Assignment: DoDEA, LEAs Education, and Support

Journal of Family Psychology © 2011 American Psychological Association

2011, Vol. 25, No. 4, 508–520 0893-3200/11/$12.00 DOI: 10.1037/a0024395

Page 508

A Meta-Analytic Review of Internalizing, Externalizing, and Academic Adjustment Among Children of Deployed Military Service Members

Noel A. Card, Leslie Bosch, Deborah M. Casper, Christine Bracamonte Wiggs, Stacy Ann Hawkins, Gabriel L. Schlomer, and Lynne M. Borden University of Arizona

Parental deployment during military conflicts has the potential to impact child adjustment. As increased numbers of military Service members have children, it is critical to understand the association between military deployment and child adjustment. In order to resolve inconsistencies in the existing literature, we performed a meta-analytic review of 16 studies that report associations of military deployment with internalizing, externalizing, and academic adjustment among children. Results indicate a small association between deployment and poorer adjustment. This association varied across several features of the studies. Age moderation was such that the associations are strongest in middle childhood and weakest during adolescence. The method that adjustment was assessed also moderated this association, such that maladjustment was evident primarily with parents’ reports. Study design also moderated associations, such that comparisons to civilian controls indicated associations with maladjustment, whereas comparisons to nondeployed military and prepost comparisons did not. These findings summarize the existing quantitative literature to indicate that parental deployment has a negligible association with child maladjustment and provide a foundation for future research.

Keywords: academic adjustment, child psychopathology, externalizing, internalizing, metanalysis, military deployment

This article was published Online First June 27, 2011. Noel A. Card, Leslie Bosch, Deborah M. Casper, Christine Bracamonte Wiggs, Stacy Ann Hawkins, Gabriel L. Schlomer, and Lynne M. Borden, Family Studies and Human Development, University of Arizona. This work was supported by U.S. Department of Agriculture and the U.S. Department of Defense under Award #2010-4886920693. Correspondence concerning this article should be addressed to Noel A. Card, Family Studies and Human Development, 650 North Park Avenue, University of Arizona, Tucson, AZ 857210078. E-mail: [email protected]

As of September 2009, nearly one million families with at least one child had a parent in the U.S. military. These one million families represent nearly 2 million children who live in a household with a military parent (U.S. Department of Defense, 2009). Indeed, recent data show that the number of military dependents (i.e., spouses and children) outnumbers military Service members (Chartrand & Siegal, 2007; Military Homefront, 2009). Nearly half of military Service members have children (in this article, the term “children” generally refers to offspring under the age of 18 years, unless explicitly stated otherwise), and current military families have experienced the longest, most frequent, and most cumulative number of deployments in U.S. history (Tanielian & Jaycox, 2008). Deployment impacts many domains of a Service member’s life, potentially including the wellbeing and adjustment of a Service member’s child. Given the presence of global military conflicts, the effect of parental deployment on children is of special interest.

For instance, since 2001 and 2003 (respectively), the United States has engaged in major military operations in Afghanistan (Operation Enduring Freedom) and Iraq (Operation Iraqi Freedom). Approximately 1.8 million U.S. Military Service members have been deployed to these regions with many of these men and women having been deployed multiple times. Although deployment characteristics vary across branches and individuals, general estimates indicate typical lengths of deployment among U.S. Military Service members around 12 to 15 months and typical number of deployments around 2.2 times (Chandra, Burns, Tanielian, Jaycox, & Scott, 2008). Deployment is the strategic movement and positioning of military forces from a home base to an area of military combat or peacetime operations (U.S. Army, 1992); this movement of military personnel inevitably results in the Service member being separated from family members. As tours lengthen and multiple deployments to combat zones increase, families of Service members face difficult and multiple transitions that may have consequences for children. There is a clear distinction between being a military Service member and being a deployed military Service member; therefore, an important question to consider is whether, and to what extent, military deployment, particularly combat related, is associated with psychological and academic adjustment for these children.

Page 509

Potential Impact of Deployment on Children

There are reasons to expect variously that military deployment would have either a positive or negative impact, or no average impact, on children of Service members. One possibility is that parental deployment fosters new opportunities for children to take on more mature roles, suggesting a positive impact of deployment. Alternatively, deployment may negatively impact the child either indirectly (by fostering poorer quality parenting behaviors) or directly. Finally, it is possible that the average impact of deployment is negligible; even an absence of effect is meaningful. We consider each of these possibilities next.Arguments for the potential positive impact of parental deployment on children of Service members have focused on the resiliency developed through either the military lifestyle in general or during the parental deployment in particular. One aspect of the military lifestyle experienced by many military families is the frequent relocations during the career of the Service member. Weber and Weber (2005) provided evidence that relocation frequency fostered resilience among adolescent children of Service members. Specifically, parents reported fewer behavioral issues among their children as the frequency of relocations increased. A second, distinct, potential source of resilience comes from deployment. The experience of parental deployment (which is distinct from relocation in that it involves the Service member being deployed to a combat situation away from the family, rather than the family relocating together) may necessitate that the child take on additional responsibilities in the family. Huebner, Mancini, Wilcox, Grass, and Grass (2007), focusing on adolescents, argued that these additional responsibilities during deployment are opportunities for adolescents to grow in terms of skills and maturity. These arguments would suggest that parental deployment is associated with improved adjustment, perhaps especially in comparison to civilian families, during adolescence, or both.

However, there is also the possibility that deployment has a negative impact on adjustment. One possibility is that deployment could negatively impact Service members’ parenting behaviors (and in turn, their children’s adjustment) in two ways. First, deployment necessarily means that the deployed parent will not have as direct and regular contact with their child as they normally do, and such absence might negatively impact the child (for a review of the impact of father absence on adolescents, see East, Jackson, & O’Brien, 2006). A recent review of work and family research reported that high job involvement, long hours, job-related stress, and low decision-making control were associated with more work-family conflict (Bianchi & Milkie, 2010). These job characteristics are similar in description to the job-related stressors experienced by Service members, especially during deployment. The impact of these stressors on the Service member, as well as parental absence on the child during deployment, may increase conflict in the home (i.e., increased conflict between parents, parent and child, or both) and may negatively affect parenting behaviors. The influence of job-related stressors can have a direct effect on child adjustment. Second, the stress of deployment may cause both short- and long-term difficulties for the deployed Service member that might impact their parenting behaviors directly or more indirectly through their marital relationship (e.g., Hiew, 1992; Huebner et al., 2007; Jensen, Martin, & Watanabe, 1996; Schumm, Bell, & Gade, 2000; for reviews of similar processes within civilian families, see Buehler, Anthony, Krishnakumar, & Stone, 1997; Erel & Burman, 1995). The intense and oftentimes dangerous nature of military deployment causes physical, psychological, or both stresses among many Service members. For instance, Service members may contend with alcohol- and substance abuse issues after returning from deployment (Hoge, Auchterlonie, & Milliken, 2006; Prigerson, Maciejewski, & Rosenheck, 2002) and almost one third of Service members who returned from Iraq or Afghanistan experience depression, posttraumatic stress disorder (PTSD), or traumatic brain injury (Hoge et al., 2006; Hosek, Kavanagh, & Miller, 2006). These problems of adjustment among some Service members are likely to impact their marital relations as well as their parenting behaviors (Galovski & Lyons, 2004), which in turn might negatively impact their children’s adjustment (Buehler, Anthony, Krishnakumar, & Stone, 1997; Papp, Goeke-Morey, & Cummings, 2004; Wieland & Baker, 2010).

