Measuring Experiences of Minority Stress for Transgender Women: Adaptation and Evaluation of Internalized and Anticipated Transgender Stigma Scales

Abstract

Purpose: Transgender women (TGW) experience high rates of stigma based on their gender identity. Research has documented how transgender stigma and other discrimination negatively contribute to health inequities, including higher rates of depression compared with the general population. However, few scales measuring transgender stigma exist, even fewer that specifically assess anticipated or internalized transgender stigma. We sought to validate an adapted transgender stigma scale in a diverse sample of TGW.

Methods: We adapted an existing stigma measure to capture experiences of anticipated and internalized transgender stigma for TGW. Adapted measures were completed by 213 diverse TGW. Factor analysis was completed to reduce the number of items in each scale and sociodemographic differences in each construct were explored.

Results: The final nine items comprising anticipated transgender stigma and the five items for internalized transgender stigma both showed evidence of adequate model fit, unidimensionality, and internal consistency. The two constructs were moderately correlated with one another (r=0.36, p<0.001). We identified educational and HIV status differences in anticipated transgender stigma but no sociodemographic differences in internalized transgender stigma.

Conclusion: We developed brief measures of internalized and anticipated transgender stigma through initial adaptation by TGW themselves and subsequent psychometric evaluation, demonstrating evidence of unidimensionality and internal consistency. These subscales were only moderately associated with one another and may provide unique insights in future research on minority stress among TGW.

Introduction

Transgender women (TGW) experience higher rates of health inequities, social and financial instability than the general population,¹ including disproportionately higher rates of HIV infection.^2–4 TGW also experience mental health difficulties with diagnosis rates higher than the general population—around half of TGW experience a lifetime diagnosis of major depression and lower psychological well-being across different stages of their lives.^5–7 Overall, TGW are more likely to experience suicidal ideation, self-harm, and intimate partner violence, and engage in behaviors that may place them at risk for poorer health outcomes.^8,9 They are also more vulnerable to experiencing employment and housing discrimination.¹⁰ These disproportionate rates of negative experiences are likely related to the various forms of stigma surrounding transgender identity that TGW experience.

There is an array of various types of stigma that can impact an individual's health outcomes. According to Earnshaw and Chaudoir,¹¹ experienced or enacted stigma refers to people's perception about having experienced stereotyping, prejudice, and discrimination directed at them from others. Anticipated stigma is the extent to which people expect to experience or worry about experiencing stereotyping, prejudice, and discrimination directed at them from others in the future.¹² Finally, internalized stigma refers to experiencing feelings of shame, guilt, or negative emotions directed at oneself that are related to one's stigmatized status.¹³ Previous research has also shown that the process of stigmatization has implications in health research because it meets all the criteria to be considered a fundamental cause of health inequalities.¹⁴ Hatzenbuehler et al.¹⁴ reported that access to resources, social isolation, stress, and psychological and behavioral responses to stigma are simultaneously linked to experiences of stigma and act as mediators in the association between stigma and health outcomes. This may be because stigma is a fundamental social cause of these outcomes, meaning that in a significant number of people, stigma influences multiple health-related outcomes through multiple risk factors. These theoretical frameworks that delineate the definition and process of stigmatization and its association to health provide the general framework for understanding the complex and widespread effects of stigma. However, it is important to understand the unique experience of stigma among gender-diverse populations.

Meyer¹⁵ describes the association between stigma, prejudice, and discrimination in terms of the Minority Stress Model, whereby the constant threat of rejection and other negative experiences based on sexual minority status leads to mental health issues, which can help explain the heightened risk of mental illness among these populations. These stressful experiences can lead to internalized stigma about or attempts to hide one's minority status. With respect to “gender minorities,” including TGW, Mizock and Lewis¹⁰ define transphobia as the prejudice, discrimination, and gender-related violence as a result negative attitudes toward transgender identity. For this article, we will use the term transgender stigma to describe the negative experiences and attitudes toward transgender identity, which approximates the definition of stigma applied to TGW. The experience of transgender stigma, whether on more distal (i.e., external factors) or proximal levels (i.e., internal factors), is one of the potential risk factors accounting for disproportionately high rates of trauma within the transgender community.¹⁰

Various studies have provided evidence about the impact of stigma on TGW health outcomes. Nemoto et al.¹⁶ found that suicidal ideation was specifically associated with frequent exposure to transgender stigma for TGW. Experiencing transgender stigma is also associated with social exclusion, which impacts and limits access to significant resources, social services, employment, and stable housing.¹⁷ Finally, transgender stigma has been associated with sexual risk factors for HIV acquisition, substance use, harmful mental health outcomes, and traumatic experiences.^10,18–20 As a result, transgender stigma, which occurs simultaneously at individual, interpersonal, structural, and other levels, affects the physical and mental health of TGW.

