Attachment, Communication, and Relationship Functioning Among College Student Veterans and Nonveterans

In higher education settings, veterans are a distinct demographic group that is more likely to be older, male, and married compared with nonveteran students (Whiteman, Barry, Mroczek, & MacDermid Wadsworth, 2013; Widome, Nelson Laska, Gulden, Fu, & Lust, 2011). While balancing the demands of school and family, student veterans may face a number of other challenges, including military‐related psychological distress, suicidal ideation (Rudd, Goulding, & Bryan, 2011), and risk behaviors related to safety, tobacco use, and alcohol use (Widome et al., 2011). With one in four student veterans withdrawing from college (Rein, 2015), further exploration of factors influencing veteran success in higher education is warranted.

Evidence suggests that, like traditional college students (Hefner & Eisenberg, 2009), student veterans with higher levels of support from family and peers enjoy better mental health and academic outcomes (Romero, Riggs, & Ruggero, 2015; Whiteman et al., 2013). However, research shows that military deployment is associated with a decline in marital satisfaction (McLeland, Sutton, & Schumm, 2008), and postdeployment mental health concerns may exacerbate stress by negatively affecting communication and intimacy (Sayers, Farrow, Ross, & Oslin, 2009). As a biologically evolved strategy for managing stress (Kobak, Cassidy, Lyons‐Ruth, & Ziv, 2006), the attachment system is activated by separation from attachment figures and is particularly salient among military couples who have been separated by deployments (Vormbrock, 1993). Given the clinical relevance of attachment processes and the potential impact of relationship distress on student veterans’ success, additional research in these areas can inform college programming and treatment for student veterans. Our purpose in this article is to explore how veteran status and attachment style are related to relationship functioning in a college sample.

Attachment and Romantic Relationships

Adult attachment is most commonly manifested in romantic relationships and has numerous implications for couple functioning and communication. Adult romantic attachment comprises two dimensions of anxiety and avoidance that can be combined to form four categories of attachment (Brennan, Clark, & Shaver, 1998). According to Bartholomew and Horowitz (1991), secure individuals have positive models of self and others, which reflect low anxiety and avoidance, respectively. Secure adults show a balance between a healthy connection to others and self‐reliance, demonstrating trust and open communication in relationships and adaptive coping skills, such as help seeking (Lopez, Melendez, Sauer, Berger, & Wyssmann, 1998). In contrast, dismissing adults have a positive model of self (low anxiety) but a negative model of others (high avoidance) and consequently are highly independent, are distrustful, and tend to avoid closeness and intimacy. Preoccupied adults have a positive model of others (low avoidance) and negative model of self (high anxiety), so they question their self‐worth and can be overly dependent. Fearful adults have a negative model of both self and others and thus experience high levels of attachment anxiety and avoidance, which may contribute to contradictory approach–avoidance behaviors in personal relationships (Simpson & Rholes, 2002).

Adult attachment plays a significant role in important relationship domains among dating and married couples (Mikulincer & Shaver, 2007). Research consistently demonstrates that, compared with insecure attachment, secure attachment is related to better romantic relationship functioning and satisfaction (e.g., Banse, 2004). Secure individuals demonstrate effective conflict management strategies and communication, with frequent positive regard for partners (Mikulincer & Shaver, 2007). In contrast, insecure adults often have difficulty negotiating intimacy and closeness in romantic relationships. Individuals with high attachment avoidance prefer interpersonal distance and tend to disengage in both conflictual and supportive marital interactions (Barry & Lawrence, 2013), whereas individuals with high attachment anxiety seek excessive closeness in their relationships (Roberts & Pistole, 2009).

Relationship Communication and Interaction

Communication is strongly tied to romantic relationship satisfaction and is a critical component of couple health, particularly in terms of interaction patterns during conflict resolution (Driver, Tabares, Shapiro, Nahm, & Gottman, 2003; Eldridge & Christensen, 2002). Gottman's (1993) work suggests that satisfied couples display positive communication at a ratio of 5:1 when compared with negative communication, whereas unhappy couples make fewer positive than negative comments (at a ratio of 0.8:1). Moreover, both the absence of positive affect and the presence of negative affect predict divorce (Gottman & Levenson, 2000).

