Grit is associated with decreased mental health help-seeking among student veterans

Abstract

Objective:

To examine the role of grit as a mediator of the relationship between student-veteran status and not seeking mental health help.

Participants:

A diverse and nationally representative sample of students (8,203 women, 4,934 men) from 18 U.S. colleges and universities included in the Understanding Student Distress and Academic Success study was used.

Method:

Data were analyzed using structural equation modeling to assess associations between military service, grit, and help seeking.

Results:

Military service was positively associated with both facets of grit: consistency of interests (CI) and perseverance of effort (PE). CI, but not PE, mediated the relationship between military service and never having sought mental health help.

Conclusions:

These results suggest that grit mediates the relationship between military service and not seeking mental health help. Interventions for student-veterans that emphasize the utility of mental health treatment may be useful to diminish the negative influence of grit on help-seeking.

After the establishment of the Post-911 GI Bill in 2008¹ there has been a substantial increase in the number of veterans seeking higher education.² In 2017, approximately 600,000 veterans utilized education benefits.³ In the same year, 11% of all U.S. college graduates were student veterans.⁴ It has been projected that there will be more than 5 million student-veterans enrolled in colleges and universities by the year 2020.⁵ Access to the Post-911 GI Bill—which covers in-state tuition and provides housing and material stipends—makes going to college an increasingly attractive and viable option for veterans. At the same time, given the growing number of degree-seeking veterans, it is important to understand the challenges they face when transitioning to a college environment.

Student-veterans encounter challenges in returning to higher education. Stresses, such as combat trauma, and injury, such as injury from concealed explosives, have left this population with both visible and hidden wounds.^6,7 Further, even among veterans not personally exposed to combat or other life-threatening situations, rates of PTSD are elevated due to other forms of exposure, such as having friends be critically injured or killed.⁸ Of those servicemembers who have experienced combat, estimates of lifetime prevalence of PTSD ranges between 12% and 30% with a past-year prevalence of 12% to 15%.⁹ In a comparative study of college students, student-veterans have been found to present with a higher prevalence than their nonveteran peers with depression, PTSD, and suicidal ideation.¹⁰ Further, military service members also report high rates of sexual harassment and assault (WILSON). Furthermore, in a survey of 628 student-veterans, 35% reported experiencing anxiety and 46% reported thoughts of suicide—of which 20% reported having a plan for attempting suicide.¹¹ These studies indicate high rates of mental health distress among U.S. student-veterans and demonstrate the need for mental health services among student veterans.

With a proliferation of evidence-based mental health treatment and the increasing availability of care, veterans have access to psychological services. Mental health care services are available via the Veteran’s Administration hospital system (VA), though many veterans prefer to not use the VA for varied reasons.^12–14 For student-veterans, college counseling centers also offer a low-cost option for mental health care that is not connected to their VA records.¹⁵ However, reluctance to seek mental health services remains a significant barrier to care for student veterans.^16,17 This reluctance, characterized as stigma toward help-seeking, is associated with less help-seeking behavior and thus the persistence of untreated mental health concerns.¹⁸ Reluctance to seek mental health help may result in persistence of symptoms that arose due to military experience (e.g., combat trauma) and may also produce reluctance to seek help for mental health issues that arose before military service (e.g., childhood abuse). U.S. veterans are more likely to perceive stigma than nonveterans¹⁰ and as such are more reluctant to seek mental health services even when they also perceive the need for such services.¹⁰ Despite this stigma, some studies have found that student veterans are more likely than non-veteran students to receive therapy,^10,19 with rates of help-seeking as high as 98%,²⁰ though other studies have found no difference.²¹ With higher rates of mental illness and a greater reluctance to seek care, these findings illustrate the importance of addressing barriers to help-seeking behavior within the veteran population.

