The Deployment Life Study

Short abstract

In 2009, RAND launched the Deployment Life Study to study military family readiness. This article presents analyses on marital relationships, family environment, psychological and behavioral health, child well-being, and military integration.

Abstract

In 2009, RAND launched the Deployment Life Study, a longitudinal study of military families across a deployment cycle in order to assess family readiness. Family readiness refers to the state of being prepared to effectively navigate the challenges of daily living experienced in the unique context of military service. The study surveyed families at frequent intervals throughout a complete deployment cycle---before a service member deploys (sometimes months before), during the actual deployment, and after the service member returns (possibly a year or more after she or he redeployed). It assessed a number of outcomes over time, including:

• the quality of marital and parental relationships

• the psychological, behavioral, and physical health of family members

• child and teen well-being (e.g., emotional, behavioral, social, and academic)

• military integration (e.g., attitudes toward military service, retention intentions).

This culminating paper briefly reviews the study design and data collection procedures, presents results from analyses of the longitudinal data collected from some 2,700 military families, and offers recommendations for programs and future research related to military families. The research was jointly sponsored by the Office of the Surgeon General, U.S. Army, and by the Defense Centers of Excellence for Psychological Health and Traumatic Brain Injury.

Armed conflict between nations and groups usually requires those who fight to leave behind their families and loved ones. From the Greeks mounting their siege of Troy to the “Greatest Generation” preparing to enter World War II, the three phases of the deployment cycle---preparation, separation, and reintegration---have been a regular part of military life throughout the ages. Although many service members anticipate deployments, eager for the opportunity to defend their country and utilize their training, few look forward to time spent separated from spouses and children. Indeed, the separations required by deployments have long been described as one of the most stressful aspects of being a military family.

How are military families affected by the challenges of deployment? To what extent are service members and their loved ones able to maintain their individual health and interpersonal relationships in the face of lengthy separations? To what extent does the well-being of service members and their families change across the phases of the deployment cycle? The Deployment Life Study was designed to address these questions. The broad goal of the study was to gather data to evaluate the effects of deployment on service members, spouses, and their children. In particular, this study sought to identify which families are best able to withstand the strains of deployment and the kinds of coping strategies that characterize these families, and which families are most vulnerable to the negative consequences of deployment so that those families might be targeted for extra support.

The Deployment Life Study examined these issues by interviewing married service members, their spouses, and (when available) one child, multiple times across a three-year period, and across all phases of a deployment cycle. At baseline, the study recruited 2,724 Army, Air Force, Navy, and Marine Corps families. The sampling frame was restricted to cohabiting, married service members who were deemed eligible for deployment (by their respective services) within six to 12 months of the baseline survey. In each family, a service member, his or her spouse, and a child between the ages of 11 and 17 (if there was one in the household) each provided information independently. Interviewers obtained consent from each participating family member and conducted the baseline interviews by phone. Every four months thereafter, respondents logged into the study website and completed a follow-up survey online. Data collection included a total of nine waves spanning the entire deployment cycle, including periods of predeployment, deployment, and postdeployment. Each interview included instruments measuring a wide range of variables relevant to understanding how the social, economic, and psychological well-being of military families change across the deployment cycle. As a consequence of this breadth, the Deployment Life Study data set can serve as a resource for researchers from multiple disciplines and, more importantly, provide robust data with which to assess the association between deployment and military family health and well-being.

The data accumulated through these efforts are suitable for addressing a wide range of questions about the health and well-being of military families, but the following questions were priorities that guided the design of the study:

• Controlling for initial conditions and functioning, how is deployment related to military family outcomes? This study evaluated the association between deployment and a range of outcomes, including those relevant to the functioning of the entire family unit and outcomes relevant to understanding the well-being of each individual family member.

• Across the phases of the deployment cycle---preparation, separation, and reintegration---how do the outcomes of military families change or remain stable over time? This study is the first of which we are aware to describe changes before, during, and after deployment on a wide range of outcomes for multiple family members at once.

• Across families that experience deployment, what accounts for variability in families' postdeployment outcomes? Of the factors accounting for such variability, which are the most important? The implications of deployment are likely to vary significantly across families, and even across individuals within families. The Deployment Life Study assessed and analyzed the individual and family traits, resources, and circumstances that account for which families thrive after a deployment and which families suffer.

