Effects of Poly-Victimization on Adolescent Social Support,Self-Concept,and Psychological Distress

Abstract

Past research has demonstrated the particularly damaging effects of exposure to multiple forms of victimization, or “poly-victimization,” on youth mental health. The primary objective of the present study is to begin to identify the mechanisms that help explain its powerful impact. Analyses are based on two waves of longitudinal data from the National Survey of Children’s Exposure to Violence (NatSCEV), conducted in 2008 and 2010, that comprised a telephone sample of 1,186 youth ages 10 to 17. Using structural equation modeling, we examine direct and indirect effects on distress symptoms of increased, decreased, and stable high poly-victimization between Waves 1 and 2 compared to no or low victimization in both waves. Specifically, we consider the extent to which reductions in core psychosocial resources, including family support, peer support, self-esteem, and mastery, mediate the relationship between these poly-victimization conditions and distress. Relative to stable low victimization, both increased poly-victimization and stable high poly-victimization were associated with declines in all four resources. However, only self-esteem and mastery significantly mediated the association between poly-victimization and distress, with mastery showing the strongest effect. Although significant indirect effects were evident, poly-victimization still had a strong direct effect on distress with resource factors controlled. Findings support the hypothesis that the potent effect of poly-victimization on youth mental health is, in part, due to its damaging influence on core psychosocial resources.

Keywords

poly-victimization maltreatment bullying self-concept social support mental health

Introduction

Poly-victimization refers to the experience of multiple victimizations of different kinds, such as sexual victimization, physical abuse, bullying, witnessing family violence, and exposure to community violence, not just multiple episodes of the same kind of victimization. Recent research on poly-victimization has shown it to have especially damaging effects on child well-being. Studies by the researchers who originally developed the poly-victimization model (Finkelhor, Ormrod, & Turner, 2007a, 2007b, 2009; Finkelhor, Ormrod, Turner, & Hamby, 2005a; Turner, Finkelhor, & Ormrod, 2010c) found that: (a) a significant portion of children in the United States who identify as victims of child abuse or bullying or other single forms of violence are in fact poly-victims who have experienced many different types of victimization; (b) poly-victimization is more highly related to trauma symptoms than experiencing repeated victimizations of a single type, even repeated serious forms of victimization; and (c) poly-victimization explains most of the psychological consequences of individual forms of victimization.

Additional research has confirmed these original findings. For example, in a study using longitudinal data from the Project on Human Development in Chicago Neighborhoods (PHDCN), investigators examined the effects of exposure to school violence, community violence, child abuse, and parental intimate partner violence, as well as the cumulative effects across these multiple domains. Findings indicated that the accumulation of exposure to violence across the different types was most predictive of both future alcohol and marijuana use (Wright, Fagan, & Pinchevsky, 2013). In a large sample of violence-exposed children, Hickman et al. (2013) measured violence exposure in three different ways: total frequency of all lifetime exposure, total frequency of lifetime exposure by broad category (i.e., assault, maltreatment, sexual abuse, and witnessing violence), and poly-victimization, defined as exposure to multiple violence categories. The results indicated that only poly-victimization (exposure to two or more categories of violence exposure) emerged as a consistent predictor of negative symptoms (posttraumatic stress disorder [PTSD] and behavior problems). In a study of youth involved in the juvenile justice system, investigators conducted latent class analysis derived from 19 types of adversity, including different forms of violence exposure. Among the three unique classes of youth that emerged, the poly-victims group was distinct from other adversity groups in the likelihood of exposure to multiple types of traumatic victimization, and in the presence of severe emotional and behavioral problems (Ford, Grasso, Hawke, & Chapman, 2013). Several additional studies further document strong links between poly-victimization and negative outcomes in children and adolescents (Cyr et al., 2013; Elliott, Alexander, Pierce, Aspelmeier, & Richmond, 2009; Ford, Elhai, Connor, & Frueh, 2010; Holt, Finkelhor, & Kaufman Kantor, 2007; Soler, Kirchner, Paretilla, & Forns, 2013). Taken together, evidence strongly suggests that poly-victimization—that is, exposure to multiple different forms of victimization—is a more powerful predictor of negative child outcomes than exposure to any individual type of victimization, even when it occurs repeatedly.

Why the Powerful Effects of Poly-Victimization?

Given the growing body of research on the damaging effects of poly-victimization, it becomes increasingly important to identify the psychosocial mechanisms that may help to explain its especially powerful effects. As Turner et al. (2010c) point out, poly-victimization reflects a substantial level of adversity, comprised of a variety of stressful events that occur across multiple life domains. Consistent with research on Adverse Childhood Experiences (ACE; Felitti et al., 1998), poly-victimization represents a diverse set of potentially traumatic adverse experiences that accumulate in their detrimental effects on health and well-being. We know from developmental, behavioral, and biological research that a strong link exists between childhood exposure to multiple stressful events and conditions, and impaired neurological, physiological, and psychosocial systems that contribute to a wide array of mental and physical health problems (Shonkoff, Boyce, & McEwen, 2009). Yet the poly-victimization model also highlights the particular importance of children’s cross-context exposure to violence and victimization. Researchers have speculated that the widespread cross-context victimization that occurs when children are exposed to many different forms of victimization creates a life condition where children have no “safe haven”—no context or location where they can be free from the threat of harm. Poly-victims are likely to experience victimization by peers at school, by family members at home, and by a variety of individuals within their neighborhoods and communities.

It has been suggested that these conditions not only reflect substantial adversity but also are likely to damage resources that might otherwise encourage effective coping and resilience and reduce psychological distress (Turner et al., 2010c). A very substantial body of research on stress processes has shown that social resources, such as social support from family and peers, as well as personal resources, such as self-esteem and mastery, have direct positive effects on psychological well-being and, in some cases, moderate or buffer the negative effects of adversity (Pearlin & Bierman, 2013; Thoits, 2013; Turner & Turner, 2013). To the extent that poly-victimization erodes these resources, this may represent a crucial mechanism by which poly-victimization exerts its substantial influence on youth mental health.

Poly-Victimization and Social Support

There is little doubt of the importance of social support for human health and well-being. For children and youth, social interactions and environmental contexts associated with family and caregiving relationships represent a crucial part of the social support needed for healthy development (Mercy & Saul, 2009; Turner et al., 2012). Positive peer relationships have also been clearly linked to psychological well-being, while peer rejection and low peer support in adolescence are associated with distress (Newman, Newman, Griffen, O’Connor, & Spas, 2007; Zimmer-Gembeck, Hunter, & Pronk, 2007).

Research has clearly demonstrated the link between violent family contexts, including inter-parental violence and maltreatment, and reduced perceptions of family social support (Gabalda, Broth, Thompson, & Kaslow, 2009; Merrill, Thomsen, Sinclair, Gold, & Milner, 2001; Pepin & Banyard, 2006). Similarly, we know that victimization at the hands of peers is associated with reduced levels of peer social support (Demaray & Malecki, 2003; Flaspohler, Elfstrom, Vanderzee, Sink, & Birchmeier, 2009; Holt & Espelage, 2007; Pepin & Banyard, 2006; Rigby, 2000). Findings show that when youth are exposed to multiple forms of peer victimization at school, the negative effects on peer social networks are especially pronounced (Furlong & Chung, 1995).