Deployment might also impact children by impacting the parenting behaviors of the nondeployed parent (i.e., the spouse or other co-parent of the deployed Service member). These parents face additional parenting and household responsibilities while simultaneously being faced with lost support (emotional, financial, or both) from the spouse and worries about the deployed Service member. These stressors might negatively impact parenting behaviors, such as reducing parental monitoring and warmth and increasing parental negativity, which might in turn negatively impact children’s adjustment (see Simons, Lorenz, Wu, & Conger, 1992; Wieland & Baker, 2010). In addition, deployment may influence the functioning of the family system dynamics, which may result in boundary breaching (e.g., child serves as confidant, child takes on responsibility for care of siblings) between the parent and children who remain at home (Huebner, Mancini, Wilcox, Grass, & Grass, 2007). However, it is also possible that deployment could increase closeness in the parent–child relationship that might promote more positive adjustment. Finally, parental deployment could have a direct impact on children, beyond the diminished contact with the deployed parent and changes in parental behaviors. As mentioned, children (perhaps especially adolescents) might be asked to take on additional responsibilities within the family, which has the potential to increase maturity (Huebner et al., 2007). Alternatively, parental deployment might directly impact children because the children worry about the safety of their parent and might have difficulty finding support outside the home if they feel that others cannot understand their situation.

Page 510

Although we do not investigate these potential mechanisms by which military deployment might impact children in this review, these possibilities present several rationales to expect that parental deployment is associated with child adjustment. To date, the direction and magnitude of associations between parental deployment and children’s adjustment is unclear. Previous narrative reviews of this association have consistently described the findings as inconclusive, widely varying, or both. For example, Chandra et al. (2008) stated “there are conflicting findings in the literature pertaining to the well-being of children and adolescents from military families . . . ” (p. 13). Even a casual perusal of this literature reveals markedly inconsistent findings (which we review in more detail in qualitative reviews below). One reason for this inconsistency could be that the studies have too often used small samples and have been inconsistent in their methodological choices. In this article, we present results of a meta-analytic review that provides a quantitative summary of the existing literature on military deployment and children’s psychological and academic adjustment.

Goals of Meta-Analysis

In this article, we present results of a meta-analytic review, supplemented with brief qualitative reviews, of associations between parental deployment and their children’s internalizing, externalizing, and academic adjustment. Meta-analysis is a statistical methodology of systematically coding existing quantitative (vs. qualitative), multiple participant (vs. single participant) empirical research to extract effect sizes (here, of the association between deployment and child adjustment) that represent the direction and magnitude of associations. These effect sizes are placed on a common metric (e.g., correlation coefficients), which can then be combined and/or systematically compared across studies through well-established statistical principles (Hedges & Olkin, 1985). Research (see Rosenthal, 1991) has shown that reviews of literature that use meta-analytic techniques have greater statistical power (i.e., ability to identify effects when they exist) and are more consistent than reviews performed that use qualitative approaches (which is a limitation of prior reviews of the impact of deployment).

The first goal of this review was to summarize the existing empirical literature to determine the existence, direction, and magnitude of associations between parental deployment and children’s adjustment. Specifically, we considered adjustment in terms of internalizing symptoms (e.g., depression, anxiety), externalizing symptoms (e.g., aggression, rule breaking), and academic adjustment (e.g., grades, achievement test scores). These three domains represent a wide range of psychological and academic adjustment and are also the domains most commonly studied in the existing literature. It is worth mentioning the numerous ways in which deployment and nondeployment were operationalized varied across studies. For example, deployment was described in terms of the total number of months the Service member was deployed in the previous 3 years (Chandra, Lara-Cinisomo, et al., 2010); the average number of deployments over a specified period of time (Chandra et al., 2008); being currently deployed (Horton, 2007; Lester et al., 2010); having been deployed in the last 6, 12, or 15 months (Flake, Davis, Johnson, & Middleton, 2009; Jensen, Grogan, Xenakis, & Bain, 1988; Morris & Age, 2009), retrospective reports of being deployed (Pierce, Vinokur, & Buck, 1998; Rosen, Teitelbaum, & Westhuis, 1993); peacetime versus wartime deployment (Kelley, 1994); Service members facing deployment in the near future (Kelley et al., 2001); and by the amount of hazardous duty pay the Service member earned (Lyle, 2006). The term nondeployment was used most often to describe the nondeployed, at-home caregiver who was not deployed and who may or may not be an active duty Service member. No studies in the current review used the term nondeployment to describe Service members who had never been deployed. By met analytically combining associations between deployment and adjustment across existing studies, we hope to provide an authoritative statement in regard to the existence and magnitude of these associations that overcomes the frequent reliance on small samples in this area of study. This information is valuable not only in helping us understand the average adjustment experienced by these children, but it also provides a foundation for investigating differences in adjustment across children of deployed Service members.

The second goal of this review was to investigate characteristics of the studies that might lead to them finding larger or smaller associations between deployment and adjustment (i.e., moderator analyses). As we describe later, several seemingly relevant characteristics were not reported consistently enough to use for formal statistical analysis. However, three study characteristics were evaluated in this review: age of child, the informant of the child’s adjustment, and the design of the study. The mean age of children in study samples was examined in relation to associations between deployment and adjustment to evaluate whether there is evidence of greater vulnerability or resilience during early childhood (0 to 6 years), middle childhood (6 to 12 years), or adolescence (12 to 18 years). We also evaluated whether associations differ depending on whether children, teachers, deployed or nondeployed parents, or standardized test scores are used to measure children’s adjustment. These findings might inform whether any changes in adjustment associated with deployment exist only in some contexts or in the eyes of some beholders. Finally, we investigated whether studies yield different associations depending on their design. Specifically, we compared studies contrasting children of deployed parents with children of civilians, studies contrasting children of deployed parents with children of nondeployed Service members, and studies contrasting children of deployed parents during versus prior to deployment. The latter two designs provide the most valid information (because these account for any potential differences between children of civilians vs. Service members). It is important to understand whether the designs differ in their findings in order to guide future research.

