Making Meaning of Traumatic Events: Youths’ Strategies for Processing Childhood Sexual Abuse are Associated With Psychosocial Adjustment

Abstract

The need to make meaning of childhood sexual abuse (CSA) is common and often persists long after the abuse ends. Although believed to be essential for healthy recovery, there is a paucity of research on how youth process their CSA experiences. The current study identified individual differences in the ways youth process their CSA and examined associations with psychosocial adjustment. A sample of 108 youth with confirmed abuse histories enrolled in the study within 8 weeks of abuse discovery, when they were between 8 and 15 years old. Six years later, they participated in interviews about their CSA experiences, reactions, and perceived effects. Using a coding system developed for this study, youths’ CSA narratives were reliably classified with one of three processing strategies: Constructive (13.9%), Absorbed (50%), or Avoidant (36.1%). Absorbed youth reported the highest levels of psychopathological symptoms, sexual problems, and abuse-specific stigmatization, whereas Constructive youth tended to report the fewest problems. Avoidant youth showed significantly more problems than Constructive youth in some but not all areas. Interventions that build healthy processing skills may promote positive recovery by providing tools for constructing adaptive meanings of the abuse, both in its immediate aftermath and over time.

It is widely believed that optimal recovery following child sexual abuse (CSA) requires that youth process their abuse experiences (Brewin, Dalgleish, & Joseph, 1996; Cohen, Mannarino, & Deblinger, 2006; Ehlers & Clark, 2000; Foa & Rothbaum, 1998). Processing allows youth to elaborate trauma memories into organized accounts and better tolerate negative emotions associated with abuse events (Brewin et al., 1996; Ehlers & Clark, 2000). These activities pave the way for meaning making, a type of conceptual processing whereby youth can reappraise cognitive or emotional reactions to the abuse that undermine well-being and construct more adaptive meanings (Brewin et al., 1996; Ehlers & Clark, 2000; Horowitz, 1986; Janoff-Bulman, 1992). By re-storying the abuse, youth can develop narratives of their CSA experiences that allow for psychological comfort and healthy development.

Even when some resolution is achieved, most adults with CSA histories continue to process their experiences in search of meanings (Silver, Boon, & Stones, 1983; Wright, Crawford, & Sebastian, 2007). Unfortunately, these efforts are often unsuccessful. Up to 50% of adults remain unable to make sense of CSA experiences years and even decades later (Silver et al., 1983; Wright et al., 2007). For many, processing CSA experiences is complicated by distressing memories, difficult emotions (e.g., shame, anger, and sorrow), and negative appraisals of the self (e.g., blame) and others (e.g., untrustworthy) that contextualize the abuse and its sequelae (Deblinger & Runyon, 2005; Feiring, Taska, & Chen, 2002; Feiring, Taska, & Lewis, 2002). Although not uncommon, these reactions increase risk for various adjustment problems, including posttraumatic stress disorder (PTSD), depression, externalizing behaviors, dissociation, and relationship difficulties (Feiring, Cleland, & Simon, 2010; Feiring, Miller-Johnson, & Cleland, 2007; Feiring, Simon, & Cleland, 2009).

Clarifying how youth attempt to process CSA may help identify those at risk for developing problems. Articulating different processing strategies and their links to adjustment could inform the development of interventions that are sensitive to youths’ particular processing difficulties. The goal of the current study was to identify systematic differences in youths’ strategies for processing CSA experiences and examine their associations with a range of problems known to be common in this population. The broader literature on trauma-related coping styles suggested three potential strategies for processing CSA-related memories, cognitions, and emotions (Brewin et al., 1996; Ehlers & Clark, 2000; Holmes, 1996; Horowitz, 1986; Janoff-Bulman, 1992). We labeled these primary strategies “Constructive,” “Absorbed,” and “Avoidant” and sought to detect their presence in the trauma narratives of youth with confirmed cases of CSA.

Our conceptualization of Constructive processing derives from the writings of Horowitz (1986), who suggested that successful adaptation requires gradually working through trauma in manageable doses. This entails an effortful balance between attending to and disengaging from abuse-related memories, emotions, and cognitions, which protects youth from becoming overwhelmed. Although not without discomfort, this strategy allows youth to gradually construct more adaptive meanings that alleviate distress and promote positive adjustment (Brewin et al., 1996; Ehlers & Clark, 2000; Janoff-Bulman, 2006). Iterative in nature, constructive processing allows meanings to be updated and revised in accordance with changing life circumstances (Harvey, Mishler, Koenen, & Harney, 2000). We expected that Constructive CSA narratives would evidence deliberate processing; willingness to revisit past and present emotions and appraisals; efforts to pace and regulate exposure to CSA material; and openness to new perspectives on CSA experiences.

Our Absorbed and Avoidant strategies capture what Horowitz (1986) and others (e.g., Holmes, 1996) describe as two types of imbalance in attention to and disengagement from trauma processing. The Absorbed strategy involves excessive attention to and engrossment in abuse-related memories, emotions, or appraisals. In youths’ CSA narratives, we expected this strategy to manifest as perseverative thinking about the abuse, inability to productively reflect on CSA experiences, and difficulty regulating attention and responses to CSA material (Ehlers & Clark, 2000; Holmes, 1996; Roth & Newman, 1993). Consistent with this idea, individuals who are intensely preoccupied with stressful or traumatic events report more severe symptoms of posttraumatic stress, dissociation, and depression (Michael, Halligan, Clark, & Ehlers, 2007; Nolen-Hoeksema & Morrow, 1991). In contrast, the Avoidant strategy involves a pattern of habitual disengagement from CSA material. Although distancing may provide temporary relief from the upset of revisiting painful aspects of the abuse, the use of avoidance as a primary processing strategy may hinder the elaboration, reflection, and cognitive restructuring required to construct adaptive meanings of the abuse (Ehlers & Clark, 2000; Foa & Rothbaum, 1998; Roemer, Litz, Orsillo, & Wagner, 2001). Thus, it is not surprising that self-reported use of avoidant coping strategies is associated with more psychological distress among individuals with CSA histories (Bal, van Oost, de Bourdeaudhuij, & Crombez, 2003; Merrill, Thomsen, Sinclair, Milner, & Gold, 2001). In the context of youths’ CSA narratives, we expected the Avoidant strategy to manifest in speakers’ active attempts to restrict attention to CSA experiences and minimize the importance of abuse-related events, affects, or cognitions (Holmes, 1996; Roth & Newman, 1993).

