Challenging and understanding gendered narratives: the development and validation of the transnormativity measure (TM)

Abstract

Background: Transnormativity refers to the accountability structure that regulates the acceptable gender presentations, narratives, and ways of being of trans and nonbinary (TNB) individuals.Aims: The present research extends prior qualitative research on transnormativity to develop and validate the Transnormativity Measure (TM). Methods: The initial developed pool of potential items was presented to a focus groups and three content experts for review. In Study 1 (N = 497), the 69 initial items of the TM hypothesized to underly a six-factor structure were analyzed using Exploratory Factor Analysis (EFA) and construct and discriminant validity were assessed. In Study 2 (N = 540), an independent sample of TNB participants’ TM responses were subjected to Confirmatory Factor Analysis (CFA), invariance testing, and construct and predictive validity. Finally, in Study 3 (N = 107), an Interclass Correlation Coefficients 2-way mixed-effects model of the TM was assessed.Results: EFAs conducted in Study 1 revealed a two-factor structure as the best fit with 50 items removed. Conceptually there was considerable overlap in the items comprising the two factors and it was decided that one general factor should be utilized. Study 1 also provided preliminary construct and discriminant validity due to expected relations between the TM and existing measures of heteronormativity and internalized transphobia. In Study 2 findings from correlational tests of the remaining items revealed that four items were highly correlated and were removed. Subsequent CFA indicated that the one factor model fit the data well. Configural invariance was supported however metric noninvariance was found. Additionally, Study 2 results supported construct and predictive validity through correlations between the TM and measures of TNB community belonginess and mental health outcomes. Finally, Study 3 provided support for test-retest reliability. Discussion: Across three studies, the TM was found to be a valid measure of transnormativity.

Austin Johnson (2016) defines transnormativity as “the specific ideological accountability structure to which transgender people’s presentations and experiences of gender are held accountable” (pp. 465–466). These accountability structures regulate the acceptable gender presentations, narratives, and ways of being for trans and nonbinary (TNB) individuals (Johnson, 2016; McIntyre, 2018). As a result, transnormativity as an ideology regulates who gets to be TNB and is enforced through gatekeeping practices around who is “trans enough.” Transnormative ideologies and practices create a hierarchy in which TNB individuals who assimilate and project cisgender identities and behaviors are rewarded with conditional acceptance into society–those who do not are subjugated and obscured (Johnson, 2016; McIntyre, 2018; Vipond, 2015). Qualitative research has indicated that transnormativity can impact the psychological distress of TNB individuals through anxiety and worry as well as decreased belongingness (Bradford & Syed, 2019; Garrison, 2018; Johnson, 2016). However, a means to quantify the ways in which transnormativity affects the psychological health of TNB individuals has yet to be developed given the unwritten nature of this construct. In response, the current study aims to develop and validate the Transnormativity Measure (TM) to understand how this accountability structure impacts TNB individuals.

Roots of transnormativity

Cis-heteronormativity represents the unspoken rules regarding the acceptable ways to do and be a (cis)man or (cis)woman (Ansara & Berger, 2016). Individuals who “do gender” according to cis-heteronormative script (e.g., assigned male at birth individuals performing hegemonic masculinity) are given acceptance, whereas those who are nonconforming are held accountable through social systems (i.e., stigma, violence, and ostracization; Hollander, 2013). Whereas cis-heteronormativity regulates how to perform a cisgender identity, transnormativity designates the “correct” way to be TNB. Linking directly to cis-heteronormativity, transnormativity imposes rules onto TNB individuals that bestows rewards to individuals who achieve cisgender norms (often through medical transitions) and punishes those whose experiences and behaviors do not align to cisgender expectations of gender performance. While a complete discussion of the linkage between cis-heteronormativity and transnormativity is beyond the scope of the current study (see Pasley et al., 2018), it is important to note that transnormativity is enforced both by cisgender individuals through anti-trans rhetoric and violence as well as within TNB communities through gatekeep practices.

Regulation of transnormativity

Like cis-heteronormativity, transnormativity is rigidly enforced and codified through medicine, media, and social practices. Transnormativity is rooted in the medicalization of TNB identities and bodies, where individuals who represented the most privileged in society (i.e., white, rich, binary gender individuals) are the normative standard (see Riggs et al., 2019 for a full discussion of this history). Medicalization refers to the process by which human experiences become subjected to medical authority and the experience becomes something to be diagnosed and treated (Conrad, 1992; Zola, 1972). In the context of TNB individuals, their identities are medicalized and their bodily experiences are expected to be cured (Johnson, 2016). As a result, the authority writing about what is true of TNB experiences becomes the cisgender medical establishment which then dictates who is “clinically” transgender (Johnson, 2016; Riggs et al., 2019; Yeadon-Lee, 2009).