Grit as a barrier to help-seeking

Grit is typically defined as a passion and perseverance with regard to long-term goals.²² However, grit may be associated with the reduced help-seeking tendencies observed among veterans. The grit construct includes two subfacets: consistency of interests (CI) refers to maintaining focus and working toward a consistent goal over time, and perseverance of effort (PE) refers to persistently working toward one’s goals and overcoming setbacks. The two facets of grit are related but not redundant, including among college and military samples. The correlation between the two subscales was r = .55 among a sample of US high school students,²³r = .48 among a sample of Canadian college students,²⁴ and r = .40 among a sample of US military college students.²⁵

Grit has traditionally been viewed as a desirable trait due to its positive associations with outcomes such as academic retention,²⁶ productivity,²⁷ and academic achievement.²⁸ These findings have led to widespread interest and investment in fostering grit in students, including an official recommendation by the U.S. Department of Education that grit be taught in schools.²⁹

However, grit has recently come under scrutiny due to its similarity to other motivational and personality constructs including self-discipline and conscientiousness, which may better account for many of grit’s purported benefits.^23,30–33 Studies also increasingly highlight the “dark side”³⁴ of grit and raise the possibility that being “gritty” may result in a form of stubbornness or resistance to change that can ultimately be detrimental. Grit may decrease cognitive flexibility³⁵ and lead individuals to persist in tasks for longer than is advantageous.³⁶ This excessive persistence may lead individuals to engage in detrimental behaviors, such as suffering large financial losses, for the sake of not changing course.³⁷ These findings may have implications for help-seeking behaviors, in that individuals may persist in increasingly maladaptive coping strategies and not “change course” toward help-seeking when needed.

If military veterans are unlikely to seek help and view help-seeking as a personal failure, these effects may be mediated by high levels of grit. Traditional military cultural values promote stoicism and putting the needs of the group or mission above personal needs.¹⁷ Indeed, military documents describe soldiers as “stubborn, bold figures” who “camouflage their emotions”.³⁸ Military personnel with these characteristics are also advantaged in training and selection for elite forces. Grit predicts retention at the United States Military Academy,³⁹ and soldiers high in grit are more likely to complete the Army Special Operations Forces selection course, after controlling for intelligence and physical fitness.⁴⁰ Additionally, the U.S. military has invested in large-scale programs such as Comprehensive Soldier Fitness, which seek to increase psychological strength by targeting highly related constructs such as resilience.⁴¹ In sum, it is likely that military personnel may be higher in grit, which may in turn predict help seeking behaviors.

The present study

The goal of the present study was to explore the role of grit as a mediator of never seeking psychological help among student veterans. We had the following hypotheses:

Hypothesis 1a: Military service will be associated positively with grit-CI.

Hypothesis 1b: Military service will be associated positively with grit-PE.

Hypothesis 2a: Grit-CI will be associated positively with never having sought mental health help.

Hypothesis 2b: There will be a significant indirect effect from military service to never having sought mental health help, via grit-CI.

Hypothesis 3a: Grit-PE will be associated positively with never having sought mental health help.

Hypothesis 3b: There will be a significant indirect effect from military service to never having sought mental health help, via grit-PE.

Method

Participants

The analytic sample consisted of 13,137 college students; 8,203 (62.4%) identified as women, 4,934 (37.6%) identified as men. Participant age ranged from 18 to 80 (M = 22.78, SD = 6.49). A total of 322 (2.5%) of the participants reported having served in the military. The majority of participants (7,768; 59.1%) identified as White, followed by Asian (1,566; 11.9%), Hispanic/Latina/o (1,198; 9.1%); multiracial (1,070; 8.1%), Black/African American (834; 6.3%), Middle Eastern (430; 3.3%), or another group 1.9%). A total of 6,627 (50.4%) of participants reported never receiving counseling; 6,510 (49.6%) reported having received counseling at some point.

Measures

Military service

Military service was assessed using the item, “Have you served in the military?” Responses were coded as 1 = No, 2 = Yes.

Grit

The Short Grit Scale (GRIT-S) was developed by Duckworth and Quinn⁴² as a more efficient way to measure grit than the original twelve-item scale. The GRIT-S has a two-factor structure (CI and PE) consisting of four items on each subscale. The eight items (sample item: “I am a hard worker”) are scored on a 5-point Likert scale (1 = Not at all like me, 5 = Very much like me). In prior research response to items on the subscales have yielded Cronbach’s alphas of .73 to .79 and .60 to .78, respectively.⁴²

Help-seeking behavior

Never having sought mental health help was measured using the item, “From which of the following have you ever received counseling or mental health services? (Select all that apply.)” Participants who selected the option, “I have never received counseling or mental health services” were coded as 1, and participants who selected any other option were coded as 0. Thus, higher scored reflect not receiving mental health counseling in order to model risk for not receiving counseling.