If trauma processing is essential to posttraumatic recovery, individual differences in processing strategies should be associated with adjustment. Accordingly, we hypothesized that those with a Constructive strategy would report the lowest levels of various problems frequently associated with CSA, including PTSD, dissociation, depression, and sexual problems. Because both abuse-specific self-blame and shame reflect self-disparaging ways of evaluating CSA experiences, we expected that those with a Constructive strategy would exhibit lower levels of both as compared to those with Absorbed or Avoidant strategies. To the extent that these strategies are specific to CSA, they were not expected to be related to individuals’ overall attribution style.

Most studies of trauma processing focus on adults, with few studies examining sexual abuse. Furthermore, the variable-centered, questionnaire approach that typifies many of these studies may obscure important individual differences in trauma processing. For example, by focusing on general coping patterns, most existing questionnaires fail to capture the trauma-specific reactions that are most predictive of posttraumatic adaptation (Coffey, Leitenberg, Henning, Turner, & Bennett, 1996; Feiring et al., 2009; Feiring et al., 2010; Simon & Feiring, 2008). In addition, measures that focus on single variables may cluster different types of processing styles at the scale ends. For example, low scores on a continuous measure of absorbed processing may reflect either high levels of avoidance or high levels of healthy processing. Similarly, low scores on a measure of avoidant processing could indicate high use of either absorbed or healthy processing strategies.

To clearly distinguish primary strategies for processing CSA experiences, we analyzed the abuse narratives of 108 youth with confirmed CSA histories. Narrative methods are uniquely suited for studying how individuals process trauma because they provide a window into speakers’ cognitive and emotional lives (Bruner, 1986; McAdams, 1993). When invited to freely relate their accounts and evaluations of CSA experiences, youths’ narratives reflect their current strategies for processing abuse-related memories, emotions, and cognitions. Rather than a factual recounting of events, narratives reveal the constructive process by which individuals organize and evaluate past events in light of current conditions (Riessman, 1993). In the CSA literature, narrative methods have been used to identify common content themes in individuals’ attempts to make meaning of CSA (e.g., Harter, Erbes, & Hart, 2004; Liem, O’Toole, & James, 1996) and assess the extent to which resolution of CSA themes are related to PTSD (Newman, Riggs, & Roth, 1997). Although not without merit, these approaches cannot discriminate those who are more or less successful in their attempts to process abuse experiences. The current study uses a structural analysis that focuses on how youth approach and organize their narrative accounts to identify individual differences in the ways youth attend to and evaluate CSA memories, affects, and cognitions. This approach makes a distinction between how youth approach the task of processing CSA experiences and the particular meanings made of CSA events. It also differs from questionnaire strategies by evaluating implicit or automatic modes of processing rather than relying on youths’ ability to accurately reflect upon and report their processing strategies.

Method

Sample Selection and Characteristics

Participants were part of a prospective longitudinal study of the consequences of child sexual abuse. The majority of the sample (95%) was referred directly by Child Protective Services (CPS) offices or regional child abuse medical clinics working with CPS. Project staff first reviewed intake logs to identify eligible cases. To qualify for the study, children had to be between 8 and 15 years of age, in the custody of a nonoffending parent or caregiver, and identified as a CSA case within 8 weeks from the date CPS opened the case. Caseworkers then contacted 185 families to obtain permission for project staff to contact them to discuss the study. All but three families agreed to be contacted by project staff, and of the 182 families contacted by project staff, 160 families agreed to and did participate in the study.

Sexual abuse was defined as sexual involvement with a juvenile or an adult perpetrator by coercion and typically involved physical contact (e.g., fondling or oral, anal, or vaginal penetration). Even in the few cases of nonphysical contact, other indicators of severity, such as a higher frequency of abuse or abuse by a close relative (e.g., forced to watch a parent masturbate on multiple occasions) were present. The final recruited sample comprised of children with confirmed cases of sexual abuse, with confirmation determined from CPS or court records indicating at least one of the following criteria: specific medical findings, confession by the offender, abuse validated by an expert, or conviction of the offender in family or criminal court.

Children and their families were assessed at abuse discovery (T1) and again 1 (T2) and 6 years later (T3). At T1, participants included 117 (73%) girls and 43 (27%) boys. Of these, 88 were children aged 12 years and below (M = 9.5, SD = 1.1) and 72 were adolescents aged 13 years and older (M = 13.5, SD = 1.1). A third assessment was obtained approximately 6 years following abuse discovery (M = 6.2, SD = 1.2; range = 4.3–10.1). The sample for the current study consists of the 121 youth who completed the T3 assessment. The vast majority of participants (76%) were females, which precluded a thorough consideration of gender differences in CSA processing. The age distribution of the sample was evenly divided between those who were adolescents (ages 13 to 17 years; 54%) and young adults (ages 18 to 23 years; 46%) at T3. The majority of participants came from single-parent families (70%) and had an income of $25,000 or less (71%). Ethnicity was self-reported as African American (39%), White (31%), Hispanic (21%), and Other (9% including Native American and Asian American). The sample for the current analyses did not differ on demographics, abuse characteristics, or adjustment levels compared to those without trauma interviews at T3 or from those seen for the T1 and T2 but not the T3 assessment.