Although research has provided substantial evidence about the negative impact of transgender stigma on TGW, this construct has been measured differently across many studies using adaptations of existing measures or creating new ones that may not appropriately capture the experience of transgender stigma. The measurement of transgender stigma is critical, because the process of measuring the experience of stigma can be distinct among TGW compared with sexual or other minorities for a range of reasons.²¹ Testa et al.²² created a tool called the Gender Minority Stress and Resilience Measure (GMSR) based on theory, the input of transgender and gender nonconforming focus groups, and adapting or including prior measures, such as the Transgender Identity Survey.¹ This measure contains a total of nine subscales addressing constructs that include stigma, such as gender-related discrimination, gender-related rejection, gender-related victimization, nonaffirmation of gender identity, internalized transphobia, negative expectations for future events, and nondisclosure.²² Despite the validity and reliability of the measure, it has a large number of related subscales and a total of 53 items. For researchers and practitioners hoping to capture stigma in work for which it may not be the primary purpose, brief measures of internalized and anticipated transgender stigma might be particularly useful.

Given the body of research documenting health issues of TGW, it is increasingly important to examine the possible sources of these inequalities, including the experience of stigma based on transgender identity. It is critical to develop valid and sensitive measurements to document the prevalence of stigma in transgender populations, especially TGW. As such, the goal of this study was to work with TGW to adapt existing stigma measures to capture the experience of anticipated and internalized transgender stigma for TGW, to refine the two measures using factor analysis, and to explore sociodemographic differences in these two measures.

Methods

Data for this study came from T-Talk, a two-phase study that began with a cross-sectional in-office observational study that also simultaneously screened TGW for potential participation in a randomized controlled trial of an intervention to address sexual risk and substance use. Data for this study were taken from the cross-sectional, observational study, which collected between May 2014 and September 2016. Recruitment approaches included collaborative partnerships with community-based organizations, outreach at community events, tabling at a health center serving TGW, outreach to venues and bars, and distribution of recruitment materials to service providers. We also conducted social media advertisement on various online sites and sent out project newsletters to TGW who had expressed interest in the project. We used day and evening drop-in hours for the initial baseline appointment to maximize flexibility for participants.

A total of 487 TGW were phone screened for participation, of whom 382 (78%) were preliminarily eligible. Eligibility criteria for the observational study included being at least 18 years of age, assigned male at birth, identified as a transgender or female and reporting either one or more sexual acts (regardless of condom use) or one or more days of drug use (excluding alcohol) in the past 60 days; participants were also required to provide contact information, reside in the New York City (NYC) metropolitan area, and be able to complete the assessment in English. In total, 214 TGW completed the assessment, which included a computer-assisted self-interviewing delivered in Qualtrics, and participants received $40 for their participation. Further description of study procedures can be found in Parsons et al.²³ All study protocols were approved by institutional review board of the City University of New York.

Measures

Background characteristics

Demographics as given in Table 1 included race and ethnicity (black/African American, Latina/x, white, or multiracial/other), current gender identity (open-ended question recoded as female/woman/girl, trans, or other), sexual orientation (lesbian/gay, bisexual, queer, or heterosexual), relationship status (single or partnered), income (<$20,000, $20,000–$49,000, or $50,000 or more), education (high school or less, some college, bachelor degree, or graduate degree), HIV status (positive or negative), current age, and age of first starting to transition or live as a woman (which may or may not have initially involved medical transition-related services).

Table 1.