The medical establishment legitimatizes certain TNB identities through creating diagnostic criteria (e.g., the Diagnostic and Statistical Manual for Mental Health Disorders [DSM]; American Psychiatric Association (APA), 2013) that prescribes the ways a TNB individual should understand their body and their gender. Thus, in order to be viewed as clinically transgender and validated in their identity by medical establishments, TNB individuals learn to shape the narrative of their gender identity experiences to fit the medically prescribed and cisgender normed understandings of gender diversity (Borba, 2019; Davy, 2010; Vipond, 2015). For example, describing personal gender expression in traditional masculine/feminine (i.e., cis-normative) ways, wearing clothing reflective of a strict gender binary, and describing a childhood preference for gender affirming toys (Bradford & Syed, 2019; Linander, 2018). This then creates a cyclical process where TNB individuals describe their experiences in ways which map onto the DSM criteria to access care and medical providers and researchers take this to mean that their criteria are correct (Riggs et al., 2019). However, this cyclical process of the medical establishment dictating criteria of gender dysphoria and TNB individuals repeating this narrative to their provider is not the fault of TNB individual who are utilizing transnormativity to receive necessary medical care.

This authoritative ownership of gender dysphoria by the cisgender-majority medical establishment has constrained the ways TNB individuals can describe their experiences with gender dysphoria due to a fear of being denied medical transition care (Borba, 2019). TNB individuals are held hostage to the medicalized narrative that highlights and centers struggle with identity and embodiment (e.g., wrong body trope; Johnson, 2015). This rigid accountability structure does not allow for TNB experiences that fall outside of the distress narrative and construes those individuals as “not trans enough.” This enforcement extends beyond medical offices and impacts TNB communities through intracommunity sanctioning of individuals who do not reaffirm clinical criteria through social ostracization and cisgender enforcement of TNB identity affirmation though limiting access to medical recognition (Johnson, 2015; Schleifer, 2006). For example, Hollander (2013) notes that this doing of gender often “demands that they [people] describe or explain their behavior with reference to shared normative expectations (p. 10)” and thus TNB people will continue to assess and enforce expectations based on cisnormative medical ideals that are then translated into new but rigid assessments within community.

Like the ways medical establishments have legitimized certain TNB identities (i.e., those validated by the DSM), cisgender-controlled mass media also strictly regulates which TNB stories are told. Numerous sociological studies have documented the narrow depictions of TNB lives that are discussed in cisgender created movies, daytime talk shows, and documentaries (e.g., Johnson, 2016; McBride, 2020; McIntyre, 2018; Miller, 2019). The stories that are told in movies and documentaries depict a trans narrative, or timeline, where an individual recognized in childhood that they were “born in the wrong body” and seek out medical transitions to bring their bodies into “alignment” with their gender identity (Johnson, 2016; McIntyre, 2018). Further, the stories that some TNB individuals themselves tell via video blogs depict coming out stories centered on “wrong body” tropes and transition times lines following the impact of hormone therapies on an ability to be read as cisgender (Miller, 2019). Yet, these transnormative stories are often focused on the experiences of white, binary identified individuals (Horak, 2014; Miller, 2019; Raun, 2016) who seek to validate themselves within cisgender and medically legitimized understandings of TNB identities and experiences (Miller, 2019). In response, tensions exist between TNB communities who align themselves with trans medicalized understandings and those who reject it, and subsequently reject transnormativity (Miller, 2019). Those who benefit from these tensions are the ones who have internalized the norms and therefore enact and reinforce these norms both within and outside of TNB communities.

Considering that aligning oneself with transnormativity allows for a conditional acceptance into society and legitimization by cisgender society, TNB community members themselves may also act as gatekeepers around who is “trans enough” and deserving of validation (Johnson, 2016). Since TNB identities are not connected familiarly like other marginalized identities, TNB individuals often look to community groups to find validation of their experiences. As such, transnormativity provides the guidebook on how to be TNB and the “boundaries of legitimacy” (p. 322; Bradford & Syed, 2019). TNB individuals can easily locate and describe the aspects or benchmarks of transnormativity and will utilize them to gauge their own and others transness (Bradford et al., 2019; Bradford & Syed, 2019). Individuals whose personal narrative or experiences of gender do not align with these unspoken rules may face criticism or questioning of their identities from other TNB individuals as well as cisgender individuals (Garrison, 2018). This questioning, or gatekeeping, within TNB communities is a form of protection of past political and social activism that has created a space for some level of societal acceptance (Bradford & Syed, 2019; Garrison, 2018). Individuals who align closely with transnormativity believe that allowing anyone to claim a TNB identity weakens the legitimacy of TNB identities and challenges the provisional acceptance that is granted from cisgender individuals (Garrison, 2018). This within community gatekeeping is most visually accessible on YouTube where TNB individuals attack other TNB people who they view as being deceiving or not really trans (Miller, 2019).

