Psychometric Properties of the Institutional Betrayal Questionnaire,Version 2: Evidence for a Two-Factor Model

Abstract

Institutional betrayal reflects the failings of a trusted institution to prevent or respond appropriately to negative experiences. Following sexual assault, survivors who encounter institutional betrayal may experience greater distress and poorer functioning. The current study sought to assess the construct validity of the Institutional Betrayal Questionnaire, Version 2 (IBQ.2) and evaluate its factor structure. Survivors of sexual assault (N = 426) were recruited via Amazon Mechanical Turk and completed various questionnaires related to mental health, disclosure and assault characteristics, world beliefs, and rape myth adherence. The IBQ.2 demonstrated convergent validity with disclosure to formal support providers, assault severity, turning against reactions, and beliefs about self-control, and evidenced discriminant validity with disclosure timing, rape myth adherence, and beliefs about randomness and controllability of outcomes. Notably, the IBQ.2 was unrelated to measures of distress, including symptoms of stress, depression, anxiety, and posttraumatic stress disorder, providing mixed evidence for the IBQ.2’s construct validity. Confirmatory factor analyses failed to replicate the single-factor model of institutional betrayal found in a previous study, and, instead, suggested a two-factor structure of the IBQ.2 that delineates between the promotion of and response to sexual victimization. Post hoc analyses revealed that only one of the two factors (Response to Sexual Victimization) evidenced convergent and discriminant validity largely consistent with the single-factor model. The novelty of these relationships and factor structure of the IBQ.2 found in the current study warrants replication in future research.

Keywords

institutional betrayal trauma betrayal trauma sexual assault psychometrics

Introduction

Following a sexual assault, many survivors choose to disclose the experience to informal support providers such as friends, family members, or significant others (Ahrens, Campbell, Ternier-Thames, Wasco, & Sefl, 2007). Theoretical models of disclosure posit that survivors disclose to those whom they believe will be most helpful (Golding, Siegel, Sorenson, Burnam, & Stein, 1989), including disclosing to receive needed aid, feel better, or obtain justice (Bachman, 1993; Feldman-Summers & Norris, 1984; Golding et al., 1989), yet most survivors do not disclose to formal support sources whose professions exist in part to provide such services (e.g., law enforcement, legal professionals, medical professionals, universities; Starzynski, Ullman, Filipas, & Townsend, 2005). Disclosure to formal support sources are more likely to occur following assaults committed by strangers that involve physical injury, perceived life threat, and/or peritraumatic fear (Golding et al., 1989; Ullman & Filipas, 2001). Unfortunately, these same characteristics that predict disclosure to formal support providers also seem to increase the likelihood that disclosure recipients will respond negatively; research has found that survivors of more violent, life-threatening, and fear-evoking assaults are more frequently met with negative reactions from others (Ullman & Siegel, 1995; Ullman, Starzynski, Long, Mason, & Long, 2008). Generally, the literature differentiates positive reactions to sexual assault survivors from the more negative reactions of unsupportive acknowledgment and turning against (Relyea & Ullman, 2015). Whereas positive reactions confer marginal improvements, if any, in postassault outcomes (Filipas & Ullman, 2001; Ullman, 1996, 2014), negative reactions have been found to worsen perceived health (Ullman, 1996), posttraumatic stress disorder (PTSD; Andrews, Brewin, & Rose, 2003), depression (Campbell & Raja, 2005), self-esteem (Ullman, 2000), alcohol use (Ullman et al., 2008), self-blame, use of avoidant coping (Ullman, 1996), and feelings of guilt, anxiety, distrust of others, and reluctance to seek further help (Campbell & Raja, 2005). Moreover, the receipt of negative reactions may silence survivors from future disclosures out of fear of further negative treatment (Ahrens, 2006).

Often, unsupportive acknowledgment and turning-against reactions are provided by formal support providers who represent larger institutions (e.g., dean of students representing a university, a commanding officer representing a military branch, or a priest representing a church; Filipas & Ullman, 2001). In these cases, mistreatment of assault survivors and mishandling of reported cases is referred to as institutional betrayal (IB). Empirically, IB is a fairly recent construct originating from betrayal trauma theory (Freyd, 1996). According to the theory, abuse perpetrated within the context of a close relationship (e.g., family member) has more severe effects on the victim because accompanying the abuse is a violation of trust. Recent anecdotal evidence regarding abuse perpetrated within the context of institutions, including the military, universities, and the Catholic Church, combined with research on the impact of systemic patterns of inadequate response to trauma (e.g., Campbell, 2006; Campbell & Raja, 2005; Carr et al., 2010) paved the way for betrayal trauma theory to extend to application of institutions.

IB can include acts of both commission and omission. Acts of omission by institutions may include a failure to take protective or preventative steps or failing to be responsive to allegations. Conversely, acts of commission by institutions may include more active and intentional behaviors, such as covering up or denying the experience, retaliation, or mishandling the case, if reported (Smith & Freyd, 2013). As with abuse perpetrated within a close relationship, victims feel betrayed by institutions because they have placed a level of trust in the institution, or because they are unavoidably dependent on the institution (Smith & Freyd, 2013). For example, a military service member may be sexually abused by a higher ranking official in his or her chain of command who would typically be responsible for responding to reports of abuse, leaving the victim with limited options. College students, military service members, and, more recently, Hollywood actors have come out publicly with claims that institutions had threatened retaliation following acts of sexual violence to silence victims and maintain the institution’s reputation (e.g., Felton, 2018; A. Shapiro, 2017; Ziering, Barklow, & Dick, 2012). Increased attention to these and other IB experiences highlighted the need for research examining the predictors and outcomes of IB.

Smith and Freyd (2013) created the Institutional Betrayal Questionnaire (IBQ) to begin to standardize how IB experiences are measured. IB has been linked to several adverse outcomes, including dissociation, anxiety, PTSD, sexual dysfunction, depression, suicide attempts, treatment noncompliance and disengagement, delayed help-seeking, and poorer perceived physical health (Altice, Mostashari, & Friedland, 2001; Armstrong et al., 2006; Monteith, Bahraini, Matarazzo, Soberay, & Smith, 2016; Smith, 2014; Smith & Freyd, 2013). The authors stated that principle components factor analysis did not suggest multiple dimensions for the IBQ—the eigenvalue for only one component was 1.96, which explained 28.03% of the variance in the measure (Smith & Freyd, 2013). However, the one-factor model left nearly 72% of unexplained random error, suggesting perhaps that the inclusion of multiple factors may better explain the variance in the measure.

The IBQ was updated to include five additional betrayal experiences, and, thus, psychometric analysis of the IBQ, Version 2 (IBQ.2), including convergent and discriminant validity, is lacking. Moreover, it is unclear whether the addition of five additional betrayal experiences altered the dimensionality of the measure. In one study, the five additional items were endorsed at comparable rates to the previously existing items and were similarly associated with physical health problems and dissociation. Analyses were run with and without the additional items and found that although the results did not meaningfully change, effect sizes were stronger with the inclusion of the additional items (Smith & Freyd, 2017). Taken together, these findings suggest that the five additional items are valid additions to the IBQ, however, their impact on the dimensionality of the measure remains unknown.

To date, only one study has examined the dimensionality of the IBQ.2 items. Tamaian and Klest (2017) developed and tested the psychometric properties of the Institutional Betrayal in the Medical System Questionnaire (IBQ-MS), a measure of betrayal perpetrated by healthcare systems. The wording of IBQ.2 items was altered to specifically relate to the medical system, and 30 additional items were added, resulting in 42 total items. The authors conducted exploratory factor analysis (EFA) with principal axis factoring and promax (oblique) rotation, resulting in a three-factor solution that explained 73.68% of the variance in healthcare betrayal prior to rotating. The three items were labeled (a) negative cognitive-affective reactions to healthcare providers/system (61.57% of variance explained); (b) provider/system factors that may lead to negative healthcare experiences and, ultimately, feelings of betrayal (7.33% of variance explained); and (c) system factors in response to a negative healthcare experience that contributes to feelings of betrayal (4.79% of variance explained). Of note, factor 1 was comprised entirely of newly created items unique to the IBQ-MS.¹ Although items were altered, and additional items included, findings from the IBQ-MS may help inform examination of two-factor models of the IBQ.2.