A common maladaptive cycle among unhappy couples is the demand–withdraw pattern, which often emerges when one partner desires change for a happier relationship (Gottman & Levenson, 2000). To induce change, the dissatisfied spouse may begin by asking or discussing with the partner but then escalates to criticizing, nagging, pressuring, or complaining, while the partner withdraws or avoids the interaction, usually with either defensiveness or passivity (Eldridge & Christensen, 2002). Demand–withdraw interactions are associated with high rates of hyperarousal, intrusive thoughts and feelings, and disruption of activities (Malis & Roloff, 2006). In community and college student samples, couples in which at least one partner has an insecure attachment style use more demand–withdraw communication, whereas couples with two secure partners demonstrate significantly more constructive communication (Domingue & Mollen, 2009).

Attachment, Communication, and Relationship Functioning of Veterans

Military couples often experience difficulty communicating when a spouse is deployed (Greene, Buckman, Dandeker, & Greenberg, 2010), and the stress of subsequent reintegration may contribute to or further exacerbate communication problems (Houston, Pfefferbaum, Sherman, Melson, & Brand, 2013). Military spouses note the need to get reacquainted, reconnect emotionally, and reestablish a sexual relationship with their partner following deployment (Baptist et al., 2011; Knobloch & Theiss, 2012). In some cases, couples struggle with increased conflict; difficulty understanding their partner's experience; and issues of trust, commitment, or infidelity (Baptist et al., 2011; Knobloch & Theiss, 2012).

A relatively small but growing literature has examined attachment processes among veterans, and higher rates of insecure attachment styles have been found among veterans relative to nonveteran college students or the general population (Dieperink, Leskela, Thuras, & Engdahl, 2001; Renaud, 2008). Although adaptive in a wartime environment, military personnel tend to suppress fears, which can create problems in the family (Riggs & Riggs, 2011). In particular, these coping strategies and other psychological avoidance defenses, such as denial and emotional numbing, are associated with avoidant attachment strategies (dismissing and fearful styles) and can affect the capacity for intimacy with romantic partners (Scaturo & Hayman, 1992). Theoretically, a service member with high attachment avoidance may appear unemotional and use distancing strategies in close relationships, which could trigger fears of abandonment and clinginess in a secure or anxiously attached spouse—a situation conducive to the development of a maladaptive demand–withdraw marital pattern (Riggs & Riggs, 2011). In one of the few studies examining attachment and relationship functioning among veterans, Cafferty, Davis, Medway, O'Hearn, and Chappell (1994) found support for the benefits of secure attachment, reporting that for both spouses, secure attachment was significantly associated with higher positive affect, greater marital satisfaction, and less marital conflict several months after deployment.

The Current Study

More research is needed to identify specific relational characteristics that may be fruitful targets for counseling intervention with veterans. The current study aims to examine attachment and relationship processes among college students, exploring differences between student veterans and nonveteran students that will inform treatment planning for student veterans. Although the conceptualization of adult attachment along two dimensions versus four categories has been debated (Fraley & Spieker, 2003; Sroufe, 2003), we chose to use both dimensional and categorical measures of attachment, the latter of which are better known among counselors, who often rely on an attachment framework to conceptualize and treat clients (Shi, Wampler, & Wampler, 2013, 2014).

We tested the following hypotheses: Compared with nonveteran students, student veterans are more likely to endorse dismissing attachment (Hypothesis 1), fearful attachment (Hypothesis 2), and relationship problems (Hypothesis 3). We also expected that secure attachment is related to better relationship functioning than insecure attachment (Hypothesis 4), and this association is moderated by veteran status. The final hypothesis predicted that, controlling for relevant demographic differences, the combination of veteran status, attachment dimensions, and communication patterns significantly contributes to overall relationship adjustment, with each step contributing to a significant change in variance accounted for by the model (Hypothesis 5).

Method

Participants

Participants were drawn from a larger project examining the functioning of college student veterans. Data were collected from college students enrolled in psychology courses at one large state university, as well as from U.S. military veterans enrolled at two large state universities and a small private university. The psychological and academic functioning of the student veterans in our sample are described in Campbell and Riggs (2015). For the current study, participants in a committed romantic relationship of 6 months or more (i.e., marriage, cohabitation, committed partnership) were eligible. To better match the veteran and nonveteran student groups, we analyzed only participants 20 to 40 years of age.