As described in this study, analyses focused on identifying the characteristics of more (or less) successful families over time, where success is defined in terms of a range of important outcomes that were measured at every assessment, including:

• quality of marital and parental relationships

• psychological, behavioral, and physical health of family members

• child and teen well-being (e.g., emotional, behavioral, social, and academic)

• military integration (e.g., attitudes toward military service, retention intentions).

Strengths of the Deployment Life Study

The Deployment Life Study has several unique characteristics that allow it to address issues that no other study to date can address as effectively.

Anchoring on Deployment

The primary rationale for conducting longitudinal research on military families is that longitudinal designs allow for estimates of changes in family outcomes after a deployment, controlling for preexisting differences between families prior to deployment. The greatest benefits of this design, therefore, arise from the power to observe families before and after a deployment cycle, and compare them with matched families that did not experience a deployment during the same period. The Deployment Life Study was designed to take advantage of the power of such analyses. Whereas the other ongoing longitudinal studies sampled from the general military population (some of whom may already be deployed and others of whom may never deploy over the course of a study), the Deployment Life Study sampled exclusively from the population of married service members eligible for deployment within the next six to 12 months. Moreover, to be eligible for the study, the service member could not be deployed at baseline. Thus, baseline assessment of the Deployment Life sample was a true baseline for a sample highly likely to experience a deployment over the course of the study.

Repeated Assessments over Three Years

Military families may cope in different ways across different stages of the deployment cycle (i.e., functioning at predeployment may not be the same as functioning during deployment or postdeployment). Moreover, the consequences of deployments may evolve or fade as families adjust and restore their predeployment equilibrium. Even longitudinal studies risk missing or mischaracterizing these effects if the interval between assessments is too long or the duration of the entire study is too short. The Deployment Life Study was designed to produce higher-resolution descriptions of the deployment cycle than has been available in other studies by assessing families every four months over three years (nine assessments total). Because all families were recruited prior to a deployment, not only did this design allow for assessments of family functioning at each stage of the deployment cycle; in most cases, it allowed for multiple assessments at each stage, including multiple assessments shortly after a deployment. Thus, the Deployment Life Study is able to describe how families change within each stage of the cycle, how long changes associated with deployment last, and what characteristics of families prior to deployment predict a family's postdeployment return to equilibrium.

Data from Multiple Family Members

The vast majority of existing research on military families assesses family functioning exclusively through the reports of a single family member, usually the service member or the spouse. There are several reasons that reliance on the reports of a single family member offer only a limited window into family functioning across a deployment cycle. First, service members and spouses may use different resources to cope with deployment. Second, coping strategies that are effective for one member of a family may not be effective for another. Third, to the extent that mental health is an outcome of major importance and the rates of mental health problems in this population are relatively high, the perceptions of any single family member may be distorted in ways that are impossible to estimate. Fourth, to the extent that one family member is asked to describe outcomes for another family member (e.g., parents reporting on their children, service members reporting on their spouses), associations among responses can be inflated by shared method variance. The Deployment Life Study avoided these problems by obtaining responses from service members, their spouses, and one child between the ages of 11 and 17, if available.

The Historical Context of the Deployment Life Study---A Caveat

Interpretations of the results reported here must keep in mind the specific period in which this sample was assembled and the data were gathered. Support for this study was motivated by the severe increase in the pace of deployments for service members in the middle to later part of the 2000s. By the time this study was commissioned and data collection began, however, the period of peak deployments for the U.S. military had passed. The surge in Iraq, for example, took place in 2007--2008 and was followed by a gradual reduction in troop strength through the formal end of operations there in December 2011. Deployments to Afghanistan peaked in 2010, and have been declining ever since. Initial recruitment of the initial sample for the Deployment Life Study began in March 2011 and ended in August 2012 for the Army, Air Force, and Marine Corps sample, and began in November 2012 and ended in February 2013 for the Navy sample. The subsequent three years of longitudinal data collection ended for all samples by the summer of 2015. Thus, although the Deployment Life Study took place following a period of frequent and lengthy deployments for the U.S. military, those deployments were becoming less frequent during the data collection phase of the study.