Page 511

Method

Selection of Studies

We identified relevant literature in three ways. First, we performed keyword searches of several electronic databases (PsycInfo, ERIC, Sociological Abstracts, Inform World, EconLit, Google Scholar, Science Direct, and papers of the Rand Corporation). We began with the keyword deployment and then added additional terms during subsequent searches (family, child, or adolescen) to enhance the specificity of the search. Second, we examined several previous narrative (i.e., qualitative) reviews of the impact of deployment on children. Finally, we identified other works cited in the reports identified through the first two strategies (i.e., backward searches; Card, 2011). In total, as a result of these techniques we located and examined 159 reports for potential inclusion in this review.Studies were included in this meta-analytic review if they met five criteria. First, the studies needed to include quantitative measures of at least one of three aspects of children’s adjustment: (a) internalizing symptoms, (b) externalizing symptoms, or (c) academic adjustment. Second, the studies needed to compare adjustment during parental deployment to one of four types of control conditions: (a) a group of children of civilian parents, (b) a group of children of nondeployed military parents, (c) pre-deployment data for the same children, or (d) standardized norms for the instruments. Third, the studies needed to measure adjustment during or shortly after the deployment period. Fourth, we limited studies to those sampling families of U.S. Military Service members. Finally, we necessarily limited our sample to reports that provided sufficient information to compute effect sizes (see below). We also combined results of multiple reports using the same sample to meet statistical assumptions of independence (see Card, in press). Using these five inclusion criteria, 16 studies from the 159 identified from our search were included in this meta-analytic review.

Coding of Studies

Coding of study characteristics. We systematically coded these 16 studies to record basic descriptive information (e.g., study authors, year, whether the study was published), sizes of the sample (used for weighting, as described below), and several study characteristics that might be systematically related to variations in results (effect sizes). All studies were coded by either the second or third author, and then coding was checked by the first author. Because several seemingly relevant study characteristics (e.g., deployment stage, length of deployment, how many times the parent was previously deployed, nature of deployment) were frequently omitted in existing reports, they were not included in this meta-analysis. Ultimately, we coded three study characteristics used in analyses.

First, we coded the mean age of the sample, which we then categorized into three developmental periods: early childhood (prior to 5.5 years, corresponding approximately to the preschool period), middle childhood (from 5.5 to under 12 years; corresponding approximately to elementary school age), and adolescence (12 to 18 years). When studies reported results separately by age, we recorded effect sizes for each age group.

Second, we recorded the informant used to measure children’s adjustment. Child reports represent the child’s own view of functioning; these might be most accurate for some internal experiences, but might also be subject to bias (e.g., a child not wanting to report being depressed or acting out). Teacher reports represent an external perspective of an adult who is able to compare the child’s functioning to others, but whose observations are generally limited to the school context. Parents also provide important adult perceptions of adjustment but may also be biased by what they are willing to report, by their own adjustment to deployment, and potentially limited to the home context. We differentiated between reports of the nondeployed parents (i.e., the caregiver who was not the deployed Service member of focus in the study and who was most often the child’s mother), who have more direct contact with the child but might themselves be impacted by the deployment, and reports of the deployed parents, who have less (and often indirect) knowledge of the child’s adjustment and may themselves be experiencing unique stressors of deployment. Finally, some studies have relied on standardized tests to measure children’s adjustment; these might be the most objective measure, but they have typically been limited to measures of academic adjustment in this literature.Third, we coded the study design used to evaluate the association of deployment with child adjustment. Some studies contrasted children of deployed parents with children of civilians either directly by measuring a control group of children, or else indirectly by comparing children of deployed parents to normative data. The majority of studies contrast children of deployed parents to children of nondeployed Service members. We view studies using this design as strong in that they account for any potential differences in children with military versus civilian parents. The third type of design commonly used in these studies are longitudinal comparisons of the same children during versus prior to their parents’ deployment. We also view studies using this design as strong because they provide a direct comparison to the same child’s adjustment before deployment.

Coding of effect sizes. An effect size is a number that represents the direction and magnitude of associations. For this meta-analysis, we used the correlation coefficient, r, as an effect size that represents the association between deployment and children’s adjustment. Positive values indicate that children of deployed parents have more problems than controls, whereas negative values denote that these children had fewer problems than controls. A common standard in social sciences for interpreting the magnitudes of these correlations is that values around positive or negative ().10 are considered small, values around .30 are considered medium, and values around .50 are considered large (Cohen, 1969). Although we generally interpret effect sizes by using these common standards, it is important to note that these standards have been criticized (e.g., McCartney & Rosenthal, 2000), and we will reinterpret results in terms of practical importance later.

Page 512

We coded the effect sizes from included studies by either recording the correlations reported or computing these correlations from other reported data using common metanalysis equations (see Card, 2011; Rosenthal, 1991). Two of the 16 studies coded yielded effect sizes that could be considered outliers: Medway, Davis, Cafferty, Chappell, and O’Hern (1995) yielded correlations that were considerably more positive than other studies, and Nice (1978) yielded correlations that were considerably more negative than other studies. Supplemental analyses excluding these two studies (available upon request) yielded generally comparable results. Given the small number of studies in this review, these two studies were retained in the analyses presented, and instances where findings differed if these studies were excluded are noted.Given that the correlation coefficient is skewed in the population, we transformed these effect sizes via Fisher’s transformation to obtain Zr. Analyses were performed by using this transformed effect size. Results (e.g., average Zr) were back transformed to the more familiar r for reporting

Statistical Analysis

The general analytic strategy for each of the metanalyses reported below was to first compute the weighted1 random-effects mean effect size of the association between deployment and adjustment (note that the random-effects model would simplify to the fixed-effects model in the absence of heterogeneity; for information about these models see Card, in press; Hedges & Vevea, 1998 Raudenbush, 1994). We next evaluated the heterogeneity of effect sizes across studies (Q), followed by analysis of study characteristics that predict this variability in effect sizes (i.e., moderator analyses; see Card, 2011; Lipsey & Wilson, 2001). We evaluated three study characteristics that were reported consistently enough in the literature (age, informant, and design) as potential moderators. When evaluating these moderators, we considered multiple effect sizes per study when the effect sizes contributed to different levels of the moderators (see Card, 2011); for example, the separately reported results by age of Rosen et al., 1993, were used to compute three separate effect sizes for early childhood, middle childhood, and adolescence. Given the small number of studies in each analysis, it was necessary for us to evaluate each moderator separately rather than simultaneously. Also, we noted when there are only a small number of studies at some levels of these moderators, as these conclusions should be interpreted with caution.