The current study sought to provide further evidence for the IBQ.2 as a valid measurement of IB by examining the convergent and discriminant validity of the measure. In addition, we sought to factor analyze the IBQ.2 to determine whether the scale represents a unidimensional construct or whether, like the IBQ-MS, the IBQ.2 is better represented by multifactor models. Based on theory and previous literature, with respect to convergent validity, we predicted that IB experiences would relate to more severe depression and anxiety symptom distress (Smith & Freyd, 2014) and PTSD (Monteith et al., 2016). Moreover, because the experience of IB may involve an institutions’ inadequate response to reported assaults (Smith & Freyd, 2014), IB was expected to positively correlate with disclosure to formal support providers (i.e., representatives of institutions) and greater endorsement of unsupportive acknowledgment and turning-against reactions, as well as less endorsement of positive reactions (Filipas & Ullman, 2001). Similarly, IB was expected to correlate with assaults that do not conform to rape myths (i.e., more severe IB correlated with less assault severity, less perpetrator use of threatened or actual physical force, greater survivor intoxication; Estrich, 1987; Frazier & Haney, 1996; Shaw, Campbell, Cain, & Feeney, 2017; Venema, 2016), given evidence that such factors may lead disclosure recipients to respond negatively to victims. Because assault experiences may impact certain assumptions about the self and the world (Janoff-Bulman, 1989), IB was expected to be related to more negative beliefs regarding self-worth, luck, self-control, benevolence of the world, and justice (Elklit, Shevlin, Solomon, & Dekel, 2007; Foa, Ehlers, Clark, Tolin, & Orsillo, 1999). With respect to discriminant validity, the timing of an individual’s first disclosure was not expected to be associated with IB due to insufficient empirical or theoretical links between these variables. We similarly predicted that the experience of IB would be unrelated to one’s adherence to rape myths, nor world beliefs regarding controllability of outcomes and randomness (Elklit et al., 2007; Foa et al., 1999; Janoff-Bulman, 1989). Last, we hypothesized that a one-factor model would yield a better fit for the data compared with competing two-factor models derived from findings from the IBQ-MS (Tamaian & Klest, 2017).

Method

Participants

Informed consent was obtained from 1,088 participants recruited for a larger study of sexual assault disclosure (see Pinciotti, Allen, Milliken, Orcutt, & Sasson, in press) from the crowdsourcing website, Amazon Mechanical Turk (MTurk). A total of 601 participants completed the survey, 52.6% (n = 495) of whom reported an unwanted sexual experience after the age of 14. In total, 69 participants were removed from analyses because they responded incorrectly to more than one of three forced response items (e.g., “If you are still reading, please select moderately agree”), leaving a final sample of 426. The final sample of survivors of sexual trauma ranged in age from 18 to 74 (M = 35.60, SD = 12.10) and identified as 73.7% female, 26.1% male, and 0.2% Other. The reported racial breakdown of the final sample was as follows: White (79.0%, n = 334), Black (8.5%, n = 36), Asian (5.2%, n = 22), American Indian or Alaskan Native (2.3%, n = 10), Native Hawaiian or Pacific Islander (0.7%, n = 3), Mixed Race (3.1%, n = 13), and Other racial background (1.2%, n = 5). An additional 8.4% identified as Hispanic or Latino/a (n = 35).

Measures

Acceptance of Modern Myths About Sexual Aggression Scale (AMMSA)

The AMMSA (Gerger, Kley, Bohner, & Siebler, 2007) is a 30-item self-report measure of acceptance of modern rape myths (e.g., “Nowadays, men who really sexually assault women are punished justly”). Items are measured on a 7-point Likert-type scale (1 = completely disagree to 7 = completely agree). The AMMSA has yielded very high internal consistencies in previous research (α = .90 to .95) and mean rape myth acceptance scores are higher on the AMMSA than a previous measure of rape myth acceptance, suggesting that the AMMSA captures more subtle beliefs than past rape myth scales (Gerger et al., 2007). In the current study, the AMMSA yielded strong internal consistency (α = .95), and participants rated their adherence to rape myths as 82.50 on average (SD = 30.45, range = 27 to 171).

Depression Anxiety and Stress Scale (DASS-21)

The DASS-21 (Lovibond & Lovibond, 1995) is a 21-item self-report scale of symptoms of depression, anxiety, and stress. Comprised of three related subscales, a total score on the DASS-21 represents a combination of these symptoms that reflects overall distress in these areas (e.g., “I felt down-hearted and blue”). Items are measured on a 4-point Likert-type scale (0 = did not apply to me at all to 3 = applied to me very much, or most of the time). The DASS-21 has yielded strong internal consistency (α = .93) and convergent validity with other measures of depression and anxiety (Henry & Crawford, 2005). In the current study, the DASS-21 also yielded strong internal consistency (α = .96), and participants rated their depression, anxiety, and stress as 18.61 on average (SD = 14.62, range = 0 to 63). It is recommended that the DASS-21 total score be doubled to produce levels that are equivalent to those derived from the full version (Henry & Crawford, 2005; Lovibond & Lovibond, 1995). In the current study, doubled scores for the DASS-21 produced an average total score of 37.22 (SD = 29.24). When compared with normative data, this score falls within the 89th percentile (Crawford & Henry, 2003), and is higher than those previously reported in nonclinical adult samples (Henry & Crawford, 2005; Sinclair et al., 2012). Thus, the present sample reported elevated levels of depression, anxiety, and stress.

Institutional Betrayal Questionnaire, Version 2 (IBQ.2)

The IBQ.2 (Smith, 2014) is a 15-item self-report measure of 12 IB experiences, with 3 additional items assessing relationship to and identification with the offending institution. Participants are asked to indicate, by selecting yes (1) or no (0), which of the 12 IB experiences they personally experienced, ranging from acts of omission (e.g., “Not taking proactive steps to prevent this type of experience?”) to acts of commission (e.g., “Covering up the experience?”). A total of 11.8% (n = 50) of participants reported experiencing at least one instance of IB (M = 0.48, SD = 1.31). This is lower than rates of IB reported in previous studies (Smith & Freyd, 2017; Wright, Smith, & Freyd, 2017) and may reflect our use of an MTurk sample. For instance, studies have found that MTurk workers report higher rates of unemployment than the general population (D. N. Shapiro, Chandler, & Mueller, 2013). Thus, this sample may have less institutional involvement and, thus, be at lower risk of encountering IB than samples used in previous IB research (e.g., college students, military service members). Consistent with previous research (Monteith et al., 2016; Smith & Freyd, 2013) and other checklists of traumatic experiences (e.g., Life Events Checklist; Gray, Litz, Hsu, & Lombardo, 2004), the current study computed a continuous variable of IB by summing all 12 experiences to indicate more severe IB. The three additional items were excluded so that the total IB score reflected only endorsement of betrayal experiences, consistent with previous studies.

Sexual Assault Severity Scale (SASS)

The SASS (Swinson, 2013) assessed assault characteristics. Although the SASS provides comprehensive information about assault characteristics, in the current study, variables were chosen based on relations with betrayal and negative reactions in previous literature. These variables included a single item assessing survivors’ perception of assault severity (0 = not at all severe to 8 = extremely severe; M = 4.10, SD = 2.20), a single item assessing survivors’ rating of their level of drug or alcohol intoxication at the time of the assault (0 = not at all drunk/high to 8 = black out drunk or extremely high; M = 2.13, SD = 3.02), and two yes/no items assessing whether the perpetrator threatened to use, and actually used, physical force. Threatened and actual use of physical force were combined into one dichotomous item of physical force (0 = did not threaten/use force, 1 = threatened/used force; M = .51, SD = .50). Among survivors in the current study, 38.8% used alcohol or recreational substances at the time of the assault, and 24.6% of perpetrators threatened to use, or actually used, physical force during the assault.