Demographic Characteristics of the Sample of Transgender Women (N=213)

	n (%)		Anticipated transgender stigma		Internalized transgender stigma
			M	SD	M	SD
Overall	213 (100)		2.59	0.79	1.54	0.71
Race and ethnicity
Black/African American	68 (31.9)		2.63	0.74	1.61	0.73
Latina/x	60 (28.2)		2.58	0.93	1.51	0.69
White	52 (24.4)		2.58	0.72	1.45	0.57
Multiracial/other	33 (15.5)		2.55	0.77	1.58	0.87
Gender identity
Female/woman/girl	126 (59.2)		2.60	0.76	1.53	0.67
Transa	80 (37.6)		2.59	0.84	1.55	0.78
Other	7 (3.3)		2.41	0.79	1.51	0.55
Sexual orientation
Lesbian/gay	41 (19.2)		2.53	0.88	1.52	0.8
Bisexual	40 (18.8)		2.51	0.75	1.53	0.66
Queer	31 (14.6)		2.70	0.66	1.43	0.51
Heterosexual/straight	101 (47.4)		2.62	0.81	1.59	0.74
Relationship status
Single	109 (51.2)		2.62	0.73	1.51	0.62
Partnered	104 (48.8)		2.56	0.86	1.56	0.79
Income
<$20K	171 (80.3)		2.62	0.80	1.57	0.71
$20K—$49,999K	35 (16.4)		2.43	0.76	1.33	0.49
$50K or more	7 (3.3)		2.84	0.61	1.77	1.20
Education
High school or less	89 (41.8)		2.44	0.85	1.58	0.78
Some college	62 (29.1)		2.82	0.73	1.52	0.63
Bachelor degree	49 (23)		2.61	0.73	1.45	0.70
Graduate degree	13 (6.1)		2.52	0.73	1.65	0.56
HIV status
Negative	139 (65.3)		2.72	0.78	1.51	0.67
Positive	74 (34.7)		2.36	0.77	1.60	0.77
	M	SD
Age (range: 18–65)	34.3	11.7
Age first transition or living as a woman (range: 1–64)	22.4	10

^aTrans includes all reported identities along the trans feminine spectrum.

Sample size for mean comparisons was n=212 because of missing data.

M, mean; SD, standard deviation.

Transgender stigma

Both the anticipated transgender stigma and internalized transgender stigma scales were adapted from the HIV stigma scale (Berger et al.²⁴), which originally used four subscales of personalized stigma, disclosure concern, negative self-image, and concerns with public attitudes. Initially, four TGW staff members adapted the 40 items to be more relevant to stigma experiences affecting many TGW. Responses were 4 points on a Likert scale from strongly disagree (1) to strongly agree (4).

Data analysis

We began by characterizing the sample in terms of sociodemographic characteristics using descriptive statistics. After this, we used Mplus version 8 to iteratively conduct a series of confirmatory factor analyses to refine the two scales. The models for anticipated and internalized stigma were run separately and began by including all original items (with the exception of one item from each that was designed to be reverse coded), with the ultimate goal of reducing the scales to a brief set of items that best captured the two constructs. To do this, we ran the initial model and iteratively refined it by removing items based on a successive set of steps with specific criteria. First, items were removed with standardized factor loadings <0.6, which was the threshold we set for a conceptually meaningful factor loading. Next, we consulted the modification indices to identify potential misfit resulting from residual correlation among the items—conceptually, these residual correlations suggest the items overlap in content and may be redundant, and thus we sought to minimize the presence of residual correlation to refine the item pool. Across models, the Likert-type responses were specified as categorical indicators and the models were estimated using the Mplus default of weighted least squares. Upon arriving at the final factor solutions, subscale scores were calculated based on the average of the final item pools and we analyzed internal consistency by calculating Cronbach's alpha statistics. Finally, we conducted one-way analysis of variance models and Pearson's correlations to examine demographic differences and associations.

Results

The demographic characteristics of the sample are given in Table 1. As can be seen, the majority were women of color, with nearly one-third identifying as Black and more than one-quarter identifying as Latina/x. Nearly two-thirds of the sample had a gender identity of female and more than one-third identified as trans. Approximately half had a heterosexual identity, with the remaining sample being nearly evenly split between lesbian/gay, bisexual, and queer. The sample was nearly evenly split by relationship status, and the vast majority was making <$20,000/year and had less than a bachelor's degree. Approximately two-thirds were HIV negative. The average age of the sample was 34.3 years and, on average, they first began their transition at age 22 years.