This historical context has several important implications for understanding the Deployment Life Study data. First, by the time recruitment for the Deployment Life Study began, most married service members eligible for the study had already been deployed at least once before to support the current conflicts. As such, the consequences of deployment may have already been felt for most of the population from which the study sampled. This means that the military families vulnerable to experiencing the worst consequences from deployments may have left the population (either through separation from the military or through divorce) before the study began. By studying couples who remained married and serving in the military during this period, the Deployment Life Study was able to examine a sample of military families that had already endured and survived the worst of the recent operations. As a consequence, although the results reported here can be generalized to the current married population of the military (i.e., a population that has survived prior, and sometimes multiple, deployments), these results may not generalize to populations that have yet to be deployed and may underestimate the negative effects of deployment for the population of first-time deployers.

Second, during the period of the Deployment Life Study, deployments were, on average, shorter across all services and arguably less dangerous (because the combat zones were less volatile) than they had been earlier in the decade. During the period of data collection, the number and rates of injury and death among American troops were much lower than in prior years, and certainly significantly less than during the peak of combat operations (Fischer, 2015). Thus, the effects of deployment reported here may not reflect the effects experienced during the period of greatest demand on the military in the early to mid-2000s.

Summary of Findings

Marital Relationships

We examined marital satisfaction, positive and negative affect, and psychological and physical aggression as reported by service members and spouses. Key findings include:

• Across the entire deployment cycle, couples, on average, become significantly less satisfied with their marriages and engage in less psychological and physical aggression than they reported prior to the deployment.

• These changes in marital outcomes across the deployment cycle do not significantly differ from the changes experienced during the same period by matched couples that did not deploy.

• For spouses, more frequent communication with the service member during deployment predicts greater marital satisfaction postdeployment, controlling for baseline characteristics of the couple.

• Service members' exposure to physical (e.g., being injured) and psychological (e.g., seeing injured noncombatants) trauma during deployment predicts spouses' reports of higher levels of psychological and physical aggression postdeployment, while service members' exposure to combat trauma (e.g., engaging in hand-to-hand combat or experiencing explosions) during deployment predicts lower levels of psychological aggression from both the service member and the spouse postdeployment, as reported by the spouse.

• For spouses, separating from the military during the postdeployment period is associated with lower marital satisfaction, and lower positive affect and higher negative affect after communicating with the service member, during the postdeployment period.

Family Relationships

We examined the family environment, satisfaction with parenting, and financial distress as reported by service members and spouses. Key findings include:

• Family outcomes change over the deployment cycle, but these changes occur primarily during the deployment itself and are not always the same for all family members.

• Service members report better family environments during deployments than before or after deployments, possibly because they are removed from the day-to-day challenges associated with family life. Spouses, in contrast, report no significant changes in family environment across the deployment cycle.

• Service members' and spouses' reports of parenting satisfaction both change over a deployment cycle, but service members report higher parenting satisfaction during the deployment while spouses report declines in parenting satisfaction over the entire deployment cycle.

• Financial distress, as felt by both service members and spouses, declined during deployment.

• Service members who report engaging in more predeployment preparation activities report higher satisfaction with parenting postdeployment. Similarly, spouses who report more preparation activities and greater satisfaction with the frequency of communication with the service member during deployment reported higher parenting satisfaction postdeployment.

• The negative psychological consequences of deployment, including psychological trauma and stress reported by the service member, are associated with increased risk for negative family outcomes for both service members and spouses postdeployment.

Psychological and Behavioral Health

We examined several psychological and behavioral health outcomes, including posttraumatic stress disorder (PTSD), depression, binge drinking, and the perceived need for counseling or therapy as reported by service members and spouses. Key findings include:

• We found no overall significant effect of deployment on persistent psychological or behavioral health outcomes for service members or spouses. This is consistent with the fact that we studied an experienced military population, serving during a period of the conflict with comparatively low levels of reported deployment trauma.

• However, the deployed service members who experience deployment trauma during the study did show a persistent increase in depression, PTSD, and anxiety symptoms relative to their predeployment levels. When the service member experienced physical trauma (i.e., injury) during the study deployment, their spouses also showed persistent increases in those symptoms, as well as in binge drinking.

• Mean levels of psychological symptoms showed substantial variation across the deployment cycle. During the deployment period, service members showed heightened depressive symptoms; spouses showed elevated depression, PTSD and anxiety symptoms. In contrast, service members showed significantly less binge drinking during deployments, perhaps because of limited access to alcohol.