Sexual Assault Inventory of Disclosure (SAID)

The SAID (Pinciotti et al., in press) measures the content and context of sexual assault disclosures. However, the current study only utilized items examining survivors’ timing of first disclosure (1 = same day; 2 = 1 to 6 days later; 3 = 1 to 3 weeks later; 4 = 1 to 3 months later; 5 = 4 to 6 months later; 6 = 7 to 12 months later; 7 = 1 to 2 years later; 8 = 3 to 5 years later; 9 = 6 to 10 years later; 10 = more than 10 years later) and whether survivors reported disclosing to a formal support provider (1 = yes, 0 = no). In the current sample, 12% reported disclosing to at least one formal support provider (n = 51), most commonly to mental health professionals (8.5%). Survivors disclosed their assault on the same day 24% of the time, 51% of the sample disclosed within 6 days after the assault, and 70% of the sample disclosed within a year.

Sexual Experiences Survey–Short Form Victimization (SES-SFV)

The SES-SFV (Koss et al., 2007) measures sexual victimization history since age 14. The SES-SFV includes 17 items measuring behaviorally specific sexual assault experiences ranging from unwanted touching or clothing removed to completed rape (i.e., penetration). Participants endorse on a 4-point Likert-type scale how often each of the experiences happened to them since age 14 from 0 to 3+ times. All participants included in analyses endorsed at least one victimization experience since age 14, and 63.1% reported contact on the SES-SFV meeting the legal definition of rape.

PTSD Checklist (PCL-5) for Diagnostic and Statistical Manual of Mental Disorders–Fifth Edition (DSM-5)

The PCL-5 (Weathers et al., 2013) is a 20-item self-report measure of PTSD severity administered after the Life Events Checklist (LEC; Weathers et al., 2013). Following the LEC, participants are instructed to answer PCL items based on the event that they consider their worst event (i.e., is most bothersome currently) out of those listed on the LEC. The PCL items, anchored to this index trauma, reflect the four diagnostic symptom clusters of PTSD, including re-experiencing (e.g., “Repeated, disturbing, and unwanted memories of the stressful experience”), avoidance (e.g., “Avoiding memories, thoughts, or feelings related to the stressful experience”), hyperreactivity (e.g., “Feeling jumpy or easily startled”), and negative alterations in cognitions and mood (e.g., “Blaming yourself or someone else for the stressful experience or what happened after it”). Participants were asked to indicate on a 5-point Likert-type scale how much they were bothered in the past month by each symptom (0 = not at all to 4 = extremely). A total PCL-5 symptom severity score is calculated by summing all items and ranges from 0 to 80. In previous research, the PCL-5 has demonstrated very strong internal consistency (α = .94) and good test-retest reliability (r = .82; Bovin et al., 2016; Weathers et al., 2013). In the current study, internal consistency was strong (α = .96), and participants reported a mean PCL-5 score of 18.50 (SD = 18.87). This mean score is comparable yet slightly higher than those found in previous studies using the PCL-5 in trauma-exposed undergraduates (Blevins, Weathers, Davis, Witte, & Domino, 2015) and soldiers deployed to Iraq or Afghanistan (Hoge, Riviere, Wilk, Herrell, & Weathers, 2014).

Social Reactions Questionnaire (SRQ)

The SRQ (Ullman, 2000) is a 48-item self-report measure of assault-specific reactions received from others following victimization. The SRQ is comprised of three subscales: Positive Reactions (e.g., “Saw your side of things and did not make judgments”); Unsupportive Acknowledgment (e.g., “Told you to go on with your life”); and Turning Against (e.g., “Avoided talking to you or spending time with you”; Relyea & Ullman, 2015). Participants indicate how often they experienced each reaction from others following their assault on a 5-point Likert-type scale (0 = never to 4 = always). Each of the three subscales has demonstrated strong internal consistency in previous research (Positive Reactions [α = .92], Unsupportive Acknowledgment [α = .85], Turning Against [α = .92]) and convergent validity with perceived social support, coping, depression, PTSD, and self-blame (Relyea & Ullman, 2015). In the current study, internal consistency was also strong: Positive Reactions (α = .96), Unsupportive Acknowledgment (α = .90), Turning Against (α = .95); 79.9% of participants reported experiencing positive reactions, 76.3% reported experiencing unsupportive acknowledgment, and 69.4% reported experiencing turning-against reactions.

World Assumptions Scale (WAS)

The WAS (Janoff-Bulman, 1989) is a 32-item self-report measure of assumptions about the world, based upon the three principles of meaningfulness of the world, benevolence of the world, and worthiness of the self (e.g., “The good things that happen in this world far outnumber the bad”). The measure is comprised of 8 subscales: Benevolence of the World, Benevolence of People, Justice, Controllability, Randomness, Self-Worth, Self-Controllability, and Luck. Participants are asked to indicate the degree to which they agree or disagree with each statement on a 6-point Likert-type scale (1 = strongly disagree to 6 = strongly agree), with higher scores indicating greater endorsement of the assumption. The internal consistency for the total measure is good (α = .85; Lilly, Howell, & Graham-Bermann, 2015), and test–retest reliabilities for the subscales ranged from .66 to .78 in previous research (Janoff-Bulman, 1989). Cronbach’s alphas for these subscales have been in the marginal to adequate range: Benevolence of the World = .82, Benevolence of People = .63, Justice = .65, Controllability = .70, Randomness = .65, Self-Worth = .83, Self-Control = .72, and Luck = .84 (Harris & Valentiner, 2002). In the current study, Cronbach’s alphas were similar: Benevolence of the World = .84, Benevolence of People = .76, Justice = .66, Controllability = .77, Randomness = .68, Self-Worth = .83, Self-Control = .77, and Luck = .85.

Procedure

Participants were recruited to complete the Human Intelligence Task (HIT) on MTurk. Participants were informed that they would be asked sociodemographic questions as well as questions regarding their sexual experiences, and that their responses may make them eligible to complete additional items regarding stressful life experiences and mental health. The study was approved by the Institutional Review Board at the university at which the study was conducted.

After providing informed consent, participants reported their sociodemographic characteristics and history of adult sexual assault (i.e., after age 14). Participants who endorsed no history of adult sexual assault were debriefed and US$0.15 was credited to their MTurk account. Participants who endorsed a history of adult sexual assault were asked additional items about IB, assault-specific social reactions, world beliefs, and symptoms of depression, anxiety, stress, and PTSD; because the current study is part of a larger study on sexual assault disclosure (Pinciotti et al., in press), participants also answered items about sexual assault disclosure. Participants were then debriefed and credited a bonus of US$0.60 (total payment = US$0.75) for completing the additional items. All participants were provided a list of national sexual assault resources.

Data Analysis Plan

All variables were assessed for linearity and normality. Then, two-tailed bivariate correlations were performed to assess the relationship between the 12 IB experiences from the IBQ.2 and all other study variables. Given evidence of nonnormality, the current study computed Pearson’s correlation coefficients using 1,000 bootstrap samples with bias-corrected and accelerated (BCa) 95% confidence intervals (CI).² Data screening and correlations were performed in the Statistical Package for the Social Sciences (SPSS) Version 22.

Next, a series of confirmatory factor analyses (CFA) with maximum likelihood estimation were performed to produce robust estimators and standard errors using the statistical program, R.³ The following indices were used to assess model fit: Robust comparative fix index (CFI), robust Tucker–Lewis index (TLI), robust root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR).⁴ First, a single-factor model was tested in which the 12 IB experiences were indicators of one factor. Second, a single-factor model was tested for the original IBQ (i.e., without the new IB experiences introduced in the IBQ.2; Items 8, 9, 10, 11, and 12). Our third model examined two factors adapted from a medical system-specific IB questionnaire developed by Tamaian and Klest (2017). Although Tamaian and Klest (2017) found a three-factor structure for their measure, one of their three factors (“negative cognitive-affective reactions”) consisted of novel items not included in the IBQ.2. Thus, our third model tested their other two factors with modified factor names to reflect differences in content; Factor 1: “system factors that may lead to sexual victimization and ultimately feelings of betrayal” (e.g., not taking proactive steps to prevent the experience); and Factor 2: “system factors in response to sexual victimization that contribute to feelings of IB” (e.g., denying an individual’s experience; cf. Tamaian & Klest, 2017, p. 8). In addition, given that Tamaian and Klest (2017) removed Item 5 of the IBQ.2 from analyses (“Responding inadequately to the experience, if reported”) due to nonsignificant factor loadings, we tested this two-factor model with and without this item. Factor 1 and Factor 2 will hereafter be referred to as Leading to Sexual Victimization and Response to Sexual Victimization, respectively. Post hoc correlations were then performed to examine the construct validity of these two factors.

