Self-Esteem in Children Exposed to Intimate Partner Violence: a Critical Review of the Role of Sibling Relationships and Agenda for Future Research

Abstract

This paper provides a critical analysis of research on self-esteem in children exposed to intimate partner violence (IPV), with particular emphasis on the role of sibling relationships. Over 100 empirical and review articles published from 2000 to 2015 examining self-esteem in children exposed to IPV, the role of sibling relationships in the development of self-esteem, and the sibling relationships of children exposed to IPV, were reviewed. Research findings show that self-esteem is an important component of understanding children’s positive adjustment and their adjustment difficulties. A substantial research literature also documents that siblings significantly influence one another’s self-esteem; however, the potential risk or protective function of sibling relationships in the development of self-esteem in children and adolescents exposed to IPV has only been minimally investigated. This is a serious gap in the literature, as this information has significant potential for informing prevention and intervention programming. Recommendations for future research were provided.

Introduction

According to a recent estimation, as many as 275 million children worldwide are exposed to intimate partner violence (IPV) every year (Pinheiro 2006). It is well documented that children exposed to IPV are at elevated risk for a variety of psychological, emotional, social, behavioral, and cognitive problems, both in the short term and over the life course (Hungerford et al. 2012; MacDonnell 2012). However, it is also widely acknowledged that there is considerable variability in adjustment among children exposed to IPV (Skopp et al. 2005). That is, although a relevant proportion of these children appear to experience moderate to severe adjustment problems, many show only mild levels of distress or no problems at all (Graham-Bermann et al. 2009; Grych et al. 2000; Piotrowski 2011).

Given the high degree of heterogeneity in adjustment of children exposed to IPV, researchers have recently begun to investigate key mechanisms that may account for this variability (Miller-Graff et al. 2015). Self-esteem has long been identified as a pivotal factor in children’s development that can mediate or moderate adjustment problems, as well as promote resilience in the face of adversity. Positive self-esteem has been linked to many important life outcomes, while low self-esteem is often related to adjustment difficulties (Ziegler-Hill 2013). For example, low self-esteem has been linked to internalizing problems in children, including depression, eating disorders (DeHart et al. 2013), anxiety (Sowislo and Orth 2013), and suicidal ideation (R. McGee and Williams 2000), to highlight only a few. Interestingly, the relationship between self-esteem and externalizing problems such as aggression, delinquency, and drug or alcohol abuse is less clear and still subject to debate (Bos et al. 2010). It is, therefore, surprising that self-esteem is not more widely studied in relation to children exposed to IPV, since variability in children’s self-esteem clearly has short and long-term potential implications for their individual adjustment. Little is currently known about the factors that affect self-esteem in children exposed to IPV, or the role self-esteem may play in mediating or moderating their adjustment difficulties, or contributing to their well-being and resilience.

Therefore, we conducted the present review of the literature with a threefold purpose: 1) to provide an overview of the nature of self-esteem in children and its linkages to adjustment difficulties, 2) to critically review what is known about the self-esteem of children exposed to IPV, and 3) to provide an in-depth critical review on the influence of sibling relationships on self-esteem of children exposed to IPV. Specifically, our third aim was to synthesize and integrate findings concerning the influence of siblings and self-esteem in children exposed to IPV, with the ultimate goal of informing directions for future interventions and research concerning children exposed to IPV. We chose to focus on the sibling relationship rather than on the parent-child relationship because IPV research conducted to date has focused almost exclusively on the mother-child relationship (Gewirtz et al. 2011). While we acknowledge the crucial significance of the parent-child relationship for children’s adjustment in general, and for their self-esteem in particular, recent work has emphasized the important influence of siblings in the adjustment of children exposed to IPV (Horn et al. 2013; Piotrowski et al. 2014). Therefore, we limited our review to studies examining links between the quality of sibling relationships and children’s self-esteem in families affected by IPV and those not affected.

Search Strategy

The literature review was structured to identify and synthesize research findings about the role of self-esteem in children’s adjustment, self-esteem in children exposed to IPV, and the influence of siblings on children’s self-esteem. We searched several online databases, including Psyc-INFO, Medline (PubMed), Google Scholar, ScienceDirect, and Scopus, for the following key terms: “self-esteem,” or “psychological adjustment,” or “adjustment,” and “intimate partner violence,” or “domestic violence,” or “violen*,” and “sibling*,” or “sibling* relationship*,” and “child*,” and “adolescen*.” The asterisk (i.e., the truncation symbol) allowed for the inclusion of alternate words ending of the search term (e.g., child* yielded articles containing child, children, and childhood). Finally, all reference lists of the articles included in the manuscript were reviewed in order to find additional articles that met the following criteria. We included relevant peer-reviewed journal articles, review articles, book chapters, meta-analyses, and student theses published between 2000 and 2015 in order to focus on the state of current knowledge on the topic. In total, 103 publications were included in the present review.

Self-Esteem

Generally, self-esteem is considered to be a person’s subjective evaluative self-knowledge that reflects the extent to which people view themselves as worthy competent individuals (Ziegler-Hill 2013). Researchers call this form of self-esteem global or trait self-esteem, as it is relatively stable across time and situations (Kuster et al. 2012). High global self-esteem characterizes people with a favorable view of the self who believe they are individuals of worth and value. Conversely, low global self-esteem characterizes people with an uncertain or negative evaluation of the self, who believe they are incompetent and have limited value.

