Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 May 31.
Published in final edited form as: J Community Psychol. 2018 Sep 12;47(2):371–384. doi: 10.1002/jcop.22126

Psychometric validation of the identity abuse scale among LGBTQ individuals

Jillian R Scheer 1, Julie M Woulfe 2, Lisa A Goodman 3
PMCID: PMC6543831  NIHMSID: NIHMS1027897  PMID: 30207588

Abstract

Identity abuse (IA) comprises a set of abuse tactics that exploit discriminatory systems including homophobia, biphobia, and transphobia (Tesch & Berkerian, 2015). This study examined the factorial validity of the IA Scale (Woulfe & Goodman, 2018) with a large independent sample of lesbian, gay, bisexual, transgender, and queer (LGBTQ) individuals. Participants included 1,049 LGBTQ-identified participants (Mage = 27.3, 71.9% White, 52.6% cisgender women, and 18.7% as other nonheterosexual identity in their sexual orientation), recruited through listservs. Participants completed an online survey measuring past-year and adult exposure to identity, physical, and psychological abuse. Confirmatory factor analysis indicated that the measurement model had good fit to the data, and strong factor loadings were found across the seven items, confirming a unidimensional factor structure. Findings demonstrate the IA Scale’s validity and reliability, supporting its use to assess the frequency of IA tactics experienced within intimate partnerships among LGBTQ individuals.

1 |. INTRODUCTION

Intimate partner violence (IPV), including emotional, physical, or sexual violence, is a pervasive and devastating problem that can affect people across any gender identity or sexual orientation (Suarez et al., 2018). Although most scholarship, to date, has explored IPV exposure among cisgender (nontransgender), heterosexual females (Finneran, & Stephenson, 2013; Kubicek, McNeeley, & Collins, 2015), mounting evidence suggests that LGBTQ individuals are at equal or even greater risk of IPV exposure relative to cisgender, heterosexual individuals (Dank, Lachman, Zweig, & Yahner, 2014; Edwards, Sylaska, & Neal, 2015; Stotzer, 2009). For example, the National Violence Against Women survey found that 21.5% men and 35.4% women with histories of same-sex relationships reported physical abuse from a partner, compared with 7.1% heterosexual men and 20.4% heterosexual women (Tjaden & Thoennes, 2000).

IPV patterns may be distinct across various LGBTQ subgroups. For example, Goldberg and Meyer (2013) demonstrated that women who identified as bisexual reported higher rates of IPV in the past year (27.9%) and over the lifespan (52.0%) compared with lesbian women with IPV exposure in the past year (10.2%) and across the lifespan (31.9%). Moreover, although it is difficult to accurately assess IPV prevalence rates among transgender and gender nonconforming (TGNC) individuals, given that IPV is underreported in these communities, researchers have documented that they may have heightened levels of risk of IPV exposure compared with cisgender, heterosexual and sexual minority men and women (National Coalition of Anti-Violence Programs [NCAVP], 2012).

1.1 |. IPV forms unique to LGBTQ individuals

The nature of IPV against LGBTQ individuals may be characteristically different from that used against cisgender, heterosexual individuals, given their experiences of structural and interpersonal stigma. For example, LGBTQ individuals contend with larger heterosexist and transphobic contexts, as well as chronic minority stressors related to their LGBTQ identity, including discrimination, actual and perceived rejection, and internalized stigma (Balsam & Szymanski, 2005; Meyer, 2003; Otis, Rostosky, Riggle, & Hamrin, 2006; Reuter, Newcomb, Whitton, & Mustanski, 2017). Taken together, this broader context of stigma may fuel abusive power dynamics in LGBTQ intimate relationships (Guadalupe-Diaz, & Anthony, 2017). However, there is limited research to date focusing specifically on the unique dynamics of IPV among LGBTQ individuals (Ristock, 2011).

A few researchers have used the sparse extant literature on LGBTQ IPV to suggest unique patterns of abuse that might be salient among LGBTQ communities (Tesch & Bekerian, 2015). These authors have focused specifically on identity abuse (IA). IA refers to the ways that abusers may use homophobic, biphobic, and transphobic societal and structural norms against their LGBTQ partner, discrediting, undermining, or devaluating their already stigmatized sexual or gender identity (Ard & Makadon, 2011; Goodmark, 2013; Guadalupe-Diaz & Anthony, 2016; Tesch & Berkerian, 2015; Woulfe & Goodman, 2018). Relevant to transgender individuals, for example, Guadalupe-Diaz and Anthony (2016) documented how tailoring abuse toward discrediting or manipulating transgender individuals’ gender identity can be detrimental to self-worth, identity integration, and overall well-being. Abusers who use IA may themselves also identify as LGBTQ. Indeed, having negative feelings about being LGBTQ (e.g., internalized discrimination) has been shown to contribute to greater violence against LGBTQ partners (Balsam & Szymanski, 2005; Renzitti, 1998).

Many health care organizations recommend routine screening in general for experiences of IPV (Waalen, Goodwin, Spitz, Petersen, & Saltzman, 2000). In addition, it is also critical for providers to assess for IA among LGBTQ individuals. Currently, however, there exists only one measure that has been previously psychometrically evaluated to assess for LGBTQ-specific IA, the Identity Abuse Scale (IA Scale; Woulfe & Goodman, 2018); and this measure has only been subject to exploratory factor analysis (EFA). Given the preliminary nature of EFA (Limoubpratum, 2015), psychometric validation requires confirming the IA Scale factor structure in a new, independent sample using confirmatory factor analysis (CFA; Innamorati et al., 2017). This current project aims to address this limitation in previous literature by confirming the validity of the IA Scale for use in a large sample of diverse LGBTQ individuals.

1.2 |. Identity abuse in LGBTQ communities

Although IA can be used against any marginalized group, such as people of color (i.e., non-White), in the LGBTQ context, it involves cissexist and heterosexist tactics that emphasize LGBTQ IPV individuals’ marginalized position in society. In a recent article, Woulfe and Goodman (2018) delineated four domains of IA tactics that can be broadly applied to the LGBTQ community: (a) disclosing a partner’s LGBTQ status; (b) undermining, attacking, or denying a partner’s LGBTQ identity; (c) using slurs or derogatory language regarding a partner’s LGBTQ status; and, (d) isolating a partner from LGBTQ communities (Woulfe & Goodman, 2018).

