, and ES = 0.15, with estimated 1 – β error probability = .82, indicating adequate power to detect significant effects for this sample. Missing data analysis was performed using SPSS, Version 22, confirming that 93.81% of the data were present before the analysis was conducted. Results from Little's missing completely at random test (χ2 = 456.84, df = 590, p = 1.0) suggested that data were missing completely at random; therefore, multiple imputation was used as it is considered an appropriate method when missing data values are not extreme (Enders, 2010; Schlomer, Bauman, & Card, 2010).

Results

Results of the correlational analysis revealed GPA to have nonsignificant and negligible relationships with all other variables. Moderate to strong relationships were detected between self‐stigma and quality of life (r = –.54), self‐stigma and depression/anxiety symptoms (r = .34), and self‐stigma and vocational expectations (r = –.27). Vocational outcome expectations were unrelated to any other variable aside from self‐stigma. Depression/anxiety symptoms and quality of life were significantly and negatively related (r = –.48). See Table 1 for full correlational results.

Table 1

Two‐Tailed Pearson Intercorrelations Between Independent and Dependent Variables

Note. N = 80.

*p < .05. **p < .001.

We determined our regression model on the basis of the results of our correlation matrix in conjunction with prior research findings, as psychological health and quality of life have been connected to self‐efficacy and self‐stigma in prior research (Corrigan et al., 2016; Corrigan, Larson, & Rusch, 2009). Because we relied on a single‐item assessment for quality of life and a brief four‐item screener for psychological health, these variables were entered as controls in Step 1 of our model to determine the relationship between self‐stigmatizing tendencies and vocational expectations. Both models were significant (p = .000), with the final model containing all variables accounting for 34% of the variance in self‐stigma. Quality of life contributed uniquely and significantly to the model, B = –.49, t(80) = –4.56, p = .000, whereas anxiety/depression symptoms, B = .11, t(80) = 1.0, p = .327, failed to make a significant contribution overall. Positive vocational outcome expectations, B = –.19, t(80) = –2.00, p =.049, significantly contributed to the model as well. We did not include GPA in the model because it was unrelated to any variable in the correlation matrix. See Table 2 for the results of regression models.

Table 2

Predictors of Self‐Stigma

Note. N = 80. R2 = .34.

Discussion

A successful transition from postsecondary education to work is imperative for many students in recovery. Even though CRPs can benefit students in recovery, pervasive stigma has been a critical challenge facing this group (Dearing et al., 2005; Perron et al., 2011). Given the relevance of employment beliefs and the influence of self‐stigma in career development for students in recovery, the critical goal of this study was to examine the relationship between vocational expectations and self‐stigma while considering other factors that may relate to these constructs among students in recovery. We expected to find a stronger relationship between vocational expectations and self‐stigma on the basis of prior research connecting these constructs in other samples and on the basis of the “why try” model (Corrigan et al., 2016). Although vocational expectations did significantly relate to self‐stigma, the control variables made more significant contributions to the model explaining self‐stigma.

This is the first study we are aware of that investigated self‐stigmatizing tendencies and vocational expectations among students pursuing higher education who are affiliated with a CRP. This is important as the growing trend of CRPs on college campuses continues. The stated aims of CRPs are to provide social and academic support to students. CRPs also generally aim to connect students in recovery with additional supports on campus, such as mental health providers, campus career centers, academic supports, residential options, and social outlets designed to support recovery efforts. Each of these additional services is available to a varying degree and is of varying quality depending on university resources. The unique strength of CRPs is to provide support to students that other collegiate services do not typically provide. Although it is essential for developing CRPs to design their services and aims on the basis of research findings, this endeavor is challenging because of the minimal research available on this topic. The findings from this study contribute to closing the gap and can serve CRP directors and other university and CRP stakeholders as they continue to develop programmatic supports designed to consider the barriers to, and solutions most likely to enhance, recovery for students pursuing higher education.

On the basis of the findings from our sample, we recommend that CRPs be aware of the relationships among self‐stigmatizing tendencies, vocational expectations, quality of life, and psychological health. Continued support for student psychological health is recommended, as our results align with prior research indicating high rates of psychological disorders in populations of students with SUDs (Laudet et al., 2015). The correlation results further suggest that higher overall psychological health is related to lower self‐stigmatizing tendencies and higher overall quality of life. Many factors contribute to quality of life, and our results indicate that CRPs should aim to better understand what specific factors explain high quality of life among students in recovery.

The results of the regression model indicate that self‐stigma is significantly explained by poorer vocational outcome expectations, meaning that students who hold reduced expectations about their future career are more likely to have self‐stigmatizing tendencies, which can further demotivate them to work on academic and career‐related goals. The self‐defeating beliefs related to one's substance abuse can be a vicious cycle for some. This is the first study to our knowledge that assessed these constructs among students affiliated with CRPs in recovery. To our knowledge, addressing self‐stigmatizing views is not explicitly included as a specific aim of CRPs; however, these issues are essential to consider to help students attain their long‐term career goals. We suggest that one unique goal for CRPs could be to include methods to help students identify self‐stigmatizing views, understand how they connect to adverse life outcomes, and work to combat them. For example, self‐stigma is connected to reduced motivation (Corrigan, Larson, & Rusch, 2009) and lower quality of life (Corrigan et al., 2013). We also recommend that CRPs focus on strengthening students' vocational expectations, because this enhances the motivation needed to complete academic training.