Results

We report findings from this meta-analytic review in five parts. We first describe the sample of 16 studies in this meta-analysis in terms of coded study characteristics. We then consider associations between deployment with overall adjustment and then three specific aspects: internalizing symptoms, externalizing symptoms, and academic adjustment. Within each section, we report three sets of results. First, we report the average correlation that represents the single best index of the association between deployment and child adjustment. We also report 95% confidence intervals of this mean, which can be interpreted as the plausible range of actual effect sizes. Second, we report results of a statistical test (denoted by Q) of the variability of effect sizes across studies: Findings of homogeneity indicate that the studies appear to estimate a single population correlation with any variations across studies due to sampling fluctuation alone, whereas findings of heterogeneity indicate that the studies produce more variable results than expected by sampling fluctuation alone. Third, if the set of studies are found to be heterogeneous (variable in their results), then we perform analyses that investigate whether coded study characteristics predict these variations across studies.

Descriptions of Studies Included in the Meta-Analysis

Characteristics of the 16 studies included in this metaanalytic review, including the effect sizes found, are summarized in Table 1. These studies were published (or otherwise completed) from 1978 to 2010, with 88% published after 1992. Most (81%) of the studies were published in peer-reviewed literature. The large majority (14) of the samples were recruited on military installations, specifically through on-base schools, family support groups, or through on-base military housing. Two studies used archived Service member personnel records matched with relevant child outcomes, and in one study data were collected from children at Operation Purple Camp. The sample sizes generally ranged from 42 to 1,601 children (plus an outlier containing 14,448 children), with a median sample size of 142. The median average sample age of studies was 8.9 years. There was also considerable range of ages of children represented among these studies, ranging from less than 1 year to over 15 years, although as described later, samples of early childhood and adolescence were less common than those during middle childhood. A variety of informants, including child reports, teacher reports, deployed or nondeployed parent reports, and standardized tests were used to measure children’s psychosocial adjustment. When measuring internalizing and externalizing symptoms, five studies made comparisons to relevant clinical cutoffs, and three studies made comparisons to community norms or a civilian sample. The most frequently (seven studies) used measure was the Child Behavior Checklist (CBCL; Achenbach & Edelbrock, 1983). The second most common measure was the Strength and Difficulties Questionnaire (SDQ; Goodman, 2001) with two studies using this measure. Other studies used a variety of other measures including the Behavior Assessment System for Children–2nd ed. (BASC-2; Reynolds & Kamphaus, 2004), the California Test of Personality (CPT; Tiegs, Clark, & Thorpe, 1941), the Pediatric Symptom Checklist (PSC; Jellinek et al., 1999), and the Child Health Inventory (CHI; Rosen, Moghadam, & Bain, 1992). Only one study used nonstandard items generated by the authors.

Page 514 Most studies either (a) compared children of deployed parents to those of nondeployed Service members, or (b) compared children of deployed parents during versus prior to deployment; though a small number of other studies used other comparisons (e.g., to children of civilians). Specifically, of the 16 studies described in this analysis, 11 made direct comparisons between pre-deployment and current deployment. Of these 11 studies, seven used a cross-sectional approach, wherein children with a deployed parent were compared to children without a deployed parent. An additional three studies used longitudinal assessments, wherein children, parents, or both were assessed prior to deployment and then again during or after deployment. One study used retrospective reports of child functioning prior to and during deployment. Several study characteristics that we intended to code were reported too infrequently. Deployment stage was ambiguous in 69% of studies, and other details of deployment history (e.g., length of the current deployment, number of previous deployments) were rarely reported. The percentage of parents who were officers versus enlisted was reported in only 38% of studies. Although most (81%) studies reported the percentage of deployed parents who were fathers or mothers, there was generally not enough range to evaluate this study characteristic as a potential moderator (the median percentage of deployed parents who were fathers in studies was 92%).

Associations With Overall Adjustment

A total of 16 studies, consisting of 19,172 participants, provided information in regard to associations between deployment and any aspect of children’s adjustment. Across these studies, the average association between deployment and child problems was r .08 (p .05), with a 95% confidence interval that ranged from .02 to .15. This correlation can be considered small in magnitude according to Cohen’s (1969) standards. In sum, children with deployed military parents, on average, have a small amount of overall maladjustment. These, and all subsequent results, are summarized in Table 2. The 16 studies that reported this association, however, were heterogeneous (see Table 2) in their effect sizes. Three study characteristics (i.e., age of child, informant of child’s adjustment, study design) were used as predictors of this variability in effect size. Age of child (early childhood, middle childhood, and adolescence) moderated the association between parental deployment and overall child problems. As shown in Table 2, this association was small in early and middle childhood but virtually nonexistent during adolescence. However, these results should be viewed with caution given the small number of studies using early childhood (k = 3) or adolescent (k = 3) samples.

Among these studies, adjustment has been reported by a variety of sources, including through child reports, teacher (or similar childcare provider) reports, reports of the deployed parent (most commonly fathers), reports of the nondeployed parent (most commonly mothers), and through standardized instruments (e.g., standardized achievement tests). The source of information moderated the magnitude of association between deployment and overall child adjustment, such that both deployed and nondeployed parents perceived worse adjustment associated with deployment (as stated earlier, nondeployed parents were the spouses of deployed Service members, and may or may not themselves be Service members). When adjustment (primarily academic) is measured through standardized instrument, children of deployed parents exhibit a statistically significant but trivial magnitude degree of poorer adjustment. In six studies that rely on children’s own reported adjustment deployment was not significantly associated with overall problems, and there was no detectable association in three studies relying on teacher reports, though this finding should be viewed with caution given the small number of studies that use teacher reports. Overall, there exists little evidence of deployment that has a meaningful relation to adjustment from the perspective of children, teachers, and standardized instruments, whereas parents (deployed and nondeployed) perceive a small degree of poorer adjustment. Studies of the association between deployment and adjustment have used one of three designs: (a) comparisons of children of deployed parents relative to children of nondeployed Service members, (b) comparisons of children of deployed parents relative to children of civilian parents (either as a control group within the study or by comparing scores to standardized norms, which presumably contain a relatively small number of military children), and (c) comparisons of the same children during versus prior to deployment. Associations between deployment and overall adjustment differ depending on this methodology. Studies that use children of civilians yielded small-to-medium associations between deployment and overall problems, whereas comparisons to children of nondeployed Service members indicated a statistically significant but trivial association, and comparisons during versus pre-deployment did not indicate any association between deployment and overall problems. These findings suggest little association between deployment and overall adjustment problems in methodologically rigorous studies.