In his classic work on global self-esteem, William James described it as a ratio between successes and pretensions, namely people’s accomplishments and their aspirations (James 1890). Following this theorization self-esteem can also be defined as a self-evaluative emotional reaction to significant events that yield feelings of self-worth. Many researchers use the term state self-esteem to describe such feelings of self-worth. Feeling proud or pleased after a job promotion on the positive side, or humiliated and ashamed after a divorce on the negative side, are examples of state self-esteem feelings.

Morris Rosenberg added that self-esteem involves feelings of self-respect and self-acceptance, and that it can be understood in terms of worthiness (Rosenberg 1965). Specifically, he defined self-esteem as a positive or negative attitude toward the self. According to Rosenberg, individuals with high self-esteem respect themselves, consider themselves worthy, recognize their limitations, and expect to grow and improve. In contrast, those with low self-esteem lack respect for the self and show feelings of rejection, contempt, and dissatisfaction with the self.

The construct of self-esteem is also used to describe the way individuals evaluate their various abilities and attributes (Brown and Marshall 2006). These evaluations or appraisals of life aspects are domain-specific, as they concern evaluations of particular domains or facets of the self, such as academic or athletic competence. Therefore, the term domain specific self-esteem is used to describe a specific type of self-esteem that reflects the way individuals evaluate their competence or value in a specific life domain.

Even though researchers continue to debate the relative utility of a unidimensional versus a multidimensional perspective of self-esteem, there is a widespread agreement that self-esteem represents a fundamental human need. According to this perspective, individuals try to reach high-level self-esteem across the life course in all domains of life. Indeed, boosts to self-esteem are reported to be one of the reasons why individuals classify their life events as satisfying and meaningful (Ziegler-Hill 2013).

Self-Esteem Development from Childhood to Adolescence

Current perspectives view the development of self-esteem from childhood through adolescence as an ongoing complex interaction of processes between children’s genetic make-up and their family, social, and cultural context (Trzesniewski et al. 2013). In fact, even though many classic developmental theories emphasize the role of environmental factors, such as relationships with parents and peers, in the construction of a mature sense of self (Harter 2012); a biological component of self-esteem has also been increasingly recognized. Generally, heritability estimates for self-esteem range widely, from 29% to 73% (Saphire-Bernstein et al. 2011), with some twin studies showing that genetic factors account for as much as 40% of the observed variability in self-esteem among same age individuals (Neiss et al. 2002). In studying the development of self-esteem, researchers have adopted two main measurement approaches that provide a helpful background for understanding current findings relating to self-esteem in children exposed to IPV: rank order stability and mean level changes (Trzesniewski et al. 2013).

Rank Order Stability

Rank order stability reflects the degree to which the relative ordering of individuals on a scale that measures self-esteem is maintained over time, and it is typically assessed estimating the correlation between self-esteem scores obtained at two points in time. It is the extent to which individual differences in self-esteem levels remain constant over-time. Researchers have found a robust curvilinear trend in the rank order stability of self-esteem. Specifically, stability is relatively low during childhood, increases throughout adolescence and early adulthood, and then declines during midlife and later stages of life (Trzesniewski et al. 2013). This curvilinear trend has been replicated over the years across gender, ethnicity, nationality, and the use of different self-esteem measures.

Despite the persistence of a debate (Donnellan et al. 2012), the findings about self-esteem rank order stability across the life span have led researchers to conclude that self-esteem is a stable personality trait in adulthood, but not in childhood (Trzesniewski et al. 2013). Specifically, a meta-analysis that examined the rank order stability of self-esteem across the life span found moderate test-retest correlations for self-esteem comparable to those found for personality traits (e.g., the Big Five) (Trzesniewski et al. 2003). The evidence provided by research shows a lower rank order stability for self-esteem during childhood and a higher rank order stability for self-esteem during adolescence and later in life, supporting the conclusion that self- esteem is more stable in adulthood. The higher variability in children’s self-esteem has important implications for both research and interventions concerning self-esteem of children exposed to IPV.

Mean Level Changes

Mean level changes in self-esteem are evaluated by comparing mean differences between groups obtained on self-esteem scales, and can be investigated using a variety of research designs including cross-sectional, longitudinal, or cohort-sequential designs. Since rank order stability and mean level changes are quite independent from each other, they can function very differently within the same sample. For instance, in the transition from childhood to adolescence, mean levels of self-esteem can decrease significantly, but if everyone decreases by the same amount during that period, individuals’ rank order is maintained.

Research findings show significant mean level differences in self-esteem among individuals of different ages (Huang 2010). Specifically, mean levels in self-esteem across the life span follow a reliable quadratic curvilinear trend from childhood to adulthood (Harter 2012; Orth et al. 2012, 2010). Self-esteem levels are higher during childhood, decline during the transition to adolescence, increase during adolescence and adulthood reaching a peak in middle adulthood, and decrease again in late adulthood. Some researchers believe the high mean levels of self-esteem observed in childhood are the result of inflated self-evaluation due to children’s cognitive limitations that affect the accuracy of their assessment (Harter 2012). According to this perspective, the high levels of self-esteem observed in young children are due to their egocentrism and limited understanding of the complex and multidimensional self-esteem construct. Meanwhile, the exceptional physical, intellectual, emotional, and social changes that take place during adolescence delineate a period of sudden change and challenge that explain the decrease in self-esteem observed in the transition from childhood to adolescence (Trzesniewski et al. 2003). Greater fluctuations in self-image are more common in early adolescence and gradually diminish (Shahar and Henrich 2010).