As given in Table 2, the model fit for the first confirmatory factor analysis of the 27 anticipated transgender stigma items was poor, with no indicators below the standard threshold for adequate fit. We next conducted several rounds of examining modification indices to reduce items leading to misfit and improve model fit indices, with the goal of reducing to a brief scale of 10 items or fewer. The initial step was to remove items with standardized factor loading <0.60, which led to the removal of items 1, 5, 9, 10, 17, and 30. Next, we sought to remove items with significant residual correlations; we did this by focusing in each step on those with a residual correlation >0.20 and then re-running, so as to recalibrate the model and re-examine remaining items before removing those that might ultimately reach nonsignificance. This led to a series of five steps, with four steps removing those items with residual correlations >0.20 and the final step removing any remaining, statistically significant correlations. There were nine items remaining that all met our criteria for retention (i.e., factor loading >0.60 and no significant residual correlation) and the model fit across all indicators ranged from adequate to good. These final nine items were retained for subsequent scoring and analyses.

Table 2.

Initial and Final Confirmatory Factor Models for Items Capturing Anticipated Transgender Stigma

Item content	Initial factor loadings			Final factor loadings
	Unstd.	SE	Std.	Unstd.	SE	Std.
In many areas of my life, no one knows that I am transgender [1]	1.00	0.00	0.17	—	—	—
Telling someone I am transgender is risky [4]	4.25	1.70	0.70	—	—	—
People who are transgender lose their jobs when their employers find out [5]	3.14	1.28	0.52	—	—	—
People who are transgender are treated like outcasts [9]	3.18	1.35	0.52	—	—	—
Most people believe that a person who is transgender is dirty [10]	3.35	1.40	0.55	—	—	—
Most people think that a transgender person is disgusting [14]	3.69	1.57	0.61	—	—	—
Most people who are transgender are rejected when others find out [16]	4.16	1.70	0.69	1.00	0.00	0.64
I am very careful who I tell that I am transgender [17]	3.55	1.37	0.58	—	—	—
Some people who know I am transgender have grown more distant [18]	4.29	1.77	0.71	1.13	0.09	0.73
Since coming out as transgender, I worry about people discriminating against me [19]	4.25	1.75	0.70	—	—	—
Most people are uncomfortable around someone who is transgender [20]	4.16	1.72	0.69	1.07	0.09	0.69
I worry that people may judge me when they learn I am transgender [22]	4.33	1.74	0.71	1.08	0.09	0.70
I have been hurt by how people reacted to learning I was transgender [24]	3.91	1.57	0.64	1.07	0.09	0.69
I worry that people who know I am transgender will tell others [25]	4.81	1.96	0.79	—	—	—
I regret having told some people that I am transgender [26]	4.56	1.85	0.75	—	—	—
Some people avoid touching me once they know I am transgender [28]	4.26	1.76	0.70	—	—	—
People I care about stopped calling after learning I was transgender [29]	4.72	1.93	0.78	—	—	—
People have told me that being transgender is what I deserve for how I lived my life [30]	2.83	1.14	0.47	—	—	—
Some people close to me are afraid others will reject them if it becomes known that I am transgender [31]	4.55	1.85	0.75	—	—	—
People don't want me around their children once they know I am transgender [32]	4.31	1.76	0.71	—	—	—
People have physically backed away from me when they learn I am transgender [33]	4.49	1.86	0.74	1.18	0.09	0.76
Some people act as though it's my fault I am transgender [34]	4.65	1.91	0.77	1.27	0.09	0.82
I have stopped socializing with some people because of their reactions to me being transgender [35]	4.75	1.94	0.78	1.27	0.09	0.82
I have lost friends by telling them I am transgender [36]	4.77	1.95	0.79	—	—	—
I have told people close to me to keep the fact that I am transgender a secret [37]	4.04	1.62	0.67	—	—	—
When people learn I am transgender, they look for flaws in your character [40]	2.49	1.77	0.71	—	—	—
	0.03	0.02	—	0.41	0.06	—
Estimated factor variance	Model fit			Model fit
CFI	0.85			0.98
TLI	0.84			0.98
Model χ2 (df)	1096.31 (324), p<0.001			43.90 (20), p=0.002
RMSEA	0.11			0.08
Probability RMSEA ≤0.05	<0.001			0.08
WRMR	1.62			0.66

Note: N=213.

Items adapted from Berger et al.²⁴ HIV stigma; Numbers in parenthesis are the original item numbers from Berger's HIV stigma scale.