Components of transnormativity

As noted, transnormativity represents the unwritten rules that govern TNB individuals’ experiences. A review of the literature demonstrates that transnormativity encompasses the following domains: a) emphasis on anatomical gender dysphoria; b) discovery narrative; c) medicalization; d) assimilation; e) victimization; and f) social interactions. The first, an emphasis on anatomical gender dysphoria, represents the belief that anatomical dysphoria and physical/bodily experiences are necessary components of TNB identification (Bradford & Syed, 2019; Garrison, 2018; Johnson, 2016; McIntyre, 2018; Miller, 2019). Often this physical manifestation of bodily distress is described in the trope of being “born in the wrong body” (Bradford & Syed, 2019; Garrison, 2018). The second domain, discovery narrative, represents the belief that gender is nascent, that an individual is born TNB and must have known since childhood (Bradford & Syed, 2019; Garrison, 2018; Johnson, 2016; McIntyre, 2018). Individuals who can locate memories of childhood “cross gender” behavior are believed to more “trans” than folks who cannot. A common example of this in recent culture and research is the idea of “trans-trenders” and “rapid onset gender dysphoria” which has led to societal understandings of newly out trans individuals as pleading for attention (Garrison, 2018). The third domain, medicalization, reflects the belief that TNB individuals should desire to medically transition (Bradford & Syed, 2019; Garrison, 2018; Johnson, 2016; McIntyre, 2018; Miller, 2019). An aspect of this belief is that only through medical interventions can one bring their gender into “alignment” with their body and thus medical transition is a life-or-death matter and that a medical transition trajectory should be uniformly followed by binary trans individuals (Bradford & Syed, 2019). Additionally, part of this belief is that nonbinary folks should be ambivalent around medical transition.

While the first three domains of transnormativity are clearly aligned with trans medicalization, the final three reflect the ways in which TNB individuals are expected to behave or understand themselves. The fourth domain, assimilation, represents the belief that TNB individuals desire to assimilate into cisgender culture and be read as cisgender (Bradford et al., 2019; Bradford & Syed, 2019; McBride, 2020; McIntyre, 2018). Inherent to this believe is the linkage of gender presentation to gender identity and a valuing of rigid binary gender roles. As a result, nonbinary expressions are considered less valid and any semblance of fluidity is devalued (Bradford & Syed, 2019). The fifth domain, victimization, reflects the belief that TNB individuals are inherently victims, that struggle and distress are key aspects of TNB experience (Bradford & Syed, 2019; Garrison, 2018; Johnson, 2016). This victimization understanding often takes the form of viewing oneself as prone to discrimination or psychological distress (Bradford & Syed, 2019). Lastly, the sixth domain, social interactions, centers on the belief that trans folks should behave in certain ways in social contexts. This belief involves gatekeeping around the kinds of friends one should have (e.g., only queer and TNB), the social-political views held (e.g., liberal), and the interests one should hold (e.g., artistic).

Impacts of transnormativity

Trans and nonbinary individuals are often aware of the rules of transnormativity and will utilize them in utilitarian ways (Bradford & Syed, 2019; Johnson, 2019; Pasley et al., 2018). For example, as described above TNB individuals often utilize transnormative narratives regarding the wrong body trope when accessing medical transition services (Bradford & Syed, 2019; Linander, 2018). Still for others, engaging with transnormativity allows for societal acceptance through avoiding the consequences that can occur for transgressing societal (i.e., cis-heteronormative) norms of doing gender (Miller, 2019). Although TNB individuals can utilize transnormativity to their advantage, transnormativity also imparts negative psychological impacts.