Procedure

Data were collected as part of the Understanding Student Distress and Academic Success study run by the National Research Consortium of Counseling Centers in Higher Education. The study included a large, nationally representative sample of U.S. undergraduate students across 18 institutions. Additional details about the larger data collection may be obtained from https://www.researchgate.net/project/Understanding-Student-Distress-and-Academic-Success. The larger survey included assessments of academic success, academic identity, and perfectionism. Stratified random sampling was used and selected students were contacted by the Consortium via email. This study was reviewed and approved by the Institutional Review Board (IRB) at The University of Texas at Austin. ].

Analysis plan

The student veteran sample was more likely to be male (75.8% male) than the non-veteran sample (33.6% male; χ² [1] = 205.59, p < .001). The veteran group was also older (M = 32.42, SD = 10.79) than the non-veteran group (M = 22.54, SD = 6.15; t (10787) = 25.15, p < .001. There was no difference between the veteran and non-veteran samples by race (χ² [9] = 13.88, p = 0.13). Skew and kurtosis for the Grit items were all within acceptable parameters (greatest skew = −0.983, greatest kurtosis = 0.651).

Data were analyzed using structural equation modeling in Mplus using maximum likelihood estimation to account for the complex survey design, to model the two grit subscales as parallel mediators in the same model, and to model grit subscales as latent variables.⁴³ Missing data were low, with less than 0.3% of data points in the set missing; missing data were modeled in Mplus using full information maximum likelihood estimation.⁴⁴ Grit subscales were modeled as latent variables using the respective items as indicators. Because gender and age are known to be associated with help-seeking, they were entered as control variables. Mediation was estimated using 5000 bootstrapped samples to estimate bias-corrected bootstrapped confidence intervals for the indirect effects.

Results

Correlations, means, and standard deviations are presented in Table 1. he hypothesized path model is displayed in Figure 1. Note that because the dependent variable (help-seeking) is dichotomous, paths to help-seeking are odds ratios and thus coefficients are significant if the value of the confidence interval does not overlap with 1.

Table 1.

Correlations and descriptive statistics.

	2	3	4	5	6	M	SD
1. Gender	−.03*	−.13*	.06*	.05*	−.12*	1.62	0.48
2. Age		.24*	.13*	.08*	−.11*	22.78	6.49
3. Military service			.05*	.05*	.00	0.03	0.16
4. Grit-CI				.44*	.10*	3.12	0.80
5. Grit-PE					.08*	3.77	0.65
6. Never received counseling						0.50	0.50

Note. CI = Consistency of interests. PE = Perseverance of effort.

^*p < .01.

Figure 1.

Structural equation model with path coefficients. R² for never receiving counseling was.06.

The hypothesized model Akaike information criterion (AIC) was 220324.700, and the Bayesian information criterion was 220594.29.368. Compared to the measurement model (i.e., with all covariances among variables estimated), the hypothesized model fit statistics were lower (AIC = 277927.944, BIC = 278506.487; AIC_diff = 10097.004, BIC_diff = 10421.119). Thus, the hypothesized model was superior to the measurement model.

Consistent with hypothesis 1a and 1b, military service was associated positively with both CI, B = 0.02, SE = 0.01, p < .05, and PE, B = 0.03, SE = 0.01, p < .01. Consistent with hypothesis 2a, Grit-CI was associated with never having sought mental health help, OR = 1.74, 96% confidence interval = 1.57, 1.94. Consistent with hypothesis 2b, the indirect effect from military service to never having sought mental health help via Grit-CI was significant, B = .009, SE = 0.002, bootstrapped 95% confidence interval = 0.005, 0.012, p < .001. Contrary to hypothesis 3a and 3b, Grit-PE was not associated with not seeking mental health help, OR = 0.93, 95% confidence interval = 0.77, 1.12, and the indirect effect from military service to never having sought mental health help via Grit-PE was not significant, B = .000, SE =.001, bootstrapped 95% CI = −0.002, 0.002, p = 0.889.

We conducted a post-hoc Monte Carlo analysis in Mplus to compute achieved power. Power was 1.00 for all relationships in the hypothesized model, with the exception of the relationship between Grit-PE and never going to counseling. For this association, achieved power was only 0.048, which is to be expected since the unique relationship between these two variables in the structural equation model was near-zero.