CFI, comparative fit index; RMSEA, root mean square error of approximation; SE, standard error; Std., standardized; TLI, Tucker-Lewis Index; Unstd., unstandardized; WRMR, weighted root mean residual.

As given in Table 3, the model fit for the first confirmatory factor analysis of the 12 internalized transgender stigma items was poor. Based on modification indices, we went through two rounds of item reduction to improve model fit. The initial step was to remove items with standardized factor loading <0.60; however, no items met this criterion and this step was skipped. Next, we sought to remove items with significant residual correlations, which led to removal of items 38 and 39; another run showed additional items meeting this criterion, which were subsequently removed (items 3, 6, 11, 13, and 27). This led to a model with five items with all but the root mean square error of approximation (RMSEA) value showing evidence of adequate to strong model fit, and this five-item version of the scale was retained for scoring and subsequent analyses.

Table 3.

Initial and Final Confirmatory Factor Models for Items Capturing Internalized Transgender Stigma

Parameter	Initial factor loadings			Final factor loadings
	Unstd.	SE	Std.	Unstd.	SE	Std.
I feel guilty because I am transgender [2]	1.00	0.00	0.81	1.00	0.00	0.80
People's attitudes about transgender people make me feel worse about myself [3]	0.79	0.05	0.64	—	—	—
I work hard to keep my gender identity a secret [6]	0.78	0.05	0.64	—	—	—
I feel I am not as good a person as others because I am transgender [7]	0.92	0.06	0.74	0.96	0.06	0.78
It is easier to avoid new friendships than worry about telling someone that I am transgender [11]	0.82	0.05	0.67	—	—	—
Being transgender makes me feel unclean [12]	1.08	0.04	0.88	1.10	0.06	0.88
Since coming out as transgender, I feel set apart and isolated from the rest of the world [13]	0.88	0.05	0.71	—	—	—
Being transgender makes me feel that I'm a bad person [15]	1.05	0.05	0.85	1.12	0.05	0.90
Being transgender is disgusting to me [23]	0.85	0.07	0.68	0.94	0.07	0.75
As a rule, telling others that I am transgender has been a mistake [27]	0.83	0.06	0.68	—	—	—
People who know I am transgender tend to ignore my good points [38]	0.87	0.05	0.70	—	—	—
People seem afraid of me once they learn I am transgender [39]	0.88	0.05	0.72	—	—	—
Estimated factor variance	Model fit			Model fit
CFI	0.90			0.99
TLI	0.88			0.98
Model χ2 (df)	270351 (54), p<0.001			17.31 (5), p=0.01
RMSEA	0.14			0.11
Probability RMSEA ≤0.05	0.00			0.04
WRMR	1.40			0.50

Note: N=213.

Items adapted from Berger et al.²⁴ HIV stigma; Numbers in parenthesis are the original item numbers.

The two measures showed evidence of strong internal consistency—the Cronbach's alpha for the nine anticipated stigma items was 0.86 and was 0.84 for the five internalized stigma items. The two measures were significantly, moderately correlated with one another (r=0.36, p<0.001).

We next sought to examine whether there were any sociodemographic differences in either construct, and the group means for each sociodemographic breakdown are given in the right-hand columns of Table 1. Average scores for anticipated transgender stigma were slightly below the midpoint (mean=2.59, median=2.63, standard deviation [SD]=0.79) and relatively normally distributed, whereas internalized transgender stigma were right skewed with a modal score of 1 (mean=1.54, median=1.20, SD=0.71). For anticipated stigma, we observed two significant differences. There were differences by educational attainment, with those participants with some college having the highest levels of anticipated transgender stigma and those with high school or less having the lowest levels, F(208,3)=2.94, p=0.03. There were also differences by HIV status, with HIV-negative participants reporting significantly more anticipated transgender stigma than HIV-positive participants, F(210,1)=10.67, p=0.001. We observed no significant group differences by any sociodemographic variable on internalized transgender stigma. We also found that anticipated (r=−0.22, p=0.001) transgender stigma was correlated with age, with older age being associated with lower levels of stigma, although internalized was not (r=−0.10, p=0.16). We did not observe associations with age of transition for anticipated or internalized stigma (r=−0.06, p=0.43 and r=−0.10, p=0.15, respectively; n=206 because of missing data on the transition age).