Poly-victimization, involving victimization in multiple life domains, however, may be even more damaging to support resources. Youth who are victimized by peers at school, but are not poly-victims, may be able to seek advice, support, or sympathy from their parents. However, when children are also victimized at home, they may not be able or desire to receive support from parents. Moreover, in households where children have witnessed inter-parental violence, parents may not be emotionally available to provide support to the child because they are dealing with victimization themselves (Holt & Espelage, 2007). Similarly, although positive peer relationships can help to mitigate the negative effects of maltreatment at home (Ezzell, Swenson, & Brondino, 2000; Schultz, Tharp-Taylor, Haviland, & Jaycox, 2009), when youth also experience victimization at school, they will be less likely to have a supportive peer relationship that they can turn to (Demaray & Malecki, 2003; Flaspohler et al., 2009; Rigby, 2000). Further, neighborhood contexts where children are exposed to community violence are often characterized by low levels of trust of residents and low perceived support from both family and neighbors (Aneschensel & Sucoff, 1996; Turner, Shattuck, Hamby, & Finkelhor, 2013). In sum, poly-victimized children who are exposed to violence in multiple domains of life may also experience declines in levels of support in multiple life domains, including parents, family members, and peers.

Poly-Victimization and Self-Concept

In addition to the potential consequences of poly-victimization for social support, children’s exposure to multiple forms of victimization may have particularly detrimental effects on self-concept. We focus on two core aspects of self-concept—self-esteem and mastery. According to Morris Rosenberg’s (1965) classic definition, self-esteem refers to a “favorable or unfavorable attitude towards the self” (p. 15). In other words, self-esteem is an individual’s sense of his or her value or worth, or the extent to which a person values, approves of, appreciates, prizes, or likes himself or herself (Blascovich & Tomaka, 1991).

Mastery has been defined as “the extent to which one regards one’s life chances as being under one’s own control in contrast to being fatalistically ruled” (Pearlin & Schooler, 1978, p. 5). Other constructs that share conceptual ground with mastery include locus of control (Rotter, 1966), personal control (Ross & Mirowsky, 2013), self-efficacy (Bandura, 1982), and fatalism (Bandura, 1982; Wheaton, 1983). Although not identical, all these constructs incorporate the notion that personal agency—perceiving that one is causally relevant in life outcomes—is an important factor in human development and functioning (Turner & Roszell, 1994).

Research on the impact of stress exposure in general, and the effects of child victimization in particular, suggests the significance of changes in self-processes that stem from adverse conditions and experience in the social environment. Significant stressors, such as victimization, are presumed to damage normal psychosocial processes responsible for the development of positive self-concept. Given that children’s experience with successes and failures in the social environment and their perceptions of how they are viewed by others would inevitably help to shape mastery and self-esteem, exposure to victimization is likely to be damaging to both these aspects of self-concept.

Research in this area has tended to consider the effects of only a single type or category of victimization at a time. Child sexual abuse, for example, has been linked to low mastery, external locus of control, and poor self-efficacy (Bolstad & Zinbarg, 1997; Simmons & Weinman, 1991; Valentine & Feinauer, 1993), as well as low levels of self-esteem (Bagley & Ramsay, 1986). In a longitudinal study, Bolger and Patterson (2001) found that the effect of child maltreatment on children’s distress was partially mediated through declines in perceived personal control (i.e., mastery). Low self-esteem has been found to be associated with peer victimization or bullying in several studies (Benas & Gibb, 2007), and Turner, Finkelhor, and Ormrod (2010b) found that self-esteem significantly mediated the effect of adolescent sexual victimization on symptoms of distress.

Despite the above findings, almost no research has examined the effects of children’s exposure to multiple types of victimization on self-concept (see Soler et al., 2013, for exception). There is reason to suspect, however, that poly-victimization may be especially impactful. Poly-victims often experience victimization by peers at school, by family members at home, and by a variety of individuals within their communities (Turner et al., 2010c). Faced with the inability to avoid such pervasive threats, these youth may be especially likely to develop feelings of powerlessness and/or to come to believe that they are personally to blame for their circumstances; in other words, poly-victimization may often lead to reductions in mastery and self-esteem.

Study Objectives

Using a nationally representative longitudinal sample of youth ages 10 to 17, this research sought to examine the effects of poly-victimization on changes in levels of social and personal resources, including family social support, peer social support, self-esteem, and mastery. We also wished to determine whether declines in resources, in turn, help to explain the effect of poly-victimization on psychological distress. Specifically, we compared four groups of respondents: (a) those who scored low on poly-victimization in both Wave 1 (W1) and Wave 2 (W2; stable low group), (b) those who moved from high levels of poly-victimization at W1 to low levels at W2 (decline group), (c) those who moved from low levels of poly-victimization at W1 to high levels at W2 (increase group), and (d) those who scored high on poly-victimization in both W1 and W2 (stable high group). These groups were compared on demographic characteristics, changes in resource levels between the two waves, and W2 distress. Using structural equation modeling, we then examined direct and indirect effects of membership in the stable low group versus each of the other three groups on psychological distress. Specifically, we consider the extent to which W2 resource levels may mediate associations between the different poly-victimization conditions and W2 distress, controlling for W1 resource levels.

Method

Participants

These analyses use data from the first NatSCEV I, a two-wave longitudinal study of a representative sample of U.S. children and adolescents. This study focuses on 1,179 children who participated in both waves and who were aged 10 to 17 in Wave 1.

NatSCEV I was designed to obtain incidence and prevalence estimates of a wide range of childhood victimizations. The Wave 1 survey was conducted between January 2008 and May 2008 with a nationally representative sample of 4,549 children age 0 to 17 living in the contiguous United States. The Wave 2 survey was conducted approximately 2 years later between January 2010 and November 2010. Interviews with parents and youth were conducted over the phone by the employees of an experienced survey research firm.

The primary foundation of the Wave 1 design was a nationwide sampling frame of residential telephone numbers from which a sample of telephone households was drawn by random digit dialing (RDD). This nationally representative cross-section represented 67% of the completed interviews. To ensure that the study included a sizable proportion of minorities and low-income respondents for more accurate subgroup analyses, there was also an over-sampling of U.S. telephone exchanges that had a population of 70% or more of African American, Hispanic, or low-income households. This “over-sample” yielded 33% of the completed interviews. Sample weights were calculated for Wave 1 to adjust for differential probability of selection due to (a) study design, (b) demographic variations in non-response, and (c) variations in within household eligibility.

In Wave 2, efforts were made to re-contact Wave 1 respondents and solicit their participation. A total of 2,497 children who were aged 0 to 17 at Wave 1 took part in both waves (55% of the original sample). The remaining 45% of the original sample who did not participate consisted of respondents at Wave 1 who did not wish to be contacted for Wave 2 (5%), telephone numbers that were no longer active residential numbers (9%), numbers that were no longer associated with the original household (8%), refusals (9%), and failure to reach respondents at scheduled callbacks (13%).

Response Rates and Non-Response Analyses and Weighting

The Wave 1 cooperation rate for the RDD cross-section portion of the survey was 71% and the response rate was 54%. The cooperation and response rates associated with the smaller over-sample were somewhat lower at 63% and 43%, respectively. We compared parent reports of adolescents who completed the interview to parent reports for adolescent non-responders for 10- to 17-year-old participants. Non-responders were not systematically different from respondents on factors related to victimization risk (details of the non-response analyses can be obtained from the authors).