The first set of IA domains involves intentionally threatening to disclose a partner’s marginalized sexual orientation or gender identity without their consent (Woulfe & Goodman, 2018). LGBTQ individuals’ concealment of their sexual orientation or gender identity may present as a barrier to seeking support from family, friends, or law enforcement, further isolating them in abusive relationships (Ard & Makadon, 2011; Kulkin, Williams, Borne, de la Bretonne, & Laurendine, 2007; Marcotte, 2016). However, it is important to note that concealing a stigmatized identity may also be used as a coping mechanism, for example, by avoiding negative consequences of stigma, such as workplace discrimination (Meyer, 2003; Miller & Major, 2000). Further, LGBTQ individuals may decide to conceal their stigmatized sexual orientation or gender identity based on actual or anticipated rejection, as a method of protection against shame or feeling unsafe (Meyer, 2003). When an abuser threatens to or does reveal their partner’s sexual orientation or gender identity, this undermines LGBTQ individuals’ agency to choose to expose their stigmatized identity, potentially making them feel unsafe even outside of the relationship.

According to Woulfe and Goodman (2018), a second dimension of IA includes actively undermining, attacking, or denying a partner’s sexual orientation or gender identity (Balsam & Szymanski, 2005; FORGE, 2014; Guadalupe-Diaz & Anthony, 2016; National Center on Domestic & Sexual Violence [NCDSV], 2014; Roche, Richie, & Morton, 2010). For example, one study identified tactics of IA such as questioning the authenticity of their partner’s sexual orientation or gender identity authenticity (e.g., for not acting like a “real” lesbian; Bornstein, Fawcett, Sullivan, Senturia, & Shiu-Thornton, 2006; Woulfe & Goodman, 2018). Further, many individuals who have experienced IPV and who identify as transgender have reported experiencing partners who have insulted them specifically because of their gender identity or transition status (Roche et al., 2010; Woulfe & Goodman, 2018).

A third component of IA involves using slurs or derogatory language regarding their partner’s sexual orientation or gender identity by using words such as “tranny,” “fag,” and “dyke” (FORGE, 2014; NCDSV, 2014; Woulfe & Goodman, 2018). In the context of IA, stigma (i.e., the co-occurrence of labeling, stereotyping, and discrimination; Hatzenbuehler, Phelan, & Link, 2013; Valdiserri, Maulsby, & Holtgrave, 2018) is used as an abuse tactic against LGBTQ individuals. Scholars have noted that many LGBTQ individuals experience their partner using their own as well as their partner’s (if they identify as LGBTQ) feelings of shame or internalized discrimination as a tactic of abuse in effort to control the relationship (Allen & Leventhal, 1999; Erbaugh, 2007; Woulfe & Goodman, 2018). Taken together, although there has been limited empirical literature, to date, examining the prevalence of the rates of LGBTQ-specific forms of derogatory language as a unique form of IPV experienced by LGBTQ individuals, these assaults have been identified as a tactic of verbal abuse in the clinical literature (FORGE, 2014; NCDSV, 2014; Woulfe & Goodman, 2018).

Finally, a fourth IA dimension involves isolating individuals from the LGBTQ community (Woulfe & Goodman, 2018). Although isolation from family and friends is a common abuse tactic among cisgender, heterosexual relationships, it is particularly damaging in insular LGBTQ communities where partners may share the same support network of LGBTQ peers (Bergen, 1998). Also, many perpetrators who have not disclosed their LGBTQ identity may deter their partners from forming close friendships and openly discussing their intimate relationship in effort to continue to conceal their own stigmatized identity (Walters, 2011). Moreover, intentionally limiting LGBTQ individuals’ access to support may increase their dependence on abusive partners, another known IPV risk factor (Bornstein et al., 2006; Martin, Cromer, DePrince, & Freyd, 2013).

1.3 |. Development of the IA scale

In an effort to create a measure of IA, Balsam and Szymanski (2005) added five items to an already established and validated domestic violence scale, the Conflict Tactics Scale, Revised Edition (Straus & Douglas, 2004), to highlight IA tactics such as disclosing, undermining, and belittling a marginalized sexual orientation (Woulfe & Goodman, 2018). Although Balsam and Szymanski (2005) provided novel information about the ways in which sexual minority women may experience these forms of abuse and their relation to internalized homophobia and relationship quality, psychometric information, including validity and reliability of the items that were created, was not reported. In addition, the items did not incorporate homophobic or transphobic language as tactics of abuse, nor did they include tactics involving isolation of individuals from the LGBTQ community, which are important dimensions of IPV experienced by many LGBTQ individuals (Tesch & Bekerian, 2015). Finally, the items focused exclusively on IA among sexual minority women, using language that did not generalize or apply to the broader LGBTQ community.

Given these limitations, Woulfe and Goodman (2018) built on not only the original five items but also a broad literature review of clinical research (e.g., FORGE, 2014; NCDSV, 2014), theoretical papers (e.g., Hart, 1986; West, 1998), and qualitative and quantitative research (e.g., Bornstein et al., 2006; McClennen, Summers, & Vaughn, 2008; NCAVP, 2015), to create a more comprehensive scale of IA exposure among various demographic subgroups of a large sample of LGBTQ individuals (Woulfe & Goodman, 2018).

Specifically, Woulfe and Goodman (2018) made the following revisions to the original scale: First, two items were added to highlight unique aspects of IA that were not previously examined in the measure, created by Balsam and Szymanski (2005). Second, the item’s language, “I used my partner’s age, race, class, or religion against her,” was changed to reflect using a person’s gender identity or sexual orientation against them (Woulfe & Goodman, 2018). Third, language in all the items was changed in an effort to be more inclusive of sexual and gender minorities (e.g., “The person told me I deserve what I get because of my sexual orientation or gender identity”; Woulfe & Goodman, 2018). Finally, response options on the IA Scale were changed to account for past-year abuse and adult lifetime experiences so that the scores for IA prevalence could be comparable to the scores for other measures of abuse (Woulfe & Goodman, 2018).