Although the study did not include treatment, treatment recommendations were made at T1 and T2 to the agencies from which families were recruited, if caregivers and children gave permission. T1 assessments occurred before any participants received treatment. By T2, 68% of youth had received some form of treatment, typically from community-based agencies, because of referrals subsequent to abuse discovery. Individual therapy was the primary modality and the average length of treatment was 5.4 months (SD = 4.7). Between the second and third assessments, 39% of the sample reported having received some form of treatment though not necessarily because of the abuse. Individual therapy was the primary modality and the average length of treatment was 8 months (SD = 8.5).

Specific characteristics of the abuse incidents that qualified participants for inclusion were determined using a checklist specifically designed for the study. After children’s T1 assessment, project staff reviewed records from law enforcement and CPS and systematically completed the checklist for the following information: relationship of the perpetrator to the victim; frequency and duration of the abuse; how the abuse was discovered; types of abusive acts experienced (e.g., fondling and penetration); use of force; medical findings; and how the case was confirmed. Based on the most serious form of contact abuse reported, 67% experienced genital penetration. Almost all perpetrators were known to their victims with 35% being a parent figure, 25% a relative, 37% a familiar person who was not a relative, and 3% a stranger. Forty-three percentage of participants lived with the perpetrator at the time of the abuse. Frequency of the reported abusive events was once for 30% of the sample, 2–9 times for 40%, and 10 times or more for 30%. The abuse lasted for a year or longer in 33% of cases. Use of force was reported in 25% of the sample, threat of force in 19%, and in 56% of the cases, no force or threat was reported. The time lapse from the last abusive act to the time of discovery was 2 weeks or less (45%), more than 2 weeks to 6 months (33%), and 7 months or more (22%).

Procedure

All procedures for this study were approved by the institutional review boards of the academic institutions where the research took place. At each of the three assessment points, when the participant was a minor, children provided written informed assent, and their parents/guardians provide written informed consent. At T3, participants who were 18 or older supplied written informed consent. Assessment data were gathered via interview, questionnaire, and computer-assisted methods by a trained clinician in a private office. Participants received a total of $250 for completing the three assessments.

Measures

Abuse characteristics

After the T1 assessment, trained staff members copied information on specific abuse characteristics from law enforcement agencies and CPS records to a checklist. Based on records of 20 participants, two staff members copied information from the same case files onto the checklist with 100% or near 100% accuracy for each category of information. Coding of abuse severity information from the checklist (e.g., identity of the perpetrator as a stranger = 1, familiar person = 2, relative = 3, and parent figure = 4) was completed by trained project personnel, among whom acceptable interrater reliability was obtained (κ = .73–1.0).

CSA processing strategies

Individual differences in youths’ strategies for processing CSA experiences were assessed using narratives generated from a semistructured interview conducted by project staff known to participants. The interview was designed to elicit representations of abuse experiences rather than veridical accounts of events. Participants were asked to describe their abuse experiences; express their thoughts and feelings about the abuse and its discovery at the time it happened as well as over time; and to explain the perceived effects of their CSA experiences. The protocol allowed participants to respond with as much or as little detail as desired, with standard probes provided as needed. Interviews ranged from 4 to 10 min in length, averaging about 5 min. They were audiotaped, transcribed verbatim by project staff unfamiliar with participants or study hypotheses, and checked for accuracy by an independent reader. All participants assessed at T3 agreed to the interview. Thirteen interviews were excluded from the current analyses due to technical problems (e.g., equipment failure, poor recording; n = 11) or interviews deemed too brief to be reliably coded (n = 2). The remaining 108 narratives were coded using a system that was first developed on an independent set of 35 CSA narratives from a separate study (Simon, Feiring, Noll, & Trickett, 2005). Consistent with a structural approach, our system considers both the content and style of participants’ discourse. The specific detail provided about abuse events was less important than the coherence of the narrative, its structure, and the manner in which youth conveyed abuse-related memories, affects, and cognitions. As with other structural approaches, narrative coherence served as a starting point for examining how youth process CSA. The coherence of a narrative involves the extent to which it is structured as a sensible, internally consistent, and unified whole (Fiese et al., 1999; Grice, 1989; Main, Goldwyn, & Hesse, 2002). Although coherence is critical for constructing adaptive meanings, it is insufficient for differentiating variations in processing strategies (Simon et al., 2005). For example, Constructive narratives are inherently more coherent than Avoidant or Absorbed narratives, but the latter two may be equally incoherent. Thus, our coding system considered coherence but centered on the identification of three specific strategies for processing CSA experiences: Constructive, Absorbed, and Avoidant.

A Constructive narrative is characterized by the speaker’s effortful processing of abuse memories, affects, and cognitions. In addition to providing a coherent account of abuse events, Constructive youth recount vulnerable thoughts and feelings from a child’s point of view while maintaining a present day perspective. These memories, thoughts, and feelings provide a basis for youth to construct an understanding of the causal circumstances of abuse events, as they contemplate their own or others’ actions or reactions and evaluate the impact of abuse events at the time they occur and subsequently. Either explicitly or implicitly, these understandings are viewed as active constructions that are open to revision in the face of changing circumstances. For example, Constructive youth may acknowledge changes in maturation, environment, or experience since the time of the abuse, which altered the meanings and implications of the abuse. As such, abuse experiences are clearly contained in the past, but the significance of these events for past and current functioning remains open for consideration. Similarly, Constructive youth may acknowledge some degree of continued distress as they continue to process abuse-related memories, thoughts, and feelings; yet, they remain open to continued processing and may even describe explicit strategies for regulating their attention or distress that keep them from feeling overwhelmed. The following quote from one of our participants aptly captures many of the features of the Constructive strategy:

At the time, and for a while after, I felt dirty, different from other kids. I stopped, um, hanging around my friends because I really didn’t want to be around anyone. I didn’t know how to express what I was feeling and I like held it all inside. But eventually I just couldn’t do that anymore and, you know, I had to start dealing with it, cause if I didn’t it would just keep getting bigger and bigger. Not like all at once you know, but bit by bit. I’m not sure I’ll ever totally understand why it happened. I mean it’s like a sentence for a crime I didn’t commit. But I have to, you know, accept it. So I’m figuring out how to deal with it—not completely, well not yet, but in some ways. So I can be a teenager and continue on with my life.