Overall, 55% percent of Wave 1 respondents aged 0 to 17 completed an interview in Wave 2. This rate was slightly higher (57%) for the subsample used in this study (respondents aged 10 to 17 in Wave 1). To adjust for differential attrition by demographic factors, victimization, and trauma symptom levels at Wave 1, a new set of sample weights was calculated for Wave 2 using both the Wave 1 sample weights and propensity scores based on the likelihood of each Wave 1 case returning for Wave 2. This method of adjusting for non-response is outlined by Wun, Ezzati-Rice, DiGaetano, Goksel, and Hongsheng (2005).

Procedure

In Wave 1, a short interview was conducted with an adult caregiver (usually a parent) in each household to obtain family demographic information. One child was randomly selected from all eligible children living in a household by selecting the child with the most recent birthday. If the selected child was 10 to 17 years old, the main telephone interview was conducted with the child, after obtaining consent from both the parent and the child. The Wave 2 study followed the same protocol. In both waves, respondents were paid US$20 for their participation. The interviews, averaging around 45 min in both waves, were conducted in either English or Spanish. All study procedures were approved by the University of New Hampshire Institutional Review Board for the Protection of Human Subjects.

Measures

Distress

Wave 2 distress was measured with the anger/aggression, depression, anxiety, dissociation, and posttraumatic stress scales of the Trauma Symptoms Checklist for Children (TSCC) (Briere, 1996). For the purpose of this study, the instruments were shortened for a total of 28 items. Respondents are asked to indicate how often they have experienced each symptom within the last month. Response options are on a 4-point scale from 1 = not at all to 4 = very often. For the bivariate analysis shown in Table 1, all item responses for the five scales together were summed to create an aggregate distress symptom score. In the structural equation model shown in Figure 1, distress was modeled as a latent variable with mean scores on each of the five subscales as observed indicators. The TSCC has shown very good reliability and validity in both population-based and clinical samples (Briere, 1996; Briere et al., 2001). In this study, the alpha coefficient for this scale was .93.

Table 1.

Demographic Characteristics and Change in Resources by Change in Victimization Across Waves (N = 1,179).

		Victimization Change Across Waves
		1	2	3	4	χ² or F Test Sig Level
	All	Stable Low	Declining	Increasing	Stable High	χ² or F Test Sig Level
Male (%)	51.1	48.2	58.2	46.6	54.8
Age at Wave 1 (M)	13.3	13.4	13.2	13.2	13.4
Socioeconomic status—Wave 1 (M)	0.08	0.123	0.103	0.075	−0.056
Family structure—Wave 1 (%)
Two parents	61.9	68.3	61.0	57.6	47.5	**
Parent and stepparent/partner	12.2	10.2	12.1	16.3	15.4
Single parent	21.0	17.8	23.6	23.0	26.2
Non-parent adult	4.9	3.6	3.3	3.1	10.8
Race/ethnicity (%)
White, non-Hispanic	61.7	65.1	58.2	60.3	56.8	^†
Black, non-Hispanic	14.7	10.1	17.8	19.4	21.8
Other, non-Hispanic	5.3	5.4	6.4	7.1	3.0
Hispanic, any race	18.3	19.5	17.6	13.3	18.5
Mean change in resources Wave 1 to Wave 2
Family social support	−0.18	−0.08³	0.15³	−0.95 ^1,2	−0.33	***
Friend social support	0.51	0.54	0.71	−0.01	0.49
Self-esteem	−0.07	−0.09	0.05	−0.09	−0.11
Mastery	0.10	0.11²	0.74^1,3,4	−0.47²	−0.27²	***
Mean distress score at Wave 2	42.8	39.0^2,3,4	41.5^1,3,4	47.7^1,2,4	51.7^1,2,3	***

Note. Percentages and means are weighted. N of 1,179 is unweighted.

Significance tests: Chi-square tests for percentages and ANOVA for means with post hoc Bonferroni tests of pairwise comparisons.

For pairwise comparisons, superscripts indicate that the value differs significantly from the value in the column numbered with that superscript. M = mean.

†

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

Figure 1.

SEM of change in past year poly-victimization and resources across two waves and distress at Wave 2.

Victimization

Both waves of this survey used an enhanced version of the Juvenile Victimization Questionnaire (JVQ), which obtained reports on specific types of youth victimization covering five general areas of interest: conventional crime, maltreatment, victimization by peer and siblings, sexual victimization, and witnessing and indirect victimization (Finkelhor, Ormrod, Turner, & Hamby, 2005b). The JVQ also includes follow-up questions that collect additional information about victimizations including perpetrator, weapon use, injury, and whether the event occurred in the past year. For this study, we constructed measures of children’s total burden of past year victimization at Waves 1 and 2 by summing the number of different victimization types they had experienced in the past year out of a total of 36 possible types. The mean number of past year victimizations in this sample was 2.5 (SD = 2.8, range = 0-16) at Wave 1 and 2.1 (SD = 2.8, range = 0-20) at Wave 2. Next, children were classified into four groups by whether and how their poly-victimization level changed from Wave 1 to Wave 2: (a) the stable low poly-victimization group includes youth whose number of past year victimizations fell below the mean in both waves, (b) the declining poly-victimization group had an above mean number of victimization types in Wave 1 but fell below the mean in Wave 2, (c) the increasing poly-victimization group increased from below the mean in Wave 1 to above the mean in Wave 2, and (d) the stable high poly-victimization group includes youth whose number of past year victimizations fell above the mean at both waves.

Resources

Four resources that are hypothesized to be impacted by poly-victimization were measured at both Waves 1 and 2: family support, friend support, self-esteem, and mastery. A modified version of the Multidimensional Scale of Perceived Social Support (Zimet, Dahlem, Zimet, & Farley, 1988) was used to assess perceived family support and friend support. Family social support and friend social support were each assessed using four items scored on a scale of 1 = never to 4 = always. The family social support items were as follows: “My family really tries to help me,” “My family lets me know that they care about me,” “I can talk about my problems with my family,” and “My family is willing to help me make decisions.” Cronbach’s alpha coefficients for the family support measure were .67 and .73 at Waves 1 and 2, respectively. The friend social support items were as follows: “My friends really try to help me,” “I can count on my friends when things go wrong,” “I have friends with whom I can share my good times and bad times,” and “I can talk about my problems with my friends.” The reliability coefficients for the friend support measure were .76 and .77 at Waves 1 and 2, respectively.

Self-esteem

Self-esteem was measured in both waves with a modified version summary score of an instrument developed by Rosenberg (1965); given the younger respondents in our sample, item wording was slightly simplified. The current scale is composed of 3 of the original 10 items reflecting different ‘‘self-statements’’ or beliefs. Respondents rate each statement on a 3-point scale ranging from very true to not true. Items include, “You are happy with yourself,” ‘‘You have a lot to be proud of,’’ and “You take a positive attitude toward yourself.” Cronbach’s alpha coefficients for these items were .70 in Wave 1 and .80 in Wave 2.

Mastery

Mastery was assessed using a shortened 5-item version of a 7-item scale developed by Pearlin and Schooler (1978); again, modifications were made to simplify the language and response categories, making items more appropriate for youth. The items were as follows: “You often feel helpless in dealing with problems,” “You sometimes feel that you are being pushed around in life,” “You cannot change important things in your life,” “You have little control over the things that happen to you,” and “There is no way you can solve some of the problems that you have.” Cronbach’s alpha coefficients for the mastery scale were .69 in Wave 1 and .73 in Wave 2.

