Predictors of Help‐Seeking Attitudes in College Student Service Members and Veterans

As more student service members and veterans (SSM/V) utilize educational benefits, it is important to understand the unique mental health needs and challenges of this population as they negotiate a collegiate environment (Barry et al., 2014). Psychological concerns, including depression, anxiety, substance abuse, suicidality, and posttraumatic stress disorder (PTSD), are common for SSM/V (Bonar et al., 2015; Campbell & Riggs, 2015; Rudd et al., 2011). For example, Schonfeld et al. (2015) found that 28.3% of SSM/V report adjustment difficulties upon entering higher education, including acclimating to civilian life and managing emotional problems. Among a sample of SSM/V (N = 189), Love et al. (2015) found that 36.1% of the participants reported high levels of stress, 15.1% reported high levels of anger, 17.3% reported current PTSD symptoms, and 27.1% screened positive for alcohol issues compared with less than 15% of the general college student population using alcohol at this level (American College Health Association, 2019).

Suicidality is also relatively high for SSM/V. Rudd et al. (2011) found that 46% of student veterans (n = 525) had experienced suicidal thinking in the past and 20% of student veterans reported suicidal thoughts with a plan. These rates are higher than those of the general student population in which 13% of university (i.e., nonmilitary) students report suicidal ideation (American College Health Association, 2019; Liu et al., 2019). Recent data from the Center for Collegiate Mental Health (2019) indicate that 33% of SSM/V who seek university counseling services report having seriously considered attempting suicide and 12% have made at least one attempt in their lifetime.

Stigma and Psychological Help Seeking

Given the high acuity in SSM/V, it is incumbent upon university counseling centers to facilitate access to services and attend to the help‐seeking barriers among this student population. Unfortunately, negative beliefs about mental health are often a barrier to psychological help seeking for SSM/V, thus decreasing utilization rates for psychological services such as counseling (for reviews, see Corrigan 2004; Corrigan & Penn, 1999). Only 1.5% of university counseling center clients nationwide self‐identify as SSM/V (Center for Collegiate Mental Health, 2019). This is a relatively low percentage given the total number of SSM/V on campuses, estimated to be at nearly 1 million, along with their documented mental health needs (Borsari et al., 2017). Additionally, Bonar et al. (2015) found that SSM/V in their sample had self‐reported unmet mental health needs, with only 46% of SSM/V who screened positive for mental health issues seeking psychological services in the past year. In contrast, the utilization rate for non‐SSM/V in their sample was 56%.

Yet, there is a paucity of research investigating the relationship between psychological help seeking and mental health attitudes of SSM/V. Stigma about mental illness is theorized as one explanation for low utilization (i.e., seeking psychological services in the face of mental health difficulties) rates among SSM/V. Two types of stigma for SSM/V include public stigma and self‐stigma (Corrigan, 2004). Public stigma constitutes the negative beliefs society holds about people with mental illnesses. For example, individuals with mental health issues are often perceived as dangerous, unpredictable, psychotic, and responsible for their own issues (Angermeyer & Dietrich, 2006; Brockington et al., 1993; Fuller et al., 2000). These attributes are widely accepted and form the basis of public stigma (Corrigan, 2004; Corrigan & Watson, 2002). Self‐stigma is the internalization of negative messages about mental illness (Vogel et al., 2006). This internalization contributes to a person negatively judging themselves should they need to seek necessary psychological treatment (Vogel et al., 2005).

Although limited, some evidence exists to support the connections between stigma and psychological help‐seeking intentions among non‐SSM/V and SSM/V college students. For example, Topkaya (2014) found that self‐stigma negatively predicts attitudes toward psychological help seeking, and Held and Owens (2013) found that self‐stigma fully mediates the relationship between public stigma and attitudes toward psychological help seeking in SSM/V. Furthermore, Currier et al. (2017) found that student veterans who report a need for mental health treatment are less likely to seek professional help and more likely to endorse stigma‐related barriers compared with a matched group of non‐SSM/V peers with similar mental health needs. The most frequently cited reason among SSM/V for not seeking psychological services, even when there is a need, is not wanting treatment to be documented in their military records (Bonar et al., 2015). Other reasons include embarrassment, concerns about treatment harming one's career, and concerns about confidentiality (Bonar et al., 2015).