In the total group of 165 student veterans, all were 20 years or older, and 60% (n = 99) reported being in a committed relationship, with 48.5% (n = 80) married, 3.0% (n = 5) cohabitating, and 8.5% (n = 14) in a committed partnership. After ineligible cases and outliers were removed, the student veteran subsample (n = 82) consisted of 64 men and 18 women with a mean age of 29.42 years (SD = 4.25). Of this subsample, 75.6% (n = 62) reported being married, with an average relationship duration of 62 months. In terms of race/ethnicity, 73.2% (n = 60) identified as Caucasian, 8.5% (n = 7) as Hispanic/Latino/Mexican American, 7.3% (n = 6) as African American, and 3.7% (n = 3) as Asian/Pacific Islander; the remainder indicated multiracial heritage (4.9%, n = 4) or did not endorse a race/ethnicity (2.4%, n = 2). Almost 41.5% (n = 34) of participants reported serving in the U.S. Army, 25.6% (n = 21) in the Marine Corps, 19.5% (n = 16) in the Navy, and 11% (n = 9) in the Air Force, with 2.4% (n = 2) not reporting service branch.

In the comparison group of 430 nonveteran students, 60.9% (n = 262) were in the appropriate age range and 33.0% (n = 142) reported being in a committed relationship, with 3.0% (n = 13) married, 5.6% (n = 24) cohabitating, and 27.4% (n = 118) in a committed partnership. After ineligible cases and outliers were removed, the nonveteran student subsample (n = 85) consisted of 13 men and 72 women with a mean age of 22.51 years (SD = 3.52). Of this subsample, 11.8% (n = 10) were married, with an average relationship duration of 41 months, and 60% (n = 51) identified as Caucasian, 17.6% (n = 15) as Hispanic/Latino/Mexican American, 7.1% (n = 6) as African American, 5.9% (n = 5) as Asian/Pacific Islander, 3.5% (n = 3) as Native American, and 5.9% (n = 5) as having multiracial heritage.

Instrumentation

Background Information Questionnaire–Student Veteran version (Riggs & Campbell, 2012). Participants provided demographic information such as age, sex, race/ethnicity, education, occupation, and military background. This measure also assessed relationship status and duration, as well as psychiatric and family‐of‐origin history.

Revised Dyadic Adjustment Scale (RDAS). The RDAS (Busby, Christensen, Crane, & Larson, 1995) comprises 14 items from the original Dyadic Adjustment Scale (Spanier, 1976) measuring relationship consensus (e.g., agreement on important issues), satisfaction, and cohesion (e.g., common interests), which may be examined separately or combined to form an overall measure of relationship adjustment. Respondents indicate frequency or level of agreement on a 6‐point scale, with lower scores reflecting greater relationship distress. RDAS total adjustment scores below 48 are classified as distressed. Among clinical and nonclinical couples, the scale developers reported good construct and criterion validity, as well as good internal consistency (α = .81–.90) and strong split‐half reliability for the total scale (Guttman split‐half reliability coefficient = .94). In this study, Cronbach's alpha coefficients ranged from .70 to .85 for student veterans and .70 to .75 for nonveteran students.

Communication Patterns Questionnaire–Short Form (CPQ‐SF). The CPQ‐SF (Christensen & Heavey, 1990) is an 11‐item revision of the original Communication Patterns Questionnaire (Christensen, 1988), which assesses perceived positive and negative communication. The scale developers reported good validity and reliability (α = .70 or higher). We used the Demand–Withdraw and Positive Communication subscales identified by Futris, Campbell, Nielsen, and Burwell (2010) in a representative sample of married and remarried couples. Cronbach's alpha coefficients in the current study ranged from .71 to .80 for student veterans and .71 to .79 for nonveteran students.

Experiences in Close Relationships Scale (ECR). The ECR (Brennan et al., 1998) measures adult romantic attachment. Each item is rated on a 7‐point Likert‐type scale ranging from 1 = not at all like me to 7 = very much like me. Two 18‐item scales assess attachment‐related avoidance and anxiety, which can be conceptualized as attachment behaviors or strategies in close relationships. Crowell, Fraley, and Shaver (1999) reported good internal consistency (α = .90–.91); test–retest reliability; and high construct, predictive, and discriminant validity. In the current sample of student veterans and nonveterans, respectively, the Cronbach's alpha coefficients were .95 and .86 for the Avoidance scale and .91 and .93 for the Anxiety scale. On the basis of cluster analysis, the scale developers used the standardized coefficients for the two scales to identify four distinct attachment styles (secure, preoccupied, dismissing, and fearful). This classification scheme was used to form the four attachment categories in the current analysis.