Self-Esteem & Psychological Adjustment

A vast body of research has shown a positive correlation between self-esteem and important life outcomes (Swann et al. 2007; Zeigler-Hill and Wallace 2012). Individuals with high self-esteem report lower stress levels in response to critical life events and a more positive adjustment to transitions, showing fewer psychological difficulties. Conversely, individuals with low self-esteem report higher levels of stress in facing critical life events, and poor psychological adjustment, including depression, anxiety, eating disorders, poor school performance, aggression, delinquency and antisocial behaviors, and health-compromising behaviors such as alcohol and marijuana use (Baumeister et al. 2003; Orth et al. 2012; Sowislo and Orth 2013). Even though correlations between low self-esteem and adjustment difficulties are common, the findings are still mixed and occasionally inconclusive, and the precise nature of these correlations has not ultimately been established (Baumeister et al. 2003).

Empirical research and diagnostic criteria suggest a robust association between self-esteem and both the development and expression of psychopathology (Crocker and Park 2004; Zeigler-Hill 2011). The link between self-esteem and psychopathology is evident in the DSM-V (American Psychiatric Association 2013), where low self-esteem appears as a diagnostic criterion or concomitant characteristic of a vast range of disorders, including most mood and anxiety disorders. Due to their high prevalence and incidence in the general population, most of the research examining the connections between self-esteem and psychopathology has focused on mood and anxiety disorders, showing strong correlations between depression and anxiety symptoms and low self-esteem in children and adolescents (Abela et al. 2012; Orth et al. 2009; Sowislo and Orth 2013).

Vulnerability Model & Scar Model

The vulnerability model postulates that high self-esteem is a protective factor from potential consequences of negative life experiences, whereas low self-esteem may increase the probability of poor adjustment in the wake of stressful or negative experiences (Zeigler-Hill 2011). This pattern is believed to emerge because individuals with lower self-esteem possess fewer coping resources than those with higher self-esteem. This is known as the stress-buffering model, which proposes that high self-esteem “buffers” individuals from the deleterious consequences of stress, while low self-esteem increases vulnerability to the effects of stress. That is, those with high self-esteem are thought to be less affected by negative experiences and to recover from these sorts of experiences more quickly than individuals with low self-esteem (Orth et al. 2009; Orth et al. 2008). In order to test the vulnerability model, numerous studies have investigated the prospective effect of self-esteem on depression across the life span. A longitudinal study involving a multi-method assessment based on both self-report and informant report found that self-esteem scores in early adolescence (aged 11 to 15) predicted depression by age 26, controlling for previous depression levels, sex, and socioeconomic status (Trzesniewski et al. 2006).

In contrast to the vulnerability model, the scar model suggests that low self-esteem is a consequence of psychopathology, rather than one of its causes. According to this model, psychological disorders tend to erode psychological resources and leave “scars” that distort how individuals feel about themselves (Zeigler-Hill 2011). Studies examining the scar model suggest that various forms of psychopathology may interfere with psychological adjustment preventing people from functioning successfully in their daily lives.

It should be noted that the vulnerability model and the scar model are not mutually exclusive. In fact, it is quite possible that both processes may operate simultaneously and reciprocally (Harter 2012). The vulnerability and scar models have both been largely applied to the study of depressive symptoms over the past decades: Thus far, there is no unequivocal evidence in favor of one model or the other (Abela et al. 2006).

Secure & Fragile Self-Esteem

Secure and fragile self-esteem can also play a role in children’s adjustment problems. This perspective suggests that some positive feelings of self-worth are fragile and vulnerable, and are associated with different types of self-protection or self-enhancement strategies (Kernis and Paradise 2002). Conversely, secure and well-anchored self- esteem does not require continual validation, is not highly vulnerable to threats, and reflects assertive feelings of self-worth associated to positive psychological adjustment and well-being.

Theory and research evidence suggest that four markers distinguish between secure and fragile high self-esteem (Kernis 2003). First, it is important to discriminate between genuine and defensive high self-esteem, which traditionally have been distinguished by responses to socially desirable measures. Defensive high self-esteem represents as a combination of interior negative self-feelings publicly presented positively because of a great desire to be accepted by others.

Second, it is crucial to identify people who report high self-esteem but who simultaneously hold unconscious negative feelings of self-worth. These unrecognized feelings of self-worth have been defined as implicit self-esteem. Implicit self-esteem can be assessed through self-implicit measures, and it is theoretically and empirically distinct from the conscious feelings of self-worth assessed through self-reports and defined as explicit self-esteem. Discrepancies between explicit and implicit self-esteem have been found to be detrimental to physical health (Schröder-Abé et al. 2007), and are linked to depressive symptoms (Franck et al. 2007), anger (Schröder-Abé et al. 2007), perfectionism (Zeigler-Hill and Terry 2007), and narcissism (Bosson et al. 2008). A recent study has examined the longitudinal correlations between peer victimization and children’s damaged self-esteem, that is high implicit and low explicit self-esteem (Leeuwis et al. 2015). Results showed that damaged self-esteem at age 11 predicted an increase in internalizing problems over time and mediated the relationship between peer victimization and the development of later internalizing problems.