Avoidant and Absorbed narratives each lack coherence but differ markedly in the underlying strategy for processing abuse experiences. Whereas Constructive strategies involve the active processing of manageable doses of abuse-related material, Absorbed strategies reflect an engrossment in abuse-related memories, affects, or cognitions in ways that impede meaning making. Confusing or overly detailed descriptions of abuse experiences suggest difficulty constructing an incisive story of abuse experiences and convey a sense of being overwhelmed by abuse-related material. “In-the-moment” accounts indicate a struggle to maintain a current day perspective on past events. Each of these narrative features conveys a fundamental difficulty containing and regulating abuse-related reactions in the service of discerning key information from which useful meanings can be constructed. In short, abuse experiences are viewed as important, but efforts to evaluate them are largely unproductive and lack insight. As a result, Absorbed speakers seem unable to benefit from attempts to process abuse experiences. Such efforts may appear rigid or futile (e.g., excessive rumination or preoccupying anger) and convey a sense that whatever meanings have been derived from abuse experiences are not readily amenable to new or discrepant input. The following quote illustrates an Absorbed speaker’s difficulty relating an account of her CSA experiences:

It was one of the kinds of things that happened but it was over with and nobody was there. I don’t remember when it started. Ok, what happened, yes. My cousin—um—God—Lord—(30sec). When I was, it’s hard because when it first happened, well not at first, but I told then. It was just that, but after that, I never talked about it anymore. I don’t know what to say. Where do you even begin? It’s just too much, I don’t know.

This Absorbed speaker finds it difficult to maintain a present-day perspective on past events:

… and when I tell somebody, I am mad that the one person I mean I was extremely mad, I still am to this day, I mean the one person that’s supposed to be there for me 24-7 for the rest of my life til whenever you know, she lives—she doesn’t believe me, my mother and this is my mom, and she doesn’t believe me and I am eight years old.

Unlike the approach orientation of the Constructive and Absorbed strategies, Avoidant meaning making is evidenced by a notable restriction of attention to abuse-related memories, affects, and cognitions. Avoidant narratives are characterized by a marked lack of effort to understand why abuse events occurred or to evaluate the impact of the abuse on the self or relationships. The story provided of the abuse events may be extremely brief or fact driven, with little spontaneous discussion of thoughts or feelings. Questions regarding abuse-related affect or cognition are typically met with marked distancing or minimization of their importance, and evidence of personal vulnerability is either absent or largely limited to the past. Through content and manner, youth with Avoidant meaning-making strategies suggest that their abuse experiences are of little relevance to their present life or identity. The following quote is illustrative of this sentiment:

Ok. Well, I had a molestation and it entered court and they took care of it. You know, all that typical big mess that you see on TV. I was like a kid at the time. Of course, it was his fault, so I mean, it didn’t really bother me.

Each narrative was independently coded by two of the authors (V.S. and S.M) to determine the presence of a primary processing strategy: Constructive, Absorbed, or Avoidant. Narratives were classified with a primary processing strategy using a two-stage approach. First, coders used a 9-point scale to rate the presence and strength of each processing strategy in a given transcript. Higher scores reflect a stronger and more pervasive use of that strategy. Next, the coder used the constellation of processing style scale scores along with the narrative’s fit to the overall descriptions of the three processing strategies to determine whether a primary strategy was evident. A narrative was classified with one primary processing strategies when (a) the narrative was a good fit to only one of the overall strategy descriptors; (b) one of the processing strategies scale scores was at or above the scale midpoint (i.e., score of 5 or above) and was higher than the two other scale scores; and (c) the highest processing strategy scale score corresponded with the overall strategy fit. Any transcript not meeting these criteria was to be designated as “Unclassifiable” with a descriptive explanation. This method reflects a person-oriented approach to understanding how youth process CSA experiences whereby individuals rather than variables are treated as the unit of analysis, and the patterning of variables (e.g., processing style scores) rather than the linear relationships are used to capture the essential features of a system (von Eye & Bergman, 2003).

Concurrent adjustment

Indicators of concurrent (T3) adjustment included measures of PTSD, dissociation, depression, abuse-specific self-blame and shame, general self-blaming attribution style, and sexual difficulties (i.e., sexual concerns and dysfunctional sexual behavior). Symptoms of PTSD, dissociation, and sexual problems were assessed with the Trauma Symptom Inventory (TSI; Briere, 1995). Items on the TSI are rated on a 4-point Likert-type scale ranging from “never” to “often” and index problems during the previous 6 months. Scale scores are created by summing items, such that the higher the score, the more symptoms reported. The scales for intrusive experiences (e.g., flashbacks of upsetting things, suddenly remembering something upsetting from the past), defensive avoidance (e.g., trying to forget about a bad time in your life, staying away from places or people that remind you of something), and anxious arousal (e.g., feeling jumpy, being startled or frightened by sudden noises) index different types of PTSD symptoms. These scales (8 items each) demonstrated acceptable internal consistency in the current sample (Intrusive experiences α = .89; Defensive avoidance α = .87; Anxious arousal α = .81). The TSI Dissociation scale taps symptoms such as derealization, out-of-body experiences, depersonalization, and emotional numbing. This 9-item scale showed adequate internal consistency (α =.86). Two of the TSI scales were administered to index sexual difficulties. The Sexual Concerns scale includes 9 items that tap perceptions of sexual problems in relationships, sexual dissatisfaction, and unwanted sexual thoughts and feelings (e.g., bad thoughts or feelings during sex, problems in sexual relations). The 9-item Dysfunctional Sexual Behavior subscale measures sexual behavior that is indiscriminant and the use of sex to achieve nonsexual goals (e.g., having sex to obtain love or attention). The internal consistency of these subscales in our sample was acceptable (Sexual concerns α = .82; Dysfunctional sexual behavior α = .78).