Page 515

Associations With Internalizing Symptoms

Internalizing symptoms include depression, sadness, anxiety, loneliness, and withdrawal. A total of 12 studies, which included 2,707 participants, provided information in regard to associations between deployment and internalizing symptoms. An initial qualitative examination of these studies indicates mixed evidence about the association between deployment and internalizing symptoms. Some research has shown that children of Service members experience more internalizing problems, compared to national samples, to children with nondeployed parents, and to children with civilian parents (Chandra, Martin, Hawkins, & Richardson, 2010; Jensen et al., 1988; Jensen et al., 1996; Kelley et al., 2001). Other studies, however, found no association between parent deployment and internalizing behavior problems (Chartrand, Frank, White, & Shope, 2008; Flake et al., 2008; Horton, 2007; Morris & Age, 2009). It might be, however, that factors such as child age and parent wellbeing affect the association between parent deployment and child outcomes. For example, Chartrand et al. (2008) found that children with deployed parents exhibited more internalizing symptoms than those with nondeployed parents only among children over 3 years of age.

Page 516

Meta-analytic results indicated that although the mean association was comparable to that of overall problems (r .09; see Table 2), the average effect is not statistically significant. Thus, there is not definitive evidence of an association between deployment and internalizing symptoms. Statistical evaluation indicates that the results of these studies are heterogeneous across studies, which might be one reason why definitive statements about an average association between deployment and children’s internalizing symptoms cannot be made. Age of child significantly predicted (moderated) the association between parental deployment and child internalizing symptoms among 11 effect sizes, providing information about specific This association was small to medium in middle childhood (this value was somewhat, though not completely, reduced if the two studies with extreme scores are removed) but nonsignificant in early childhood and during adolescence. Again, these results should be viewed with caution given the small number of studies using early childhood (k =2) or adolescent (k =1) samples.

Among studies that examined associations between deployment and internalizing symptoms, child reports, teacher reports, reports of the deployed parents, and reports of the nondeployed parents have been used. Associations differed across these informants. Deployment is associated with nondeployed parents’ reports of child internalizing problems (this association was inflated by the extremely high association found by Medway, Davis, Cafferty, Chappel, & O’Hern, 1995, though the association remains higher if this study is removed). Deployment was unrelated to child reports or teacher report of internalizing symptoms, though the latter result should be viewed with caution given the reliance on only two studies. The design of studies (comparison to children of civilians, comparison to children of nondeployed Service members, and comparison during vs. prior to deployment) predicted the magnitude of association between deployment and internalizing symptoms. Surprisingly, comparison across time indicates that internalizing symptoms are lower during than prior to deployment (r = -.19, p < .05), though this finding is largely due to the extreme results (r = -.50) of Nice (1978). Studies that compare internalizing symptoms of children of deployed parents with either nondeployed Service members (r= .01, ns) or of civilian

parents (r = .07, ns) do not find consistent evidence of associations between deployment and internalizing symptoms.

Associations With Externalizing Symptoms

Externalizing symptoms include aggression and rule breaking behaviors (e.g., destruction of property or theft). A total seven studies, which included 2,646 participants, provided information on the association of parental deployment with these externalizing symptoms among their children. As with internalizing behavior problems, a preliminary qualitative review of these studies yielded mixed results. Some studies have found no association between deployment and externalizing symptoms (Chartrand et al., 2008; Flake et al., 2009; Jensen et al., 1988, 1996; Morris & Age, 2009). In contrast, in a study of children with Navy mothers, Kelley et al. (2001) found that children with deployed mothers exhibited more externalizing behavior problems than those with nondeployed mothers. This study also suggested potential differences by age: Only children ages 3 to 5 years had elevated externalizing symptoms (Chartrand et al., 2008).

In a meta-analytic combination of these studies, , the mean effect size was r = .05, which was not significantly (reliably) different from zero (p = .43). The 95% confidence interval ranged from -.07 to .16, indicating that even if an association exists, it would be considered small in magnitude. In short, there appears to be little if any consistent association between parental deployment and children’s externalizing symptoms. However, these seven studies were heterogeneous, Q(6) = 47.61, p < .001, and we investigated study characteristics that might predict this variability. Age of child significantly predicted (moderated) the association between parental deployment and child externalizing symptoms among nine effect sizes. This association was small in middle childhood and nonsignificant in early childhood and adolescence, although these results are tempered by the paucity of studies in the latter two groups (ks = 2 and 1, respectively). Despite this caution, these results suggest that externalizing problems among children of deployed Service members might be of most concern during middle childhood.

Deployment has been studied in relation to child-, teacher-, and nondeployed parent- reported externalizing problems, and this association varies across reporter. Specifically, deployment has a small association with nondeployed parent-reported externalizing symptoms (r = .08, p <.01), though this association was due primarily to the strong association found by Medway et al. (1995). There were no detectable associations of deployment with child (r = -04, ns), teacher (r = .06, ns), or deployed parent (r = -.02, ns) reports of externalizing symptoms. Study design did not predict associations between deployment and externalizing symptoms, Q(2) = 2.56, p =.28.

Page 517

Associations With Academic Adjustment

Our review included five studies that reported associations between parental deployment and academic adjustment (including child reports of academic engagement, standardized test scores, and more ambiguous measures such as parent reports of “doing poorly in school”). Qualitative synthesis of these studies suggests that parent deployment is associated with poorer academic outcomes (Lyle, 2006; Medway et al., 1995; Pisano, 1996). However, the magnitude of this association is unclear from a qualitative synthesis. Meta-analytic combination, summarized in Table 2, yielded an average correlation of r = .08 (p < .05, 95% confidence interval .01 to .16), indicating that children of deployed parents had a small degree of poorer academic adjustment than controls. However, these studies were heterogeneous, Q(4) =33.79, p < .001, and study characteristics were evaluated as predictors of this variability in results. As with the other aspects of adjustment, the age of child significantly predicted associations between parental deployment and child academic adjustment among six effect sizes.

Results, which should be viewed extremely cautiously given the small number of studies in each group, suggest a small-to-medium association in middle childhood, a small association in early childhood, and essentially no association in adolescence. Deployment has been studied in relation to child reports, nondeployed parent reports, and standardized test scores of academic adjustments, and the type of informant moderated this association with academic adjustment. Nondeployed parental reports of academic problems had a small association with deployment (the value in Table 2 is only slightly smaller with the exclusion of Medway et al., 1995). For the two studies that measure academic adjustment through standardized tests, the association was statistically significant but trivial in magnitude. The single study that used child reported academic adjustment did not find an association with deployment. Associations between deployment and academic adjustment have been examined in studies that use comparison to nondeployed Service members’ children as well as comparisons to pre-deployment scores. These two types of studies did not differ in the associations found

Discussion

This meta-analytic review synthesized existing research to provide conclusions about what is currently known in regard to associations between parental military deployment and children’s internalizing, externalizing, and academic adjustment. At the same time, this review also pointed to areas of future study in regard to these associations. Next, we summarized the findings from this review and then noted the limits of our knowledge and future research that might illuminate these issues.