A third relevant distinction exists between contingent high self-esteem and true high self-esteem. Contingent self-esteem refers to feelings about oneself that result from and are dependent on standards of excellence, and are strictly dependent on the achievement of such standards and on the endorsement of others. People with high contingent self-esteem are highly concerned about their accomplishments and about where they stand on specific evaluative dimensions, and they place great importance on how they are viewed by others (Kernis 2003). Research has found associations between contingent self-esteem and several psychological problems, including depression in adolescence (Burwell and Shirk 2006), anxiety (Bos et al. 2010), and narcissism (Zeigler-Hill et al. 2008).

A fourth marker to distinguish between secure and fragile high self-esteem is the extent to which a person’s feelings of self-worth fluctuate across time and situations (Kernis 2003). These short-term fluctuations in one’s contextually based feelings of self-worth reflect the degree to which one’s self-esteem is stable or unstable. Self-esteem instability is characterized by enhanced sensitivity to evaluative events, increased concern about self-image, and over-reliance on social sources of evaluation (Kernis 2005). Self-esteem instability has been found to predict depression (Roberts 2006), aggressive behavior (Boden et al. 2007), and bipolar disorder (Knowles et al. 2007).

A considerable body of research supports the usefulness of discriminating between secure self-esteem and these four types of high fragile self-esteem (Zeigler-Hill and Wallace 2012). In a recent investigation, Zeigler-Hill and Wallace (2012) examined the psychological adjustment of over 700 young adults with stable and unstable self-esteem across three studies. Unstable self-esteem was found to moderate the associations between self-esteem levels and global distress, depression, hopelessness, rejection sensitivity, negative affect, anxiety, hostility, interpersonal sensitivity, hostility, verbal and physical aggression, and psychological well-being. Another recent study on adolescents examined the relationships between global, contingent, and implicit self-esteem, on one side, and psychopathological symptoms on the other (Bos et al. 2010). The interaction between contingent and global self-esteem significantly predicted depressive symptoms, suggesting that adolescents with lower global self-esteem and higher contingent self-esteem experienced relatively more depression. Furthermore, unique and interactive effects of contingent self-esteem on anxiety and eating disorders were found. That is, adolescents who showed a combination of higher contingent self-esteem and lower global self-esteem reported the highest levels of anxiety and eating problems.

Sandstrom and Jordan (2008) conducted one of the few studies investigating defensive self-esteem and aggression in childhood. They hypothesized that children with higher levels of defensive self-esteem, characterized by high explicit and low implicit self-esteem, would engage in higher levels of aggressive behaviors. Children completed measures of both implicit and explicit self-esteem, while teachers assessed children’s physical and relational aggression in the school setting. Findings showed no direct association between self-esteem and aggression when explicit self-esteem was considered on its own. However, the relation between self-esteem and aggression emerged clearly when children’s levels of explicit and implicit self-esteem were considered together. Specifically, children with higher explicit self-esteem engaged in more aggressive behaviors only when implicit self-esteem was low (Sandstrom and Jordan 2008).

Even though the most recent research on fragile high self-esteem, including defensive, implicit, contingent, and unstable self-esteem, has shown promising results toward a better understanding of the relations between self-esteem and adjustment problems in childhood and adolescence, it is important to note that considerable inconsistencies have emerged across studies (Roberts 2006). These inconsistencies may be due to the common use of cross-sectional designs and heavy reliance on correlational findings, which preclude causal conclusions. Given the subjective nature of self-esteem, it is commonly assessed through self-report instruments that, even if generally accurate, can be affected by numerous biases associated with social desirability, defensiveness, and shared method variance. Research across samples of children and adolescents often use different measures to evaluate self-esteem, making the comparison of findings across studies problematic, and there is a substantial lack of instruments that measure implicit self-esteem, self-esteem stability, and contingent self-esteem (Leeuwis et al. 2015). Studies that test the role of potential mediators and moderators in the relationships between self-esteem and adjustment in children and adolescents, such as sex and age are lacking and needed.

Exposure to Intimate Partner Violence & Self-Esteem

Given the consistent association found across studies between self-esteem and adjustment difficulties in childhood and adolescence, such as anxiety and depression, it is surprising that only limited research to date has investigated self-esteem in children and adolescents exposed to IPV (Holt et al. 2008; Howell 2011; Moylan et al. 2010; Owen et al. 2009). This is particularly unfortunate because children’s sense of agency, self-efficacy, and self-esteem are often considered to be central features of resilience, and emerge as critical elements underlying children’s ability to experience positive outcomes despite a context of risk (Luthar et al. 2000). Guille (2004) suggested that self-esteem and locus of control contribute to the children’s ability to cope with IPV, since children who feel in control of their life circumstances and who have feelings of positive self-worth may be less affected by the adversities they are facing. Moreover, Holt et al. (2008) posited that those children who live in IPV-affected families and have higher self-esteem in one domain, such as academics, may choose to focus on that area of their life in order to obtain an effective distraction from their violent family context (Holt et al. 2008).