Discussion

Given the demonstrated impact of stigma and other forms of minority stress on health disparities, the aim of this study was to develop and refine brief measures of anticipated and internalized transgender stigma for TGW. Items from the Berger HIV Stigma Scale were adapted by TGW to capture multiple subtypes of stigma experienced by TGW because of their gender identities and then completed by a diverse sample of 213 TGW in NYC. Starting with relatively large item pools, we were able to refine the measures to a total of nine items capturing anticipated transgender stigma and five capturing internalized transgender stigma, with the final item pools showing evidence of adequate model fit, unidimensionality, and internal consistency. In addition to their psychometric properties, two measures have a high degree of face validity—examining the final item pools, anticipated stigma is characterized primarily by items capturing perceptions of and fears about interpersonal rejection and internalized stigma is characterized by items tapping shame and low self-image. We found that the two forms of stigma were significantly, moderately correlated.

As is often found, endorsement of internalized stigma was low, with the average score falling only a half-point higher than the scale's lowest response. For anticipated stigma, we found both higher scores—more than a full point higher than internalized stigma—and slightly greater variability between participants. It is important to point out the concept of internalized stigma draws heavily on the literature around concealable stigmas like sexual identity and HIV-positive status—these low scores may simply represent difficulty in obtaining informative self-report data on internalized stigma more broadly, or may specifically speak to the possibility that the construct of internalized stigma is less applicable to transgender identity. As other recent published studies have pointed out,²⁵ difficulties in understanding the minority stress model for TGW and other gender-diverse populations is complicated by a lack of common measures. Given there are now several validated scales, future studies should include several measures of minority stress to better disentangle measure-specific issues from findings that might hold meaning for the underlying constructs of interest.

We had no preconceived hypotheses regarding sociodemographic group differences in the two measures and indeed found very few. We found that anticipated stigma was higher among those with higher levels of education and those with HIV-negative status. We also found that levels of anticipated stigma were somewhat lower among TGW of older ages. More research is needed to determine whether the relatively few differences are because of the homogeneity of the sample, the ubiquity of stigma for TGW, or features of the measures themselves.

Strengths and limitations

Although the sample was highly diverse and represented relatively broad eligibility criteria (i.e., TGW who were either sexually active or engaged in drug use), this was nonetheless a community-based sample from NYC and may be limited in generalizability to experiences outside major metropolitan areas. For the study, items were adapted by TGW, but were nonetheless taken from a prior scale of HIV stigma—some items were initially deemed inappropriate for adaptation for the experience of transgender stigma, but there may also have been experiences not captured by the prior scale that may have been warranted for inclusion were the measure newly developed rather than adapted. Future research is needed to examine the correspondence of this scale with other validated scales, like the GMSR,²² which would help to delineate whether this abbreviated scale captures similar or distinct constructs as some of those contained in earlier measures.

Conclusions

The newly developed scale is a brief measure of internalized and anticipated transgender stigma that was adapted by TGW themselves. From a large initial item pool of multifaceted forms of stigma, we reduced the scale to a total of only 14 items capturing two forms of transgender stigma. The two subscales were found to have evidence of unidimensionality and internal consistency, and we identified some initial sociodemographic differences in the experience of stigma. Future research is needed to examine how the psychometric properties of this brief scale compare with those of other validated scales, like the GMSR,²² and how they predict relevant outcomes, like mental health and health behaviors. Finally, although the scales focus on self-report from the individual perspective, it remains critical to highlight that these represent psychological manifestations of stigma that are the result of structural injustice, social exclusion, and interpersonal experiences of discrimination and victimization.

Abbreviations Used

GMSR

Gender Minority Stress and Resilience Measure

NYC

New York City

SD

standard deviation

SE

standard error

TGW

transgender women

Author Disclosure Statement

All authors declare that they have no conflict of interest.

Funding Information

This research was funded by the National Institutes on Drug Abuse (R01-DA034661); H.J.R. was supported in part by a career development award from the National Institute on Drug Abuse (K01-DA039030); and J.L.M. was supported in part by a diversity supplement from the National Institute on Drug Abuse (R01-DA041262-S1). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Cite this article as: Rendina HJ, Cain DN, López-Matos J, Ray M, Gurung S, Parsons JT (2020) Measuring experiences of minority stress for transgender women: adaptation and evaluation of internalized and anticipated transgender stigma scales, Transgender Health 5:1, 42–49, DOI: 10.1089/trgh.2019.0059.