After the initial list of items was created, Woulfe and Goodman (2018) then consulted with experts to determine the cultural appropriateness and inclusiveness of the items and the degree to which they reflected the underlying construct of IA. Specifically, they conducted interviews with one research scholar and five practitioners who demonstrated competence in assessing for IPV exposure among LGBTQ individuals, including LGBTQ people of color (Woulfe & Goood-man, 2018). Based on their feedback, Woulfe and Goodman (2018) revised the language of several items before finalizing the seven items of the IA Scale. This version of the scale was then tested with a sample of 734 participants who identified as LGBTQ (Woulfe & Goodman, 2018).

After the seven-item scale was initially developed, reliability analyses regarding item performance was conducted, and then the corrected item-total correlations and resulting coefficient alpha items were assessed for content validity and comparison with the purpose of the scale (Woulfe & Goodman, 2018). Specifically, Woulfe and Goodman (2018) conducted an EFA using principal axis factoring with a varimax rotation with 734 participants who identified as LGBTQ (Matsunaga, 2010). The Kaiser-Meyer-Olkin index (.85) indicated adequate sampling and Bartlett’s test of sphericity, 1697.73 (degree of freedom = 21, p < .001) suggested that the correlation matrix was appropriate for EFA (Woulfe & Goodman, 2018). The EFA of the IA Scale revealed a unidimensional factor structure (eigenvalue = 3.50; variance accounted for = 50.06%; factor loadings = .77, .77, .68, .67, .56, .53, and .52), which indicated adequate construct validity (Woulfe & Goodman, 2018). Further, communalities were acceptable (.37 to .64; mean [M] = .50) as was the measure of reliability ( = .80; Woulfe & Goodman, 2018).

To assess the construct validity of the IA Scale, Woulfe and Goodman (2018) conducted a bivariate correlation analysis comparing the relationship between identity, physical, and psychological abuse for both past-year and adulthood exposure. As expected, the variables were associated in congruent directions, providing preliminary information about adequate construct validity (r = .08, p < .05 to r = .57, p < .001). Although the IA Scale was internality valid and the items were correlated with other forms of abuse in expected directions, data-driven modifications to instruments may capitalize on chance (Bride, Robinson, Yegidis, & Figley, 2004; Jöreskog, 1993). As such, there is a need for greater assessment confirming the reliability and validity of the IA Scale with a second independent sample. The main objectives of the current study were to (a) validate the IA Scale with a large independent sample of LGBTQ-identified participants, including examining the IA Scale’s validity, reliability, and factor structure, and (b) to examine the preliminary factor structure of the IA Scale across various subgroups of the LGBTQ community, specifically across gender identity and race and ethnicity.

2 |. METHOD

2.1 |. Participants and procedures

The current study used nonprobability sampling techniques in an effort to create a specific yet inclusive sample, as suggested by Guadalupe-Diaz and Anthony (2016). Sampling techniques included marketing for potential participants through several local and nationwide email listservs, groups, and online forums (e.g., events, social media, and email broadcasts distributed by LGBTQ- and IPV-related organizations). Inclusion criteria for participation in this study were being 18 years of age or older and self-identifying as LGBTQ. Of note, the current study was part of a larger recruitment effort for a study examining LGBTQ IPV individuals’ utilization of health care services following experiences of IPV.

This study used a secure online data collection tool (Qualtrics) to collect survey responses. All potential participants received instructions that directed them to the survey website for the study, where they viewed the consent form and chose to participate. The survey included measures assessing participants’ age, gender identity, sexual orientation, race and ethnicity, and exposure to IPV within the past year and during adulthood: psychological abuse (14-item Psy-chological Maltreatment of Women Inventory; Tolman, 1999), physical abuse (six-item Conflict Tactics Scale; Straus & Douglas, 2004), and IA (seven-item IA Scale; Woulfe & Goodman, 2018). The psychological and physical abuse exposure measures were used as part of the investigation of the convergent validity of the IA Scale. Participants could elect to enter themselves into a raffle to win Amazon gift cards (fifteen $10.00 cards, ten $20.00 cards, or three $50.00 cards).

A total of 1,147 people began the survey; of these individuals, 1,049 (91.4%) completed the survey assessing for demographic information and IPV exposure during the past year and in adulthood. In effort to detect and minimize Internet research fraud, as suggested by Teitcher et al. (2015), all potential participants were assessed through external validation methods (e.g., checking data for same email addresses and/or fake addresses, as well as through computer information methods, e.g., IP addresses). All study procedures received approval from the institutional review boards at the host institution.

2.2 |. Measures

2.2.1 |. Demographics

Participants reported their current age and several other characteristics. Gender identity was assessed with the following question: “What is your current gender identity?” Response options were cisgender woman, cisgender man, and transgender or gender nonconforming. Race and ethnicity was assessed with the following question: “What is your race/ethnicity?” Response options were as follows: African American/Black, African, Asian/Asian American, His-panic/Latino, Native Hawaiian/Pacific Islander, Native American/Alaska Native, Middle Eastern, Biracial/Multiracial, White, and other. Because of small sample sizes within each race and ethnic category, a dichotomous variable was created to reflect White LGBTQ individuals and LGBTQ people of color. Sexual orientation was assessed with the following question: “How do you identify your current sexual orientation?” Response options were lesbian, gay, bisexual, pansexual, queer, asexual, and other.

2.2.2 |. Identity abuse

The IA Scale (Woulfe & Goodman, 2018) is a self-report measure that evaluates exposure to IA in intimate partnerships. Response options ranged from 0 (did not occur) to 7 (occurred more than 20 times in the past year). The IA produces two scores: a total scale score for past-year frequency with higher total scores representing greater exposure to IA over the past year, and a dichotomous adult exposure score indicating whether participants reported any of the items as having occurred in adulthood (Woulfe & Goodman, 2018). Woulfe and Goodman (2018) found adequate reliability as the Cronbach’s alpha for past-year IA was = .79. The internal consistency estimate for IA during the past year among the current sample wasα = .90.