Given the importance of ensuring conceptual clarity between our outcome measure, distress, and each of the resource measures (friend support, family support, self-esteem, and mastery), we ran a series of exploratory factor analyses (EFA) using data from another large national sample of 10- to 17-year-olds who had been asked identical survey questions. The EFAs showed that the four resources and the five subscales of distress (anxiety, anger, depression, dissociation, and posttraumatic stress) comprised distinct dimensions. However, there were three items that loaded similarly on two factors among the depression, self-esteem, and mastery dimensions. Those three items were dropped for this study.

Demographics

Demographic information for each child was collected in the initial parent interview, including the child’s gender, age (in years), race/ethnicity (coded into four groups: White, non-Hispanic; Black, non-Hispanic; Other, non-Hispanic; and Hispanic, any race), and socioeconomic status (SES). SES is a composite based on the sum of the standardized household income and standardized parental education (for the parent with the highest education) scores, which was then re-standardized. Family structure, defined by the composition of the household, was categorized into four groups: children living with (a) two biological or adoptive parents, (b) one biological parent plus partner (spouse or non-spouse), (c) single biological parent, and (d) other non-parent caregiver.

Data Analysis

Data analysis was conducted in three phases. First, bivariate analyses were conducted showing sample demographics, changes in resources across waves, and Wave 2 distress scores by the four poly-victimization groups described above (Table 1). For these analyses, Wave 1 and 2 scores for family and friend social support, self-esteem, and mastery were calculated as sums of the individual question scores for each resource variable. Changes in resources across waves were calculated by subtracting Wave 1 scores from Wave 2 scores. Distress scores shown in Table 1 were calculated as a sum of the 28 TSCC items. Differences across groups were tested for significance using chi-square tests for nominal variables and ANOVA with post hoc pairwise Bonferroni tests for continuous variables.

Next, Wave 1 and 2 resources were modeled in two confirmatory factor analyses (CFA). Each CFA contained the four resources modeled as latent variables with their associated survey questions as observed indicators. Each of the four latent resource variables was allowed to co-vary with the other three. The Wave 1 and 2 resource CFAs showed excellent fit—Wave 1: comparative fit index (CFI) = .98, Tucker–Lewis index (TLI) = .97, root mean square error of approximation (RMSEA) = .03, χ²(92) = 177, p < .001; Wave 2: CFI = .98, TLI = .98, RMSEA = .03, χ²(93) = 191, p < .001. Factor loadings were all significant at p < .001 and ranged from .49 to .73 in the Wave 1 CFA and from .48 to .79 in the Wave 2 CFA. Wave 2 distress scores were also modeled as latent variables in a CFA with children’s mean scores on each of the five subscales as indicators. The distress CFA also showed excellent fit, CFI = .99, TLI = .99, RMSEA = .03, χ²(3) = 6.13, p = .104, with loadings ranging from .73 to .80, all significant at p < .001. Finally, all latent and observed variables were combined into a single structural equation model (SEM) using Stata 13 (Figure 1).

The fit of the CFAs and the main structural model was assessed using the Bentler (1990) CFI, the TLI (Tucker & Lewis, 1973), and the Steiger-Lind RMSEA (Steiger, 1990). We also report overall χ² goodness-of-fit statistic for the model shown in Figure 1, but this statistic was significant at p < .001. Non-significant χ² statistics indicate better model fit; however, in large samples, “it is possible that rather small model-data discrepancies can result in a statistically significant value” of the model χ² statistic (Kline, 2011, p. 201). Given our sample size of 1,179, we chose not to reject our model based solely on the χ² statistic and rely instead on the above fit indices.

Results

Bivariate Analyses

Table 1 compares youth in the four groups of poly-victimization stability or change across waves. Group comparisons are shown for demographic characteristics, changes in resources (difference scores before vs. after victimization exposure), and Wave 2 distress levels. There were no significant differences in gender, age, or SES across the four groups. However, Black youth were overrepresented in the stable high poly-victimization group while White youth were overrepresented in the stable low group (p < .10). Significant differences in family structure were also evident; greater percentages of youth from single parent households and households with no biological parent were in the stable high group, while youth from two-parent households fell disproportionately into the stable low poly-victimization group.

Most relevant to the goals of the study, there were significant differences in the amount and direction of change in resources across the groups between Waves 1 and 2. Family social support declined for all youth on average, but there was variation among the four groups: Those who had experienced an increase in poly-victimization showed the largest decline in family social support—significantly different from the decrease among youth who experienced low poly-victimization at both waves (p < .01) and significantly different from the increase reported by youth whose poly-victimization level decreased across waves (p < .001). Changes in level of mastery also differed across groups: Youth whose poly-victimization level had increased or remained high across waves showed declines in mastery that were significantly different from the increase in mastery reported by youth in the declining poly-victimization group (p < .001 for both comparisons). Finally, as expected, distress symptoms at Wave 2 differed across groups, with youth in the stable low group having the lowest distress scores, and those in the stable high poly-victimization group exhibiting the highest.

Structural Equation Models

As shown in Figure 1, we estimated a model of the relationships between each poly-victimization condition across waves (declining, increasing, or stable high) and each of the hypothesized resources (family support, friend support, self-esteem, and mastery) at Wave 2, controlling for Wave 1 levels of each resource. The stable low poly-victimization group serves as the reference category in the model. Thus, we are assessing the effects of poly-victimization stability and change on changes in levels of each resource between Waves 1 and 2. We also model the effects of Wave 2 resource levels on Wave 2 levels of distress, as well as the direct paths between youths’ poly-victimization condition and Wave 2 distress. All demographic factors, including age, gender, SES, race/ethnicity, and family structure are controlled in the model. Paths and coefficients for demographic controls are omitted from the diagram for simplicity. Also for simplicity, disturbance terms and correlations between them are not shown.

This model fits the data well, especially given the number of parameters being estimated, χ²(1015) = 2,289, p < .001; CFI = .91; TLI = .90; RMSEA = .03. As Figure 1 shows, youth whose number of past year victimization types increased from Wave 1 to Wave 2 and youth who had high victimization levels at both waves showed significant reductions in each type of resource across waves, relative to youth whose number of victimizations were low at both waves. In contrast, youth who experienced decreases in poly-victimization did not differ from the stable low group on any of the resources. Mastery showed the strongest reduction among the increasing and stable high groups (Beta = −.18 and −.23, respectively, p < .001) followed by family social support (Beta = −.16, p < .001 for both groups). Only the self-concept resources were significantly associated with distress at Wave 2. Controlling for Wave 1 levels of resources, higher levels of Wave 2 self-esteem and, especially, mastery were substantially related to lower levels of distress at Wave 2 (Betas = −.25 and −.51, respectively; p < .001). There were also significant direct effects of both increased and stable high poly-victimization on Wave 2 distress (Beta = .11 and .19, respectively; p < .001). The full model explains 64% of the variance in distress symptoms.