Despite being generally neglected, some retrospective research has investigated the long-term consequences of childhood exposure to IPV on adult self-esteem (MacDonnell 2012; Wood and Sommers 2011). For example, Paradis et al. (2009) found that exposure to physical IPV by age 18 was significantly correlated to low self-esteem by age 30 (Paradis et al. 2009). Huth-Bocks et al. (2001) found that preschoolers exposed to IPV reported lower self-esteem than non-exposed preschoolers (Huth-Bocks et al. 2001), while others have found lower self-esteem in school age children exposed to IPV as compared to children who were not exposed (McGee 2000).

Two different studies have focused on self-esteem variability in children exposed to IPV using cluster analysis to identify distinct patterns of adjustment (Graham-Bermann et al. 2009; Grych et al. 2000). Grych et al. (2000) investigated children residing in women’s shelters and identified five distinct patterns of adjustment, with approximately 30% of children characterized as having high self-esteem and no clinically significant adjustment problems. Graham-Bermann et al. (2009) investigated children recruited from the community and also found that a significant minority of children showed high self-esteem and few, if any adjustment problems. Conversely, across studies, children with the lowest levels of self-esteem showed multiproblem-internalizing patterns of adjustment (Grych et al. 2000) and high levels of depression (Graham-Bermann et al. 2009). Interestingly, some children with high self-esteem demonstrated externalizing problems (Grych et al. 2000), perhaps indicative of defensive responding discussed earlier. In fact, as suggested by previous work on children’s aggressive behavior (Sandstrom and Jordan 2008), high self-esteem here may reflect fragile self-esteem, high explicit self-esteem coupled with low implicit self-esteem. It is also noteworthy that some characteristics of exposure to violence were associated with lower self-esteem in both studies. Children with the lowest levels of self-esteem in the study by Grych et al. (2000) were exposed to significantly higher levels of father-to-mother violence and father-to-child aggression, while lower self-esteem was associated with a longer history of exposure to IPV in the study by Graham-Bermann et al. (2009). These findings highlight the importance of investigating specific characteristics of IPV that may mediate or moderate the association between exposure to IPV and children’s self-esteem.

Shen (2009) examined the joint impact of exposure to IPV and child maltreatment on young adults’ self-esteem (Shen 2009). Participants exposed to IPV reported significantly lower self-esteem compared to those who did not experience IPV, while those participants who experienced both types of violence reported the lowest levels of self-esteem. Lower self-esteem was also predicted by self-blame, which is an important finding as self-blame has been linked to feelings of guilt and shame that deeply affect children’s self-esteem (Muris and Meesters 2014)..

Some studies have suggested that violent families are more likely to be characterized by negative parenting behaviors that contribute to children’s experiencing debilitating levels of shame (Bennett et al. 2005; Stuewig and McCloskey 2005). Cross-sectional and prospective studies with children and adolescents suggest that feelings of shame are associated with parental indifference, rejection, and abandonment (Han and Kim 2012), authoritarian parenting style (Mills 2003), parental conditional positive regard (Assor and Tal 2012), and devaluation and overt shaming (Mills et al. 2010). The use of verbal disapproval, hostility, contempt, and physical abuse poses a serious threat to children’s self-esteem as it conveys a strong message that they fail to live up to parental expectations and are a disappointment and unlovable as persons (Feiring 2005). In support of this notion, research has shown strong linkages between shame proneness and children’s adjustment problems. For example, Bennett et al. (2010) investigated the relationships between guilt and shame proneness and depressive symptoms in a sample of 7-year-old children, some of whom had a documented history of neglect. Neglected children reported higher levels of shame proneness and depressive symptoms as compared to the control group. Further, shame proneness (but not guilt) was significantly related to depression for these children, highlighting the specificity of shame as pertinent for children’s adjustment (Bennett et al. 2005). Given these findings, it is unfortunate that so little work to date has addressed the degree of shame proneness in children exposed to IPV: Shame proneness may be higher in children who engage in more self-blame concerning IPV, contributing to lower or more fragile self-esteem, depressive symptoms, and other adjustment difficulties (Miller et al. 2014).

In summary, this body of work clearly demonstrates that exposure to IPV can have a significant and potentially long lasting effect on children’s self-esteem. Given the extensive literature demonstrating the role of self-esteem in adjustment difficulties, such as anxiety and depression, it is clear that more work is needed to address the interaction between exposure to IPV and children’s self-esteem. A normative developmental perspective suggests that in childhood mean levels of self-esteem are high and rank order self-esteem is variable over time; however, it is unknown if this is also the case for children exposed to IPV. It is also unknown if the normative decline in self-esteem observed during the transition to adolescence applies equally to children exposed to IPV, and if so, what this decline means for adjustment difficulties such as depression, anxiety, or suicide ideation.