2.2.3 |. Physical abuse

The Conflict Tactics Scale (CTS-2; Straus & Douglas, 2004) is a 20-item short form is used to evaluate the convergent validity of the IA Scale (Woulfe & Goodman, 2018). The CTS-2 assesses victimization across four domains: assault, injury, psychological aggression, and sexual coercion. The current survey excluded the psychological aggression items and combined the four physical assault items and two sexual assault items to form one physical abuse scale (Woulfe & Goodman, 2018). Evidence for this measure’s previous reliability and validity across racial and ethnic groups as well as gender identity is strong ( = .86; Black, Sussman, & Unger, 2009; Straus & Douglas, 2004). A total scale score for past-year frequency ranged from 0 (did not occur) to 7 (occurred more than 20 times in the past year). Higher total scores represented greater exposure to physical abuse over the past year. The internal consistency estimate for past-year CTS-2 for the current sample was α = .83.

2.2.4 |. Psychological abuse

The short form 14-item version of the Psychological Maltreatment of Women Inventory (PMWI; Tolman, 1999) was used to assess convergent validity of the IA Scale (Woulfe & Goodman, 2018). The PMWI assesses psychological violence in relationships and contains two subscales: Domination/Isolation and Emotional/Verbal Abuse. Response options ranged from 0 (did not occur) to 7 (occurred more than 20 times in the past year). The PMWI has demonstrated strong internal consistency across both subscales: = .95 for Domination/Isolation and = .93 for Emotional/Verbal. The PMWI also has demonstrated adequate support for its construct validity (Beck et al., 2011; Tolman, 1999). The reference time period and response options were changed to reflect the CTS-2 (Straus & Douglas, 2004). The wording of the measure was also adjusted to be applicable to individuals and abusers of all gender identities. A total scale score for past-year frequency was created and response options ranged from 0 (did not occur) to 7 (occurred more than 20 times in the past year). Higher total scores represented greater exposure to psychological abuse. For the current study, the Cronbach’s alpha for past-year PMWI was α = .94.

2.3 |. Analytic plan

CFA using AMOS IBM (version 24) was conducted to test whether the factor structure of the IA Scale (Woulfe & Goodman, 2018) could be repeated among a larger sample of LGBTQ individuals. CFA for scale development is commonly used to examine a factor structure that had been initially tested through EFA (Clark & Watson, 1995). A maximum likelihood estimate was used to assess whether the model adequately fit the covariance matrix of the current sample. As described by Wolfe and colleagues (2001), initial fit was examined using Satorra and Bentler’s (1990) scaled chi-square statistic. The chi-square statistic reflects a comparison between the covariance matrices of the model and the sample, with a lower chi-square yielding a higher convergence between the matrices, thus indicating a better fit. Further, the scaled chi-square was used due to its robustness to violations of normality (Satorra & Bentler, 1990).

Covariances among factors were free to be estimated and measurement errors were not allowed to correlate for the initial measurement model. All indicators were constrained to load on the respective factor (IA). The following fit indices were used to assess their goodness of fit to the data: goodness of fit index (GFI), comparative fit index (CFI), normed fit index (NFI), and root mean square error of approximation (RMSEA). Values of at least .90 for the GFI, CFI, and NFI indicate that the model is a good fit to the data (Kline, 1998), whereas RMSEA values of .08 or lower are acceptable (Hu & Bentler, 1999). The items that remained stable after the CFA was conducted were retained. After confirming the factor structure of the IA Scale, more evidence for the validity of the IA Scale (i.e., convergent validity) was explored.

All other analyses were completed using SPSS (version 24). There were moderate missing data (10.2% on demographic items and 30.2% across the abuse measures), and based on Little’s (1988) missing completely at random (MCAR) test, there were no significant differences on these variables for cases with missing data and those without missing data prior to imputation (i.e., the data were missing completely at random). Missing values were computed using SPSS with plausible simulated values based on the actual data. This method is preferable to listwise deletion or mean substitution, given that these latter methods can introduce statistical bias (Schlomer, Bauman, & Card, 2010). Statistical significance was determined at the alpha .05 level.

Pearson’s r correlations were conducted to determine bivariate relationships among the abuse variables (psychological abuse, IA, and physical abuse) to assess for convergent validity. EFA was performed using the principal axis factoring with varimax rotation to assess for discriminant validity between IA and the other abuse measures. Bartlett’s test of sphericity and the Kaiser–Meyer–Olkin statistic were used to examine the possibility of performing factor analysis (Wang et al., 2017). Factors with eigenvalues greater than 1 were extracted, and factor loadings of more than 0.50 were considered satisfactory (Wang et al., 2017). Finally, in an effort to determine whether the construct of IA is relevant to LGBTQ individuals who identify as people of color and White, as well as across gender identity (i.e., comparing initial factor structure between TGNC individuals, cisgender sexual minority males, and cisgender sexual minority females), separate EFA models of the IA Scale were analyzed for each subgroup.

3 |. RESULTS

3.1 |. Preliminary analyses

Participants were 1,049 self-identified LGBTQ adults aged 18–69 years (Mage = 27.32, standard deviation = 9.28; see Table 1). Participants identified their sexual orientation as other nonheterosexual identity (18.7%), gay (18.6%), queer (18.0%), bisexual (17.1%), lesbian (16.9%), pansexual (7.4%), and asexual (3.3%). Most participants identified their gender as cisgender women (52.6%), followed by TGNC (28.0%) and cisgender men (19.4%). Participants identified their race and ethnicity as White (71.9%), biracial or multiracial (15.4%), Hispanic or Latino/a (5.2%), African American or Black (3.0%), Asian or Asian American (3.0%), Middle Eastern (.7%), other (6%), and Native American or Alaska Native (.2%). Of the full sample, 64.4% reported having some form of exposure to IA over the past year, compared with 30.8% who reported that they had some form of IA exposure in adulthood but not in the past year. See Table 2 for the prevalence of each IA item in the past year and adulthood.

TABLE 1.