Given that transnormativity is a relatively new concept, there has been limited research into how it impacts TNB individuals’ psychological health. Further, transnormativity has predominantly been discussed within the fields of sociology and women and gender studies who appear to be focused on how transnormative messages are spread rather than on how it directly impacts TNB individuals, with the notable exception of the work of Bradford and Syed (2019). Although the research is in its infancy, psychological implications can be hypothesized based on qualitative data. For example, those who experience greater pressure to believe transnormativity ideology may experience anxiety and distress around being “trans enough” (Garrison, 2018). These individuals may feel anxiety or decreased belongingness when entering TNB community spaces as they question their own performance or communication of their gender within the boundaries of transnormativity. This anxiety and worry may be increased among nonbinary individuals, who cannot easily situate themselves within transnormative narratives (Bradford et al., 2019; Garrison, 2018). As such, social isolation from trans communities may result for individuals who disagree with transnormative ideology and do not find belonging in these spaces (Miller, 2019). Alternatively, individuals who agree with the ideology of transnormativity are likely to feel empowered or validated which contributes to a sense of belongness within TNB communities (Johnson, 2016). However, this is more likely to be an affect for binary identified transgender individuals who can more easily align their gendered performance with the core tenants of transnormativity. Thus, the development of a quantitative measure of transnormativity would allow for a better understanding of the direct impacts of this ideological accountability structure.

Study 1

In study 1, the Transnormativity Measure I was developed and refined, as well as its factor structure and initial validity tested with a community sample of TNB individuals. The development of the TM followed the nine steps of DeVellis and Thorpe (2021). The operational definition of TM is that transnormativity comprises the attitudes and beliefs of TNB communities regarding acceptable gender presentations, behaviors, histories, and identities of TNB individuals. Our review of the available qualitative research is that transnormativity is composed of six domains of central attitudes or beliefs. In response, the proposed TM was hypothesized to have the following scale structure based on the six domains: a) emphasis on anatomical gender dysphoria, b) discovery narrative, c) medicalization, d) assimilation, e) victimization, and f) social interactions (see Supplementary Information for operational definitions).

To generate the items that comprise the measure, the authors reviewed relevant literature and qualitative descriptions of TNB individuals’ experiences with transnormativity and discussed the phenomenon with TNB members of the Trans Care collaborative. This resulted in the generation of 60 items. These sample items were sent to three content experts, the majority of whom openly identify as TNB, who provided feedback regarding relevance as well as clarity of each item and wrote additional items to consider. Next, the sample items that were modified in response to expert review were presented to a focus group of TNB individuals who provided feedback on clarity, applicability, and accuracy of the items to their lived experience. The results of feedback from both experts and TNB focus group members resulted in a total of 67 items spread across the six proposed scales.

Hypothesis

It was hypothesized that the TM would reveal a six-factor structure during Exploratory Factor Analysis (EFA) and that the items within each factor, as well as the scale as whole, would be internally consistent as they measure the same construct, coping. Additionally, it was expected that the scales of TM would be related to the Heteronormative Attitudes and Beliefs Scale (NB-HABS; Habarth, 2015), demonstrating convergent validity. To ensure divergent validity, it was expected that the scales of the TM would be minimally related to the IT subscale of the Gender Minority Stress and Resilience Measure (GMSRM; Testa et al., 2015; see Supplementary Information for full study hypotheses).

Study 1 method

Recruitment

Participants were recruited from announcements posted to social networking websites targeting sexual and gender minority communities. These online resources included Reddit threads focused on trans identities, Facebook pages dedicated to trans research or support groups, and community organizations from 25 states (randomly selected via excel sort with random numbers provided to each state) in the United States. The recruitment flyer included information suggested by the Transgender Research Informed Consent Disclosure Policy (Winters et al., 2019) such as: the purpose of the study (i.e., to develop measures of trans specific coping and transnormativity), inclusion criteria (i.e., at least 18 years old and identified as trans, nonbinary, or with a trans history or status), information regarding compensation (i.e., 1 in 6 chance of winning a $25 gift card until 420 participants complete the survey), as well as the contact information of the principal investigator.

Participants

For the initial participant count, 1,378 participants initiated the survey; however, 198 participants did not answer any questions beyond demographics, 21 did not complete all of the items of the TM, 531 did not pass the engagement checks, two indicated that they were under the age of 18, and an additional 61 indicated that their gender was the same as their sex assigned at birth and were removed. Free responses of participants were also reviewed for potential bots with 61 potential participants entering “I’m sorry I don’t want to answer that question” to all free response and an additional seven entering unusual responses to the final prompt (e.g., “Low self-esteem is a state of mind that temporarily loses balance”) and were removed. This left a final sample size of 497 who identified as women (n = 204), men (n = 109), and nonbinary (n = 184). Participants ranged in age from 18 to 65 (M = 27.88, SD = 6.30) and most were residing in the United States (90.0%). Regarding race, participants were able to select multiple identities and 50.3% of the sample identified endorsed white and 64.8% of the sample endorsed a marginalized racial identity. Table 1 includes further demographic information including education and sexual identity.