Results

Convergent and Discriminant Validity

See Table 1 for descriptive statistics, bivariate correlations, and BCa CIs. Correlations revealed partial support for our hypotheses. As predicted, the IBQ.2 demonstrated convergent validity through significant positive relations with disclosure to a formal support provider, survivors’ perception of assault severity, receipt of turning-against reactions, and world beliefs about self-controllability. However, the IBQ.2 surprisingly did not evidence convergent validity with perpetrators’ threatened or actual use of physical force, survivors’ level of intoxication, positive reactions, unsupportive acknowledgment reactions, world beliefs regarding self-worth, luck, world benevolence, and justice, and symptoms of depression, anxiety, stress, and PTSD. As predicted, the IBQ.2 evidenced discriminant validity with timing of first disclosure, rape myth adherence, and world beliefs regarding randomness and controllability of outcomes.

Table 1.

Descriptive Statistics, Bivariate Correlations, and 95% Confidence Intervals.

Scale	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21
1. IBQ.2	—
2. Leading to Victimization	.725^† [.550, .947]	—
3. Response to Victimization	.909^† [.735, .967]	.372^† [.115, .698]	—
4. SAID Formal Disclosure	.303^† [.062, .475]	.152 [–.074, .392]	.317^† [.120, .465]	—
5. SASS Physical Force	.165 [–.004, .292]	.071 [–.115, .260]	.180 [.039, .281]	.179 [.001, .357]	—
6. SASS Assault Severity	.301^† [.116, .430]	.147 [–.014, .338]	.317^† [.143, .427]	.263^† [.097, .409]	.408^† [.241, .553]	—
7. SASS Victim Intoxication	−.136 [–.231, .006]	−.065 [–.191, .097]	−.143 [–.221, –.028]	−.287^† [–.434, –.133]	−.160 [–.338, .034]	−.119 [–.309, .074]	—
8. SRQ Positive Reactions	−.008 [–.247, .348]	.020 [–.172, .228]	−.023 [–.281, .380]	.081 [–.107, .265]	−.047 [–.241, .139]	.029 [–.192, .271]	−.042 [–.235, .150]	—
9. SRQ Unsupportive Reactions	.144 [–.121, .413]	.058 [–.132, .279]	.159 [–.125, .472]	.337^† [.140, .537]	.244* [.071, .416]	.250* [.073, .403]	.019 [–.182, .217]	.204* [.015, .371]	—
10. SRQ Turning Against	.322^† [.013, .531]	.175 [–.008, .385]	.328^† [–.033, .549]	.279^† [.058, .476]	.266^† [.082, .422]	.381^† [.194, .533]	.088 [–.122, .292]	−.243* [–.417, –.044]	.658^† [.479, .804]	—
11. WAS Self-worth	.110 [–.260, .397]	.078 [–.177, .279]	.101 [–.302, .362]	.080 [–.116, .255]	.074 [–.126, .279]	.050 [–.172, .258]	−.060 [–.258, .148]	.156 [–.031, .342]	−.016 [–.218, .194]	−.033 [–.291, .202]	—
12. WAS Luck	−.009 [–.304, .215]	−.046 [–.258, .111]	.016 [–.317, .230]	−.042 [–.250, .143]	−.019 [–.203, .172]	−.126 [–.332, .086]	.020 [–.177, .217]	.138 [–.049, .332]	−.006 [–.206, .185]	−.075 [–.291, .143]	.749^† [.630, .834]	—
13. WAS Self-control	.229* [.047, .387]	.173 [.047, .228]	.204* [–.032, .360]	.214* [.037, .358]	.140 [–.041, .337]	.146 [–.081, .358]	−.141 [–.339, .064]	.141 [–.055, .327]	.209* [.033, .361]	.220* [.067, .360]	.695^† [.555, .809]	.322^† [.087, .527]	—
14. WAS World Benevolence	−.041 [–.398, .259]	−.085 [–.311, .146]	−.004 [–.407, .271]	.068 [–.150, .264]	−.047 [–.242, .167]	−.020 [–.236, .213]	.050 [–.123, .223]	.128 [–.081, .325]	−.160 [–.357, .048]	−.101 [–.357, .133]	.418^† [.176, .614]	.461^† [.259, .635]	.111 [–.115, .318]	—
15. WAS Justice	−.059 [–.261, .132]	.009 [–.175, .162]	−.085 [–.312, .136]	.019 [–.186, .209]	.153 [–.048, .346]	−.135 [–.344, .092]	−.018 [–.242, .202]	−.055 [–.285, .182]	.200* [–.007, .392]	.127 [–.100, .346]	−.011 [–.221, .216]	.001 [–.198, .215]	.101 [–.141, .324]	.081 [–.093, .249]	—
16. PCL	.119 [–.106, .383]	.138 [–.132, .400]	.077 [–.096, .351]	.152 [–.068, .358]	.127 [–.066, .311]	.126 [–.067, .302]	−.007 [–.187, .180]	.036 [–.158, .227]	.452^† [.251, .618]	.389^† [.167, .582]	−.336^† [–.511, –.130]	−.247* [–.432, –.048]	−.025 [–.233, .187]	−.378^† [–.539, –.209]	.080 [–.126, .274]	—
17. DASS	.007 [–.208, .265]	.009 [–.195, .215]	.004 [–.190, .288]	.014 [–.177, .216]	.059 [–.120, .243]	.055 [–.135, .248]	−.040 [–.214, .143]	−.088 [–.305, .121]	.306^† [.061, .518]	.304^† [.064, .499]	−.496^† [–.631, –.341]	−.278^† [–.463, –.085]	−.169 [–.332, –.001]	−.418^† [–.546, –.291]	.117 [–.070, .293]	.793^† [.700, .865]	—
18. SAID First Disclosure Timing	−.049 [–.209, .151]	−.011 [–.181, .171]	−.059 [–.207, .175]	−.062 [–.249, .130]	−.077 [–.265, .114]	.120 [–.082, .311]	−.071 [–.249, .121]	.172 [–.023, .363]	.026 [–.146, .219]	−.102 [–.277, .084]	−.114 [–.317, .123]	−.112 [–.302, .085]	−.064 [–.267, .128]	−.089 [–.269, .098]	.099 [–.120, .306]	.167 [–.020, .340]	.197* [.011, .365]	—
19. AMMSA Rape Myths	−.125 [–.281, .045]	−.164 [–.296, .003]	−.069 [–.214, .123]	.083 [–.080, .253]	.117 [–.062, .285]	.102 [–.096, .294]	−.110 [–.310, .080]	.027 [–.188, .234]	.357^† [.161, .530]	.313^† [.107, .494]	.001 [–.183, .176]	.109 [–.082, .293]	.060 [–.127, .241]	−.036 [–.241, .159]	.402^† [.222, .565]	.144 [–.061, .335]	.214* [.013, .383]	.090 [–.107, .268]	—
20. WAS Control of Outcomes	−.081 [–.281, .092]	.041 [–.151, .215]	−.134 [–.355, .013]	−.015 [–.190, .159]	−.026 [–.216, .149]	−.222* [–.413, –.023]	.052 [–.156, .258]	−.193* [–.398, .030]	.080 [–.115, .276]	.161 [.009, .310]	−.041 [–.249, .157]	.039 [–.189, .257]	.012 [–.235, .237]	.035 [–.176, .236]	.660^† [.552, .749]	.049 [–.133, .245]	.121 [–.079, .314]	−.065 [–.246, .115]	.461^† [.313, .596]	—
21. WAS Random	−.176 [–.317, .026]	−.040 [–.178, .118]	−.213* [–.360, –.018]	.005 [–.211, .221]	−.092 [–.271, .092]	.018 [–.210, .239]	−.034 [–.223, .168]	−.233* [–.444, –.017]	−.293^† [–.439, –.116]	−.184 [–.316, –.036]	−.147 [–.325, .036]	−.076 [–.295, .131]	−.263^† [–.455, –.050]	.059 [–.165, .258]	.146 [–.059, .349]	−.122 [–.267, .044]	−.070 [–.235, .096]	.031 [–.157, .193]	−.044 [–.209, .121]	.223* [–.053, .466]	—
M	0.48	0.22	0.26	0.26	0.51	4.10	2.13	1.68	0.84	0.71	48.30	14.25	17.48	16.65	11.95	18.50	37.22	3.83	82.50	12.58	11.74
SD	1.31	0.59	0.97	0.44	0.50	2.20	3.02	0.93	0.68	0.84	9.22	4.51	3.59	4.04	3.70	18.87	29.24	2.93	30.45	3.89	3.76