Sociodemographic characteristics of sample (N = 1,049)

Characteristic Frequency Percent
Age: Mean, standard deviation (27.32, 9.28)
Gender identity:
 Cisgender male 204 19.4%
 Cisgender female 552 52.6%
 Transgender or gender nonbinary 293 28.0%
Sexual orientation:
 Gay 195 18.6%
 Lesbian 177 16.9%
 Bisexual 179 17.1%
 Queer 189 18.7%
 Pansexual  78  7.4%
 Asexual  35  3.3%
 Other 196 18.7%
Race and ethnicity:
 White 754 71.9%
 Biracial or multiracial 157 15.4%
 Hispanic or Latino/a  54  5.2%
 African American or Black  31  3.0%
 Asian or Asian American  31  3.0%
 Middle Eastern or Middle Eastern  7  .7%
 Native American or Alaska Native  2  .2%
 Other  13  6.0%
Open in a new tab

TABLE 2.

Frequencies of identity abuse items

% Past-year exposure (n = 1,049) % Adult exposure (n = 1,049)
The person threatened to tell my employer, family, or others about my sexual orientation or gender identity  8.9% 28.6%
The person forced me to show physical orsexual affection in public, even though I didn’t want to 14.6% 38.5%
The person used my sexual orientation or gender identity against me 18.7% 42.9%
The person questioned whether my sexual orientation or gender identity was ‘real’ 24.8% 51.9%
The person told me I deserve what I get because of my sexual orientation or gender identity 11.3% 30.8%
The person called me pejorative names that have to do with my LGBTQ status 15.6% 34.0%
The person prevented me from seeking support within the LGBTQ community 12.8% 33.7%
Total % with any exposure 64.4% 30.8%
Open in a new tab

3.2 |. Convergent and discriminant validity

Acceptable convergent validity is determined by a measure that is moderately to strongly correlated with related variables (Campbell & Fiske, 1959). As such, claims of convergent validity for the IA Scale were supported, given that scores on the IA Scale with past-year exposure were positively associated with respondent ratings of psychological abuse exposure over the past year (PMWI), r = .68, p < .001, and physical abuse exposure over the past year (CTS), r = .67, p < .001 (see Table 4). Also, scores on the IA Scale with exposure during adulthood were positively associated with respondent ratings of psychological abuse exposure during adulthood (PMWI), r = .72, p < .001, and physical abuse exposure during adulthood (CTS), r = .70, p < .001. These bivariate correlations also demonstrated that those who were more likely to experience one form of IPV were likely to experience other forms as well, both in the past year and during adulthood (Thompson et al., 2006). In addition, because the correlations were moderate in size, these findings reflect that the construct of IA is distinct from the construct of psychological abuse and physical abuse.

TABLE 4.

Correlations among the abuse measures

1 2 3 4 5 6
 1. Identity abuse (PY) –
 2. Psychological abuse (PY)  .68*** –
 3. Physical abuse (PY)  .67 ***  .51*** –
 4. Identity abuse (Lifetime)  .80***  .57***  .37*** –
 5. Psychological abuse (Lifetime)  .56***  .72***  .29***  .72*** –
 6. Physical abuse (Lifetime)  .62***  .56***  .45***  .72***  .70*** –
 M (SD) 1.64 (.86) 2.40 (1.33) 1.17 (.55) 1.37 (.37) 1.59 (.36) 1.36 (.36)
Open in a new tab

Note. PY = past year exposure to abuse; Lifetime = adult exposure to abuse.

*

p < .05.

**

p < .01.

***

p < .001.

Moreover, findings from the EFA demonstrated that the indicators for the IA factor do not cross-load with the other abuse measures, and the results yielded one distinct factor for IA, three factors for psychological abuse, and one factor for physical abuse, further confirming claims of discriminant validity between each of these three constructs of abuse. Also, an EFA determined that the items from the IA Scale represented a unidimensional factor for LGBTQ people of color (eigenvalue = 4.68, variance accounted for = 66.78%, factor loadings = .82, .79, .79, .78, .77, .77, and .74) and for White LGBTQ people (eigenvalue = 4.73, variance accounted for = 67.59%, factor loadings = .90, .87, .82, .81, .75, .71, and .66).

Finally, an EFA determined that the items from the IA Scale represented a unidimensional factor for TGNC individuals (eigenvalue = 4.53, variance accounted for = 69.72%, factor loadings = .86, .83, .82, .81, .79, .78, and 74) and cisgender sexual minority women (eigenvalue = 3.35, variance accounted for = 54.61%, factor loadings = .83, .83, .73, .68, .63, .60, and .48). However, EFA findings indicated that the items from the IA Scale represented a two-dimensional factor for cisgender sexual minority males (Factor 1: Eigenvalue = 3.35, variance accounted for = 47.80%, factor loadings = .89, .71, .70, and .61; and Factor 2: Eigenvalue = 1.31, variance accounted for = 66.46%, factor loadings = .74, .60, and, .50).

3.3 |. Factorial validity

Figure 1 presents the CFA run on the current sample (n = 1,049). As shown in Figure 1, the factor for the IA Scale was specified to load on all seven items. The fit indices yielded a good model fit (see Table 3; CFI .977; IFI = .977; NFI = .973; RMSEA .081; Oishi, 2007). The normed χ2 statistic for the proposed modified model was χ2 (14) = 110.24.12, p < .01. Standardized estimates were high in all cases with statistical significance for which the path was not constrained equal to 1 (Khan, 2017). Strong factor loadings were found across the seven items, ranging from 0.65 to 0.85.

FIGURE 1.

FIGURE 1

Open in a new tab

Confirmatory factor analysis of the one-factor, seven-item identity abuse model

TABLE 3.

Goodness of fit for the confirmatory factor analysis

Fit indices
model χ2 df IFI CFI NFI RMSEA
Measurement model 110.24 14 .977 .977 .97 .08
Open in a new tab

Note. df = degree of freedom; IFI = incremental fit index; CFI = comparative fit index; NFI = normed fit index; RMSEA = root-mean-square error of approximation.

4 |. DISCUSSION

Many LGBTQ individuals negotiate their self-concepts or sense of self according to their identities, their perceptions of how others react to their identities, and the internalization of these perceptions (Altheide, 2000; Zhao, Grasmuck, & Martin, 2008). Recent research has focused on understanding how manipulation and control of stigmatized LGBTQ identities are unique tactics of abuse in LGBTQ intimate relationships that have long-term effects on self-esteem and self-worth (Guadalupe-Diaz & Anthony, 2017). Until recently, however, there has been no psychometrically validated scale to assess for the presence of IA among LGBTQ individuals.