One gap in previous research is understanding how multiple forms of stigma, such as both public stigma and self‐stigma, relate to help‐seeking behaviors in SSM/V. This is an important omission given that SSM/V attend college with elevated mental health needs but utilize services at university counseling centers at low rates. Understanding and working to eliminate mental health stigma for SSM/V is critical for university counseling center staff and other higher education professionals who provide support for SSM/V on campus.

Study Aims and Hypotheses

The two primary study aims are (a) to examine the extent to which self‐stigma and public stigma predict attitudes toward psychological help seeking in a sample of SSM/V and (b) to determine whether the effect of public stigma on attitudes toward psychological help seeking is mediated by, or contingent on, the variable of self‐stigma. The two study hypotheses are as follows:

Hypothesis 1: Public stigma will be a significant, negative predictor variable for the outcome variable of attitudes toward psychological help seeking for SSM/V.

Hypothesis 2: The significant relationship between public stigma and attitudes toward psychological help seeking proposed in Hypothesis 1 will be fully mediated by the variable of self‐stigma.

Method

Participants

The final analytical sample included 103 SSM/V at a large, land‐grant university in the Mid‐Atlantic region. The target sample was current and former service members enrolled in a degree‐seeking program at the institution. The mean age of the sample was 28.96 years old (SD = 8.23, range 19–54). The sample was 25.2% female (n = 26) and 74.8% male (n = 77). In terms of racial and ethnic identity, 88.3% of the participants were White (n = 91), 5.8% were multiracial (n = 6), 1.9% were Black/African American (n = 2), 1.9% were Hispanic/Latino/a (n = 2), 1% were Asian American/Asian (n = 1), and 1% did not specify (n = 1). (Percentages do not total 100 because of rounding.) Participants completed an average of 6.94 years of military service (SD = 6.05, range 1–30). Regarding military status, 59.2% were veterans (n = 61), 39.8% were National Guard or reservists (n = 41), and 1% were active duty military (n = 1). Among the participants, 47.6% reported that they had been deployed to a combat or armed conflict area (n = 49), and 52.4% indicated that they had not (n = 54).

Measures

Stigma Scale for Receiving Professional Psychological Help. We used the Stigma Scale for Receiving Professional Psychological Help (SSRPH; Komiya et al., 2000) to measure public stigma, the predictor variable in our study. The SSRPH is a five‐item instrument that assesses the degree to which an individual perceives mental health treatment as stigmatizing and socially unacceptable. Items are rated on a 4‐point Likert scale ranging from 0 (strongly disagree) to 3 (strongly agree), with a possible total score of 20. Higher scores on the SSRPH are indicative of greater perceptions of public stigma. Komiya et al. (2000) reported a coefficient alpha of .72 and acceptable internal reliability (α = .78) for the SSRPH. Regarding validity, Komiya et al. reported a strong negative correlation between the SSRPH and the Attitudes Toward Seeking Professional Psychological Help Scale (Fischer & Turner, 1970; r = –.40, p < .01) in which greater public stigma was associated with more negative attitudes toward psychological help seeking. In the current study, the SSRPH demonstrated acceptable internal reliability (α = .78).