Note. Two-tailed Pearson’s correlation coefficients with bias-corrected and accelerated (BCa) 95% confidence intervals (CI) represented in brackets. Unless otherwise noted, bootstrap results are based on 1,000 bootstrap samples. DASS-21 scores were doubled to render levels equivalent with the full version; IBQ.2 = Institutional Betrayal Questionnaire, Version 2 (sum of 12 institutional betrayal experiences); Leading to Victimization = system factors that may lead to sexual victimization and ultimately feelings of betrayal (sum of IBQ.2 Items 1, 2, and 3; Factor 1); Response to Victimization = system factors in response to sexual victimization that contribute to feelings of institutional betrayal (sum of IBQ.2 Items 4, 5, 6, 7, 8, 9, 10, 11, and 12; Factor 2); SAID = Sexual Assault Inventory of Disclosure, Formal Disclosure (0 = no, 1 = yes); First Disclosure Timing (1 = same day to 10 = over 10 years later); SASS = Sexual Assault Severity Scale, Physical Force (0 = perpetrator did not threaten/use force; 1 = perpetrator threatened/used force); Assault Severity (0 = not at all severe to 8 = extremely severe); Victim Intoxication (0 = not at all drunk/high to 8 = black out drunk or extremely high); SRQ = Social Reactions Questionnaire (0 = never to 4 = always); WAS = World Assumptions Scale (1 = strongly disagree to 6 = strongly agree); PCL = PTSD Checklist for Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5; American Psychiatric Association, 2013) Sum Score (0 = not at all to 4 = extremely); DASS = Depression Anxiety and Stress Scale–21 Sum Score (0 = did not apply to me at all to 3 = applied to me very much, or most of the time); AMSSA = Acceptance of Modern Myths About Sexual Aggression Scale (1 = completely disagree to 7 = completely agree); PTSD = posttraumatic stress disorder.

p < .05. †p < .01.

Factor Structure

Single-factor models

The first single-factor model that included all 12 IBQ.2 items demonstrated poor fit: χ²(54, N = 426) = 72.207, p = .05, CFI = .53, TLI = .43, RMSEA = .03 (90% CI = [.022, .034]), and SRMR = .10. The next single-factor model only included items from the original IBQ scale and also evidenced poor fit: χ²(14, N = 426) = 23.774, p = .05, CFI = .70, TLI = .55, RMSEA = .04 (90% CI = [.031, .050]), and SRMR = .10, suggesting that neither the IBQ nor IBQ.2 is well-represented by a single factor of IB.

Two-factor models

Two-factor models (i.e., Leading to Sexual Victimization and Response to Sexual Victimization) with and without item 5 (“Responding inadequately to the experience, if reported”) were then tested, and both models were a poor fit to the data; without item 5: χ²(43, N = 426) = 55.362, p = .10, CFI = .68, TLI = .59, RMSEA = .03 (90% CI = [.019, .032]), and SRMR = .11; with item 5: χ²(53, N = 426) = 64.651, p = .13, CFI = .69, TLI = .61, RMSEA = .02 (90% CI = [.016, .029]), and SRMR = .106. Modification indices (MI = 92.53) suggested that the model fit might improve by adding a path predicting the Response to Sexual Victimization factor from item 4 (“Making it difficult to report the experience”). In addition, given that this factor represents responses to an individual’s experience, this item appears theoretically consistent with this factor. Indeed, this modification improved fit: χ²(53, N = 426) = 57.566, p = .31, CFI = .88, TLI = .85, RMSEA = .01 (90% CI = [.000, .022]), and SRMR = .09 (see Figure 1 for standardized factor loadings and each IBQ.2 item).

Figure 1.

Standardized factor loadings for IBQ.2 items.

Post Hoc Correlations

Based on the CFAs, post hoc correlations were performed to examine the two-factor model’s convergent and discriminant validity. The Leading to Sexual Victimization factor evidenced nonsignificant correlations with all other study variables. Conversely, the Response to Sexual Victimization factor demonstrated significant correlations with those same variables associated with the single-factor model (and in the same direction). The only variable that was uniquely related to the Response to Sexual Victimization factor was world beliefs about randomness.

Discussion

Sexual assault is thought to confer more severe health consequences when perpetrated by someone trusted due to the associated sense of betrayal (Ullman, 2007). Survivors of sexual assault may also experience betrayal following the failure of a trusted institution to prevent or respond appropriately to incidents of sexual assault. A recently operationalized construct, IB experiences have been associated with a myriad of health complaints (e.g., Altice et al., 2001; Armstrong et al., 2006; Monteith et al., 2016; Smith, 2014; Smith & Freyd, 2013). However, the recent changes to the IBQ.2 highlight the importance of establishing its psychometric properties. The current study investigated the IBQ.2’s construct validity and single-factor structure.

CFAs indicated superiority for a two-factor model, albeit adequate fit (see Figure 1). Leading to Sexual Victimization (Factor 1; 3 items) consists of “system factors that may lead to sexual victimization and ultimately feelings of betrayal” (e.g., creating an environment in which this type of experience seemed more likely to occur), and Response to Sexual Victimization (Factor 2; 9 items) reflects “system factors in response to sexual victimization that contribute to feelings of IB” (e.g., covering up an individual’s experience; cf. Tamaian & Klest, 2017, p. 8). Although this two-factor model provided only adequate fit to the data, taken together with the poor-fitting single-factor models, these findings nonetheless suggest that IB experiences may be best represented by more than one factor. The two factors suggested by Tamaian and Klest and the current study differ in that one factor focuses on system factors leading up to sexual victimization whereas the other factor focuses on system factors in response to sexual victimization. These factors may also be consistent with Smith and Freyd’s (2013) description of overt and covert forms of IB, such that system factors precipitating sexual victimization may involve covert acts of betrayal, and system factors in response to sexual victimization may involve overt acts of betrayal.

The IBQ.2’s construct validity was tested alongside several questionnaires. As expected, the IBQ.2 single-factor model showed convergence with disclosure to formal support sources, perceived assault severity, receipt of turning against reactions, and world beliefs regarding self-control. However, the IBQ.2 was not correlated with perpetrators’ physical force and survivors’ level of intoxication during the assault. Regarding assault-specific reactions, the IBQ.2 was surprisingly unrelated with positive or unsupportive reactions. In addition, the IBQ.2 was unrelated with other world beliefs regarding self-worth, luck, benevolence of the world, and justice. Also contrary to expectations, the IBQ.2 was not related with depression, anxiety, stress, and PTSD symptoms. The IBQ.2 demonstrated discriminant validity by being unrelated to timing of survivors’ first disclosure, adherence to rape myths, and world beliefs regarding randomness and controllability of outcomes, as predicted.

Based on findings from the factor analysis, we also tested the post hoc convergent and discriminant validity of the two factors, Leading to Sexual Victimization and Response to Sexual Victimization. Although the Leading to Sexual Victimization factor was unrelated to all other study variables, the Response to Sexual Victimization factor evidenced convergent validity largely consistent with the correlations of the single-factor model, apart from its significant negative relationship with world beliefs about randomness. Convergent validity findings from the two-factor model suggest that it is an institution’s inadequate response to sexual victimization, not the factors leading to sexual victimization, that may underlie changes in beliefs about the world. Given that the variables included in the current study correlated with certain assault and disclosure characteristics, it is intuitive that the Leading to Sexual Victimization factor would not evidence significant correlations, as this factor involves the pre-assault context. However, the nonsignificant relations between this factor and study variables do not imply that this aspect of IB is inconsequential. Rather, the moderately strong relationship between the two factors suggests that the contextual elements present in institutions may predict the institution’s subsequent response to sexual assault allegations and, accordingly, should perhaps be the primary focus for institutional change.