Procedure

Data for this study were collected as part of a larger, password‐protected online study approved by the university institutional review board. We emailed invitations to students receiving veteran benefits, posted announcements on electronic mailing lists and social networking sites, and posted fliers at university facilities. Interested student veterans contacted us and, if eligible, received an access code for the survey. After providing informed consent, participants completed a battery of standard questionnaires with programmed skip logic for nonapplicable questions (e.g., students not in committed relationships did not complete the RDAS).

For the comparison group, nonveteran students were recruited through the first author's departmental subject pool and received extra credit for participation. After registering, participants received an access code to complete the survey, which contained an additional skip logic for nonveteran status. All other questions and procedures were identical to those for student veterans.

Data were downloaded and transferred to SPSS Statistics (Version 20) for cleaning and analysis. Cases missing more than two measures, age, or relationship status information as well as extreme outliers were removed. No missing data were noted for RDAS satisfaction and cohesion. With less than 2% missing data for other variables, estimation maximization procedures were conducted separately for the two subsamples to replace missing data. For both subsamples, skewness and kurtosis for all key variables were within acceptable limits.

Results

Preliminary Analyses

Analyses of variance (ANOVAs) indicated no group differences in key variables based on ethnicity, but significant differences on some variables emerged for sex and marital status. In general, men and married individuals reported less optimal relationships. Compared with nonveteran students, student veterans tended to be older, F(1, 166) = 131.91, p < .001; male, χ2(1, N = 167) = 66.15, p < .001; married, χ2(1, N = 168) = 69.94, p < .001; and in longer committed relationships, F(1, 166) = 14.10, p < .001. On the basis of these findings, we controlled for age, sex, marital status, and duration of relationship in subsequent analyses.

Hypothesis Testing

The distribution of attachment style differed significantly for student veterans and nonveterans, χ2(3, N = 162) = 14.49, p = .002, with ϕ = .50 reflecting a large effect size (power = .99). Among student veterans, 52% were secure, 20% were fearful, 9% were preoccupied, and 19% were dismissing, whereas 46% of nonveteran students were secure, 20% were fearful, 27% were preoccupied, and 6% were dismissing. As predicted by Hypothesis 1, student veterans more often endorsed a dismissing attachment style, but findings did not support Hypothesis 2 of group differences in fearful attachment style. Although not hypothesized, it was also noted that student veterans were less likely to endorse a preoccupied attachment style relative to nonveteran students.

Chi‐square results supported Hypothesis 3, with student veterans more likely than their nonveteran counterparts to be RDAS classified as relationally distressed, χ2(1, N = 166) = 44.20, p < .001, with ϕ = .49 approaching a large effect size (power = .99). Controlling for age, sex, marital status, and relationship duration, we conducted a 2 (veteran status) × 4 (attachment style) multivariate analysis of covariance to examine effects on five romantic relationship qualities (see Table 1). Multivariate results partially supported Hypothesis 4, with significant main effects for veteran status, Wilks's λ = .54, F(5, 143) = 23.98, p < .001, partial η2 = .46, and attachment style, Wilks's λ = .61, F(15, 395) = 5.18, p < .001, partial η2 = .15. The predicted interaction between veteran status and attachment was nonsignificant, Wilks's λ = .87, F(15, 395) = 1.32, p = .18, partial η2 = .04. Adequate power (>.80) was present for both main effects, whereas the interaction effect was slightly underpowered (.76). Partial η2 results indicated that the effect for veteran status accounted for 47% of the group differences plus associated error variance, whereas the effect for attachment style accounted for 15% of the group differences plus associated error variance.

Table 1

Multivariate Analysis of Covariance (MANCOVA) Examining Couple Functioning in Relation to Veteran Status, Attachment Style, and Their Interaction

Note. Sample sizes varied for overall sample and subsamples because of missing data. For attachment styles, means with different subscripts differ significantly at p < .05 or greater. Nonvet = nonveteran; Preocc = preoccupied; Dismiss = dismissive; VS × A = Veteran Status × Attachment interaction; D‐W = demand‐withdraw; PC = positive communication; RDAS = Revised Dyadic Adjustment Scale; Con = consensus; Satis = satisfaction; Coh = cohesion.

^†p < .10. *p < .05. **p < .01. ***p < .001.