Findings From Meta-Analysis

This review summarized the existing empirical literature on the association between parental deployment and child adjustment by using quantitative methods of meta-analysis. We found evidence that deployment is associated with overall problems, as well as specific internalizing, externalizing, and academic problems. However, these associations were uniformly small by social science standards, with deployment typically predicting less than 1% of the variability in children’s adjustment. In sum, our results suggested that military deployment is not associated, on average, with any substantial improvement or detriment in children’s internalizing, externalizing, or academic adjustment. However, we caution against the interpretation that children who experience parental military deployment are unaffected. Instead, these results indicate that children of deployed military Service members on average do not experience greater maladjustment (or adjustment). There is likely considerable variability in reactions that merit attention. Supporting this likelihood of variability, the results of these studies were variable (i.e., heterogeneous). In an attempt to explain this between-study variability, we evaluated whether coded study features predicted these results (effect sizes). Unfortunately, some potentially relevant characteristics (e.g., length of deployment, combat vs. noncombat) are not consistently reported in the literature.

However, we were able to evaluate child age, informant of child adjustment, and study design, though it should be remembered that these analyses should be viewed as hypotheses for future studies rather than definitive answers, due both to the small numbers of studies in our review and the limits of interpreting meta-analytic moderators in general (Card, in press; Lipsey, 2003). In terms of age, deployment was most strongly associated with problems during middle childhood, somewhat less strongly and consistently associated with problems during early childhood and was generally not associated with adjustment during adolescence. However, even during middle childhood, the impact is typically small. Although this pattern of age moderation is unexpected given that several problems generally escalate during adolescence, it might be the case that adolescents who face deployment also possess greater coping abilities (e.g., sense of agency; support from peers) or positions within the family (e.g., ability to take on responsibilities that foster sense of maturity; Huebner et al., 2007) that allow them to cope with the challenge of parental deployment more successfully than younger children. Furthermore, adolescents in military families may have developed specific coping strategies through their longer experience with the military lifestyle (e.g., numerous transitions, frequent moves; Weber & Weber, 2005). These possibilities merit further attention in future developmental research.

Page 518

Results also differed depending on how adjustment was measured. We found that deployment is associated with problems only when adjustment is reported by parents (nondeployed or deployed parents). In contrast, children’s own reports as well as the reports of their teachers, are not substantially related. If these findings are confirmed in future primary work, then it will be important to conceptually distinguish adjustment from perceptions of adjustment. Perhaps the challenges faced by parents during deployment (both those deployed and those remaining home) bias their perceptions of their child’s adjustment difficulties. Alternatively, children may underreport and under present problems in an effort to appear “strong” during the deployment. Further research that uses multiple informants of adjustment is needed to disentangle these possibilities. Studies that use more in-depth interviewing (e.g., focus groups of school staff in Chandra, Martin, Hawkins, & Richardson, 2010) might also be useful in evaluating the sources of differing perceptions. The design of studies also predicted effect sizes. Studies that compare children of deployed Service members to civilian children found associations between deployment and adjustment, whereas studies that use what we consider more methodologically rigorous (i.e., comparison to children of nondeployed Service members or to adjustment prior to deployment) show no evidence of associations

Future Directions for Research

This meta-analytic review synthesized the existing multi person (vs. single subject) quantitative (vs. qualitative) research to draw conclusions about the association between deployment and children’s adjustment. It also exposed several limitations in the existing research that raise several additional questions to be answered in future research. In our view, the most pressing question is, why do children vary in their adjustment associated with parental deployment? Our attempts to predict between-study variability in results (effect sizes) by coded study characteristics (child age, informant, and design) began to provide potential answers, but it is important to note that substantial unexplained between-study variability remained (i.e., studies differed markedly in their findings even after accounting for differences in coded study characteristics). Even if it was possible to explain fully this between-study variance, there is undoubtedly variability in reactions of children within studies. A small number of studies have begun to evaluate how individual factors such as coping style (Morris & Age, 2009) and attachment style (Medway et al., 1995), as well contextual factors such as family functioning and community capacity (Huebner, Mancini, Bowen, & Orthner, 2009; Palmer, 2008), may play a role as buffering or exacerbating factors. However, more studies that systematically investigate the child (e.g., age, gender, race, ethnicity, previous adjustment), family (e.g., relationship between nondeployed and deployed parents, pre-deployment quality of parent– child relationships, presence and relationships with siblings), and other proximal (e.g., supportive relationships with peers, teachers, or both) and distal (e.g., geographic location) contextual qualities that buffer or exacerbate children from the stress of parental deployment are needed. It will also be important for further research to investigate the mechanisms by which deployment leads to positive and negative outcomes. For example, does the adjustment of the nondeployed parent or the quality of the relationship between parents account for (mediate) the associations between deployment and child adjustment? If such intervening mechanisms can be identified, these might be potent targets for prevention–intervention.

Future research also needs to investigate more deliberately the features of deployment and military lifestyle associated with better or worse adjustment. It is plausible that children’s adjustment varies with stage, length, and nature (e.g., combat intensity) of the deployment, the number of previous deployments, and other (nondeployment) aspects of their experiences as children of military Service members (e.g., number of previous transitions). It will also be important to study the impact of multiple deployments to evaluate whether these have a cumulative (e.g., each deployment has a small impact, by multiple deployments sum to a larger impact) or accelerating (e.g., initial deployment has a minimal impact, but subsequent deployments have increasingly large impact) impact on children. In total, it will be important for these studies to ensure that samples are representative of the full range of experiences faced by Service members and their families, and to more clearly and consistently define and measure deployment and stages of deployment, as the unique aspects and stressors of the different features and stages of deployment are likely to be differentially associated with adjustment. Finally, future longitudinal research is needed to distinguish between short- and long-term adjustment associated with parental deployment. It is plausible that any problems associated with deployment (even if only for some children) are strongest during or shortly after deployment, but then subside. Alternatively, there may be little short-term impact for some children, but long-term adjustment may be either promoted or hindered. Because the existing literature is inconsistent in when adjustment is measured in relation to deployment, we currently do not know what timeframes should be considered. For these reasons, systematic longitudinal research is needed.