The Child Depression Inventory (CDI; Kovacs, 1985) and Beck Depression Inventory (BDI-II; Beck, Steer, & Brown, 1996) were used to index depressive symptoms at T3. Participants 16 years and younger completed the CDI and those older than 16 years completed the BDI. Both measures use a forced-choice questionnaire to quantify a range of depressive symptoms such as disturbances of mood, hedonic capacity, vegetative functions, and interpersonal behaviors. Higher total scores indicate more depressive symptoms. Both the CDI and the BDI-II have shown acceptable convergent and discriminant validity (Beck et al., 1996; Saylor, Finch, Spirito, & Bennett, 1984). Participants’ scores on these measures were converted to T-scores make them comparable across the two instruments for analyses. Scores for this sample showed good internal consistency (CDI α =.80; BDI α =.92).

Abuse-related shame and self-blame were assessed using items developed for this study. A sample item from the shame scale reads, “I feel ashamed because I think that people can tell from looking at me what happened.” The items were rated on a 3-point scale ranging from 1 (not true) to 3 (very true). Items were summed, with higher scores indicating greater shame. For the abuse-specific self-blame items, participants rated the extent to which each of eight causal statements was true for why the abuse happened on the same 3-point scale used for shame. A sample items reads, “I was to blame for what happened.” Item ratings were summed, with higher scores indicating more self-blame. Each measure showed acceptable internal consistency (shame α = .85; self-blame α = .75).

To assess general attribution risk, participants 16 years and younger completed the Children’s Attributional Style Questionnaire (CASQ; Thompson, Kaslow, Weiss, & Nolen-Hoeksema, 1998), and those 17 years and older completed the parallel instrument for adults, the Attributional Style Questionnaire (ASQ; Peterson & Villanova, 1988). Both measures include an equal number of scenarios that describe events with positive and negative outcomes (e.g., CASQ “You get a bad grade in school;” ASQ “You meet a friend who acts hostile towards you”). The CASQ and ASQ provide three subscale scores for positive events on each dimension, internal/external, stable/unstable, global/specific, and three parallel subscale scores for negative events. From these, positive (positive outcome—internal, stable, and global) and negative (negative outcome—internal, stable, and global) composite scores are computed. The general self-blame attribution score corresponds to the positive composite score minus the negative composite score. This score indicates the extent to which a self-blaming style for negative events is balanced by a positive style for good events, with lower scores indicating a more self-blaming (i.e., internal, stable, and global) attribution style. The internal consistency of this measure was moderate (CASQ α =.73; ASQ α =.66). The general self-blame attribution scores for the CASQ and ASQ were converted to T-scores to make them comparable for analyses.

Results

Descriptive Information

Each of the two coders was able to classify every narrative with one of the three processing strategies based on both overall fit to the strategy description and consideration of an individual’s scores on the processing strategy scales. No transcript was judged unclassifiable. Coders agreed on the classification for 101 of the 108 interviews (κ = .89, p < .001). The seven disagreements varied in nature: Absorbed versus Constructive (n = 3), Absorbed versus Avoidant (n = 2), Constructive versus Absorbed (n = 1), and Constructive versus Avoidant (n = 1). All disagreements were effectively resolved through consensus scoring by the two coders.

Scores on the three processing style scales were consistent with overall classifications. The highest scale score corresponded with the overall classification for each of the 108 narratives (e.g., when a narrative was classified as Constructive, the Constructive scale score was the highest). In addition, the scale score that corresponded with the overall classification was always at or above the scale midpoint (M = 6.98, SD = 0.95) and at least 1.5 points higher than next highest scale score. In most cases (96.7%), the difference between the classification scale score and the next highest score was at least two points, with the mean difference between these scores approaching four scale points (M = 3.98, SD = 1.44).

Table 1 shows the distribution of the three processing strategy classifications for the entire sample as well as by gender and age group at abuse discovery. Classifications were not associated with youths’ ethnicity, income, or overall abuse severity. Similarly, they did not differ on any of the individual abuse characteristics that comprise the overall abuse severity score (i.e., frequency, duration, perpetrator identity, use of force, and penetration).

Table 1.

Distribution of Childhood Sexual Abuse (CSA) Processing Strategy Classifications

	CSA Processing Strategy Classifications
	Constructive	Avoidant	Absorbed	Total
Entire sample
N	15	39	54	108
% within classification	13.9	36.1	50
Age at abuse discovery
Child
n	5	27	31	63
% within classification	33.3	69.2	55.6
Adolescent
n	10	12	22	45
% within classification	66.7	30.8	44.4
Gender
Male
n	1	14	9	24
% within classification	6.7	35.9	16.7
Female
n	14	25	45	84
% within classification	93.3	64.1	83.3

Most striking is the low number of participants classified as Constructive. The vast majority of youth were either Absorbed or Avoidant. Processing strategy classifications were associated with age and gender, χ²(2, 106) = 7.19, p = .03 and χ²(2, 106) = 7.71, p = .02, respectively. Cell frequencies that were greater than expected by chance were identified using cellwise residual analyses in which adjusted residuals are treated as standardized normal deviates for the purposes of establishing levels of significance (MacDonald & Gardner, 2000). Youth with Constructive strategies were more likely to have been adolescents than children at abuse discovery (z_adj = 2.4, p < .05). Those with Avoidant strategies were more likely to be male than female (z_adj = 2.6, p < .05). No age or gender differences emerged within the Absorbed classification. Given these differences, gender and age group at discovery were included as covariates in subsequent analyses.