The IBQ.2’s relationship with disclosure to formal support sources and turning-against reactions is theoretically consistent. For instance, it may be a necessary condition to first disclose a sexual assault to a representative of a university, military branch, or church, for an institution to then perform an act of commission (e.g., retaliation). Likewise, it is precisely when such disclosures are met with negative reactions that one is at risk of experiencing IB. Moreover, the same characteristics that make survivors more likely to experience turning-against reactions from informal support sources (e.g., friends, family members, peers) may also put them at greater risk for experiencing IB from formal support sources. However, the IBQ.2’s nonsignificant relationship with other assault-specific reactions suggests no association between IB, positive reactions, and unsupportive acknowledgment, though future replication is needed to confirm whether IB is related to these assault-specific reactions.

Furthermore, although the IBQ.2 was related to perceived assault severity, it was unrelated to perpetrators’ use of physical force and survivors’ level of intoxication. Research indicates that certain assault characteristics (e.g., perceived life threat) may increase the likelihood that disclosure recipients will respond negatively (Ullman & Siegel, 1995; Ullman et al., 2008). However, when concerning IB, these findings indicate that survivors’ perceived assault severity may be a stronger determinant of receiving negative reactions than physical force or being intoxicated during the assault. This finding mirrors previous research that has found that greater assault severity predicts disclosure to formal support providers such as law enforcement and physicians (Golding et al., 1989). Thus, it appears that greater assault severity may not only make survivors more likely to disclose to formal support sources, including to those who represent institutions, but also puts them at risk for being mistreated by these sources.

Of world beliefs, the IBQ.2 was only related with stronger beliefs about self-control, and showed no significant associations with beliefs about luck, benevolence of the world, or justice. Beliefs about self-control reflect the degree to which people feel they are actively behaving to exert control over outcomes (Elklit et al., 2007). The development of different world beliefs is thought to depend on the type of trauma experienced, and it has been suggested that beliefs about self-control, specifically, may emerge following sexual victimization (Elklit et al., 2007), possibly because survivors become more vigilant after their assault experiences (e.g., “I almost always make an effort to prevent bad things from happening to me”). Thus, these findings may suggest that self-control beliefs might also be an outgrowth of IB, perhaps due to potential secondary victimization (e.g., victim blaming; see also Campbell & Raja, 2005). Alternatively, it may also be that those with greater self-control beliefs are more likely to report their assault experiences (e.g., “I take the actions necessary to protect myself against misfortune”), thereby increasing the chance of encountering IB. Taken with previous research (Wright et al., 2017), the nonsignificant findings regarding luck, world benevolence, and justice suggest that these world beliefs may not be impacted by IB.

The IBQ.2’s nonsignificant association with measures of stress, depression, anxiety, and PTSD symptoms was unexpected. Given that previous research has found links between anxiety and IB as measured by the original IBQ (Smith & Freyd, 2013), it may be that the five additional IB items in the IBQ.2 represent less distressing experiences, thereby weakening its overall relationship with psychopathology. Furthermore, although the IBQ.2 has previously been associated with PTSD and depression symptoms (Smith, Cunningham, & Freyd, 2016), one study found that, after excluding Veterans who reported experiencing exclusively military sexual harassment, IB experiences did not predict PTSD symptoms among Veterans with a history of military sexual assault (Monteith et al., 2016). Taken together with the current study’s findings, this may suggest that the association between IB and PTSD symptoms varies across samples of sexual assault survivors. Future research may wish to investigate whether IB differentially predicts psychopathology as a function of sample type (e.g., military, online, college student).

The IBQ.2 demonstrated discriminant validity with timing of first disclosure, adherence to rape myths, and world beliefs regarding randomness and controllability of outcomes. These findings indicate that IB experiences may be equally likely to occur regardless of whether survivors choose to disclose their assault to others right away or after a delay. Furthermore, IB appears distinct from survivors’ acceptance of modern rape myths, and beliefs about controllability of outcomes and randomness. As such, the belief that people’s behavior determines outcomes, or that outcomes occur as a matter of chance, may not form following sexual assault (Elklit et al., 2007), and, in turn, have no relationship with IB (Wright et al., 2017). However, when separated into two factors, beliefs about randomness were negatively related with the Response to Sexual Victimization factor. This suggests that overt institutional betrayal experiences (e.g., covering up the experience, being punished for reporting the experience) may undermine the belief that outcomes occur by chance. This relationship appears theoretically consistent, as this factor reflects how individuals may have been actively suppressed by institutions while trying to report their assaults, thus, possibly removing the inclination to view what happens to them as chance occurrences.

Limitations

The current study had several limitations. First, a small portion of participants endorsed at least one instance of IB (11.8%). The low rate of IB in the current study might be a by-product of our use of Amazon MTurk. For instance, previous research has found lower rates of employment in MTurk samples as compared with the general population (D. N. Shapiro et al., 2013). This may suggest that MTurk samples have limited exposure to institutions and subsequent IB, as opposed to previous research on IB using samples belonging to specific institutions, such as college undergraduates (Smith & Freyd, 2013, 2017) and Veterans (Monteith et al., 2016). Nonetheless, the low rate of IB may explain several nonsignificant relationships between the IBQ.2 and other variables. Relatedly, the nonsignificant relationship between the IBQ.2 and perpetrator’s threatened or actual use of physical force may be due to the latter item’s low variability (M = 0.51, SD = 0.50). Second, the current study’s cross-sectional design precludes any definitive conclusions regarding causality and may also account for the nonsignificant associations found between the IBQ.2 and mental health outcomes. Researchers may examine how IB prospectively impacts later functioning. Third, this study’s reliance on self-report measures introduces potential measurement error that may have obfuscated findings (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). Last, findings may not generalize to other samples. Despite research that suggests an overrepresentation of unemployed individuals in MTurk samples (D. N. Shapiro et al., 2013), other studies have found that MTurk samples tend to be more representative of racial trends in the U.S. population than other convenience samples (e.g., Paolacci, Chandler, & Ipeirotis, 2010). However, findings from the current study are based primarily on data from White participants (79%), which limits the generalizability of these findings to individuals from other racial backgrounds. For instance, previous research has found that being non-White is correlated with reporting more IB experiences (Tamaian & Klest, 2017). Furthermore, belief in a just world, some of which is measured by the WAS, appears to vary by race/ethnicity, with one study finding that Latinos report the highest adherence to just-world beliefs, followed by Whites and African Americans (Hunt, 2000). Future studies may investigate these relationships among more racially diverse samples.

Conclusions and Future Directions

Despite these limitations, the current study provides preliminary evidence regarding the IBQ.2’s psychometric properties. Given the discrepant, nonsignificant relationships between IB and mental health outcomes, future studies should seek to replicate current findings by continuing to investigate how the IBQ.2 relates to symptoms of distress (e.g., examining how IB mediates sexual assault and postassault functioning; Smith et al., 2016). Relatedly, the IBQ.2’s items regarding identification with and relationship to the institution may represent important mediators relating IB to mental health. In addition, research may investigate how the IBQ.2 functions within the context of other negative experiences, different trauma types, or specific institutions (e.g., Tamaian & Klest, 2017). Last, future studies may test this factor structure against alternative models of IB, though replication of a two-factor structure may suggest that there are two distinct types of IB experienced by survivors of sexual victimization. If true, it will be important to examine the prevalence, predictors, and differential impact of both types of IB, and how this two-factor scoring relates to other constructs. A clear understanding of the complex processes involved when betrayal occurs will be informative for research, trainings, and interventions targeted at reducing these betrayal behaviors, as well as for mental health professionals who may be tasked with treating the consequences of IB.

Footnotes

Declaration of Conflicting Interests

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

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Anthony N. Reffi

Notes

Author Biographies

Anthony N. Reffi, BA, is a graduate student in the clinical psychology program at Northern Illinois University. His research focuses on risk and resiliency factors related to posttraumatic stress disorder (PTSD) following traumatic exposure, such as emotion regulation, perceived coping ability, mindfulness, and sleep.