Strategies for Reducing Stigma

Authors (Corrigan et al., 2016; Mittal et al., 2012) have noted several strategies to reduce self‐stigma: (a) education, often entailing the use of cognitive behavioral techniques to address self‐stigmatizing beliefs through facts and cognitive restructuring; (b) social support, specifically from peers with shared experiences; and (c) programming related to disclosing mental illness or recovery status with others. Postdisclosure responses from peers, instructors, advisors, and professors have the potential to reinforce self‐stigmatizing views depending on the response. For this reason, CRPs should be particularly mindful of this possibility and provide opportunities to process these experiences and help students recognize when negative responses from others are reinforcing self‐stigmatizing views. CRPs can encourage self‐advocacy as one way to combat negative responses from others and self‐stigmatizing views.

Corrigan et al. (2017) also recommended facilitating educational campaigns to encourage exposure to individuals with SUDs in positive contexts and to challenge myths and assumptions related to individuals with SUDs. Increasing positive messages and positive experiences with people in recovery can reduce social stigma, which can contribute to reductions in self‐stigma. Social support is already foundational to CRPs; directors may consider enlisting the help of on‐campus counseling services to be available to address issues related to self‐stigma. College counselors should be aware of the detrimental impact that self‐stigma has on individuals in recovery from both substance‐ and mental health–related issues and should actively seek to challenge associated false assumptions.

Strategies for Increasing Career Outcome Expectations

Vocational outcome expectations and self‐efficacy are consistently and significantly related in other studies (Williams, 2010); vocational outcome expectations serve as a foundation for self‐efficacy and goal‐directed behavior related to career development. Thus, strategies to develop outcome expectations will likely have an impact on self‐efficacy and positive behaviors associated with career development. CRPs, as well as other counselors typically working with students in recovery, including in career centers or student health centers, can target vocational expectations in goal setting. Expectations can be modified by providing positive role models in combination with experiences designed to increase expectations. CRPs can provide learning experiences for students (mock interviews, résumé development, etc.) either directly or by utilizing campus resources (career center or career counselors; Lent et al., 2017). Developing a strong alumni network of individuals in recovery to serve as role models and mentors to students offers opportunities for learning, instills hope, and may provide instrumental support central to successful career attainment (Bandura, 1997; Lent et al., 2017). It is vital that CRPs establish a strong working relationship with college counselors on campus and career services to provide the supports necessary for students in recovery as they transition postgraduation. On the basis of the results of the current study, CRPs can target vocational expectations as a way to not only enhance students' career development but also possibly reduce their self‐stigma.

Future Research

We were somewhat surprised that GPA was unrelated to any study variable, including vocational expectations. We expected students with higher GPAs to have significantly higher vocational expectations and lower self‐stigma. The self‐report data could explain the nonsignificant relationships to GPA in the current sample; the range for GPA was sufficient, but self‐reported data cannot be validated. Future research should aim to validate student GPA or other academic measures to determine the relationship to self‐stigma and vocational expectations. Psychological health and quality of life demonstrated strong relationships to self‐stigma, which was not surprising because these findings are consistent with prior research findings (Corrigan et al., 2013; Evans, Banerjee, Leese, & Huxley, 2007).

Given the strong relationships between self‐rated quality of life and self‐stigma, we recommend future research focus on operationalizing components of subjective and objective quality of life for this population. Our one‐item, self‐rated assessment of global quality of life is not sufficient to understand points for intervention that could enhance the quality of life for students similar to those in our sample. Because our sample reported a range of psychological disorders, and their psychological health rating strongly related to self‐stigma and quality of life, we recommend continued research efforts focused on ways that CRPs can routinely screen for symptoms of psychological illness and ensure appropriate referrals for maintaining health and preventing worsening of symptoms and conditions. We recommend more comprehensive measures to assess psychological health rather than the brief screening instrument used in the current study. We strongly recommend continuing research on factors that have an impact on career development and vocational expectations for students in recovery in higher education settings. Future research should aim to confirm or refute our results connecting self‐stigma and vocational expectations and to conduct follow‐up studies to determine whether these factors predict employment outcomes postgraduation.

Limitations

Findings from this study should be considered in light of several limitations. An overwhelming majority of participants were White, limiting the generalizability of results. Future studies should aim to replicate these findings with larger, more diverse samples and attempt to explore reasons why White participants are often overrepresented at the level of service provision. The cross‐sectional nature of this study does not permit causal interpretations; future studies should strive to gather larger samples to better represent the population of students in recovery. Although the variable of vocational outcome expectations is a stable predictor of future success, it does not capture employment outcome; future research should examine employment outcome as well as vocational outcome expectations.

Conclusion

The results of this study indicated that self‐stigma was significantly and negatively associated with positive career outcome expectations in a sample of students who were seeking services from CRPs nationwide. As CRPs continue to develop and deliver support to students in recovery pursuing higher education, they are uniquely positioned to provide support tailored to the needs of these students. On the basis of the results of this preliminary study, CRPs may benefit students by specifically addressing self‐stigmatizing views and vocational outcome expectations among their members. Both constructs can have an impact on overall academic and career success as well as recovery success.