Some important mediators and moderators of self-esteem in children exposed to IPV have been identified, including IPV characteristics such as length and type of exposure, and children’s perception of self-blame. Although family relationships are widely regarded as a significant influence on the development and maintenance of children’s self-esteem, they are not often investigated as mediators or moderators of self-esteem in children exposed to IPV. Further, since self-esteem represents a potentially modifiable risk factor in treatment programs, work considering other potentially significant mediators and moderators, such as age, sex, and typology of violence (Haselschwerdt 2014), is urgently needed.

Family Relationships & Self-Esteem

Family relationships play a key role in children’s self-esteem and in their adjustment. Positive parent-child relationships are significantly associated with healthy adjustment in children and adolescents, including positive and stable self-esteem (Bulanda and Majumdar 2009; Harter 2012). High levels of family support, positivity within the parent-child relationship, and a consistent positive parenting style have been found to protect children and adolescents in the context of stressful negative life events (Youngstrom et al. 2003). Siblings are also a significant influence, yet very few studies to date have examined the specific association between the quality of sibling relationships and children’s self-esteem, especially in families experiencing challenging life circumstances, such as IPV (Conger et al. 2009).

Self-Esteem & Sibling Relationships

The importance of sibling relationships has been solidly established across a wide spectrum of developmental outcomes, including social skills, theory of mind, language, play and risk-taking behaviors, to name a few (Dirks et al. 2015). Sibling relationships also play a key role in the development of children’s self-esteem for several reasons. During childhood and adolescence siblings spend more time with each other than with virtually anyone else (Buist et al. 2013). Given this high degree of intimate daily contact, a wide variety of theoretical perspectives have noted the influence of siblings, including family systems theory, social learning theory, social comparison theory, identity and de-identification theories, genetic inheritance, as well as rivalry and resource-based perspectives, including parental differential treatment (PDT) (Dirks et al. 2015). Although in most cases siblings share their parents, their genetic inheritance, their home environment and their time with each other, there are significant differences in their experiences and outcomes due to age, sex, birth order, and other non-shared environmental influences.

Research has found significant similarities in self-esteem rank order and self-esteem mean levels between siblings (Gamble et al. 2011), with some studies showing that sibling relationship quality influences the degree of similarity between siblings’ self-esteem. For example, Gamble et al. (2010) found that greater warmth between siblings is linked to greater similarity in their self-perceptions, while conflict is related to greater dissimilarity (Gamble et al. 2010). Others have found that warm and supportive sibling relationships in childhood and adolescence are related to higher self-esteem, demonstrating that siblings provide support and companionship across the life span (Sherman et al. 2006). Siblings who have a close and positive relationship are more likely to interact with each other, to observe and learn from one another, and through these interactions to identify each other as a reliable source of support, help, and advice about their personal lives (Yeh and Lempers 2004). Children and adolescents with positive sibling relationships are more likely to experience warmth, encouragement, acceptance, and support from these relationships, which increases the likelihood of developing positive feelings of self-worth and self-confidence that are essential elements of high self-esteem. For example, Sherman et al. (2006) surveyed young adults about their sibling relationships and their psychological well-being by measuring self-esteem and loneliness. Sibling relationships were categorized into one of four possible typologies, including low involvement (low conflict and low warmth), harmonious involvement (low conflict and high warmth), conflicted involvement (high conflict and low warmth), and affect-intense involvement (high conflict and high warmth). Siblings in harmonious relationships reported the highest self-esteem scores, while those in affect-intense relationships reported the lowest self-esteem scores (Sherman et al. 2006). Buist and Vermande (2014) investigated how similar typologies of sibling relationships were linked with global self-esteem in middle childhood, and found that children with harmonious sibling relationships reported significantly higher self-esteem than children with either conflictual or affect-intense relationships (Buist and Vermande 2014).

Yeh and Lempers (2004) used three waves of data collected from multiple informants to investigate longitudinal associations between sibling relationship quality, self-esteem, and adjustment problems in adolescence. They found that the quality of sibling relationships predicted later self-esteem and adjustment problems: Adolescents who perceived their sibling relationships as more positive at Time 1 tended to have higher self-esteem at Time 2, which was in turn associated with less depression and fewer problematic behaviors at Time 3. Moreover, a bidirectional relationship was found between adolescents’ self-esteem and the quality of their sibling relationships, suggesting that a more positive sibling relationship helps to enhance self-esteem in adolescence, and that, simultaneously, higher self-esteem predicts a more positive sibling relationship (Yeh and Lempers 2004). This longitudinal study suggested that warmer sibling relationships help to enhance adolescent self-esteem, and that healthy self-esteem, in turns, helps adolescents to develop more satisfying sibling relationships over time.

As noted, a number of variables such as age difference, birth order, gender composition, PDT, and parent-child and marital relationship quality have all been identified as mediators and moderators of the association between sibling relationship quality and children’s self-esteem. PDT occurs when siblings perceive that their parents treat them preferentially, such as showing more affection towards, having more lenient and flexible rules for, or having less conflicts with one child than the other child/ren (Kowal et al. 2002; Shanahan et al. 2008). Although Western social norms call for equal treatment of offspring, PDT is extremely common, since most parents recognize differences between their children in behavior, personality, and needs (Atzaba-Poria and Pike 2008). However, PDT has been linked to negative emotional outcomes, as differential comparison between siblings may cause feelings of unfairness, personal insecurity, low self-esteem, and anxiety. A longitudinal study examining the reciprocal links between PDT, siblings’ perception of partiality, and self-esteem in adolescents over time found that the perception of parental favoritism towards a sibling consistently predicted a later decrease in self-esteem (Shebloski et al. 2005).