Indirect effects of poly-victimization group on Wave 2 distress were tested for significance using unstandardized coefficients and the Sobel test (Kline, 2011). The indirect paths from stable high poly-victimization to distress through self-esteem and mastery were significant. Comparing indirect effects using standardized coefficients, more than three quarters of the indirect effect of stable high poly-victimization (79.2%) is through mastery (−.23 × −.50 = .12; p < .001), which has the strongest indirect effect in the model. The next strongest indirect effect of stable high poly-victimization on distress is through self-esteem (−.08 × −.24 = .02; p < .05; 12.8%). Poly-victimization increase across waves also had significant indirect effects on distress through both mastery (−.18 × −.50 = .09; 71% of indirect effect, p < .001) and self-esteem (−.10 × −.24 = .03; 20% of the group’s indirect effect, p < .01). The indirect effects of stable high and increased poly-victimization through family support and friend support were not significant. Declining poly-victimization had no indirect effect on distress.

Finally, we tested two alternative models nested in our final model of Figure 1 that assumed no mediating effects. In the first model, the paths between each of the W2 resources and W2 distress were removed. This model’s fit was inferior to that of the model in Figure 1, χ²(1019) = 2,885, p < .001; CFI = .87; TLI = .86; RMSEA = .04, and significantly worse based on change in model chi-square, Δχ²(4) = 596, p < .001. In the second nested model, we trimmed the paths from the Wave 2 poly-victimization groups to Wave 2 resources. This model also showed a worse fit than our final model, χ²(1027) = 2,385, p < .001; CFI = .91; TLI = .90; RMSEA = .03, with a significant increase in the model chi-square, Δχ²(12) = 96, p < .001.

Discussion

A growing body of research has demonstrated the particularly damaging effects of exposure, both contemporaneous and cumulative, to multiple forms of victimization, or “poly-victimization,” on the mental health of children and adolescents. This research has pointed to the need for a shift in orientation, both in research and practice. Researchers need to acknowledge that a focus on individual types of victimization without attention to other co-occurring types is likely to mis-specify associations with child outcomes and underestimate the full burden of victimization exposure (Turner et al., 2010c). Also, practitioners in clinical, educational, juvenile justice, and child protection contexts need to assess for a broad range of victimization types when presented with a specific incident and incorporate this knowledge in intervention responses. An important next step to further these efforts is to try to better understand the mechanisms that make poly-victimization so impactful and craft interventions to mitigate these effects. The present research begins to address this objective.

Findings support the hypothesis that poly-victimization is associated with reductions of social and personal resources. SEM analyses showed that, relative to youth with low levels of victimization in both W1 and W2, those experiencing increases in poly-victimization and those with high poly-victimization in both waves reported significantly greater reductions in family social support, friend social support, self-esteem, and mastery, controlling for sociodemographic factors. In contrast, those youth who experienced reductions in poly-victimization did not significantly differ from those with stable low levels in resources at W2. Thus, while findings suggest that poly-victimization is damaging to core psychosocial resources, it appears that reducing victimization may allow youth to recover lost resources, leading to improvements in family and friend support, self-esteem, and mastery.

Although poly-victimization impacted levels of all four resources, only self-esteem and mastery had independent effects on distress symptoms. That meant that the indirect effects of poly-victimization on distress were operating only through self-esteem and mastery. Although it is not clear why the family and friend social support factors were not related to distress in these analyses, it may be that they operate primarily through self-concept. Indeed, it has been suggested that one primary mechanism by which support increases well-being is through self-esteem enhancement (Cohen, Underwood, Gottlieb, & Fetzer Institute, 2000). Thus, supporters provide support, in part, by encouraging and reinforcing positive self-appraisals when individuals are faced with adversity. Consistent with this possibility, post hoc analyses of the model in Figure 1 without W2 self-esteem and mastery in the equation (not shown) found significant associations between both types of social support and W2 distress. Family support, in particular, showed strong mediating effects between poly-victimization and distress for youth in the high stable and increasing groups. Given the attenuation of the support-distress associations when self-concept variables were controlled, it is plausible that perceptions of social support by friends and family operate largely through their effects on self-esteem and mastery. Future research should examine in more detail the links between informal support and self-concept in the context of youth poly-victimization.

Results clearly point to the particularly damaging effects of poly-victimization on self-processes. In the case of self-esteem, it seems plausible that youth who are victimized by multiple perpetrators and in multiple contexts are especially likely to believe they are unworthy and disliked by others. Such “reflected appraisals” are likely to become an important basis for self-regard (Kinch, 1963). These findings are also consistent with research on peer victimization showing that youth exposed to multiple forms of peer victimization were more likely to make self-blame attributions for their own victimization (Raskauskas, 2010), given that self-blame and low self-esteem are strongly linked (Graham & Juvonen, 1998).

Poly-victimization had the most powerful effect on mastery, which in turn had a particularly strong effect on distress symptoms. Consistent with research on the impact of chronic adversity, such as that associated with poverty (Evans & Cassells, 2013), experiencing multiple forms of victimization appears to have substantial implications for youth’s perceptions of their own ability to control life outcomes. Low perceived personal control is typically rooted in objective life experiences (Mirowsky & Ross, 2003). It may be, as has been suggested with respect to individual forms of victimization, that those victimized as children may feel powerless in the actual victimization experiences and may internalize and generalize these perceptions (Gold, Sinclair, & Balge, 1999). Experiencing multiple forms of victimization in multiple life contexts is likely to accelerate these self-processes. When youth are victimized, for example, at home by family members, peers at school, and by other individuals in their neighborhoods, they are not only faced with “evidence” that avoiding toxic experiences is beyond their control, but they are left without any context in which they may gain a sense of personal efficacy. Although cross-context victimization erodes mastery, the results suggest some good news as well. Mastery appears to be sensitive to changes in poly-victimization in both directions. When poly-victimization is stopped, youth can gain back their sense of personal control and as a result experience improvements in mental health.

Despite the significant indirect effects found through self-esteem and mastery, it is important to point out that the direct effect of poly-victimization on youth mental health, particularly when sustained over time, was considerable. Independent of the role of psychosocial resources, exposure to multiple forms of victimization, whether it was stable across the two waves or was a result of increased victimization, was in itself damaging to youth well-being. This may be due, for example, to high levels of emotional arousal, such as fear or anger, that arise directly from poly-victimization. Although other mechanisms not yet identified may be at work, poly-victimization appears to represent a highly consequential set of adverse life experiences, even in the short term.

The present research focused on the significance of poly-victimization in reducing levels of resources, which in turn helped to explain higher levels of distress among poly-victims. Thus, we focused on the mediating effects of resources. However, such resources have also been implicated as potential buffers of adversity, moderating the negative effects of stress. Moderating effects imply that resources interact with adversity levels such that their effects are stronger in the context of higher adversity. For example, social support may not have direct independent effects on distress but may instead function to reduce the negative effects of poly-victimization on distress. It would be useful in future research to examine longer causal chains involving victimization and resources over time to determine whether reductions in support and self-concept due to poly-victimization also function to reduce capacities of resources to moderate or buffer subsequent victimization exposure.

The current findings may contribute to a broader understanding of the processes that help to explain long-term impact of childhood adversity on health over the life course. Adverse childhood experiences, such as victimization, especially when they characterize life conditions across multiple domains, damage core psychosocial resources that not only contribute to mental health problems but likely also increase risk of subsequent adversity. That is, resource deficiencies, such as low family support, peer rejection, low self-esteem, and a sense of mastery, may in turn increase the likelihood of subsequent victimization. Research has shown, for example, that coping strategies that involve self-blame and that fail to utilize social support increase vulnerability to re-victimization (Arata, 2000). Moreover, mental health problems, such as youth internalizing and externalizing symptomatology, both well-documented outcomes of victimization exposure and resource deficits, are also significantly associated with subsequent victimization exposure (Turner, Finkelhor, & Ormrod, 2010a). Thus, erosion of social and personal resources, such as support and self-concept, likely represent an important part of the chain of risk factors that arise from multiple victimization and that contribute to a cycle of mental health problems and re-victimization over time.