• Those deployed members who subsequently separated or retired from the military showed increased levels of psychological symptoms. It may be that these symptoms facilitated the separation (for example, the service member no longer met retention standards), or that separation caused the increase in these symptoms (for example, as an individual experiences a loss of purpose, social support, or income). Regardless of the mechanism, this finding suggests that the period just after separation is one of increased risk and that addressing psychological problems in the critical window around the time of separation may be important for avoiding the longer-term difficulties.

Child and Teen Well-Being

We examined the emotional, behavioral, social, and academic functioning of children (as reported by spouses) and teens, across some 40 outcomes as reported by spouses (for children) and teens. Key findings include:

• Based on trajectory models of outcomes over the deployment cycle, most outcomes did not change over the course of a deployment cycle. Those that did change might reflect maturational changes among children and teens, rather than changes related to deployment. The passage of time, which underlies the trajectory models, also captures the natural process of aging through different developmental periods, which may also influence how children and teens adapt to a parental deployment. There were a few significant exceptions, however:

  • Three child outcomes (reported by spouses) showed significant changes over the deployment cycle: total difficulties (and more specifically, emotional problems), as well as depression screener scores. Spouses reported elevated symptoms in their children during deployment.

  • Two self-reported teen outcomes showed significant changes across the deployment cycle. Teen reports of drug use were very low before and during deployment (hovering around “never”), but increased afterward (to just slightly more than “never” on average). In addition, teens reported higher-quality relationships with their parents who deployed before and during deployment, with lower quality upon return.

• We generally found no significant effect of deployment on child and teen outcomes, which is consistent with the fact that we studied a relatively experienced population, serving during a period of the conflict with comparatively low levels of reported deployment trauma. However, there were some notable exceptions.

  • Spouses in families that experienced a study deployment reported elevated child difficulties at the end of the study (specifically, emotional conduct and peer problems), as well as a higher need for child mental health services, as compared with spouses in matched families that did not experience a study deployment. Interestingly, this was only found for spouse concerns about their children younger than age 11 following a study deployment; there was no such evidence of an effect for spouse concerns on teens or in teen self-reports.

• We also examined three sets of risk and resilience factors within families that experienced a study deployment.

  • The first set included parent deployment factors (e.g., deployment trauma, length of deployment, and separation from the military). Spouses reported that, the longer the study deployment, the more emotional problems and depressive symptoms children had. For teens, these factors were not related to spouse report of teen outcomes or for most of the teen self-reported outcomes. However, there were two exceptions: parental experience of trauma during the study deployment was associated with teen reports of their functioning and family relationships. The pattern of findings indicates that combat traumas were related to poorer functioning and relationships, whereas psychological traumas (in the absence of injury or combat trauma) were related to better functioning and relationships.

The second set of factors examined several measures of child or teen adjustment during the study deployment. When the same informant reported on adjustment and outcomes, there were many significant associations and many trends. That is, when teens self-reported on both adjustment during deployment and postdeployment outcomes, many significant associations were observed. The same was true when parents reported on both teen adjustment during deployment and teen outcomes during postdeployment. However, when informants were mixed---for example, when we examined parent report of teen adjustment during deployment and teen self-report on postdeployment outcomes, these associations were less robust. Thus, perceptions of problems during deployment are highly related to perceptions of problems postdeployment, especially when these perceptions are reported from the same informant. Finally, we examined three types of social support for teens. Socialization with other military children during deployment was strongly protective and associated with more positive outcomes at the end of the study.

Military Integration

Finally, we examined the military satisfaction, military commitment, retention intentions, and teen military career aspirations as reported by service members, spouses, and teens in active component, nondivorced families. Key findings include:

• Across the phases of the deployment cycle, measures of military integration varied more for service members than for spouses and teens. In fact, teen trajectories revealed no evidence of change over time. When there were changes in aspects of military integration during deployment, family members appeared to adapt well and return to predeployment levels of the outcome in the postdeployment period.

• Any communication with other military families during deployment was associated with higher service member retention intentions and greater spouse and teen military commitment postdeployment.

• Trauma experienced during the study deployment, especially physical and psychological trauma, was associated with decreased military satisfaction among service members postdeployment, as well as decreased retention intentions by service members themselves and by teens of service members postdeployment.

• Experiences of nondeployed families may also be stressful, which may minimize the gap between deployed and nondeployed families. For example, among nondeployed families, relocation was related to decreased military satisfaction, commitment, and retention intentions among service members.

Common Themes

Although each of the analyses described focused on a separate set of outcomes, several common themes emerged.