Table 1.

Participant demographics from studies 1–3.

	Study 1	Study 2	Study 3
	(N = 497)	(N = 540)	(N = 107)
Mean age (SD)	27.88 (6.30)	27.04 (8.64)	28.28 (8.65)
Gender identity
Woman	204 (41.1)	183 (33.7)	38 (35.5)
Man	109 (21.9)	175 (32.4)	29 (27.1)
Nonbinary	184 (37.0)	182 (33.7)	40 (37.4)
Sex assigned at birth n (%)
Female	191 (38.4)	311 (57.6)	60 (56.1)
Male	266 (53.6)	226 (41.9)	46 (43.0)
Intersex	38 (7.6)	1 (0.2)	1 (0.9)
Prefer not to answer	2 (0.4)	2 (0.4)	0 (0)
Sexual identity n (%)
Asexual	16 (3.2)	55 (10.2)	12 (11.2)
Bisexual	138 (27.8)	160 (29.6)	29 (27.1)
Gay/Lesbian	95 (19.1)	117 (21.7)	20 (18.7)
Heterosexual	44 (8.9)	25 (4.6)	4 (3.7)
Pansexual	46 (9.3)	71 (13.1)	18 (16.8)
Queer	156 (31.4)	106 (19.6)	24 (22.4)
Other	2 (0.4)	6 (1.1)	0 (0)
Race/ethnicity n (%)
Alaskan Native	7 (1.4)	0 (0.0)	0 (0)
Native American	16 (3.2)	16 (3.0)	2 (1.9)
Asian/Asian American	13 (2.6)	29 (5.4)	6 (5.6)
Biracial/Multiracial	13 (2.6)	22 (4.1)	0 (0)
Black/African American	179 (36.0)	12 (2.2)	1 (0.9)
Latine/Hispanic	72 (14.5)	40 (7.4)	5 (4.7)
Middle Eastern/North African	7 (1.4)	5 (0.9)	3 (2.8)
Native Hawaiian/Pacific Islander	6 (1.2)	3 (0.6)	0 (0)
White/Caucasian	250 (50.3)	467 (86.5)	96 (89.7)
Other	9 (1.8)	35 (6.5)	7 (6.5)
Education level n (%)
Did not complete High School	18 (3.6)	11 (2.0)	0 (0)
High School or GED	59 (11.6)	103 (19.1)	15 (14.0)
Some college, no degree	177 (35.6)	151 (28.0)	28 (26.2)
College Degree	185 (37.3)	186 (34.5)	46 (43.0)
Graduate Degree or Professional School	37 (7.4)	89 (16.5)	18 (16.8)
Prefer not to answer	21 (4.2)	0 (0.0)	0 (0)
Reside in the United States n (%)	448 (90.1)	386 (71.5)	76 (71.0)

Procedure and measures

Participants completed an online survey that began with an informed consent document followed by demographic questionnaire, the proposed items of the TM, Internalized Transnormativity (IT) subscale of the Gender Minority Stress and Resilience measure (GMSRM; Testa et al., 2015), Normative Behaviors Subscale of the Heteronormative Attitudes and Beliefs Scale (NB-HABS; Habarth, 2015), a question instructing participants who would like to complete a brief follow up survey in two weeks to enter a unique personal code (i.e., initials, first number of street address, and first number of their phone number), and concluded with a prompt which thanked participants and provided them an opportunity to leave suggestions to improve future studies. Eight engagement checks were dispersed throughout the survey to check the quality of the data and to ensure participants remained consistent in their responses and participant responses were reviewed to ensure they did not enter the same value for all of their responses.

Transnormativity measure

The Transnormativity Measure (TM) was developed to assess the accountability structure of attitudes and beliefs that exist with TNB communities regarding perceptions of “acceptable” gender presentations, behaviors, histories, and identities of TNB individuals. The proposed measure included 67 items across six subscales. Responses to items are recorded on a 6-point scale from 1 (strongly disagree) to 6 (strongly agree). Scale items are summed to create subscale scores. Results of tests of internal consistency are reported in the results section.

Construct validity

Heteronormativity was measured via the eight item Normative Behaviors Subscale of the Heteronormative Attitudes and Beliefs Scale (NB-HABS; Habarth, 2015). Item responses for this measure are recorded on a seven-point scale from 0 (strongly disagree) to 7 (strongly agree). Items are averaged with higher scores indicating greater endorsement of heteronormativity. The Normative Behaviors Subscale demonstrated good internal consistency when tested with a sample of cisgender and TNB individuals (α = .89; Hunter et al., 2021) and in the current study (α = 0.88).