Limitations

A number of limitations of the present study should be acknowledged. Self-report measures of victimization have the potential advantage of maximizing the reporting of events that occur across multiple life contexts, as only the respondent himself or herself is likely to be privy to experiences in all arenas. However, self-reports of both victimization exposure and symptoms can lead to inflated associations due to method covariance (McGee, Wolfe, Yuen, Wilson, & Carnachan, 1995). In addition, a couple of the resource measures used in these analyses exhibited lower than ideal levels of internal consistency. Family support, for example, had an alpha coefficient of .67 at Wave 1 and .73 at Wave 2. Thus, the lack of a strong association between family support and distress may, to some extent, reflect the measure’s lower reliability, which typically reduces the magnitude of associations (Zeller & Carmines, 1980). Mastery, which had similar sized alpha coefficients, may have shown even stronger associations with poly-victimization and distress had reliability been higher.

Another potential problem in many studies of stress exposure such as this one is that individuals may sometimes “telescope” more distal events into the study frame period. Although another study using the same victimization measures found little evidence of telescoping bias (Finkelhor, Ormrod, & Turner, 2007c), telescoping can inflate rates of past year victimization and possibly confound Wave 1 and 2 poly-victimization assessments. To the extent that distressed youth are particularly likely to telescope victimization events into the past year time frame, associations between distress and poly-victimization would also be inflated.

Practice Implications

It is clear that prevention efforts that keep youth from being exposed to victimization, particularly multiple forms of victimization, would yield the greatest benefits. However, given the substantial number of youth in the United States who experience poly-victimization (Finkelhor et al., 2007b; Turner et al., 2010c), the development of effective interventions that seek to reduce or treat its damaging effects also represent important goals. Findings confirm the importance of treatment strategies that address self-blame for victimization and focus on cognitive processes leading to low self-esteem and low perception of personal control. Efforts to mitigate the detrimental effects of poly-victimization require that young people develop ways to maintain a general sense of mastery and self-regard despite their victimization experiences. Such efforts may involve helping youth find social contexts and circumstances in which they may exercise actual control and mastery, as well as ways of viewing their circumstances that emphasize situational determinants rather than personal shortcomings.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by Grant 2010-JF-FX-0001, awarded by the Office of Juvenile Justice and Delinquency Prevention, Office of Justice Programs, U.S. Department of Justice.

Author Biographies

Heather A. Turner is a professor of sociology and senior research associate at the Crimes Against Children Research Center (CCRC) at the University of New Hampshire (UNH). Her research program has concentrated on social stress processes and mental health, including the effects of violence, victimization, and other forms of adversity on the social and psychological development of children and adolescents. She has 15 years of research experience on childhood exposure to violence, has conducted numerous national surveys, and has published more than 75 articles, many focusing on the epidemiology of childhood victimization and mental health. She is currently co-principal investigator for the Office of Juvenile Justice and Delinquency Prevention (OJJDP)–funded National Surveys of Children’s Exposure to Violence (NatSCEV I, NatSCEV II and NatSCEV III)—studies designed to obtain comprehensive estimates of children’s exposure to multiple forms of violence and victimization across the full developmental spectrum (age 0-17). She is the director of the International Conference on Social Stress Research and past chair of the Sociology of Mental Health Section of the American Sociological Association (ASA).

Anne Shattuck is a research scientist and data analyst at the University of New Hampshire’s Crimes Against Children Research Center (CCRC) where she is responsible for the data management and analysis of three cohorts of the National Survey of Children’s Exposure to Violence (2008, 2011, and 2014). She holds a BS from Georgetown University, an MA from the University of New Hampshire, and will complete her PhD in Sociology at UNH in 2015. She has received advanced training in statistical methods including structural equation modeling, longitudinal data analysis, multi-level modeling and latent class analysis. She has conducted data analysis for numerous CCRC publications. Before joining CCRC, she worked as a research assistant at the University of New Hampshire’s Carsey School of Public Policy (formerly Carsey Institute).

David Finkelhor is director of CCRC, co-director of the Family Research Laboratory, professor of sociology, and university professor at UNH. He has been studying the problems of child victimization, child maltreatment, and family violence since 1977. He is well known for his conceptual and empirical work on the problem of child sexual abuse, reflected in publications such as Sourcebook on Child Sexual Abuse (SAGE, 1986) and Nursery Crimes (SAGE, 1988). He has also written about child homicide, missing and abducted children, children exposed to domestic and peer violence, and other forms of family violence. In his recent work, for example, his book Child Victimization (Oxford University Press, 2008), he has tried to unify and integrate knowledge about all the diverse forms of child victimization in a field he has termed Developmental Victimology. This book received the Daniel Schneider Child Welfare Book of the Year award in 2009. Altogether, he is editor and author of 12 books and more than 200 journal articles and book chapters. He has received grants from the National Institute of Mental Health, the National Center on Child Abuse and Neglect, and the U.S. Department of Justice, and a variety of other sources. In 1994, he was given the Distinguished Child Abuse Professional Award by the American Professional Society on the Abuse of Children; in 2004, he was given the Significant Achievement Award from the Association for the Treatment of Sexual Abusers; in 2005, he and his colleagues received the Child Maltreatment Article of the Year award; in 2007, he was elected as a Fellow of the American Society of Criminology; and in 2014, he was awarded the National Scientific Impact Award from the Kempe Center for the prevention and treatment of Child Abuse and Neglect.

Sherry Hamby is a research professor of psychology at the University of the South and director of the Appalachian Center for Resilience Research. She is also founding editor of the American Psychological Association (APA) journal Psychology of Violence. A licensed clinical psychologist, she has worked for more than 20 years on the problem of violence, including front-line crisis intervention for domestic and other violence, involvement in grassroots domestic violence organizations, therapy with trauma survivors, and research on many forms of violence. She is co-investigator on the National Survey of Children’s Exposure to Violence, which is the United States’ primary surveillance of youth victimization and the first national effort to measure crimes against children under 12 that are not reported to authorities. She is the recipient of numerous honors and author or co-author of more than 100 works including The Web of Violence: Exploring Connections Among Different Forms of Interpersonal Violence and Abuse. Her work has appeared in the New York Times, Huffington Post, the Christian Science Monitor, and hundreds of other media outlets. Her most recent book is Battered Women’s Protective Strategies: Stronger Than You Know (Oxford University Press, 2014).

References

Aneschensel

C. S.

Sucoff

C. A.

(1996). The neighborhood context of adolescent mental health. Journal of Health and Social Behavior, 37, 293-310.

Arata

C. M.

(2000). From child victim to adult victim: A model for predicting sexual revictimization. Child Maltreatment, 5, 28-38. doi:10.1177/1077559500005001004

Bagley

Ramsay

(1986). Sexual abuse in childhood. Journal of Social Work & Human Sexuality, 4(1-2), 33-47. doi:10.1300/J291v04n01_07

Bandura

(1982). Self-efficacy mechanism in human agency. American Psychologist, 37, 122-147.