First, and most notably, the most-significant changes experienced by military families across the deployment cycle occur during the deployment itself. With respect to many of the outcomes examined in this study, the participants' status during the study deployment was significantly different from their status reported before and after the deployment. The dominant pattern over the deployment cycle was one of change or adaptation during the deployment period followed by a return to near predeployment levels of functioning.

Second, significant variation exists across family members with respect to how they experience deployment. For example, while service members rated the family environment higher and reported higher parenting satisfaction during deployment, spouses showed little change in these outcomes across the deployment cycle. It is plausible that service members are spared from day-to-day difficulties that families experience during their absences, although we have no direct evidence from our study to support this hypothesis. Given the physical absence of a service member during deployment, some of the differences we observed across family members (e.g., declines in binge drinking, as well as psychological and physical aggression) were expected. However, absence may not fully explain all the variations in outcomes across service members, spouses, children, and teens. For children and teens in particular, it may be important to ask them directly about how well they are coping with a deployment, as we found that for some outcomes (e.g., total difficulties and anxiety) parent and teen reports about teen well-being had little association with one another.

Third, the analyses reported here revealed a set of risk and readiness factors that appear to be associated reliably with multiple domains of postdeployment outcomes. With respect to factors associated with more successful adaptation to deployment, we find across several outcomes that families who engaged in predeployment readiness activities experienced more favorable outcomes postdeployment than families that did not. Given the correlational design of this study, it is not possible to know whether these activities protect or enhance family functioning across the deployment experience or whether the most-resilient families are the ones most likely to engage in these activities. Across a number of outcomes, we also found that more frequent communication and higher satisfaction with the amount of communication with the service member during deployment was associated with more-favorable outcomes postdeployment. Here, too, it is not possible to know whether communication plays a causal role in effective coping or if the most-resilient families were the ones taking the time to communicate with each other regularly. Until further research attempts to tease apart these alternatives, the fact that these associations emerge even after controlling for family characteristics at baseline is consistent with the view that more frequent communication during deployment has incremental benefits for families after the deployment.

Finally, with respect to risk factors, we identified the experience of traumatic events during the deployment as a risk factor for worse outcomes postdeployment. These findings join an emerging body of research suggesting that it may be the traumatic experience that service members are exposed to during deployment, rather than separation from family, that carries any negative effects associated with the deployment experience. Indeed, in most analyses, the length of the deployment itself showed little association with the postdeployment status of service members, spouses, or children---even when we did not control for deployment trauma.

However, the relationship between deployment experiences and postdeployment outcomes was far more complex than the current literature suggests. The measure of deployment experiences used in the study was designed to assess psychological trauma, and it included a range of descriptively different experiences that were all hypothesized to be psychologically traumatic and all shown to be associated with posttraumatic psychopathology (Schell and Marshall, 2008). As expected, we found evidence that the members' physical trauma, combat trauma, and psychological trauma (i.e., witnessing trauma or vicarious exposure to trauma) were each positively associated with postdeployment symptoms of PTSD and depression. However, these different types of service member experiences did not have homogeneous effects when looking at other service member outcomes, or when looking at outcomes for spouses or children/teens. This may suggest that the deployment experiences may have direct effects on the family that are not mediated through service member psychopathology. For example, being directly engaged in violence against the enemy during deployment may affect a service member's interactions with his or her family even in the absence of PTSD or depression. This intriguing finding deserves further exploration to see if parental deployment experiences may promote different mindsets, behaviors, and interactional styles with family members postdeployment. Unfortunately, this study (along with most others designed to look at the effects of deployment) has focused narrowly on assessing deployment events hypothesized to be traumatic, and most prior studies analyze these experiences as functionally interchangeable. New theory, measures, and analyses may be needed to better understand which deployment experiences have persistent effects on service members and their families, as well as how those effects are produced. Viewing all deployment events through the narrow lens of psychological trauma may lead us to miss, or misrepresent, the long-term effects of deployment.

Study Limitations

Despite the strengths of the Deployment Life Study, it is also limited in several ways that constrain the conclusions that these results can support. First, the baseline sample consisted of a selective sample of older, more-experienced family members who not only have remained in the military but also stayed married long enough to be included in our sample. This selectivity may have increased our odds of finding resilient families. We cannot speculate about what the results would look like for first-time deployers or newlyweds, given our sample. Nor can we speak to what deployment looks like for other types of families that were not part of our sample, including single-parent households or unmarried service members without dependents.