Discriminant validity

Internalized Transnormativity (IT) was measured via the eight item scale from the GMSRM (Testa et al., 2015). Responses for the scale are recorded on a five-point scale from 0 (strongly disagree) to 4 (strongly agree), where items are summed and higher scores indicate greater IT. All three scales demonstrated excellent internal consistency when tested with a TNB sample (α = .91; Lindley & Galupo, 2020) and good internal consistency in the current study (α = .82).

Data analyses

Data were analyzed with SPSS v26. To develop the final version of the TM, the proposed items were entered into Exploratory Factor Analyses (EFA) to determine the factor structure and reduction of items. First, data were screened for multivariate outliers as well as skewness and kurtosis. Next, items from the proposed TM were entered into EFAs with Maximum Likelihood (ML) estimation and Varimax rotation to assess the factor structures. Items which had factor loadings greater than .40 on the primary factor, less than .30 on other factors, and less than .20 between loadings were retained (Tabachnick & Fidell, 2007). Items were removed one-by-one and the EFAs were re-ran until simple structure was found. Next, Cronbach’s alphas for the TM were assessed. Finally, to demonstrate construct and discriminant validity, correlations were run between the TM and the NB-HABS (Habarth, 2015) and IT-GMSRM (Testa et al., 2015).

Study 1 results

Data screen

Results of Mardia’s multivariate estimates indicated significant multivariate skewness or kurtosis and that the data were multivariate normal. Further, the data were evaluated for multivariate outliers using Mahalanobis distances which indicated that there were 58 multivariate outliers (p < .001) across all items. At the item level, no significant skewness of kurtosis was detected. Despite the presence of multivariate non-normality, EFA with ML estimation was still chosen due to the strong relations among TM items (r > 0.8) and large sample size (Watkins, 2018). Finally, 26 participants did not complete any of the IT-GMSRM (Testa et al., 2015) items and 27 participants did not complete any of the NB-HABS (Habarth, 2015) items. In response, all correlation tests were conducted with listwise deletion.

Exploratory factor analysis

The 67 proposed items that comprised the TM were found to be acceptable for factor analysis as signified by a Kaiser-Meyer-Olkin value of .97 and significant Bartlett’s test of sphericity: χ2 (2211, N = 497) = 21,847.19, p < .001. EFA using ML factoring and Varimax rotation was utilized for all subsequent tests. Nine factors emerged with eigenvalues that exceeded 1 in the initial unconstrained EFA; however, the scree plot suggested four factors. A review of the factors that emerged from the initial EFA indicated that three factors contained at least three items with loadings > 0.40. As such, a three-factor model was assessed in subsequent EFAs (Field, 2013; Tabachnick & Fidell, 2007).

Items were removed one-by-one beginning with items which did not load onto a primary factor with a loading > 0.40, which resulted in the removal of four items. Next, items which cross loaded onto two or more factors according to the criteria outlined were removed, starting with the item with the lowest absolute maximum loading, which resulted in the deletion of 40 additional items. After the removal of these 42 items, one factor only contained two items and a two-factor EFA was ran. All items loaded > 0.40 on these factors; however, an additional four items were cross-loaded and were removed.

In total, 48 EFAs were conducted which resulted in two factors comprised of 19 of the original 67 items and explained 53.37% of the total variance. Factor 1 was named Overarching Transnormativity and consisted of 14 items that explained 42.36% of the variance. Factor 2 was named Gender Discordance Transnormativity, consisted of five items, and explained 11.00% of the variance. Communalities ranged from 0.20 to 0.62, with the average being 0.47. Items, factor loadings, and communalities values are displayed in Table 2. Although, the results of EFA analyses suggested the TM is comprised of two factors, conceptually there is considerable overlap in the items comprising each factor. As such, subsequent tests were run with TM comprising one general factor with all remaining items. The one factor model of the Transnormativity demonstrated excellent internal consistency (α = 0.92).

Table 2.

Final Exploratory analysis of the transnormativity measure from Study 1.