Benas

J. S.

Gibb

B. E.

(2007). Peer victimization and depressive symptoms: The role of body dissatisfaction and self-esteem. Journal of Cognitive Psychotherapy, 21, 107-116.

Bentler

P. M.

(1990). Comparative fit indexes in structural analysis. Psychological Bulletin, 102, 238-246.

Blascovich

Tomaka

(1991). Measures of self-esteem. In Robinson

J. P.

Shaver

P. R.

Wrightsman

L. S.

(Eds.), Measures of personality and social psychological attitudes (Vol. 1, pp. 115-160). San Diego, CA: Academic Press.

Bolger

K. E.

Patterson

C. J.

(2001). Developmental pathways from child maltreatment to peer rejection. Child Development, 72, 549-568.

Bolstad

B. R.

Zinbarg

R. E.

(1997). Sexual victimization, generalized perception of control, and posttraumatic stress disorder symptom severity. Journal of Anxiety Disorders, 11, 523-540. doi:10.1016/S0887-6185(97)00028-5

10.

Briere

(1996). Trauma Symptoms Checklist for Children (TSCC): Professional manual. Odessa, FL: Psychological Assessment Resources.

11.

Briere

Johnson

Bissada

Damon

Crouch

Gil

. . . Ernst

(2001). The Trauma Symptom Checklist for Young Children (TSCYC): Reliability and association with abuse exposure in a multi-site study. Child Abuse & Neglect, 25, 1001-1014.

12.

Cohen

Underwood

L. G.

Gottlieb

B. H.

, & Fetzer Institute (2000). Social support measurement and intervention: A Guide for health and social scientists. Oxford, UK: Oxford University Press.

13.

Cyr

Chamberland

Clément

M.-È.

Lessard

Wemmers

J.-A.

Collin-Vézina

. . . Damant

(2013). Polyvictimization and victimization of children and youth: Results from a populational survey. Child Abuse & Neglect, 37, 814-820. doi:10.1016/j.chiabu.2013.03.009

14.

Demaray

M. K.

Malecki

C. K.

(2003). Perceptions of the frequency and importance of social support by students classified as victims, bullies, and bully/victims in an urban middle school. School Psychology Review, 32, 471-489.

15.

Elliott

A. N.

Alexander

A. A.

Pierce

T. W.

Aspelmeier

J. E.

Richmond

J. M.

(2009). Childhood victimization, poly-victimization, and adjustment to college in women. Child Maltreatment, 14, 330-343. doi:10.1177/1077559509332262

16.

Evans

G. W.

Cassells

R. C.

(2014). Childhood poverty, cumulative risk exposure, and mental health in emerging adults. Clinical Psychological Science, 2, 287-296. doi:10.1177/2167702613501496

17.

Ezzell

C. E.

Swenson

C. C.

Brondino

M. J.

(2000). The relationship of social support to physically abused children’s adjustment. Child Abuse & Neglect, 24, 641-651. doi:10.1016/S0145-2134(00)00123-X

18.

Felitti

V. J.

Anda

R. F.

Nordenberg

Williamson

D. F.

Spitz

A. M.

Edwards

. . . Marks

J. S.

(1998). Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults: The Adverse Childhood Experiences (ACE) Study. American Journal of Preventive Medicine, 14, 245-258.

19.

Finkelhor

Ormrod

R. K.

Turner

H. A.

(2007a). Poly-victimization and trauma in a national longitudinal cohort. Development and Psychopathology, 19, 149-166.

20.

Finkelhor

Ormrod

R. K.

Turner

H. A.

(2007b). Poly-victimization: A neglected component in child victimization. Child Abuse & Neglect, 31, 7-26. doi:10.1016/j.chiabu.2006.06.008

21.

Finkelhor

Ormrod

R. K.

Turner

H. A.

(2007c). Revictimization patterns in a national longitudinal sample of children and youth. Child Abuse & Neglect, 31, 479-502.

22.

Finkelhor

Ormrod

R. K.

Turner

H. A.

(2009). Lifetime assessment of poly-victimization in a national sample of children and youth. Child Abuse & Neglect, 33, 403-411.

23.

Finkelhor

Ormrod

R. K.

Turner

H. A.

Hamby

S. L.

(2005a). Measuring poly-victimization using the JVQ. Child Abuse & Neglect, 29, 1297-1312.

24.

Finkelhor

Ormrod

R. K.

Turner

H. A.

Hamby

S. L.

(2005b). The victimization of children and youth: A comprehensive, national survey. Child Maltreatment, 10, 5-25.

25.

Flaspohler

P. D.

Elfstrom

J. L.

Vanderzee

K. L.

Sink

H. E.

Birchmeier

(2009). Stand by me: The effects of peer and teacher support in mitigating the impact of bullying on quality of life. Psychology in the Schools, 46, 636-649.

26.

Ford

J. D.

Elhai

J. D.

Connor

D. F.

Frueh

B. C.

(2010). Poly-victimization and risk of posttraumatic, depressive, and substance use disorders and involvement in delinquency in a national sample of adolescents. Journal of Adolescent Health, 46, 545-552.

27.

Ford

J. D.

Grasso

D. J.

Hawke

Chapman

J. F.

(2013). Poly-victimization among juvenile justice-involved youths. Child Abuse & Neglect, 37, 788-800. doi:10.1016/j.chiabu.2013.01.005

28.

Furlong

M. J.

Chung

(1995). Who are the victims of school violence? A comparison of student non-victims and multi-victims. Education & Treatment of Children, 18, 282-298.

29.

Gabalda

M. K.

Broth

M. R.

Thompson

M. P.

Kaslow

N. J.

(2009). Children’s emotional abuse and relational functioning: Social support and internalizing symptoms as moderators. Journal of Child & Adolescent Trauma, 2, 179-197. doi:10.1080/19361520903120228

30.

Gold

S. R.

Sinclair

B. B.

Balge

K. A.

(1999). Risk of sexual re-victimization: A theoretical model. Aggression and Violent Behavior, 4, 457-470.

31.

Graham

Juvonen

(1998). Self-blame and peer victimization in middle school: An attributional analysis. Developmental Psychology, 34, 587-599.

32.

Hickman

L. J.

Jaycox

L. H.

Setodji

C. M.

Kofner

Schultz

Barnes-Proby

Harris

(2013). How much does “how much” matter? Assessing the relationship between children’s lifetime exposure to violence and trauma symptoms, behavior problems, and parenting stress. Journal of Interpersonal Violence, 28, 1338-1362. doi:10.1177/0886260512468239

33.

Holt

M. A.

Finkelhor

Kaufman Kantor

(2007). Multiple victimization experiences of urban elementary school students: Associations with psychosocial functioning and academic performance. Child Abuse & Neglect, 31, 503-515.

34.

Holt

M. K.

Espelage

D. L.

(2007). Perceived social support among bullies, victims, and bully-victims. Journal of Youth and Adolescence, 36, 984-994. doi:10.1007/s10964-006-9153-3

35.

Kinch

J. W.

(1963). A formalized theory of the self-concept. American Journal of Sociology, 68, 481-486. doi:10.2307/2774427

36.

Kline

R. B.

(2011). Principles and practice of structural equation modeling (3rd ed.). New York, NY: Guilford Press.

37.