Self‐Stigma of Seeking Help Scale. The Self‐Stigma of Seeking Help Scale (SSOSH; Vogel et al., 2006) was used to measure self‐stigma regarding mental health, the predicted mediation variable in our analyses. The SSOSH is a 10‐item measure that assesses the degree to which an individual perceives mental health treatment as harmful to one's self‐satisfaction and self‐worth. Items are rated on a 5‐point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree), with a possible total score of 50. Higher scores are indicative of greater perceptions of self‐stigma. Vogel et al. (2006) reported a strong internal consistency reliability coefficient of .91 (n = 583) and a test‐retest reliability coefficient of .72 (n = 226) after 2 months. Results demonstrated negative associations between SSOSH scores and the Attitudes Toward Seeking Professional Psychological Help Scale (r = –.63, p < .01) in which greater self‐stigma was associated with more negative attitudes toward psychological help seeking. In the current study, the SSOSH demonstrated good internal reliability (α = .86).

Attitudes Toward Seeking Professional Psychological Help Scale–Short Form. We used the Attitudes Toward Seeking Professional Psychological Help Scale–Short Form (ATSPPH‐SF; Fischer & Farina, 1995) to measure attitudes toward psychological help seeking, the outcome variable in our study. This tool assesses the degree to which an individual is willing to engage in professional mental health services. The ATSPPH‐SF is a 10‐item measure that is a modification of the original 29‐item scale (Fischer & Turner, 1970). Items are rated on a 4‐point Likert‐type scale ranging from 0 (disagree) to 3 (agree), with a possible total score of 30. Higher scores on the ATSPPH‐SF are indicative of more positive perceptions of mental health services. Fischer and Farina (1995) reported a strong internal consistency reliability coefficient of .84. In the current study, the ATSPPH‐SF demonstrated good internal reliability (α = .85).

Demographic questionnaire. A nine‐item demographic questionnaire was created for the survey. Participants were prompted to report demographic information, including gender, race/ethnicity, age, class standing (i.e., year in school), military branch, military status (e.g., veteran, active duty), number of years of service, and deployment status.

Procedure

After obtaining institutional review board approval, we recruited participants by email using electronic mailing lists maintained by the university's student veterans organization and the local director of Veterans Affairs. The four measures were presented in counterbalanced order. Upon completing the survey, participants were provided with an optional opportunity to provide their email address to participate in an incentive drawing for a $25 electronic gift card to an online retailer.

Results

Preliminary Analyses

Of the 608 individuals who were sent a recruitment invitation, 118 participated in the survey, for a 19.4% response rate. Fifteen people were then excluded (for an adjusted response rate of 16.9%) because they either did not meet the study criteria or did not complete the measures in their entirety.

An a priori power analysis was conducted using G*Power (Version 3; Faul et al., 2007). To be conservative, we computed this analysis using a small effect size of .15 (Cohen, 1988), three predictor variables (public stigma, self‐stigma, and gender), a mediation model, and a statistical power of .80. Results from the a priori power analysis revealed that a sample size of 77 participants was adequate for this study. Therefore, we sought to collect data from at least 77 participants and obtained a final sample of 103 individuals after accounting for missing data.

To prevent familywise Type I error, we applied a Bonferroni correction to the alpha level in the correlation and regression analyses. Therefore, a significant p value was moved from .05 to .01. Additionally, prior to conducting the multiple regression analysis, we evaluated the completed responses across six model assumptions: (a) independence of observations; (b) linearity; (c) homoscedasticity; (d) multicollinearity; (e) absence of significant outliers, high leverage points, or highly influential points; and (f) normal distribution of residual errors. These preliminary analyses indicated that the assumptions were met.

Primary Analyses

Pearson bivariate correlation analyses. There are two hypotheses in the present study. First, it was hypothesized that public stigma as measured by the SSRPH (Komiya et al., 2000) would be a significant, negative predictor variable for the outcome variable of attitudes toward psychological help seeking for SSM/V, as measured by the ATSPPH‐SF (Fischer & Farina, 1995).