The IA Scale (Woulfe & Goodman, 2018) is an important contribution to the development of empirical knowledge about the unique dimension of IPV among LGBTQ individuals and further fills a need for a reliable and valid instrument specifically designed to measure the prevalence of IA among LGBTQ individuals. Further, the IA Scale has the desirable characteristic of being easy to administer, score, and interpret and is relevant across various subgroups of the LGBTQ community. The current study supports the psychometric integrity of the IA Scale (Woulfe & Goodman, 2018) among a large, diverse sample of LGBTQ individuals. Findings build on the existing evidence to confirm the convergent and factorial validity as well as the reliability of this scale to assess for past-year and adulthood prevalence of IA among LGBTQ individuals.

This study is among the first, to our knowledge, to provide preliminary evidence for the relevance of IA across several subgroups of the LGBTQ community. Although LGBTQ people of color hold multiple stigmatized identities (e.g., race and ethnicity, gender identity, sexual orientation), it may be the case that heterosexist and cissexist tactics that emphasize LGBTQ individuals’ marginalized position in society represent unique forms of partner abuse that affect many LGBTQ individuals, regardless of their race and ethnicity. This study is also the first to examine the factorial structure of the IA Scale across TGNC individuals, cisgender sexual minority men, and cisgender sexual minority women, who may have qualitatively different experiences of the ways in which their partners use their marginalized identities as leverage for maintaining power and control in the relationship. Findings from this study indicated that for TGNC individuals and cisgender sexual minority women, the IA Scale represented a unique construct of partner abuse. However, for cisgender sexual minority men, the IA Scale yielded two distinct components, namely, IA related to internal (e.g., pejorative name calling) and external (preventing a partner from accessing support within the LGBTQ community) dimensions. Future research should aim to validate these two potential IA subscales for cisgender sexual minority men.

4.1 |. Limitations

Despite this study’s contributions to the existing literature, there are several limitations to address. This study relied on a nonprobability sample by using listserv recruitment and snowball sampling, thus limiting the generalizability of the findings. For instance, there was a limited number of older LGBTQ individuals and LGBTQ people of color in the sample, affecting generalizability to these communities. Another limitation is that the abusive partner’s characteristics (e.g., sexual orientation, gender identity, degree of “outness”, length of time of the relationship) were not assessed in this study and could provide important information about the unique dynamics of IA within LGBTQ relationships in general and specifically in instances when LGBTQ individuals’ abusive partners also identify as LGBTQ (Balsam & Szymanski, 2005). This study also did not assess for the bidirectional prevalence of IA in LGBTQ relationships, which could be useful in understanding factors that predict IA.

4.2 |. Implications for research and practice

Despite these limitations, this study provides further validation of the construct of IA among LGBTQ individuals, which has important implications for future research and practice. This measure could be used to further a program of research on the prevalence and impact of IA on LGBTQ individuals. Some possible future directions could include longitudinal research on the co-occurrence of IA with other forms of IPV over time, continued exploration of the relationship between IA and mental and physical health, and identifying risks and protective factors that might moderate or mediate these associations among LGBTQ individuals. The IA Scale could also be modified for use with other sub-groups of the LGBTQ community who are at risk of experiencing structural and interpersonal stigma, including older adults, immigrants, individuals with disabilities, and low-income individuals.

The study’s findings also have important implications for health care providers working with LGBTQ individuals, given their likelihood of working with a client who identifies as LGBTQ and who has experienced some form of IPV, including IA. Assessing whether a client who identifies as LGBTQ is currently or has previously experienced IA in their intimate relationship could enable providers to offer education about IA and resources for LGBTQ survivors of IPV, specifically„ as well as determining whether their partners are part of their treatment. Also, screening for IA is critical for developing intervention and prevention efforts targeted specifically for improving overall health and well-being among LGBTQ individuals, given that IPV is associated with mental and physical health consequences (Coker et al., 2002). This measure may provide a starting point for providers and agencies serving LGBTQ individuals with IPV exposure to better identify IA and address it in their clinical practice.

4.3 |. Conclusion

This study was the first to examine the factor structure of the recently developed IA Scale (Woulfe & Goodman, 2018) and adds to our knowledge of the psychometric performance of this measure among a broad range of LGBTQ individuals. Many LGBTQ individuals experience chronic stressors related to their marginalized sexual and gender identity, including discrimination, actual and perceived rejection, and internalized stigma, all of which may contribute to distinctive tactics of abuse used against them (Balsam & Szymanski, 2005; Meyer, 2003; Otis et al., 2006; Reuter et al., 2017). Although much more work needs to be done to enable the IA measure to capture the complexity of LGBTQ individuals’ lived experiences, it marks an important step toward moving to a greater understanding of IA in this community.