Limitations and Conclusions

Our suggested directions for future research highlight the most salient limitation of this meta-analytic review: The existing research basis is small (16 studies) and has not been as systematic as desirable in investigating moderators and mediators of the links between deployment-child adjustment links within primary studies, nor in investigating these links longitudinally. It is also worth noting that the studies included in this review were naturalistic (vs. experimental); therefore, our results describe the association between deployment and child adjustment, but do not speak to causality. Despite these limitations, this article presents the first meta-analytic review of the quantitative research on the associations between military deployment and children’s internalizing, externalizing, and academic adjustment. This review goes beyond previous narrative reviews by providing quantitative conclusions in regard to the presence and magnitude of these links, as well as systematically investigating the sources of varying results across studies.

Page 519

Our results show that there exist only small associations between deployment and overall maladjustment. Our results also suggest that these associations are stronger in middle childhood than adolescence, in the perceptions of parents, and in methodologically weak study designs. It is more important to note that this review provides a foundation for further research investigating why children vary in their reactions to deployment and the mechanisms by which deployment impacts children. It is clear that additional studies are needed to have a more complete understanding of these factors that will provide foci for prevention and intervention efforts.

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

References

References marked with an asterisk indicate studies included in the meta-analysis.

Achenbach, T. M., & Edelbrock, C. S. (1983). Manual for the Child Behavior Checklist and Revised Child Behavior Profile. Burlington, VT: University Associates in Psychiatry

Bianchi, S. M., & Milkie, M. A. (2010). Work and family research in the first decade of the 21st century. Journal of Marriage and Family, 72, 705–725. doi:10.1111/j.1741-3737.2010.00726.x

Buehler, C., Anthony, C., Krishnakumar, A., & Stone, G. (1997). Interparental conflict and youth problem behaviors: A metaanalysis. Journal of Child and Family Studies 6, 223–247. doi: 10.1023/A:1025006909538

Card, N. A. (2011). Applied meta-analysis for social science research. New York, NY: Guilford Press.

Chandra, A., Burns, R. M., Tanielian, T., Jaycox, L. H., & Scott, M. M. (2008). Understanding the impact of deployment on children and families: Findings from a pilot study of Operation Purple Camp participants. Santa Monica, CA: RAND Corporation. Retrieved from http://www.rand.org/pubs/working_papers/ WR566.html

*Chandra, A., Lara-Cinisomo, S., Jaycox, L. H., Tanielian, T., Burns, R. M., Ruder, T., & Han, B. (2010). Children on the homefront: The experience of children from military families. Pediatrics, 125, 16–25. doi:10.1542/peds.2009-1180

Chandra,A.,Martin,L.T.,Hawkins,S.A.,&Richardson,A.(2010).The impact of parental deployment on child social and emotional functioning: Perspectives of school staff. Journal of Adolescent Health, 46, 218–223. doi:10.1016/j.jadohealth.2009.10.009

*Chartrand, M. M., Frank, D. A., White, L. F., & Shope, T. R. (2008). Effect of parents’ wartime deployment on the behavior of young children in military families. Archives of Pediatric and Adolescent Medicine, 162, 1009–1014.

Chatrand, M. M., & Siegel, B. (2007). At war in Iraq and Afghanistan: Children in U.S. military families. Ambulatory Pediatrics, 7, 1–2. doi:10.1016/j.ambp.2006.11.004

Cohen, J. (1969). Statistical power analysis for the behavioral sciences. New York, NY: Academic Press.

Department of Defense. (2008). Demographics 2008: Profile of the military community. Washington, DC: Department of Defense. Retrieved from http://cs.mhf.dod.mil/content/dav/mhf/ QOL-Library/Project%20Documents/MilitaryHOMEFRONT/ Reports/2008%20Demographics.pdf

East, L., Jackson, D., & O’Brien, L. (2006). Father absence and adolescent development: A review of the literature. Journal of Child Health Care, 10, 283–295. doi:10.1177/1367493506067869

Erel, O., & Burman, B. (1995). Interrelatedness of marital relations and parent-child relations: A meta-analytic review. Psychological Bulletin, 118, 108–132. doi:10.1037/0033-2909.118.1.108

*Flake, E. M., Davis, B., Johnson, P. L., & Middleton, L. S. (2009). The psychosocial effects of deployment on military children. Journal of Developmental and Behavioral Pediatrics, 30, 271–278. doi:10.1097/DBP.0b013e3181aac6e4

Galovski, T., & Lyons, J. A. (2004). Psychological sequelae of combat violence: A review of the impact of PTSD on the veteran’s family and possible interventions. Aggression and Violent Behavior, 9, 477–501. doi:10.1016/51359-1789(03)00045-4

Goodman, R. (2001). Psychometric properties of the Strength and Difficulties Questionnaire. Journal of the American Academy of Child and Adolescent Psychiatry, 40, 1337–1345.

Hedges, L. V., & Olkin, I. (1985). Statistical methods for metaanalysis. San Diego, CA: Academic Press.

Hedges, L. V., & Vevea, J. L. (1998). Fixed- and random-effects models in meta-analysis. Psychological Methods, 3, 486–504. doi:10.1037/1082-989x.3.4.486

Hiew, C. C. (1992). Separated by their work: Families with fathers living apart. Environment and Behavior, 24, 206–225. doi: 10.1177/0013916592242004

Hoge, C., Auchterlonie, J., & Milliken, C. (2006). Mental health problems, use of mental health services, and attrition from military service after returning from deployment to Iraq or Afghanistan. Journal of the American Medical Association, 295, 1023. doi:10.1001/jama.295.9.1023

*Horton, D. (2007). The impact of deployment on children in military families. Unpublished dissertation, Walden University. Retrieved from http://gradworks.umi.com/32/49/3249681.html

Hosek, J., Kavanagh, J., & Miller, L. L. (2006). How deployments affect Service members. Santa Monica, CA: RAND Corporation.

Huebner, A., Mancini, J., Bowen, G., & Orthner, D. (2009). Shadowed by war: Building community capacity to support military families. Family Relations, 58, 216–228. doi:10.1111/ j.1741-3729.2008.00548.x

Huebner, A. J., Mancini, J. A., Wilcox, R. M., Grass, S. R., & Grass, G. A. (2007). Parental deployment and you in military families: Exploring uncertainty and ambiguous loss. Family Relations, 56, 112–122. doi:10.1111/j.1741-3729.2007.00445.x

Jellinek, M. S., Murphy, J. M., Little, M., Pagano, M. E., Comer, D. M., & Kelleher, K. J. (1999). Use of the pediatric symptom checklist to screen for psychosocial problems in pediatric primary care: A national feasibility study. Archives of Pediatrics & Adolescent Medicine, 153, 254–260.