	Factor
Item	1	2	Comm.
Being trans is a result of abuse or victimization	.812	−.117	.623
Trans people must be heterosexual	.795	−.075	.611
Binary trans people must dress in traditionally masculine or feminine ways	.746	.159	.579
Binary trans people must style their hair in traditionally masculine or feminine ways	.743	.105	.544
Nonbinary identities are less valid than binary trans identities	.742	.036	.545
Children cannot be nonbinary	.732	.130	.567
Nonbinary people cannot make up their mind about their gender identity	.729	.089	.555
Nonbinary identities are a phase	.714	.120	.528
Trans people should have wanted to socially transition since they were very young	.712	.122	.544
A trans person’s gender presentation and gender identity must match	.672	.207	.496
Nonbinary people should not want to participate in gendered activities (e.g., sports, men’s/women’s social clubs)	.670	.193	.496
Binary trans people must conform to traditional gender stereotypes of gendered behavior (e.g., walking, gesturing, speaking patterns)	.661	.097	.461
Nonbinary people must use gender neutral pronouns (i.e., they/them, ze/hir/hirs, etc.)	.634	.184	.440
Trans people cannot be happy unless they medically transition	.548	.155	.380
Trans people must find their body incongruent with their gender identity prior to gender affirming medical care	.294	.663	.447
Trans people must find their body incongruent with their gender identity	.287	.694	.424
A trans identity is something you are born with	.074	.412	.293
Trans people were born in the wrong body	.049	.421	.195
People are born trans	−.222	.458	.203

Note. Bolded values represent factor loadings > .40.

Construct and discriminant validity

To test for construct validity, the finalized TM was correlated with the (NB-HABS; Habarth, 2015) with results indicating a large positive relation (r = 0.71, p < 0.001). As such, construct validity of the TM was supported. To demonstrate discriminant validity, it was expected that TM and IT-GMSRM (Testa et al., 2015) would correlate with either small or moderate correlations. The TM and IT-GMSRM produced a moderate correlation (r = 0.41, p < 0.001), supporting discriminant validity (Rönkkö & Cho, 2022).

Study 2

In study 2, the factor structure of the TM was confirmed and the construct and predictive validity were tested with an independent sample of TNB participants.

Hypotheses

It was expected that the finalized TM would reveal a stable factor structure during CFA analysis and that the items within the factor, as well as the scale as whole, would be internally consistent as they are measuring the same construct, transnormativity. Further, it was expected that configural, metric, and scalar invariance would be supported, ensuring that the factor structure, factor loadings, and item intercepts were equal across trans women, trans men, and nonbinary individuals. To demonstrate construct validity it was expected the TM would be negatively related to a measure of TNB community belonginess (i.e., the Transgender Community Belongingness Scale [TCBS]; Barr et al., 2016). Finally, it was expected the scales of the TNCM would be negatively related to mental health outcomes (i.e., the Depression, Anxiety, and Stress Scales [DASS-21]; Henry & Crawford, 2005), providing support for predictive validity (see Supplementary Information for full study hypotheses).

Study 2 method

Recruitment

Recruitment was the same as the processes outlined in Study 1; however, additional community organizations from the 25 states (randomly selected via excel sort with random numbers provided to each state) in the United States not selected in Study 1 were contacted. Additionally, no compensation was provided for participation in Study 2.

Participants

Six hundred and eighty-seven participants initiated the survey; however, 143 participants did not answer any questions beyond demographics, did not complete all of the items of the TM, or did not pass the engagement checks, one participant indicated that they were under the age of 18, and three participants indicated that their gender was the same as their sex assigned at birth and were removed. This left a final sample size of 540 participants who identified as women (n = 183), men (n = 175), and nonbinary (n = 182). Participants ranged in age from 18 to 75 (M = 27.04, SD = 8.64) and most were residing in the United States (71.5%). Regarding race participants were able to select multiple identities and 86.5% of the sample endorsed white and 30.0% of the sample endorsed a marginalized race identity. Table 1 includes further demographic information including education and sexual identity.

Procedure and measures

Participants completed an online survey that began with an informed consent document followed by demographic questionnaire, the reduced items of the TM from study 1, the TCBS (Barr et al., 2016), and the DASS-21 (Henry & Crawford, 2005), a question instructing participants who would like to complete a brief follow up survey in two weeks to enter a unique personal code (i.e., initials, first number of street address, and first number of their phone number), and concluded with a prompt which thanked participants and provided them an opportunity to leave suggestions to improve future studies. Four engagement checks were dispersed throughout the survey to check the quality of the data and to ensure participants remained consistent in their responses.

Modified transnormativity measure

The Modified Transnormativity Measure was developed in response to EFAs conducted in Study 1 and included 19 items. Responses to items are recorded on a 6-point scale from 1 (strongly disagree) to 6 (strongly agree). Scale items are summed to create subscale scores. Results of tests of internal consistency are reported in the results section.