REFERENCES

  1. Allen C, & Leventhal B (1999). History, culture, and identity: What makes GLBT battering different. In Leventhal B & Lundy SE (Eds.), Same-sex domestic violence: Strategies for change (pp. 73–81). Thousand Oaks, CA: Sage. [Google Scholar]
  2. Altheide DL (2000). Identity and the definition of the situation in a mass-mediated context. Symbolic Interaction, 23(1), 1–27. [Google Scholar]
  3. Ard KL, & Makadon HJ (2011). Addressing intimate partner violence in lesbian, gay, bisexual and transgender patients. Journal of General Internal Medicine, 26, 930–933. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Balsam K, & Szymanski DM (2005). Relationship quality and domestic violence in women’s same-sex relationships: The role of minority stress. Psychology of Women Quarterly, 29, 258–269. [Google Scholar]
  5. Beck JG, McNiff J, Clapp JD, Olsen SA, Avery ML, & Hagewood JH (2011). Exploring negative emotion in women experiencing intimate partner violence: Shame, guilt, and PTSD. Behavior Therapy, 42(4), 740–750. [DOI] [PubMed] [Google Scholar]
  6. Bergen RK, (Ed.). (1998). Issues in intimate violence London: Sage. [Google Scholar]
  7. Black DS, Sussman S, & Unger JB (2010). A further look at the intergenerational transmission of violence: Witnessing interparental violence in emerging adulthood. Journal of Interpersonal Violence, 25(6), 1022–1042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bornstein DR, Fawcett J, Sullivan M, Senturia KD, & Shiu-Thornton S (2006). Understanding the experiences of lesbian, bisexual and trans survivors of domestic violence: A qualitative study. Journal of Homosexuality, 51, 159–181. [DOI] [PubMed] [Google Scholar]
  9. Bride BE, Robinson MM, Yegidis B, & Figley CR (2004). Development and validation of the secondary traumatic stress scale. Research on Social Work Practice, 14(1), 27–35. [Google Scholar]
  10. Campbell DT, & Fiske DW (1959). Convergent and discriminant validation by the multitrait-multimethod matrix. Psycholog-ical Bulletin, 56(2), 81–105. [PubMed] [Google Scholar]
  11. Clark LA, & Watson D (1995). Constructing validity: Basic issues in objective scale development. Psychological Assessment, 7(3), 309–319. [Google Scholar]
  12. Coker AL, Davis KE, Arias I, Desai S, Sanderson M, Brandt HM, & Smith PH (2002). Physical and mental health effects of intimate partner violence for men and women. American Journal of Preventive Medicine, 23(4), 260–268. [DOI] [PubMed] [Google Scholar]
  13. Dank M, Lachman P, Zweig JM, & Yahner J (2014). Dating violence experiences of lesbian, gay, bisexual, and transgender youth. Journal of Youth and Adolescence, 43, 846–857. [DOI] [PubMed] [Google Scholar]
  14. Edwards KM, Sylaska KM, & Neal AM (2015). Intimate partner violence among sexual minority populations: A critical review of the literature and agenda for future research. Psychology of Violence, 5, 112–121. [Google Scholar]
  15. Erbaugh EB (2007). Queering approaches to intimate partner violence. In O’Toole LL, Schiffman JR, & Kiter Edwards ML (Eds.), Gender violence: Interdisciplinary perspectives (pp. 451–459). New York: New York University Press. [Google Scholar]
  16. Finneran C, & Stephenson R (2013). Intimate partner violence among men who have sex with men: A systematic review. Trauma, Violence, & Abuse, 14(2), 168–185. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. FORGE. (2014). Trans-specific Power and Control Tactics Retrieved from https://forge-forward.org/publications-resources/anti-violence-publications.
  18. Goldberg NG, & Meyer IH (2013). Sexual orientation disparities in history of intimate partner violence: Results from the California Health Interview Survey. Journal of Interpersonal Violence, 28(5), 1109–1118. [DOI] [PubMed] [Google Scholar]
  19. Goodmark L (2013). Transgender people, intimate partner abuse, and the legal system. Harvard Civil Rights-Civil Liberties Law Review, 48, 51–104. [Google Scholar]
  20. Guadalupe-Diaz XL, & Anthony AK (2017). Discrediting identity work: Understandings of intimate partner violence by transgender survivors. Deviant Behavior, 38(1), 1–16. [Google Scholar]
  21. Hart B (1986). Lesbian battering: An examination. In Lobel Kerry (Ed.), Naming the violence: Speaking out about lesbian battering (pp. 173–189). Berkeley, CA: Seal Press. [Google Scholar]
  22. Hatzenbuehler ML, Phelan JC, & Link BG (2013). Stigma as a fundamental cause of population health inequalities. American Journal of Public Health, 103, 813–821. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hu LT, & Bentler PM (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. [Google Scholar]
  24. Innamorati M, Imperatori C, Lamis DA, Contardi A, Castelnuovo G, Tamburello S, … & Fabbricatore M (2017). Weight bias internalization scale discriminates obese and overweight patients with different severity levels of depression: The Italian version of the WBIS. Current Psychology, 36(2), 242–251. [Google Scholar]
  25. Jöreskog KG (1993). Testing structural equation models. Sage Focus Editions, 154, 294–294. [Google Scholar]
  26. Khan R (2017). Sibling violence: Validating a two-factor model of severity in nonoffender populations. Psychology of Violence, 7(4), 498. [Google Scholar]
  27. Kline RB (1998). Software review: Software programs for structural equation modeling: Amos, EQS, and LISREL. Journal of Psychoeducational Assessment, 16(4), 343–364. [Google Scholar]
  28. Kubicek K, McNeeley M, & Collins S (2015). “Same-sex relationship in a straight world” individual and societal influences on power and control in young men’s relationships. Journal of Interpersonal Violence, 30(1), 83–109. [DOI] [PubMed] [Google Scholar]
  29. Kulkin HS, Williams J, Borne HF, de la Bretonne D, & Laurendine J (2007). A review of research on violence in same-gender couples: A resource for clinicians. Journal of Homosexuality, 53(4), 71–87. [DOI] [PubMed] [Google Scholar]
  30. Limoubpratum C (2015). Factors influencing sustainable distribution: A framework of co-opetition, freight consolidation, and collaborative freight distribution in the Thailand’s Newspaper Industry (Doctoral dissertation) Victoria University. [Google Scholar]
  31. Little RJ (1988). A test of missing completely at random for multivariate data with missing values. Journal of the American statistical Association, 83(404), 1198–1202. [Google Scholar]
  32. Marcotte AA (2016). Experiences of victimization and health care access among non-metropolitan LGBTQ+ Individuals. The-ses and Dissertations (Comprehensive) 1883. http://scholars.wlu.ca/etd/1883