Second, some families that participated in the baseline assessment dropped out during the course of the study, which could have biased estimates. However, it is not entirely clear which direction the bias is in. Most likely, these families are worse on our outcomes, which would upwardly bias our estimates.

Third, our analyses include a large number of statistical tests and thus the results presented should be viewed in terms of patterns, rather than select, individually significant associations.

Fourth, as with any survey that attempts to measure phenomena that cannot be observed directly, measurement error is a potential problem. To the extent possible, our surveys used well-validated measures that have been used in other studies of both civilian and service members and their families.

Fifth, self-report data, even if they are collected outside of direct human contact (i.e., over the phone) may be subject to social response bias. That is, service members, spouses, and teens may have overestimated the positive aspects of their lives and underestimated the negative aspects.

Finally, although we have done a significant amount of work to minimize bias from observed differences between the deployed and nondeployed samples in our doubly robust analyses, a limitation of doubly robust methods is that they cannot guard against bias from unobserved factors whose effect on an outcome is not captured via the included confounders.

Taken together, these potential limitations should not overshadow the uniqueness of the Deployment Life Study or the contribution it makes to understanding the consequences of deployment for military families. Rather, they should be viewed as a caution for how broadly the results apply to the overall population of military families.

Policy Implications

Our findings have a number of implications for programs and policies aimed at improving the well-being and quality of life of service members, spouses, and their children across the deployment cycle.

Programs, services, and policies should target families that experience deployment trauma, especially during the postdeployment phase. To the extent that traumatic experiences during deployment are associated with a host of negative consequences in the postdeployment period, these experiences, when documented during deployments, can be used to target families for extra support upon the return of the service member. Programs that target families based on documented experiences, regardless of self-reported symptoms, might help mitigate problems before they metastasize, or at least before they have time to affect multiple family members. The challenge for providing the programs, services, and supports to these families, of course, is identifying families before problems occur.

Addressing psychological problems around the time of separation may be important for avoiding the longer-term impairments caused by these problems, such as increased morbidity, homelessness, unemployment, or substance use among veterans. Our results indicate that service members who have separated from service postdeployment have significantly elevated psychological symptoms. Regardless of whether psychological problems predate separation, the separation period appears to be a high-risk period for individuals who leave the military.

Attention to the challenges associated with deployment should not detract from supporting services that address other challenges of military life. Although they are perhaps the most salient stressor in the lives of military families, deployments are far from the only challenges that military families face. The fact that characteristics of military families measured prior to the deployment accounted for outcomes after the deployment highlights the importance of programs that support families in other ways---e.g., by reducing financial stress or addressing mental health issues.

Programs that facilitate communication both between and within military families during a deployment may promote not only greater military integration, but also better outcomes across the domains highlighted here. Family connectedness to the military is higher when families engage with other military families. In addition, interacting with other military children was protective among teens in our study, across a number of outcomes. These relationships with peers are likely to help family members share important emotional, instrumental, and informational resources that can be used to cope with a deployment and to develop strong ties within the military community. Similarly, communication among family members, and satisfaction with that communication, were associated with improved marital and parenting relationships. Despite the fact that this study was not designed to evaluate family support programs, our results show that targeted efforts, such as those surrounding communication during deployment, could have great promise as family support programs. The true impact of these programs, however, has yet to be seen.

Support to improve relationships between service members, spouses, and their teen children during the postdeployment reintegration period may improve family functioning. Our results indicated that both postdeployment family cohesion and the quality of the relationship with the nondeployed parent were worse among deployed versus nondeployed teens, and that relationship quality with the service member declined during the deployment period. Programs should focus on preventing declines in relationship quality and family cohesion after the service member returns, as opposed to waiting for families to seek help as relationships disintegrate.

Future Research on Military Families

Our findings also highlight several areas where changes to research strategies could result in improved data---in terms of both timeliness and quality---for making policy decisions. Future work on military families should explore ways in which data can be collected from multiple family members at the same time. Across the analyses presented, results based on spouses' data were more-frequently significant than results based on data from service members. Spouses' reports may be more sensitive to conditions within the family, and so may be promising sources of data on military families when data from service members are unavailable. For some outcomes, such as the family environment and anxiety, family members' (i.e., service members and spouses) reports of outcomes differed during the same period of the deployment cycle. Collecting data from multiple members of the same families can capture these differences and help tailor support for individual family members based on their relation to the service member (e.g., spouse, child, teen).