Caitlin M. Pinciotti, MA, is a doctoral candidate in the clinical psychology program at Northern Illinois University. Her research focuses on sexual assault victimization, including factors that facilitate or inhibit healing, such as the availability of social support, disclosure experiences, world beliefs, and interventions such as self-defense and cognitive bias modification.

Holly K. Orcutt, PhD, is a professor in the Department of Psychology at Northern Illinois University. Her research focuses on the broader framework of traumatic stress, including the prevention of trauma exposure, and risk and prevention factors such as emotion regulation, acceptance, and avoidance.

References

Ahrens

C. E.

(2006). Being silenced: The impact of negative social reactions on the disclosure of rape. American Journal of Community Psychology, 38, 31-34. doi:10.1007/s10464-006-9069-9

Ahrens

C. E.

Campbell

Ternier-Thames

Wasco

Sefl

(2007). Deciding whom to tell: Expectations and outcomes of sexual assault survivors’ first disclosures. Psychology of Women Quarterly, 31, 38-49. doi:10.1111/j.1471-6402.2007.00329.x

Altice

F. L.

Mostashari

Friedland

G. H.

(2001). Trust and the acceptance of and adherence to antiretroviral therapy. Journal of Acquired Immune Deficiency Syndromes, 28, 47-58. doi:10.1097/00042560-200109010-00008

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.

Andrews

Brewin

C. R.

Rose

(2003). Gender, social support, and PTSD in victims of violent crime. Journal of Traumatic Stress, 16, 421-427. doi:10.1023/A:1024478305142

Armstrong

D. K.

Rose

Peters

Long

J. A.

McMurphy

Shea

J. A.

(2006). Distrust of the health care system and self-reported health in the United States. Journal of General Internal Medicine, 21, 292-297. doi:10.1111/j.1525-1497.2006.00396.x

Bachman

(1993). Predicting the reporting of rape victimizations: Have rape reforms made a difference? Criminal Justice and Behavior, 20, 254-270.

Bandalos

D. L.

Finney

S. J.

(2001). Item parceling issues in structural equation modeling. In Marcoulides

G. A.

Schumacker

R. E.

(Eds.), New developments and techniques in structural equation modeling (pp. 269-296). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.

Blevins

C. A.

Weathers

F. W.

Davis

M. T.

Witte

T. K.

Domino

J. L.

(2015). The Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5): Development and initial psychometric evaluation. Journal of Traumatic Stress, 28, 489-498. doi:10.1002/jts.22059

10.

Bovin

M. J.

Marx

B. P.

Weathers

F. W.

Gallagher

M. W.

Rodriguez

Schnurr

P. P.

Keane

T. M.

(2016). Psychometric properties of the PTSD Checklist for Diagnostic and Statistical Manual of Mental Disorders–Fifth Edition (PCL-5) in veterans. Psychological Assessment, 28, 1379-1391. doi:10.1037/pas0000254

11.

Browne

M. W.

Cudeck

(1993). Alternative ways of assessing model fit. Sage Focus Editions, 154, 136-136.

12.

Campbell

(2006). Rape survivors’ experiences with the legal and medical systems: Do rape victim advocates make a difference? Violence Against Women, 12(1), 30-45. doi:10.1177/1077801205277539

13.

Campbell

Raja

(2005). The sexual assault and secondary victimization of female veterans: Help-seeking experiences in military and civilian social systems. Psychology of Women Quarterly, 29, 97-106. doi:10.1111/j.1471-6402.2005.00171.x

14.

Carr

Dooley

Fitzpatrick

Flanagan

Flanagan-Howard

Tierney

. . . Egan

(2010). Adult adjustment of survivors of institutional child abuse in Ireland. Child Abuse & Neglect, 34(7), 477-489. doi:10.1016/j.chiabu.2009.11.003

15.

Crawford

J. R.

Henry

J. D.

(2003). The Depression Anxiety Stress Scales (DASS): Normative data and latent structure in a large non-clinical sample. British Journal of Clinical Psychology, 42, 111-131. doi:10.1348/014466503321903544

16.

Efron

Tibshirani

(1993). An introduction to the bootstrap. Boca Raton, FL: Chapman & Hall/CRC.

17.

Elklit

Shevlin

Solomon

Dekel

(2007). Factor structure and concurrent validity of the world assumptions scale. Journal of Traumatic Stress: Official Publication of The International Society for Traumatic Stress Studies, 20(3), 291-301. doi:10.1002/jts.20203

18.

Estrich

(1987). Real rape: How the legal system victimizes women who say no. Cambridge, MA: Harvard University Press.

19.

Feldman-Summers

Norris

(1984). Differences between rape victims who report and those who do not report to a public agency. Journal of Applied Social Psychology, 14, 532-573. doi:0.1111/j.1559-1816.1984.tb02260.x

20.

Felton

(2018). How colleges foretold the #MeToo movement. Retrieved from https://www.theatlantic.com/education/archive/2018/01/how-colleges-foretold-the-metoo-movement/550613/

21.

Field

(2013). Discovering statistics using IBM SPSS statistics (4th ed.). Thousand Oaks, CA: SAGE.

22.

Filipas

H. H.

Ullman

S. E.

(2001). Social reactions to victims of sexual assault from social network members. Violence and Victims, 16, 673-692.

23.

Foa

E. B.

Ehlers

Clark

D. M.

Tolin

D. F.

Orsillo

S. M.

(1999). The posttraumatic cognitions inventory (PTCI): Development and validation. Psychological Assessment,11(3), 303.

24.

Frazier

P. A.

Haney

(1996). Sexual assault cases in the legal system: Police, prosecutor, and victim perspectives. Law and Human Behavior, 20, 607-628. doi:10.1007/BF01499234

25.

Freyd

J. J.

(1996). Betrayal trauma: The logic of forgetting childhood abuse. Cambridge, MA: Harvard University Press.

26.

Gerger

Kley

Bohner

Siebler

(2007). The Acceptance of Modern Myths About Sexual Aggression (AMMSA) Scale: Development and validation in German and English. Aggressive Behavior, 33, 422-440.

27.

Golding

J. M.

Siegel

J. M.

Sorenson

S. B.

Burnam

M. A.

Stein

J. A.

(1989). Social support sources following sexual assault. Journal of Community Psychology, 17, 92-107. doi:10.1002/1520-6629(198901)17:1<92::AID-JCOP2290170110>3.0.CO;2-E

28.

Gray

M. J.

Litz

B. T.

Hsu

J. L.

Lombardo

T. W.

(2004). Psychometric properties of the Life Events Checklist. Assessment, 11, 330-341. doi:10.1177/1073191104269954

29.

Harris

H. N.

Valentiner

D. P.

(2002). World assumptions, sexual assault, depression and fearful attitudes toward relationships. Journal of Interpersonal Violence, 17, 286-305. doi:10.1177/0886260502017003004

30.

Henry

J. D.

Crawford

J. R.

(2005). The short form version of the Depression Anxiety Stress Scales (DASS-21): Construct validity and normative data in a large non-clinical sample. The British Psychological Society, 44, 227-239. doi:10.1348/014466505X29657

31.

Hoge

C. W.

Riviere

L. A.

Wilk

J. E.

Herrell

R. K.

Weathers

F. W.

(2014). The prevalence of post-traumatic stress disorder (PTSD) in US combat soldiers: A head-to-head comparison of DSM-5 versus DSM-IV-TR symptom criteria with the PTSD checklist. The Lancet Psychiatry, 1, 269-277. doi:10.1016/S2215-0366(14)70321-9

32.

L. T.

Bentler

P. M.

(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6, 1-55. doi:10.1080/10705519909540118

33.

Hunt

M. O.

(2000). Status, religion, and the “belief in a just world”: Comparing African Americans, Latinos, and Whites. Social Science Quarterly, 81, 325-343. Retrieved from https://www-jstor-org.web.bisu.edu.cn/stable/42864385

34.

Janoff-Bulman

(1989). Assumptive worlds and the stress of traumatic events: Applications of the schema construct. Social Cognition, 7, 113-136. doi:10.1521/soco.1989.7.2.113

35.