Follow‐up univariate ANOVAs provided additional support and refinement for Hypotheses 3 and 4. Per Hypothesis 3, student veterans reported significantly lower levels of consensus, cohesion, and satisfaction in their romantic relationships than did nonveteran students. A nonsignificant trend (p = .06) emerged for positive communication, but there was no significant difference in demand–withdraw communication between student veterans and nonveterans. Secure participants reported higher positive communication and satisfaction than all insecure participants, supporting Hypothesis 4. For other relationship processes, findings varied by insecure category. Specifically, secure participants reported less demand–withdraw communication but greater consensus than fearful and dismissing (but not preoccupied) participants and greater cohesion than dismissing (but not fearful or preoccupied) participants. Despite a nonsignificant multivariate interaction, univariate findings partially supported predicted moderation, indicating a significant interaction effect for consensus. Specifically, secure and preoccupied student veterans reported significantly better consensus than fearful and dismissing student veterans, whereas differences for nonveteran students were minimal.

The hierarchical regression model testing Hypothesis 5 in the whole sample was significant, F(9, 156) = 58.46, p < .001, accounting for 76% of the variance (see Table 2). To predict dyadic adjustment, Step 1 included veteran status and other demographics (i.e., sex, age, marital status, and duration of relationship), Step 2 added attachment anxiety and attachment avoidance, and Step 3 added demand–withdraw communication and positive communication. Supporting Hypothesis 5, each step significantly increased the variance accounted for, and all four hypothesized predictors contributed uniquely to total dyadic adjustment. However, attachment anxiety was unexpectedly associated with better adjustment.

Table 2

Multiple Regression Models for Full Sample, Student Veterans, and Nonveteran Students

*p < .05.**p < .01.***p < .001.

Similar to the whole sample, the regression for student veterans was significant, F(8, 73) = 10.76, p < .001, adjusted R2 = .49, with a significant F change in each step and all four independent variables uniquely predictive of dyadic adjustment. Attachment anxiety was again associated with better adjustment. The regression for nonveteran students was also significant, F(8, 75) = 9.34, p < .001, adjusted R2 = .45. Nonsignificance in the first two steps was followed by a 43% increase in the amount of explained variance in the final step, and positive communication was the only significant independent predictor of dyadic adjustment. All three regressions demonstrated statistical power greater than .99.

Discussion

The current findings replicated previous research and partially supported the hypotheses in this study, extending recent evidence of elevated relationship distress among the general veteran population (Sayers et al., 2009) to veterans attending college. Compared with nonveteran students in our sample, student veterans reported greater difficulties in their romantic relationships, which were characterized by more disagreements, less shared time, and lower satisfaction in the relationship. In contrast to findings with the general and combat/posttraumatic stress disorder veteran populations (Currier, Holland, & Allen, 2012; Renaud, 2008), the distribution of secure (52%) and insecure (48%) attachment in the subsample of student veterans was fairly equal and comparable with the distribution in the nonveteran subsample. However, student veterans were less likely to be preoccupied and more likely to be dismissing (but not fearful as predicted) than nonveteran students, which is in line with expectations that service members learn or strengthen their ability to minimize emotions and distance themselves from close relationships while deployed to war zones (Basham, 2008). These differences support theoretical assertions that military life and culture, particularly lengthy separations, contribute to higher levels of relationship distress in military couples as opposed to civilian couples (Riggs & Cusimano, 2014).

Consistent with previous research (Mikulincer & Shaver, 2007), secure attachment in the overall sample was related to more optimal relationship processes than insecure attachment. Secure participants were more satisfied and perceived more positive communication compared with insecure participants. Of special note, individuals high in attachment avoidance (dismissing or fearful) more commonly reported lower cohesion and more demand–withdraw communication. The latter communication style occurs when partners differ in their desire for closeness, with one partner wanting greater autonomy and the other wanting more intimacy (Christensen, 1988). Couples with a veteran who is high in attachment avoidance (dismissing or fearful) and a partner who is high in attachment anxiety (preoccupied or fearful) may therefore be at greater risk for developing the demand–withdraw interaction pattern (Riggs & Riggs, 2011). Recent research suggests that mismatched attachment styles reflect discrepancies in desired intimacy and this mismatch is associated with demand–withdraw interactions (Millwood & Waltz, 2015).