As funding resources become scarcer, future work on military families should prioritize longitudinal studies. Many existing studies of military families rely on retrospective or cross-sectional reports. Unfortunately, this means that much of the existing body of work on the impact of deployment on families does not allow for anchoring of family function prior to the deployment, potentially leading to an exaggeration of any deployment effect. Longitudinal study designs offer the most methodologically robust way to assess the impact of deployment on families. Without following the same families over time, we would not have been able to observe changes in the functioning of family members (and families) during the deployment period, and we would not have been able to report that those changes do not usually lead to sustained problems in the postdeployment period.

Procedures for collecting real-time data from military families should be explored. Throughout the study, we have been sensitive to the unique historical climate in which the data collection for the Deployment Life Study took place. Had we conducted the same study in 2006 or 2007, when deployments to Iraq and Afghanistan were longer and more frequent, our results may have painted a different picture of how well military families adapt to deployment. Longitudinal data have a lot to offer researchers, but they are time-consuming to collect, and results are often not known for several months or even years after the respondents have provided their data. Having a real-time source of data---one that is capable of tracking changes in the historical, political, and social climates---would provide a huge benefit to researchers and policymakers who seek to help families navigate the stress associated with military life. There is no single type of real-time data that would be able to address all the relevant research and policy questions that may be asked. The costs and benefits associated with different types of data should also be considered when deciding what type of real-time data collection methods to invest in. Some combination of administrative data for service members (e.g., medical records, personnel data) as well as ongoing data from a representative panel of military family members could prove to be a very useful, cost-effective solution.

Develop new theories, measures, and analyses of deployment experiences that can account for apparent complexities in the relationships between deployment and postdeployment outcomes. The complex pattern of findings relating service member deployment experiences and the outcomes across the family deserves further study. It is possible, for instance, that the experience of combat during deployment produces a set of psychological and behavioral factors that make reintegration with the family more difficult. On the other hand, witnessing of wartime traumas without being in combat or being personally injured may produce a very different set of psychological and behavioral reactions that make service members more eager or able to engage with their families. Further research is needed to investigate the persistent effects of deployment, both good and bad. The existing theory (focused almost exclusively on factors leading to posttraumatic psychopathology), the existing measures (focused on traumatic events and often geared toward the experiences of infantry), and the standard analyses (focused on using these measures as forming a homogeneous scale with consistent effects) may obscure the full effects of deployment on members and their families. Such research is necessary to develop policies or interventions that can effectively minimize the harm---and maximize the benefits---of military deployments.

Examine the intersections and relative timing of military and family events. Contemporary service members are more likely than previous generations to have families. When examining outcomes, including those detailed in study, it is worthwhile for future research to examine interactions between the relative timing of military (e.g., deployment, promotion, relocation) and family (e.g., birth of a child) events. Such research would facilitate a deeper understanding of when families might be most vulnerable or resilient in the face of deployment.

Conclusion

Across all of our analyses, comparisons between families that experienced a deployment and matched families that did not experience a deployment revealed few significant differences in outcomes that occurred after a service member returned from a deployment. Yet, the Deployment Life Study data revealed that life during the deployment itself was stressful in a number of ways, albeit differently for different family members. Given the unquestionable stresses of the deployment period and the well-established negative associations between stress and family outcomes among civilians, it is remarkable that most families are able to adjust. On one hand, this apparent resilience might reveal the noteworthy success of the support programs currently in place for military families. On the other hand, the findings presented here may reflect the natural ability of many military families to thrive, even in the face of severe challenges. In reality, both stories are probably true---existing programs help military families cope, and families have natural coping abilities. However, it is important to remember that not all military families are the same. While the sample of families in the Deployment Life Study is generally more experienced, with both deployments and military life in general, some are not. Likewise, for some families, military life is a struggle. What we can learn from these more-experienced, resilient families can then be used to help those who may be less successful at adapting and adjusting to the unique demands associated with military life.

Finally, we would be remiss if we did not acknowledge the tremendous amount of data that the Deployment Life Study collected and the analyses yet to come. It is our hope that future exploration of these data will allow us to make even finer-grained recommendations about how to create and promote readiness and resilience among all military families.