Research findings across studies provide strong support for the notion that there is a mutual association between the quality of sibling relationships and children’s and adolescents’ self-esteem. However, research that addresses whether these associations are strengthened or weakened in contexts of adversity and stress, such as family violence, is only beginning to emerge (Conger et al. 2009). For example, Mota and Matos (2015) examined how the self-concept of adolescents transitioning into residential care was related to the quality of sibling relationships. These authors posited that sibling attachment played a key role in promoting resilient characteristics such as self-trust, belief in one’s own potential, and positive self-esteem. Their findings supported this view as positive sibling relationships showed both a direct and a mediated effect on positive self-esteem (Mota and Matos 2015). Other research on vulnerable youth examined how support from siblings relates to psychological adjustment and academic competence in early adolescence. Results indicate that sibling support has differential effects based on adolescents’ sex and ethnicity. In particular, support from brothers is positively related to sibling self-esteem, while support from sisters plays a protective role against negative outcomes but is not related to self-esteem (Milevsky and Levitt 2005).

Unfortunately, little is known about the nature of sibling relationships in children who have been exposed to IPV or to other types of violence within the family (Piotrowski 2011; Skopp et al. 2005; Waddell et al. 2001). Relatively few studies have addressed sibling relationships as a possible protective factor that enhances self-esteem or fosters resilience in children who have been exposed to IPV. Trying to untangle the complex set of processes that occur in families affected by IPV, researchers have hypothesized the potential presence of two distinct processes that may account for similarities and differences between parent-child and sibling relationships (Milevsky and Levitt 2005). Several studies suggest the existence of a congruous relational pattern between the two relationships, in which the quality of the parent-child relationship is extended and translated to the sibling relationship. Conversely, some studies have proposed a compensatory relational pattern, in which children who are experiencing a negative relationship with parents develops a close sibling relationship to compensate deficiencies and emotional distress.

In the context of a violent family, sibling relationships can provide a unique source of social support because siblings have ready access to each other and by their mere presence can enhance one another’s sense of security (Waddell et al. 2001). Waddell and her colleagues (2001) compared sibling relationships of children from violent and non-violent families and found that siblings from violent homes not only used less verbal and physical aggression than the comparison group, but also showed more social support behaviors with their siblings. While these findings are encouraging, they do not address whether these supportive behaviors were linked to self-esteem.

One of the few studies that investigated the association between sibling relationship quality and self-esteem in violent families included children exposed to IPV recruited from the community and used a multi-method approach to assess IPV exposure and children’s adjustment (Piotrowski et al. 2014). In this sample, sibling relationship quality significantly predicted self-esteem. Specifically, younger sibling self-esteem had a significant impact on older sibling self-esteem when the sibling relationship was high in warmth and low in disengagement. Moreover, child adjustment problems were conditionally predicted by the quality of sibling relationships, predicated upon the adjustment of the other sibling.

Employing cluster analysis, a cross-sectional study compared patterns of adjustment in siblings exposed to IPV, testing the role of family relationships as potential factors that accounted for heterogeneity in children’s adjustment, including self-esteem (Piotrowski 2011). Cluster analysis generated five distinct clusters of adjustment, including an asymptomatic cluster with high self-esteem and no adjustment problems, and a depressed cluster with low self-esteem and depressive symptoms. The quality of sibling relationships played an important role in distinguishing among the five patterns of adjustment. In particular, asymptomatic older siblings reported significantly warmer and less hostile sibling relationships than their counterparts, while multi-problem externalizing older siblings reported the most negative sibling relationships, characterized by low warmth and high hostility. These findings corroborate previous patterns of adjustment found among children exposed to IPV (Graham-Bermann et al. 2009; Grych et al. 2000) and more importantly highlight the unique importance of the quality of sibling relationships as a significant factor related to both self-esteem and adjustment of children exposed to IPV. These findings also suggest that positive sibling relationships are related to higher self-esteem and positive adjustment in children exposed to IPV; however, other studies on siblings experiencing adverse life circumstances have shown that siblings can also enhance risk (Conger et al. 2009). More research is urgently needed to replicate these findings and to investigate individual, family, and community characteristics that may contribute to creating and maintaining these positive and protective associations.

Discussion

Our review clearly suggests that investigating siblings is important for improving our understanding of the links between children’s self-esteem and their adjustment in adverse contexts. Several limitations in the literature reviewed point towards unanswered questions that need to be addressed by future research. The most evident limitation of the extant literature is a lack of studies that examine the role sibling relationships play in the development of self-esteem in children and adolescents exposed to IPV, investigating the role of siblings in the development of resilient qualities such as positive self-esteem and adjustment outcomes.