Additional findings highlighted the differences between veteran and nonveteran participants. One interaction suggested that, among student veterans but not their civilian peers, dismissing and fearful attachment styles are risk factors for lower dyadic consensus, whereas secure and preoccupied attachment styles served as protective factors against major disagreements in the relationship. Further results indicated that attachment and communication processes differentially predicted overall dyadic adjustment for student veterans and nonveterans. As expected based on previous research (Mikulincer & Shaver, 2007), for the overall sample and for the student veteran subsample, poor dyadic adjustment was significantly related to demand–withdraw interactions, low positive communication, and attachment avoidance. Contrary to predictions and past research (Treboux, Crowell, & Waters, 2004), however, higher attachment anxiety was associated with better dyadic adjustment for the overall sample and student veterans, whereas neither attachment strategy significantly contributed to dyadic adjustment for nonveteran students. Roberts and Pistole (2009) speculated that the separation–reunion cycle of long‐distance relationships would activate the attachment system and elicit concerns regarding partner availability, resulting in greater bids for proximity and closeness. Thus, among previously deployed student veterans, attachment anxiety (or preoccupied attachment) may counteract the inclination to withdraw and promote efforts to maintain agreement and intimacy with partners.

The role of attachment and communication processes in relationship functioning differed for student veterans and nonveteran students. Positive communication was the only significant contributor to dyadic adjustment for nonveteran students. Young civilian college students may have experienced fewer negative life events (e.g., deployment, trauma, loss) and currently face less stress (e.g., no family responsibilities) than student veterans, so the attachment system may play a less prominent role in their romantic relationships. Conversely, the effect of prior separations and present‐day contextual stressors on veteran couples’ interactions may increase the salience of attachment and communication processes for student veterans. The lack of significant associations between nonveteran students’ attachment styles and their relational processes contradicts previous research findings with community and college samples (Mikulincer & Shaver, 2007). Although we controlled for relationship length, it is possible that younger nonveteran students had not yet developed attachment bonds, which may take up to 2 years to develop (Fraley & Davis, 1997), making attachment strategies less relevant to these couples.

Implications for College Counseling

This study's findings suggest that therapeutic attention to attachment style may be fruitful when counselors are working with student veterans. For example, distrust of others and avoidance of emotional connections can prevent veterans with a dismissing style from seeking help and developing meaningful personal relationships. Counselor training in military culture can enhance therapeutic rapport and also overcome distrust and the mental health stigma common in the military. Initially, a professional, directive, and cognitive stance in counseling student veterans may be useful while counselors continue to build a collaborative alliance that allows greater exploration of emotion at later stages of treatment. 

Results also highlight the importance of relationship functioning. Student veterans with a dismissing attachment style in our sample showed the most troubling relational profiles of the four attachment groups. Although less likely to be preoccupied, student veterans’ attachment anxiety appeared to contribute to better relationship adjustment, which directly contradicts much of the literature. Addressing attachment behaviors may help student veterans understand and negotiate closeness distance in relationships. Although accessibility varies widely, many college counseling centers are now offering couples therapy (Reetz, Krylowicz, & Mistler, 2014). Identifying and revising demand–withdraw patterns and highlighting the importance of mutual engagement in activities and positive communication could increase satisfaction and emotional connection in the couple relationship. Emotionally focused couples therapy (Johnson, 1996, 2002) may be particularly useful in addressing attachment difficulties and has been effectively applied to couples with trauma survivors, including combat veterans.

Limitations and Conclusion

The cross‐sectional design of this study prevents causal conclusions, and the use of self‐report instruments entails the possibility of biased reporting and common method variance. Future research directions could include expanding measurement to interviews or observations and incorporating the perspectives of romantic partners. This study's findings are relevant to college students and may not generalize to nonstudents. Further research is needed to replicate the current findings and explore other possible covariates.

The study's findings underscore the need to recognize student veterans as a distinct, nontraditional subpopulation in today's college campuses. In addition to basic demographic differences, student veterans’ relational strategies and experiences appear to distinguish them from traditional students. Our results suggest that attachment strategies are highly relevant to the relationship functioning of student veterans, who are more likely than civilian students to be burdened by family responsibilities and relationship problems that could affect their academic performance. College counseling centers are in a pivotal position to help student veterans adjust to campus and other civilian settings. Training in military culture can promote greater multicultural competence and understanding for veterans’ experiences and contextual circumstances. In addition, couples therapy services can be made available to address the relationship concerns of student veterans and increase their chances of academic success.