Discussion

The present study sought to elucidate the implications of grit on help-seeking behavior in student-veterans. Results from this analysis found that military service was positively associated with both subfacets of grit. This finding is in line with prior research that suggests that the military both selects for and instills grit in personnel.^38,39,41 In turn, CI was related to reduced mental health help-seeking, such that for every unit increase in CI the chances of having never sought mental health help rose by 1.7 times. CI mediated the relationship between student-veteran status and not seeking mental health help, suggesting that CI partially explains reluctance for help-seeking among some student-veterans. PE, however, was not uniquely related to reduced mental health help-seeking.

Regarding only CI being a significant predictor or mediator in the model, CI reflects maintaining a pattern of behavior, or “staying the course”.⁴² To the degree that help-seeking may represent a departure from behavioral norms and prior personal patterns of behavior in terms of management of one’s own challenges in life, especially for veterans, this personality characteristic may inhibit engaging in new forms of problem-solving such as mental health help-seeking. PE, on the other hand, reflects an individual’s inclination to exhibit both persistence and overcoming setbacks. This trait might allow for more cognitive flexibility to bring in other tools to help meet the goal, such as help-seeking in the face of mental health concerns. This finding is consistent with prior work that suggests that PE may be more linked to success resulting from sustained effort, such as academics.⁴⁵ In contrast, the usefulness of consistency of interests depends on personal goals being clearly defined⁴⁶ and that consistency of interests may actually promote closing oneself off to considering new options.⁴⁷

The findings from this study suggest critical avenues for outreach to and health messaging toward student veterans on campus. Consistency of interests was associated with lower help-seeking. Thus, messaging to promote use of campus mental health services by veterans might include messaging about how use of those services is consistent with goals for student-veterans, such as academic success and relief from mental health symptoms, so as to construct help-seeking as consistent with goals. Further, although stigma of mental health help-seeking is common in general,⁴⁸ veterans may be at particular risk for greater stigma about help-seeking.^49,50 One critical intervention may be the use of personal testimonials from student veterans who have sought help and visibility of veterans in student counseling center advertising materials, or by emphasizing messages such as that not using counseling center resources is like leaving a base without all protective equipment available. Such marketing techniques have been demonstrated to be powerful influences of potential consumers of services and in receptivity to and uptake of health services, and can be produced relatively inexpensively.^51,52 Video-based interventions have demonstrated promise in improving attitudes toward veterans⁵³ and as such might be useful in the context of training mental health providers to provide services to student-veterans. Identifying help-seeking as promoting current goals (e.g., reducing mental health distress, graduation) may help to reconceptualize grit as consistent with mental health help-seeking. Language in advertising that focuses on problem-solving and solution-focused therapy may also help to demystify therapy and decrease stereotypes about therapy that are based in popular media representations of counseling.⁵⁴

Student-veteran status was related to PE, but PE was not uniquely related to help-seeking in the structural equation model after accounting for the influence of CI. As a bivariate relationship did exist, and as PE has sometimes been contended to be a crucial part of grit,²² the PE-help-seeking relationship remains worthy of study. PE refers to diligence and continuation of effort in the face of setbacks. The absence of a unique relationship between PE and help-seeking in the model suggests that PE may not uniquely pose a risk to lack of help-seeking.

Several limitations must be considered when interpreting study results. First, the data used in this study are cross-sectional, so we are unable to infer a causal relationship between grit and help seeking. Second, military service was scored in our data set as a binary yes/no. As such, we are unable to assess the influences of factors such as the type of training, service length, deployment status, and exposure to direct combat, each of which may influence both grit levels and help seeking behaviors. Third, help-seeking was assessed using a single item, and our study does not differentiate between different types and degrees of help-seeking behavior. Fourth, our data were collected as part of a large, multi-site study of student health; it is possible that some students self-selected out of participation or that items on the larger study impacted responses to the items used in this study. Fifth, although the overall model was significant the total amount of variance explained for help-seeking was low. Overall variance explained for behavioral data may be impacted by many factors, resulting in a numerically low R² even in models with good predictor variables.⁵⁵ Future work may examine other factors that may ultimately explain more variance in help-seeking behaviors among veterans. Finally, our study did not measure help-seeking attitudes, which may strongly influence help-seeking behavior. Future research may also explore factors associated with mental health service use among student veterans, such as lack of resources and information related to mental health for student-veterans. For specific groups of student-veterans, such as racial/ethnic minorities, campus-based services may also not consistently provide culturally-affirming care. These intersections of identities and college campus resources may be more fully explored to develop interventions targeted toward specific underserved groups.