Processing Strategies and Post-Abuse Adaptation

A significant minority of the sample reported elevated levels of psychological distress. For example, indicators of psychopathology in the clinical range on the TSI (i.e., subscale scores of 65 or above) were noted for about a quarter of the sample 6 years after abuse discovery (n = 30, 25% for defensive avoidance; n = 30, 25% for intrusive experiences; 14, 11% for anxious arousal; n = 27, 23% for dissociation).

Table 2 presents descriptive statistics and correlations among the outcome measures of posttraumatic stress, dissociation, depression, abuse-specific self-blame, attribution style, abuse-specific shame, and sexual problems. Each measure showed good variability. To compare the mean scores of youth with different processing strategies on the measures of concurrent functioning, a series of multivariate analyses of covariance (MANCOVA) were computed controlling for gender and age group at abuse discovery. Conceptually, similar variables were analyzed together to reduce the Type I error rate. Separate MANCOVAs were conducted on the symptom variables (PTSD scales, depression, and dissociation), the sexual behavior variables (sexual concerns and dysfunctions), and the shame and attribution processes (abuse-specific shame and self-blame, and general self-blame). Planned comparisons examined pairwise differences between youth in the three processing strategy groups. Table 3 shows the results of these analyses along with the group means.

Table 2.

Descriptive Statistics and Correlations Among Study Outcome Measures

	1	2	3	4	5	6	7	8	9	10
1. Intrusive Experiences	55.70 (11.41)
2. Defensive Avoidance	0.80	55.61 (10.66)
3. Anxious Arousal	0.70	0.80	52.49 (10.13)
4. Dissociation	0.74	0.73	0.79	55.95 (13.09)
5. Depression	0.63	0.62	0.65	0.66	50.18 (9.97)
6. Abuse-specific Shame	0.55	0.57	0.49	0.62	0.55	1.92 (1.89)
7. Abuse-specific self-blame	0.58	0.64	0.57	0.65	0.61	0.90	4.39 (4.04)
8. General attribution risk	−0.29	−0.32	−0.31	−0.38	−0.46	−0.33	−0.24	50.67 (9.66)
9. Sexual concerns	0.52	0.53	0.56	0.59	0.62	0.47	0.49	−0.23	49.25 (8.49)
10. Dysfunctional sexual behavior	0.41	0.48	0.43	0.46	0.47	0.39	0.47	−0.14	0.69	49.89 (9.11)

Note. For each variable, means and standard deviations (in parentheses) are on the main diagonal. Correlations with an absolute value > .20 are significant at p < .05 and correlations with an absolute value of .31 are significant at p < .01.

Table 3.

Results of multivariate analyses of covariance (MANCOVAs) and Planned Comparisons Testing for Differences in Outcomes by Processing Strategy Controlling for Gender and Age at Time of Abuse Discovery

Psychosocial Functioning	Constructive M (SD)	Avoidant M (SD)	Absorbed M (SD)	Between Subjects Effects Univariate F2
Symptoms1 MANCOVA
Intrusive experiences	48.06 (6.39)a	53.74 (10.03)b	61.73 (12.43)c	7.62** (2, 105)
Defensive avoidance	48.07 (4.99)a	54.94 (10.87)b	60.72 (12.45)c	9.89** (2, 105)
Anxious arousal	49.26 (6.84)a	50.46 (9.87)a	56.55 (10.62)b	3.86* (2, 105)
Dissociation	48.06 (7.46)a	54.84 (11.80)b	60.80 (14.15)c	4.82* (2, 105)
Depression	46.03 (6.83)a	48.53 (8.06)a	55.39 (11.10b	5.86** (2, 105)
Sexual problems1 MANCOVA
Sexual concerns	45.13 (5.53)a	47.90 (6.35)a	52.60 (9.82)b	4.39* (2, 106)
Dysfunctional sex behavior	45.40 (3.20)a	48.95 (6.68)a	54.76 (11.89)b	5.93** (2, 106)
Shame/attribution MANCOVA
Abuse-specific shame	2.02 (1.97)a	3.85 (2.62)a	5.94 (3.13)b	5.76** (2, 106)
Abuse-specific self-blame	1.80 (1.58)a	2.65 (2.16)	3.60 (3.09)b	3.37+(2, 106)
General attribution risk1	51.79 (7.29)	50.89 (11.25)	49.89 (9.04)	0.34 (2, 106)

Note.

¹indicates reported means are T-scores;

²F(df) for planned pairwise contrasts between the three processing strategies.

^abcScores with different letter superscript are significantly different from one another at p < .05;

⁺p < .10,

^*p < .05,

^**p < .01.

For symptoms, there was a significant multivariate effect for processing strategy, Pillais’ F(10, 193) = 2.54, p = .04, but not for age or gender. The univariate tests for processing strategy were significant for all outcomes. The general pattern of mean scores was for Absorbed youth to show the highest symptom levels followed by Avoidant and then Constructive youth. Results of the planned comparisons revealed that Absorbed youth reported significantly more symptoms of intrusive experiences and depression than either Constructive or Avoidant youth. Constructive youth also reported significantly fewer symptoms of intrusive experiences, defensive avoidance, and dissociation than either Absorbed or Avoidant youth, as well as fewer symptoms of anxious arousal and depression than Absorbed youth. For sexual problems, there was a significant multivariate effect for processing strategy, Pillais’ F(4, 200) = 2.77, p = .03, as well as significant univariate effects for each problem type. Results of planned comparisons revealed that Absorbed youth reported significantly more sexual concerns and dysfunctional sexual behavior than either Constructive or Avoidant youth. Neither age nor gender showed a significant multivariate effect.

For shame and attributions, there was a significant multivariate effect for processing strategy, Pillais’ F(6,196) = 2.26, p = .05, but not for age or gender. The univariate tests for type of processing strategy were significant only for abuse-specific shame and abuse-specific self-blame. Results of the planned pairwise comparisons showed that Absorbed youth reported significantly more shame than did Avoidant or Constructive youth. Constructive youth reported significantly less self-blame for the abuse than Absorbed youth.