To test the first hypothesis and to evaluate the relationships between the main study variables of public stigma, self‐stigma, and attitudes toward psychological help seeking, we constructed a Pearson bivariate correlation matrix. Results from these bivariate correlation analyses are presented in Table 1. We found a moderate inverse correlation between public stigma and attitudes toward psychological help seeking (r = –.44, p < .01), providing initial support for Hypothesis 1 of the study. This finding indicates that greater public stigma is associated with more negative attitudes toward psychological help seeking in the SSM/V sample.

Table 1

Bivariate Correlation Matrix for the Study Variables

**p < .01, two‐tailed.

Regression analyses. We conducted regression analyses to further test Hypothesis 1. The results are presented in Table 2. In Model 1, the demographic control variables were entered into the analyses. Next, the predictor variable of self‐stigma, along with the control demographic variables, was entered into the regression analyses in Model 2. In this model, self‐stigma significantly predicted attitudes toward psychological help seeking (β = –.64, p < .01). Next, the predictor variable of public stigma, along with the control variables, was entered into the regression analyses as seen in Model 3. In this model, public stigma was shown to be a significant predictor of attitudes toward psychological help seeking (β = –.43, p < .01), in addition to gender and military status (e.g., veteran, active duty). In further support of Hypothesis 1, we found a significant inverse relationship between public stigma and attitudes toward psychological help seeking. Finally, the predictor variables of self‐stigma and public stigma, along with the control variables, were entered into Model 4. Although self‐stigma (β = –.56, p < .01) and gender (β = .27, p < .01) remained significant predictors in Model 4, military status and public stigma were not.

Table 2

Summary of Multiple Regression Analyses for Predictors of Attitudes Toward Psychological Help Seeking

Note. N = 103. Model 1 = control variables, R2 = .15, ΔR2 = .12, SE = 5.71. Model 2 = control variables plus self‐stigma, R2 = .53, ΔR2 = .50, SE = 4.26. Model 3 = control variables plus public stigma, R2 = .33, ΔR2 = .29, SE = 5.07. Model 4 = control variables plus self‐stigma and public stigma, R2 = .54, ΔR2 = .51, SE = 4.23. Military status = active duty/reservist or veteran; Combat status = previous deployment to combat or armed conflict area.

*p < .05. **p < .01.

Test of mediation. The second hypothesis stated that the significant relationship between public stigma and attitudes toward psychological help seeking proposed in Hypothesis 1 would be fully mediated by SSM/V participants' self‐stigma. Initial support for Hypothesis 2 can be observed in Model 4 of Table 2. In this step, the effect of public stigma, which is significant in Model 3, is no longer present when self‐stigma is added to the model (i.e., β = –.13, p > .01). This type of finding may suggest that mediation is occurring (Baron & Kenny, 1986).

To test this possible mediation further, we used Baron and Kenny's (1986) four‐step method (see Table 3). The first three steps of this method establish the zero‐order relationships among the mediation model variables. In the fourth step, mediation is demonstrated when, with the inclusion of the mediator variable into the regression model, the initial relationship between the main predictor variable and outcome variable is no longer significant (Baron & Kenny, 1986). As seen in Table 3, when public stigma and self‐stigma were entered in the fourth model, the relationship of self‐stigma to attitudes toward psychological help seeking remained significant (β = –.56, p < .01), whereas the relationship of public stigma to attitudes toward psychological help seeking was no longer significant (β = –.13, p > .01). In support of Hypothesis 2, these findings suggest that self‐stigma fully mediates the relationship between public stigma and attitudes toward psychological help seeking (see Figure 1).

Table 3

Self-Stigma as a Mediator Between Public Stigma and Attitudes Toward Psychological Help Seeking

Note. N = 103. Model 1 = public stigma on attitudes toward psychological help seeking, R2 = .33, SE = 5.07. Model 2 = public stigma on self‐stigma, R2 = .34, SE = 6.11. Model 3 = self‐stigma on attitudes toward psychological help seeking, R2 = .53, SE = 4.26. Model 4 = public stigma and self‐stigma on attitudes toward psychological help seeking, R2 = .54, SE = 4.23.