  33. Martin CG, Cromer LD, DePrince AP, & Freyd JJ (2013). The role of cumulative trauma, betrayal, and appraisals in understanding trauma symptomatology. Psychological Trauma: Theory, Research, Practice, and Policy, 5(2), 110–118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Matsunaga M (2010). How to factor-analyze your data right: Do’s, don’ts, and how-to’s. International Journal of Psychological Research, 3(1), 97–110. [Google Scholar]
  35. McClennen JC, Summers AB, & Vaughan G (2008). Gay men’s domestic violence. Journal of Gay and Lesbian Social Services, 14, 23–49. [Google Scholar]
  36. Meyer IH (2003). Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: Conceptual issues and research evidence. Psychological Bulletin, 129, 674–697. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Miller CT, & Major B (2000). Coping with stigma and prejudice. The Social Psychology of Stigma, 243–272.
  38. National Coalition of Anti-Violence Programs. (2012). Report of LGBT and HIV-affected intimate partner violence Retrieved from https://www.avp.org/storage/documents/ncavp_2012_ipvreport.final.pdf
  39. National Coalition of Anti-Violence Programs. (2015). Lesbian, gay, bisexual, transgender, queer and HIV-affected hate violence in 2014 Retrieved from https://www.avp.org/resources/avp-resources/452-2014-report-on-lesbian-gay-bisexual-transgender-queer-and-hiv-affected-intimate-partner-violence
  40. National Center on Domestic & Sexual Violence. (2014). Gay, lesbian, bisexual, and trans power and control wheel, Retrieved from https://www.ncdsv.org/images/TCFV_glbt_wheel.pdf
  41. Retrieved from https://www.thehotline.org/is-this-abuse/lgbt-abuse/
  42. Oishi S (2007). The application of structural equation modeling and item response theory to cross-cultural positive psychology research. Oxford Handbook of Methods in Positive Psychology, 126–138.
  43. Otis MD, Rostosky SS, Riggle ED, & Hamrin R (2006). Stress and relationship quality in same-sex couples. Journal of Social and Personal Relationships, 23(1), 81–99. [Google Scholar]
  44. Reuter TR, Newcomb ME, Whitton SW, & Mustanski B (2017). Intimate partner violence victimization in LGBT young adults: Demographic differences and associations with health behaviors. Psychology of Violence, 7(1), 101–109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Ristock JL (2011). Introduction: Intimate partner violence in LGBTQ lives. In Intimate partner violence in LGBTQ lives (pp. 9–18). New York: Routledge. [Google Scholar]
  46. Roche A, Richie G, & Morton J (2010). Out of sight, out of mind? Transgender people’s experiences of domestic abuse: LGBT Youth Scotland Equality Network
  47. Satorra A, & Bentler PM (1990). Model conditions for asymptotic robustness in the analysis of linear relations. Computational Statistics & Data Analysis, 10(3), 235–249. [Google Scholar]
  48. Stotzer RL (2009). Violence against transgender people: A review of United States data. Aggression and Violent Behavior, 14(3), 170–179. [Google Scholar]
  49. Straus MA, & Douglas EM (2004). A short form of the revised conflict tactics scales, and typologies for severity and mutuality. Violence and Victims, 19, 507–520. [DOI] [PubMed] [Google Scholar]
  50. Suarez NA, Mimiaga MJ, Garofalo R, Brown E, Bratcher AM, Wimbly T, … Sullivan PS (2018). Dyadic reporting of intimate partner violence among male couples in three US cities. American Journal of Men’s Health, 1–9. [DOI] [PMC free article] [PubMed]
  51. Teitcher JE, Bockting WO, Bauermeister JA, Hoefer CJ, Miner MH, & Klitzman RL (2015). Detecting, preventing, and responding to “fraudsters, “ in Internet research: Ethics and tradeoffs. The Journal of Law, Medicine & Ethics, 43(1), 116–133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Tesch BP, & Bekerian DA (2015). Hidden in the margins: A qualitative examination of what professionals in the domestic violence field know about transgender domestic violence. Journal of Gay & Lesbian Social Services, 27(4), 391–411. [Google Scholar]
  53. Thompson RS, Bonomi AE, Anderson M, Reid RJ, Dimer JA, Carrell D, & Rivara FP (2006). Intimate partner violence: Prevalence, types, and chronicity in adult women. American Journal of Preventive Medicine, 30(6), 447–457. [DOI] [PubMed] [Google Scholar]
  54. Tjaden P, & Thoennes N (2000). Extent, nature, and consequences of intimate partner violence: Findings from the National Violence Against Women Survey (Report No. NCJ, pp. 29–31) Washington, DC: US Department of Justice, National Institute of Justice; 2000. 181867. [Google Scholar]
  55. Tolman RM (2001). The validation of the psychological maltreatment of women inventory. Violence and Victims, 4(3), 173–189. [PubMed] [Google Scholar]
  56. Valdiserri RO, Maulsby CH, & Holtgrave DR (2018). Structural factors and the national HIV/AIDS strategy of the USA. In Structural Dynamics of HIV (pp. 173–194). Cham: Springer. [Google Scholar]
  57. Waalen J, Goodwin MM, Spitz AM, Petersen R, & Saltzman LE (2000). Screening for intimate partner violence by health care providers: Barriers and interventions. American Journal of Preventive Medicine, 19(4), 230–237. [DOI] [PubMed] [Google Scholar]
  58. Walters ML (2011). Straighten up and act like a lady: A qualitative study of lesbian survivors of intimate partner violence. Journal of Gay & Lesbian Social Services, 23, 250–270. [Google Scholar]
  59. Wang J, He B, Miao X, Huang X, Lu Y, & Chen J (2017). The reliability and validity of a new professionalism assessment scale for young health care workers. Medicine, 96(25), 1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. West CM (1998). Leaving a second closet: Outing partner violence in same-sex couples. In Jasinski JL, Williams LM, & Finkelhor D (Eds.), Partner violence: A comprehensive review of 20 years of research (pp. 163–183). Thousand Oaks, CA: SAGE. [Google Scholar]
  61. Wolfe DA, Scott K, Reitzel-Jaffe D, Wekerle C, Grasley C, & Straatman AL (2001). Development and validation of the conflict in adolescent dating relationships inventory. Psychological Assessment, 13(2), 277. [PubMed] [Google Scholar]
  62. Woulfe JM, & Goodman LA (2018). Identity abuse as a tactic of violence in LGBTQ communities: Initial validation of the identity abuse measure. Journal of Interpersonal Violence, 1–21. [DOI] [PubMed]
  63. Zhao S, Grasmuck S, & Martin J (2008). Identity construction on Facebook: Digital empowerment in anchored relationships. Computers in Human Behavior, 24(5), 1816–1836. [Google Scholar]

RESOURCES