Discussion

Although thought to be central to recovery, few studies have examined the strategies CSA youth use to organize and evaluate their experiences. The goals of this study were to identify individual differences in how youth process CSA experiences and examine associations between processing strategies and adjustment. The findings showed that coders could reliably discern three processing strategies. As expected, each participant clearly displayed one primary strategy for managing abuse-related material: Constructive, Absorbed, or Avoidant.

Coding of the Constructive, Absorbed, and Avoidant strategies focused on the structure and manner of youths’ discourse about the meaning of the abuse and its personal implications rather than specific content or themes. In this way, youths’ narratives provided a window into their strategies for reflecting upon and regulating abuse-related memories, affects, and cognitions. The use of structural narrative analysis to detect variations in the ways people represent themselves and their experiences is not new. Relationship researchers, for example, have successfully used this approach to identify systematic differences in individuals’ states of mind about relationship experiences (e.g., Main et al., 2002) and their associations with interpersonal functioning across the life span (see Cassidy & Shaver, 2008). The system developed for the current study is the first to use structural analysis to index systematic differences in youths’ strategies for processing CSA experiences. Our findings add to this work by demonstrating the utility of a person-oriented approach to trauma processing. Whereas questionnaire ratings of avoidance and rumination focus on group levels of these constructs, they fail to consider their patterning within individuals or the characteristics of healthy processing. When youth narrate their experiences in their own words, they reveal implicit modes of processing that capture patterns of attending to and disengaging from abuse-related material.

One of the most striking findings from the current study is the low proportion of youth with a Constructive processing strategy (13.9%). These results are consistent with the adult literature, in which the majority of adults report continued difficulties making meaning of CSA experiences (Silver et al., 1983; Wright et al., 2007). Taken together, these studies highlight the significant challenges that CSA poses to healthy processing. Many youth either fail to process their experiences or process in ways that do not facilitate adaptive meaning making.

As expected, processing strategies were associated with psychosocial functioning. As suggested by trauma theories, Constructive narratives showed signs of effortful processing that were measured and iterative over time (e.g., Brewin et al., 1996; Ehlers & Clark, 2000; Horowitz, 1986). Constructive youth may not have “resolved” their experiences; indeed, several participants openly acknowledged continuing struggles. However, their narratives suggested that they viewed processing as an open-ended experience of evolving meaning. This strategy was associated with the lowest levels of psychopathological symptoms, sexual problems, and abuse-related stigmatization (i.e., abuse-specific shame and self-blame).

Youth with Absorbed narratives were open to processing but showed great difficulty doing so productively. Their narratives revealed problems maintaining a current perspective on past CSA experiences without becoming overwhelmed, fearful, angry, or otherwise dysregulated. Rigid conceptualizations of CSA experiences and difficulties regulating their current reactions appeared to prevent these youth from capitalizing on the passage of time and corresponding maturational advances to gain new or more helpful perspectives. In short, Absorbed youth continued to unsuccessfully process their CSA experiences. Several of them seemed to recognize that their efforts were unproductive and described equally unsuccessful efforts to avoid thinking about their abuse. Not surprisingly, Absorbed youth reported more PTSD and depressive symptoms, sexual problems, and abuse-specific shame than either Constructive or Avoidant youth. These findings suggest that this strategy is particularly problematic for youth and indicate a need for interventions that address the ways in which Absorbed youth are attending to, regulating, and processing their abuse experiences.

Like Absorbed youth, those with Avoidant strategies reported more intrusive experiences, defensive avoidance, and dissociation than Constructive youth. These findings suggest that posttraumatic recovery is related to both whether and how youth process CSA experiences over time. Even when no statistically significant differences emerged, Avoidant youth typically reported more problems than Constructive youth. It is possible that the low number of Constructive youth resulted in insufficient statistical power to detect smaller but meaningful differences between the two strategies. Nonetheless, Avoidant youth did fare better than Absorbed youth on several outcomes. Of particular interest is that compared to Absorbed youth, Avoidant youth reported less defensive avoidance. This seemingly counterintuitive finding may point to an important difference between youths’ posttraumatic behavior and processing styles. By actively attending to CSA material, Absorbed youth may be more consciously motivated to avoid reminders of their abuse in efforts to gain intermittent relief from the intensity of negative abuse-related feelings and cognitions. In contrast, Avoidant youth have established an automatic pattern of rigidly distancing from the abuse in a way that minimizes its current relevance. To the extent that this implicit strategy is “successful,” they may sense less need for deliberate efforts to avoid CSA material. Additional research into the defensive avoidance behavior of those with and without Avoidant processing styles would help clarify this issue.

The pattern of results might suggest that Avoidant processing of CSA is less healthy than Constructive processing but with fewer generalized consequences than the Absorbed strategy. This proposition should be considered tentative at best, as it is possible that Avoidant youth may experience difficulties in domains other than those assessed in the current study. Whereas our outcome measures focused on intrapsychic correlates of CSA, Avoidant youth may be more likely to experience externalizing behavior problems, such as substance use. The extent to which avoidant coping is detrimental to psychological functioning is a point of contention in the broader trauma literature (Bonanno, Keltner, Holen, & Horowitz, 1995; Ehlers & Clark, 2000; Foa & Rothbaum, 1998; Wortman & Silver, 1989). Our findings imply that for youth with confirmed CSA histories, some degree of processing is important to long-term adaptation but that the approach to processing also is critical. Use of distancing and effortful regulation of attention to CSA-related material may be helpful and even necessary for positive recovery. Some participants with Constructive strategies discussed deliberate efforts to limit their focus on CSA experiences to balance processing the past with building a present life. For these youth, however, distancing is effortful and just one component of a more balanced approach to making sense of CSA experiences. In contrast, when avoidance entails an implicit, pervasive, and automatic strategy characterized by the minimization of CSA experiences, it is associated with a moderate degree of psychosocial dysfunction.