**p < .01.

Figure 1

Mediation Model With Public Stigma, Self‐Stigma, and Attitudes Toward Psychological Help Seeking

Note. Model testing mediation of the relationship between public stigma and attitudes toward psychological help seeking by self‐stigma. Values in brackets represent the standardized relationship of public stigma to attitudes toward psychological help seeking after controlling for self‐stigma. Psychological help seeking = attitudes toward psychological help seeking.

Discussion

Despite the growing number of SSM/V on today's college campuses (National Center for Veterans Analysis and Statistics, 2016), there is a gap in knowledge concerning their willingness to engage in psychological help seeking. We sought to address this gap by examining stigma‐related predictors of psychological help seeking among SSM/V. Our findings demonstrated that public stigma and self‐stigma are associated with SSM/V help‐seeking attitudes. In addition, public stigma was found to be fully contingent on, or mediated by, self‐stigma in predicting attitudes toward psychological help seeking (see Figure 1). Our results add to the relevant past literature (Blais & Renshaw, 2014; Lally et al., 2013; Topkaya, 2014).

Our study uniquely contributes to this line of research by focusing on help seeking in the SSM/V collegiate population, a population with unmet mental health needs. We found that SSM/V endorsed mental health stigma, particularly self‐stigma, which negatively predicted attitudes toward psychological help seeking. SSM/V may resist psychological services to avoid potentially damaging labels, including being perceived as dangerous, unpredictable, psychotic, weak, and responsible for their own issues (Angermeyer & Dietrich, 2006). This avoidance may be driven by an internalized desire to avoid social discrimination, judgment, or rejection. In addition, SSM/V may avoid psychological help seeking because they fear vocational consequences, such as being perceived as unfit for duty (Bonar et al., 2015).

Besides societal expectations, the professional ethos of self‐reliance and strength entrenched in military culture may further perpetuate stereotypes about mental illness that deter service members from seeking help (Bonar et al., 2015; Held & Owens, 2013). Despite attempts by the military to destigmatize seeking services (Kessler et al., 2013), postdeployment training programs have historically dissuaded psychological help seeking, instead emphasizing individual willpower, self‐sufficiency, and resilience (Nash et al., 2009). Although self‐reliance is a highly adaptive trait in many military settings, this value is incongruent with assumptions pertaining to mental illness.

Gender is another variable to consider in future research. In the main regression analyses, gender was a significant control variable in every model, with women endorsing psychological help seeking more than men. Theoretically, traditional gender roles may affect the development of attitudes toward psychological help seeking. Societal expectations for young men about masculinity, such as toughness and self‐reliance, conflict with psychological help seeking (Helgeson, 2012). Mahalik et al. (2003) found that men who highly ascribe to traditionally masculine attributes were less likely to engage in psychological help seeking compared with men who did not highly ascribe to traditionally masculine attributes. Military culture may further reward those individuals who highly ascribe to traditional masculine traits. It is possible that a combination of societal expectations, military culture, and traditional gender roles affects the stigmas of psychological help seeking among SSM/V.

Taken together, the findings of this study are valuable to postsecondary institutions that regularly enroll military students, as well as university counseling centers that provide services for SSM/V. Prior to this study, it could be reasonably assumed that SSM/V face many of the same barriers to mental health help seeking as college students and military populations. However, our findings provide a scientific rationale for these claims and serve as a springboard for future research with SSM/V, given that mental health stigma remains a barrier to psychological treatment.