McGee

R. A.

Wolfe

D. A.

Yuen

S. A.

Wilson

S. K.

Carnachan

(1995). The measurement of maltreatment: A comparison of approaches. Child Abuse & Neglect, 19, 233-249.

38.

Mercy

J. A.

Saul

(2009). Creating a healthier future through early interventions for children. The Journal of the American Medical Association, 301, 2262-2264.

39.

Merrill

L. L.

Thomsen

C. J.

Sinclair

B. B.

Gold

S. R.

Milner

J. S.

(2001). Predicting the impact of child sexual abuse on women: The role of abuse severity, parental support, and coping strategies. Journal of Consulting and Clinical Psychology, 69, 992-1006. doi:10.1037/0022-006X.69.6.992

40.

Mirowsky

Ross

C. E.

(2003). Social causes of psychological distress (2nd ed.). New Brunswick, NJ: Aldine Transaction.

41.

Newman

B. M.

Newman

P. R.

Griffen

O’Connor

Spas

(2007). The relationship of social support to depressive symptoms during the transition to high school. Adolescence, 42, 441-459.

42.

Pearlin

Bierman

(2013). Current issues and future directions in research into the stress process. In Aneshensel

Phelan

Bierman

(Eds.), Handbook of the sociology of mental health (pp. 325-340). Dordrecht, The Netherlands: Springer.

43.

Pearlin

L. I.

Schooler

(1978). The structure of coping. Journal of Health and Social Behavior, 19, 2-21.

44.

Pepin

E. N.

Banyard

V. L.

(2006). Social support: A mediator between child maltreatment and developmental outcomes. Journal of Youth and Adolescence, 35, 612-625. doi:10.1007/s10964-006-9063-4

45.

Raskauskas

(2010). Multiple peer victimization among elementary school students: Relations with social-emotional problems. Social Psychology of Education, 13, 523-539. doi:10.1007/s11218-010-9124-0

46.

Rigby

(2000). Effects of peer victimization in schools and perceived social support on adolescent well-being. Journal of Adolescence, 23, 57-68. doi:10.1006/jado.1999.0289

47.

Rosenberg

(1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press.

48.

Ross

C. E.

Mirowsky

(2013). The sense of personal control: Social structural causes and emotional consequences. In Aneshensel

Phelan

Bierman

(Eds.), Handbook of the sociology of mental health (pp. 379-402). Dordrecht, The Netherlands: Springer.

49.

Rotter

J. B.

(1966). Generalized expectancies for internal vs. external control of reinforcement. Psychological Monographs, 80, 1-28.

50.

Schultz

Tharp-Taylor

Haviland

Jaycox

(2009). The relationship between protective factors and outcomes for children investigated for maltreatment. Child Abuse & Neglect, 33, 684-698. doi:10.1016/j.chiabu.2009.04.004

51.

Shonkoff

J. P.

Boyce

W. T.

McEwen

B. S.

(2009). Neuroscience, molecular biology, and the childhood roots of health disparities: Building a new framework for health promotion and disease prevention. The Journal of the American Medical Association, 301, 2252-2259.

52.

Simmons

J. T.

Weinman

M. L.

(1991). Self-esteem, adjustment, and locus of control among youth in an emergency shelter. Journal of Community Psychology, 19, 277-280.

53.

Soler

Kirchner

Paretilla

Forns

(2013). Impact of poly-victimization on mental health: The mediator and/or moderator role of self-esteem. Journal of Interpersonal Violence, 28, 2695-2712. doi:10.1177/0886260513487989

54.

Steiger

J. H.

(1990). Structural model evaluation and modification: An interval estimation approach. Multivariate Behavioral Research, 25, 173-180.

55.

Thoits

P. A.

(2013). Self, identity, stress, and mental health. In Aneshensel

Phelan

Bierman

(Eds.), Handbook of the sociology of mental health (pp. 357-377). Dordrecht, The Netherlands: Springer.

56.

Tucker

L. R.

Lewis

(1973). A reliability coefficient for maximum likelihood factor analysis. Psychometrika, 38, 1-10.

57.

Turner

H. A.

Finkelhor

Ormrod

(2010a). Child mental health problems as risk factors for victimization. Child Maltreatment, 15, 132-143.

58.

Turner

H. A.

Finkelhor

Ormrod

(2010b). The effects of adolescent victimization on self-concept and depressive symptoms. Child Maltreatment, 15, 76-90.

59.

Turner

H. A.

Finkelhor

Ormrod

(2010c). Poly-victimization in a national sample of children and youth. American Journal of Preventive Medicine, 38, 323-330. doi:10.1016/j.amepre.2009.11.012

60.

Turner

H. A.

Finkelhor

Ormrod

Hamby

S. L.

Leeb

R. T.

Mercy

J. A.

Holt

M. A.

(2012). Family context, victimization, and child trauma symptoms: Variations in safe, stable, and nurturing relationships during early and middle childhood. American Journal of Orthopsychiatry, 82, 209-219. doi:10.1111/j.1939-0025.2012.01147.x

61.

Turner

H. A.

Shattuck

Hamby

S. L.

Finkelhor

(2013). Community disorder, victimization exposure, and mental health in a national sample of youth. Journal of Health and Social Behavior, 54, 258-275. doi:10.1177/0022146513479384

62.

Turner

J. B.

Turner

R. J.

(2013). Social relations, social integration, and social support. In Aneshensel

Phelan

Bierman

(Eds.), Handbook of the sociology of mental health (pp. 341-356). Dordrecht, The Netherlands: Springer.

63.

Turner

R. J.

Roszell

(1994). Psychosocial resources and the stress process. In Avison

W. R.

Gotlib

I. H.

(Eds.), Stress and mental health: Contemporary issues and prospects for the future (pp. 179-210). New York, NY: Plenum Press.

64.

Valentine

Feinauer

L. L.

(1993). Resilience factors associated with female survivors of childhood sexual abuse. The American Journal of Family Therapy, 21, 216-224. doi:10.1080/01926189308250920

65.

Wheaton

(1983). Stress, personal copying resources, and psychiatric symptoms: An investigation of interactive models. Journal of Health and Social Behavior, 24, 208-229.

66.

Wright

E. M.

Fagan

A. A.

Pinchevsky

G. M.

(2013). The effects of exposure to violence and victimization across life domains on adolescent substance use. Child Abuse & Neglect, 37, 899-909. doi:10.1016/j.chiabu.2013.04.010

67.

Wun

L.-M.

Ezzati-Rice

T. M.

DiGaetano

Goksel

Hongsheng

(2005). Using data from the National Health Interview Survey (NHIS) to assess the effectiveness of nonresponse adjustment in the Medical Expenditure Panel Survey (MEPS). Proceedings of the American Statistical Association, Survey Research Methods Section, 2005, 3695-3700.

68.

Zeller

R. A.

Carmines

E. G.

(1980). Measurement in the social sciences: The link between theory and data. New York, NY: Cambridge University Press.

69.

Zimet

G. D.

Dahlem

N. W.

Zimet

S. G.

Farley

G. K.

(1988). The multidimensional scale of perceived social support. Journal of Personality Assessment, 52, 30-41.

70.

Zimmer-Gembeck

M. J.

Hunter

T. A.

Pronk

(2007). A model of behaviors, peer relations and depression: Perceived social acceptance as a mediator and the divergence of perceptions. Journal of Social & Clinical Psychology, 26, 273-302.