*Jensen, P. S., Grogan, D., Xenakis, S. N., & Bain, M. W. (1988). Father absence: Effects on child and maternal psychopathology. Journal of American Academy of Child and Adolescence Psychiatry, 28, 171–175. doi:10.1097/00004583-198903000-00004

*Jensen, P. S., Martin, D., & Watanabe, H. (1996). Children’s response to parental separation during operation desert storm. Journal of the American Academy of Child & Adolescent Psychiatry, 35, 433–441. doi:10.1097/00004583-199604000-00009

*Kelley, M. L. (1994). Military-induced separation in relation to maternal adjustment and children’s behaviors. Military Psychology, 6, 163–176. doi:10.1207/s15327876mp0603_2

*Kelley, M. L., Hock, E., Smith, K. M., Jarvis, M. S., Bonney, J. F., & Gaffney, M. A. 2001. Internalizing and externalizing behavior of children with enlisted Navy mothers experiencing military-induced separation. Journal of the American Academy of Child & Adolescent Psychiatry, 40, 464–471.

*Lester, P., Peterson, K., Reeves, J., Knauss, L., Glover, D., Mogil, C., . . . Beardslee, W. (2010). The long war and parental combat deployment: Effects on military children and at-home spouses. Journal of the American Academy of Child & Adolescent Psychiatry, 49, 310–320

Lipsey, M. W. (2003). Those confounded moderators in metaanalysis: Good, bad, and ugly. The Annals, 587, 69–81. doi: 10.1177/0002716202250791 Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. Thousand Oaks, CA: Sage.

*Lyle, D. S. (2006). Using military deployments and job assignments to estimate the effect of parental absences and household relocations on childrens’ academic achievement. Journal of Labor Economics, 24, 319–350.

*McCartney, K., & Rosenthal, R. (2000). Effect size, practical importance, and social policy for children. Child Development, 71, 173–180. doi:10.1111/1467-8624.00131

*Medway, F. J., Davis, K. E., Cafferty, T. P., Chappell, K. D., & O’Hern, R. E. (1995). Family disruption and adult attachment correlates of spouse and child reactions to separation and reunion due to Operation Desert Storm. Journal of Social and Clinical Psychology, 14, 97–118. doi:10.1521/jscp.1995.14.2.97

Military Homefront. (2009). Supporting our Troops and Their Families. 2009 demographics: Profile of the military community. Retrieved June 6, 2011 from http://www.militaryhomefront.dod .mil/content/day/mhf/QOL-Library/PDF/MHF/QOL%20Resources/ Report/2009_Demographics_Report.pdf

*Morris, A. S., & Age, T. R. (2009). Adjustment among youth in military families: The protective roles of effortful control and maternal social support. Journal of Applied Developmental Psychology, 30, 695–707. doi:10.1016/j.appdev.2009.01.002

*Nice, D. S. (1978). The androgynous wife and the military child. In E. J. Hunter & D. S. Nice (Eds.), Child of military families: A part and yet apart (pp. 25–37). Arlington, VA: Office of Naval Research.

Palmer, C. (2008). A theory of risk and resilience factors in military families. Military Psychology, 20, 205–217. doi: 10.1080/08995600802118858

Papp, L. M., Goeke-Morey, M. C., & Cummings, M. E. (2004). Mothers’ and fathers’ psychological adjustment and marital functioning: Examination of direct and interactive links with child adjustment. Journal of Child and Family Studies, 13, 469– 482. doi:10.1023/B:JCFS.0000044728.34058.c0

*Pierce, P. F., Vinokur, A. D., & Buck, C. L. (1998). Effects of warinduced maternal separation on children’s adjustment during the gulf war and two years later. Journal of Applied Social Psychology, 28, 1286–1311. doi:10.1111/j.1559-1816.1998.tb01677.x

*Pisano, M. C. (1996). Implications of deployed and nondeployed fathers on seventh graders’ California Achievement Test scores during a military crisis. Paper presented at the annual meeting of the National Association of School Psychologists. Atlanta, GA.

Prigerson, H. G., Maciejewski, P. K., & Rosenheck, R. A. (2002). Population attributable fractions of psychiatric disorders and behavioral outcomes associated with combat exposure among US men. American Journal of Public Health, 92, 59–63.

Raudenbush, S. W. (1994). Random effects models. In H. Cooper & L. V. Hedges (Eds.), The handbook of research synthesis (pp. 301–321). New York:

Russell Sage Foundation. Reynolds, C., & Kamphaus, R. (2004). Behavior Assessment System for Children (2nd ed., BASC-2). Circle Pines, MN: AGS Publishing.

Rosen, L. N., Moghadam, L. Z., & Bain, M. W. (1992). Health problems among Army children. Military Medicine, 157, 85–88.

*Rosen, L. N., Teitelbaum, J. M., & Westhuis, D. J. (1993). Children’s reaction to the Desert Storm deployment: Initial findings from a survey of army families. Military Medicine, 158, 465–469.

Rosenthal, R. (1991). Meta-analytic procedures for social research. Newbury Park, CA: Sage.

Schumm, W. R., Bell, D. B., & Gade, P. A. (2000). Effects of a military overseas peacekeeping deployment on marital quality, satisfaction, and stability. Psychological Reports, 87, 815–821. doi:10.2466/PR0.87.7.815-821

Simons, R. L., Lorenz, F. O., Wu, C., & Conger, R. D. (1992). Support from spouse as mediator and moderator of the disruptive influence of economic strain on parenting. Child Development 63, 1282–1301. doi:10.1111/j.1467-8624.1992.tb01695.x

Tanielian, T., & Jaycox, L. H. (2008). Invisible wounds of war: Psychological and cognitive injuries, their consequences, and services to assist recovery. Santa Monica, CA: Rand.

Tiegs, E. W., Clark, W. W., & Thorpe, L. P. (1941). The California Test of Personality. Los Angeles: California Test Bureau.

U.S. Army. (1992). U.S. Army Field Manual (FM 100–17, chap. 4). U.S. Army Training Doctrine and Command. Retrieved January 8, 2010, from http://library.enlisted.info/field-manuals/ series-2/FM100_17/CH4.PDF

U.S. Department of Defense. (2009). Profile of the military community: Demographics 2009. Retrieved from http://cs.mhf.dod .mil/content/dav/mhf/QOL-Library/PDF/MHF/QOL%20Resources/ Reports/2009_Demographics_Report.pdf

Weber, E., & Weber, D. (2005). Geographic relocation frequency, resilience, and military adolescent behavior. Military Medicine, 170, 638–642.

Wieland, N., & Baker, B. L. (2010). The role of marital quality and spousal support in behaviour problems of children with and without intellectual disability. Journal of Intellectual Disability Research, 54, 620–633. doi:10.1111/j.1365-2788.2010.01293.x

Received October 29, 2010

Revision received May 10, 2011

Accepted May 13, 2011