More evidence is needed regarding the effects of exposure to IPV on self-esteem at different developmental stages in childhood and adolescence. Due to the high co-occurrence of different types of violence both in and outside the family, future research should investigate the impact of exposure to IPV together with other types of violence within the family, in the community, and in the media such as child abuse and bullying (Osofsky 2003). More advanced measurements of exposure to violence is also needed, detailing severity, type, and duration of violence, where possible (Haselschwerdt 2014). From a conceptual perspective, more attention is needed to developmental issues related to the effects of exposure to IPV at different ages and to risk and protective factors that can increase the negative consequences of IPV or, conversely, improve children’s resilience capacity (Low and Mulford 2013). To date, a developmental psychopathology framework has been very useful for identifying which specific risk and protective factors operate independently or in co-occurrence as mediator and moderator variables between exposure to IPV, self-esteem, and patterns of adjustment. The next critical step for developing theory is to detect what specific mechanisms operate as a function of particular risk and protective factors to damage or promote healthy self-esteem. We advocate strongly for the inclusion of sibling relationships in this line of research.

Longitudinal research using population-based samples is needed and will help provide a developmental perspective on self-esteem rank order and mean level changes in children exposed to IPV, clarifying the long-term consequences of exposure to IPV on self-esteem changes over time. A prospective approach increases accuracy by reducing measurement error and historical bias on adverse childhood experiences (Hardt and Rutter 2004). The few longitudinal studies conducted thus far have led to encouraging results, suggesting that there are both short and long-term negative effects of childhood exposure to IPV, and that children’s adjustment problems may change over time (Litrownik et al. 2003; McCloskey and Lichter 2003; Paradis et al. 2009).

Methodologically, most studies on children exposed to IPV rely heavily on maternal reports only and are based primarily on small samples of White and North American families recruited mostly from shelters. Multi-method approaches that include self-reports, interviews, and behavioral observation from several sources such as children, mothers, fathers, other family members, and teachers, as well as formal sources such as police reports, social work case files and administrative data would reduce measurement error and reporter bias. More ethno-culturally diverse and representative samples of families that include not just one but multiple sibling relationships within families would be a giant step forward. In addition, future studies should control for genetic similarities between siblings, so that these effects can be better understood in terms of similarities and differences in self-esteem between siblings (Natsuaki et al. 2009).

For interdisciplinary scholars who pursue research on sibling relationships, our review points toward several future directions. An important path is toward greater integration of the diverse theoretical approaches used to study sibling relationships and to provide a more in depth explanation of the significant influence siblings can have on individual and family functioning. Because research conducted in the past two decades has been grounded in theoretical traditions and associated methodologies derived from different fields of study that are largely complementary, an important step toward the understanding of the complexity surrounding sibling relationships would be the development of an integrated model (Conger et al. 2009).

We argue that an important area of future research is the role that siblings play in adverse family contexts, such as in IPV-affected families (Feinberg et al. 2011). Research is currently lacking an adequate understanding of the processes underlying how siblings exacerbate or buffer each other from the negative effects of IPV, especially in terms of self-esteem. For example, do the mechanisms of sibling influence on self-esteem identified in non-violent contexts, such as sibling de-identification and social learning (Whiteman et al. 2009), operate similarly or differently in the context of IPV? Future studies should assess the individual personality characteristics of each sibling and investigate mutual bi-directional influences, considering the developmental stage of each sibling and examining more than one sibling relationship in the same family in order to improve our understanding of the complexity of multiple sibling relationships (Kramer and Conger 2009).

Above all, future research should dedicate more attention to the study of the bi-directional association between sibling relationship quality and self-esteem of children and adolescents exposed to IPV. Mediator and moderator models should be tested in order to better understand the role of risk and protective factors such as sex, age spacing, and length and severity of exposure to IPV in affecting the mutual influence of siblings on each other’s self-esteem and adjustment. Understanding these processes will help to better explain sibling similarities and differences, as well as the factors responsible for the effects of sibling relationship quality on adjustment outcomes, and the specific conditions under which spillover versus compensatory processes emerge. For example, is the influence of siblings stronger in compensatory situations, where parent-child relationships are negative or weakened due to parental stress, trauma or absence? Do siblings have less influence in congruous situations where both parent-child and sibling relationships are equally warm and positive or, more unfortunately, equally hostile and negative? Under which of these conditions is it more likely that one or both siblings will demonstrate fragile or damaged self-esteem?

Conclusion

Our main purpose was to review what is known about self-esteem in children exposed to IPV, highlighting the importance of sibling relationships as a key influence that is under-studied and not well understood. For decades, a mounting body of work has documented the overarching influence of sibling socialization on children’s development and adjustment across the life course. Future work needs to clarify the role that sibling relationships play in self-esteem development, particularly in children who live in adverse circumstances. Findings thus far strongly suggest that siblings exposed to IPV actively support each other and that sibling relationships can be protective against the detrimental consequences of exposure to IPV. Are sibling relationships an unacknowledged modifiable risk factor that could significantly improve current interventions for children exposed to IPV (Stormshak et al. 2009)? An improvement in the theoretical and methodological rigor of research on siblings exposed to IPV would provide a solid ground for the advancement of knowledge concerning the role of siblings in promoting children’s resilience in IPV-affected families. Unfortunately, prevention and intervention researchers have paid limited attention to siblings to date (Feinberg et al. 2011). It is our hope to capture the attention of family scholars, to encourage researchers to study more than one child in families affected by IPV, and to enhance prevention and intervention programs designed to lessen the negative effects of exposure to IPV on children and adolescent.