Future studies should examine the development of youths’ processing strategies. CSA processing strategies were unrelated to abuse severity but were associated with age and gender. Although participants with Constructive strategies were a small minority, they tended to be older. The extent to which this finding is due to the timing of the assessment versus the timing of the abuse is unclear. The former explanation suggests that the developmental advances of late adolescence and early adulthood may provide new skills and opportunities for developing a more adaptive understanding of CSA experiences. However, findings from the few relevant studies of adults with CSA histories suggest that the vast majority either avoid or are frustrated in their efforts to process their CSA experiences (Silver et al., 1983; Wright et al., 2007). Such findings raise questions about whether the passage of time or developmental advances in processing capacities are themselves sufficient for developing a Constructive strategy. Alternatively, developmental capacities at the time of the abuse or its discovery may better account for the observed age differences. When CSA occurs at an early age, processing strategies may be limited and contribute to greater use of dissociative coping, which may engender subsequent processing difficulties (Macfie, Cicchetti, & Toth, 2001; Putnam, 1996). Consistent with this idea, younger participants were less likely than older participants to use Constructive processing and those with nonconstructive strategies reported higher levels of dissociation. An important goal of future work will be to assess youths’ CSA processing strategies across development. Our results suggest that when abuse occurs at an earlier age, it may set in motion a reliance on less healthy processing styles. However, maturational advances in cognitive, emotional, or social capacities appear to confer little “protection” for older youth.

Gender did not distinguish among those with Constructive strategies, but Avoidant youth were more likely to be male than female. Males’ orientation toward avoiding rather than approaching emotions under stress might contribute to greater avoidance of processing emotion-laden CSA material (Stanton & Franz, 1999; Stanton, Kirk, Cameron, & Danoff-Burg, 2000). Their reluctance to discuss their CSA experiences may also be important (O’Leary & Barber, 2008). In future work, assessment of emotional approach tendencies, gender roles, and perceived reactions of others’ to CSA experiences may be helpful for understanding whether and why males might prefer avoidant CSA processing strategies (Paine & Hansen, 2002).

Understanding the development of CSA processing strategies also will require attention to the broader context in which the abuse occurs, including youths’ initial reactions to the abuse, the reactions of significant others, and the presence of other traumatic events. In addition to these abuse-specific factors, youths’ biobehavioral dispositions for stress reactivity and emotion regulation warrant careful consideration. Self-capacities for distress tolerance and modulation begin to form early in life but are also affected by abuse and may mediate distress responses when recalling abuse experiences (Alexander, 1992; Cicchetti & Rogosch, 2007; Palesh, Classen, Field, Kraemer, & Spiegel, 2007).

Our study is an initial effort to articulate individual differences in the ways youth process CSA experiences and several limitations are important for interpreting the reported results. First, the cross-sectional, nonexperimental nature of the data precludes causal interpretations about observed patterns of adjustment among youth with different processing strategies. Articulating the reciprocal relations between processing, meaning making, and adjustment over the course of development requires longitudinal research. Such work could also help identify markers of the three strategies at different developmental stages. A second consideration concerns the scope of the current study. As previously noted, our goal was to identify individual differences in processing strategies rather than explain their origins or explore the extent of their psychosocial implications. As such, many important factors, including other aspects of the abuse experience (e.g., caregiver reactions and polyvictimization), the broader scope of CSA-related outcomes, and youths’ dispositional characteristics remain important topics for future work.

Nonetheless, the current findings suggest that interventions to facilitate constructive CSA processing may be particularly helpful for long-term adaptation. Such work is compatible with early intervention efforts that unfold over time in the context of a trusted therapeutic alliance. For example, trauma-focused cognitive behavioral therapy (TF-CBT), prescribes a gradual exposure paradigm prior to trauma processing (Cohen et al., 2006). Specifically, youth first develop a storybook about their traumatic experiences and reactions. Narrating trauma-related events, thoughts, and feelings provides youth with an integrated story of their experience, desensitizes them to traumatic reminders, and encourages a metacognitive perspective of the trauma as a part rather than a defining feature of their lives. In turn, these activities set the stage for therapists to facilitate conceptual processing skills that allow youth to reappraise their cognitive or emotional reactions and develop adaptive meanings of their experiences. The carefully titrated approach of TF-CBT and other integrative treatment models (e.g., Cloitre, Chase Stovall-McClough, Miranda, & Chemtob, 2004) stand in sharp contrast to the immediate processing demands of the single session debriefing model, which is at best ineffective and at worst harmful (Rose, Bisson, & Wessely, 2003).

Healthy processing skills should prepare youth to revisit and update their understanding of the abuse in accordance with their developing cognitive, emotional, and social capacities. This may be particularly helpful for successfully negotiating developmental tasks that often pose challenges to CSA youth, such as the emergence of romantic and consensual sexual relationships or the transition to parenthood (Harvey et al., 2000; Silver et al., 1983). When viewed in this way, promoting healthy processing strategies may be at least as important as encouraging healthy understandings of CSA experiences. Youth with such strategies should have the tools for constructing adaptive meanings of the abuse, both in its immediate aftermath and over time.

Biographies

Valerie Simon is an Assistant Professor of Psychology at the Merrill Palmer Skillman Institute at Wayne State University. Her research focuses the development of adolescents’ peer and romantic relationships, the ways youth process significant life experiences, and processes related to adjustment among sexually abused youth.

Dr. Candice Feiring, is a Senior Research Scholar at The College of New Jersey. Her research focuses on processes such as shame, attribution and social support related to adjustment in sexually abused youth and adolescent romantic relationships.

Sarah Kobielski McElroy is a postdoctoral fellow in clinical child psychology at the Virginia Treatment Center for Children. Her research focuses on youths’ representations of maltreatment experiences and polyvictimization in relation to psychosocial functioning.