Implications for College Counseling

As mentioned previously, the results of our study suggest that self‐stigma is a barrier to psychological help seeking for SSM/V. These findings have important implications for colleges and universities, especially those institutions that regularly enroll SSM/V. First, it is recommended that university counseling centers continue to prioritize the needs of SSM/V, such as by using targeted outreach and developing strategic relationships across campus. For example, collaborative efforts between student services and Veterans Affairs offices could assist in this effort. Veterans advocates can be valuable conduits through which messages that destigmatize psychological help seeking can be effectively communicated to SSM/V (Borsari et al., 2017). Ultimately, this collaboration can forge an important bridge between university counseling centers, student services, and SSM/V by communicating a unified message that promotes healthy coping strategies; reduces mental health stigma; increases awareness of services; and, ultimately, facilitates access to services for SSM/V. On the basis of our findings, it is important that outreach messages focus on reducing mental health stigma and encouraging psychological help seeking among SSM/V. For example, these efforts could focus on misconceptions surrounding toughness, masculinity, and other gender roles, as well as chain‐of‐command considerations. These messages should be informed by the current literature, potentially developed in conjunction with the campus veterans advocate where possible, and disseminated in ways that are sensitive to the saliency of military identity (Borsari et al., 2017).

University counseling centers can also identify staff members who have an interest in working with this population and provide trainings to develop a strong working knowledge of the complex landscape of military culture. In collaboration with the Veterans Affairs office, specialized staff could conduct outreach for SSM/V to promote psychological help seeking and foster trust between SSM/V and counseling services. University counseling centers should also consider developing in‐house programming specific for SSM/V, including workshops and military student support groups, in addition to didactic training for staff members on best practices for providing therapeutic support to SSM/V.

In summary, SSM/V bring diversity of experience to higher education, in addition to valuable leadership training, problem‐solving skills, focus, and an orientation toward service learning. SSM/V may also deal with stigma regarding psychological help seeking as they simultaneously balance the identities of military and college student. University counseling centers can play an important role in reducing this stigma, encouraging a campus‐wide network of support, and helping SSM/V reach their goals in higher education.

Limitations

Our findings and clinical implications should be considered in light of the study's limitations. A cross‐sectional research design, like the one used in this study, provides limited information about any causal relationships between and among variables because there is limited control over the variables in the study. Furthermore, the results from this study are based on self‐report data. The nature of self‐report is susceptible to social desirability bias and bias in self‐selection into the study. Also, while convenience sampling was advantageous in terms of proximity and ease of data collection, this study is limited in terms of its generalizability because the participants were recruited from one institution in the Mid‐Atlantic region and, thus, represent only a subset of the population, with a 19.4% survey response rate.

Although the variables chosen for this study were based on existing literature, it is possible that other variables may also predict attitudes toward psychological help seeking among SSM/V. For example, despite attempts by the military to reduce stigma for psychological help seeking, there could be legitimate vocational consequences for some service members to receive mental health services. These consequences could include loss of status, loss of security clearances, informal negative evaluations from superiors, and informal pressures from peers (Coll et al., 2011). Future research should continue to parse out this experience for the unique SSM/V student population.

Directions for Future Research

Future studies could expand on our findings by using other research designs, such as qualitative or mixed methods. Qualitative methodology could deeply explore constructs such as the meanings behind self‐stigma, mental health, and what help seeking means to SSM/V. Additionally, it is recommended that future studies prompt participants to report their psychological and psychiatric treatment history to explore the effect of a person's previous treatment experiences on their perceptions and attitudes toward psychological help seeking. Finally, future research in this area may examine the development of mental health stigma in SSM/V, including the ways in which society, military culture, and gender affect the development of mental health stigma and attitudes toward psychological help seeking. For example, it will be important to investigate further the connection between gender, masculinity, military culture, and help seeking in this population. Future studies in this area could facilitate access to services for these students and could lead to improved psychological functioning.

Conclusion

The findings from this study demonstrate that SSM/V experience barriers when engaging in psychological help seeking. Specifically, our results show that self‐stigma fully mediates the relationship between public stigma and attitudes toward psychological help seeking among SSM/V. This study opens the door for greater discourse, both in science and practice, about the psychological needs of SSM/V and the stigma‐related barriers they experience. Ultimately, this discourse could facilitate access to psychological services for SSM/V to improve their well‐being and aid in the transition from military service to civilian life.