Koss

M. P.

Abbey

Campbell

Cook

Norris

Testa

. . . White

(2007). Revising the SES: A collaborative process to improve assessment of sexual aggression and victimization. Psychology of Women Quarterly, 31, 357- 370. doi:10.1111/j.1471-6402.2007.00385.x

36.

Lilly

M. M.

Howell

K. H.

Graham-Bermann

(2015). World assumptions, religiosity, and PTSD in survivors of intimate partner violence. Violence Against Women, 21, 87-104. doi:10.1177/1077801214564139

37.

Lovibond

P. F.

Lovibond

S. H.

(1995). The structure of negative emotional states: Comparison of the Depression Anxiety Stress Scales (DASS) with the Beck Depression and Anxiety Inventories. Behaviour Research and Therapy, 33, 335-343.

38.

MacCallum

R. C.

Browne

M. W.

Sugawara

H. M.

(1996). Power analysis and determination of sample size for covariance structure modeling. Psychological Methods, 1, 130-149. doi:10.1037/1082-989X.1.2.130

39.

Monteith

L. L.

Bahraini

N. H.

Matarazzo

B. B.

Soberay

K. A.

Smith

C. P.

(2016). Perceptions of institutional betrayal predict suicidal self-directed violence among veterans exposed to military sexual trauma. Journal of Clinical Psychology, 72, 743-755. doi:10.1002/jclp.22292

40.

Paolacci

Chandler

Ipeirotis

P. G.

(2010). Running experiments on Amazon Mechanical Turk. Judgment and Decision Making, 5, 411-419.

41.

Pinciotti

C. M.

Allen

C. A.

Milliken

J. M.

Orcutt

H. K.

Sasson

(in press). Reliability and findings from an instrument examining sexual assault disclosure content and context: The Sexual Assault Inventory of Disclosure (SAID). Violence and Victims.

42.

Podsakoff

P. M.

MacKenzie

S. B.

Lee

J. Y.

Podsakoff

N. P.

(2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879-903. doi:10.1037/0021-9010.88.5.879

43.

Relyea

Ullman

S. E.

(2015). Unsupported or turned against: Understanding how two types of negative social reactions to sexual assault relate to postassault outcomes. Psychology of Women Quarterly, 39, 37-52. doi:10.1177/0361684313512610

44.

Shapiro

(2017). Where does Hollywood go after numerous allegations of sexual assault and harassment? Retrieved from https://www.npr.org/2017/10/23/559618445/where-does-hollywood-go-after-numerous-allegations-of-sexual-assault-and-harassm

45.

Shapiro

D. N.

Chandler

Mueller

P. A.

(2013). Using Mechanical Turk to study clinical populations. Clinical Psychological Science, 1, 213-220. doi:10.1177/2167702612469015

46.

Shaw

Campbell

Cain

Feeney

(2017). Beyond surveys and scales: How rape myths manifest in sexual assault police records. Psychology of Violence, 7, 602-614. doi:10.1037/vio0000072

47.

Sinclair

S. J.

Siefert

C. J.

Slavin-Mulford

J. M.

Stein

M. B.

Renna

Blais

M. A.

(2012). Psychometric evaluation and normative data for the depression, anxiety, and stress scales-21 (DASS-21) in a nonclinical sample of US adults. Evaluation & the Health Professions, 35, 259-279. doi:10.1177/0163278711424282

48.

Smith

C. P.

(2014, April). Unawareness and expression of interpersonal and institutional betrayal. Paper Presented at the Western Psychological Association, Portland, OR.

49.

Smith

C. P.

Cunningham

S. A.

Freyd

J. J.

(2016). Sexual violence, institutional betrayal, and psychological outcomes for LGB college students. Translational Issues in Psychological Science, 2, 351-360. doi:10.1037/tps0000094

50.

Smith

C. P.

Freyd

J. J.

(2013). Dangerous safe havens: Institutional betrayal exacerbates sexual trauma. Journal of Traumatic Stress, 26, 119-124.

51.

Smith

C. P.

Freyd

J. J.

(2014). Institutional betrayal. American Psychologist, 69, 575-587. doi:10.1037/a0037564

52.

Smith

C. P.

Freyd

J. J.

(2017). Insult, then injury: Interpersonal and institutional betrayal linked to health and dissociation. Journal of Aggression, Maltreatment & Trauma, 26, 1117-1131. doi:10.1080/10926771.2017.1322654

53.

Starzynski

L. L.

Ullman

S. E.

Filipas

H. H.

Townsend

S. M.

(2005). Correlates of women’s sexual assault disclosure to informal and formal support services. Violence and Victims, 20, 417-432. doi:10.1891/0886-6708.20.4.417

54.

Swinson

K. C.

(2013). The sexual assault severity scale: A comprehensive measure of assault severity (Unpublished doctoral dissertation). University of Southern Mississippi, Hattiesburg.

55.

Tamaian

Klest

(2017). Institutional betrayal in the Canadian medical system: Reliability and validity of a self-report questionnaire. Journal of Aggression, Maltreatment & Trauma, 27(7), 703-719. doi:10.1080/10926771.2017.1294638

56.

Ullman

S. E.

(1996). Social reactions, coping strategies, and self-blame attributions in adjustment to sexual assault. Psychology of Women Quarterly, 20, 505-526. doi:10.1111/j.1471-6402.1996.tb00319.x

57.

Ullman

S. E.

(2000). Psychometric characteristics of the Social Reactions Questionnaire: A measure of reactions to sexual assault victims. Psychology of Women Quarterly, 24, 257-271. doi:10.1111/j.1471-6402.2000.tb00208.x

58.

Ullman

S. E.

(2007). Relationship to perpetrator, disclosure, social reactions, and PTSD symptoms in child sexual abuse survivors. Journal of Child Sexual Abuse, 16, 19-36. doi:10.1300/J070v16n01_02

59.

Ullman

S. E.

(2014). Correlates of posttraumatic growth in adult sexual assault victims. Traumatology, 20(3), 219. doi:10.1037/h0099402

60.

Ullman

S. E.

Filipas

H. H.

(2001). Correlates of formal and informal support seeking in sexual assault victims. Journal of Interpersonal Violence, 16, 1028-1047. doi:10.1177/088626001016010004

61.

Ullman

S. E.

Siegel

J. M.

(1995). Sexual assault, social reactions, and physical health. In Landrine

Klonoff

E. A.

(Eds.), Women’s health: Research on gender, behavior, and policy (Vol. 1, pp. 289-308). Hillsdale, NJ: Lawrence Erlbaum.

62.

Ullman

S. E.

Starzynski

L. L.

Long

S. M.

Mason

G. E.

Long

L. M.

(2008). Exploring the relationships of women’s sexual assault disclosure, social reactions and problem drinking. Journal of Interpersonal Violence, 23, 1235-1257. doi:10.1177/0886260508314298

63.

Venema

R. M.

(2016). Police officer schema of sexual assault reports: Real rape, ambiguous cases, and false reports. Journal of Interpersonal Violence, 31, 872-899. doi:10.1177/0886260514556765

64.

Weathers

F.W.

Litz

B.T.

Keane

T.M.

Palmieri

P.A.

Marx

B.P.

Schnurr

P.P.

(2013). The PTSD Checklist for DSM-5 (PCL-5). Scale available from the National Center for PTSD at www.ptsd.va.gov

65.

Worthington

R. L.

Whittaker

T. A.

(2006). Scale development research: A content analysis and recommendations for best practices. The Counseling Psychologist, 34, 806-838. doi:10.1177/0011000006288127

66.

Wright

D. B.

London

Field

A. P.

(2011). Using bootstrap estimation and the plug-in principle for clinical psychology data. Journal of Experimental Psychopathology, 2, 252-270. doi:10.5127/jep.013611

67.

Wright

N. M.

Smith

C. P.

Freyd

J. J.

(2017). Experience of a lifetime: Study abroad, trauma, and institutional betrayal. Journal of Aggression, Maltreatment & Trauma, 26, 50-68. doi:10.1080/10926771.2016.1170088

68.

Ziering

Barklow

T. K.

(Producer), & Dick

(Director). (2012). The invisible war [Motion picture]. USA: Chain Camera Pictures.