Construct validity

Trans and nonbinary community belonging was measured via the Transgender Community Belongingness Scale (TCBS), a modified version of the lesbian community belongingness scale (Doolin & Budge, 2015) validated for use with TNB samples by Barr et al. (2016). The TCBS is comprised of 9 items scored on a 5-point Likert scale from 1 (not at all) to 5 (all of the time) and scores are averaged with higher scores indicating greater belonging. When initially validated with a sample of TNB individuals, the scale showed excellent internal consistency (α = 0.90; Barr et al., 2016) and in the current study (α = 0.89).

Predictive validity

Mental health distress was measured via the depression and anxiety subscales of the Depression, Anxiety, and Stress Scales (DASS-21; Henry & Crawford, 2005), with each subscale composed of 7 items. Responses range from 0 (never) to 3 (always) with items summed and higher scores indicate greater distress. In previous studies with TNB individuals, the DASS-21 has demonstrated good to excellent internal consistency (α = .94; Lindley & Bauerband, 2022) and in the current study (α = .92).

Data analyses

All data analysis took place in R using the statistical package lavaan (Rosseel, 2012) and SPSS v26. First, data were screened for multivariate and univariate outliers as well as skewness and kurtosis. Next, items from the finalized TM were entered into a CFA according to the facture structure found in study 1. Overall fit for CFA was assessed via Comparative Fit Index (CFI) and the Root-mean Square Error of Approximation (RMSEA) with its 90% confidence interval (Schmitt, 2011). For the CFI, values greater than 0.90 indicate good fit and values greater than 0.95 indicate very good fit (Hu & Bentler, 1999; Kline, 2011). For the RMSEA, values less than 0.10 indicate good fit and values less than 0.05 indicate very good fit (Browne & Cudeck, 1992; Kline, 2011). To assess the invariance of the TM, the difference in CFI and Chi-Square of the restricted model against the less restricted model was calculated. A ΔCFI ≥ 0.01 and a significant Δχ2 indicates a significant decrease in model fit and suggests noninvariance (Chen, 2007). To demonstrate construct and predictive validity, Pearson’s correlations were run.

Study 2 results

Data screen

Results of Mardia’s multivariate estimates indicated significant multivariate skewness and kurtosis and that items of the TNCM were multivariate nonnormal. Further, the data were evaluated for multivariate outliers using Mahalanobis distances which indicated that there were 40 multivariate outliers (p < .001) across all items of the TM. At the item level, the Lilliefors test indicated that the items from all scales data were nonnormally distributed. In response, items were inspected with three items being identified has having skewness > 3 and kurtosis > 10; “trans people must be heterosexual,” “being trans is a result of abuse or victimization,” and “binary trans people must style their hair in traditionally masculine or feminine ways” (Weston & Gore, 2006). Additionally, more than 94% of participants responses for these items were either strongly disagree or disagree, indicating a clear floor effect which necessitated the removal of these items. To address the multivariate outliers, multivariate nonnormality, and ordered-categorical data the CFAs and invariance models were conducted using diagonally weighted least squares (DWLS). Finally, some participants did not complete any of the validity measures and all correlation tests were conducted with listwise deletion.

Review of correlation table

The correlation table of the remaining 16 TM items was reviewed to ensure that no item pairs were highly correlated and over inflating path estimates (Brown, 2015). Four item pairs were identified as being strongly correlated. A review of the wording of these item pairs revealed that each item pair is either worded quite similarly or one item of the pair is specific and the other more general. As such the following items were removed: “nonbinary people cannot make up their mind about their gender identity” (r = 0.79, p < 0.01), “trans people must find their body incongruent with their gender identity” (r = 0.74, p < 0.01), “binary trans people must dress in traditionally masculine or feminine ways” (r = 0.77, p < 0.01), and “a trans identity is something you are born with” (r = 0.68, p < 0.01).

Confirmatory factor analysis

A CFA was conducted to determine if the structure found in Study 1 was supported in an independent sample. The one factor model of 12 items demonstrated good to excellent fit the data (χ² [54] = 211.60 p < .001, CFI = .992, RMSEA = 0.074 [90% CI: 0.063, 0.084]). All items loaded greater than 0.50 on to the primary factor with the exception of the item “people are born trans,” as such this item was removed (Awang, 2014). The CFI was re-run with the remaining 11 items and the model showed excellent fit to the data, (χ² [44] = 112.67, p < .001, CFI = .997, RMSEA = 0.054 [90% CI: 0.054, 0.066]). The final items, factor loadings, and Cronbach’s alpha score can be found in Table 3 and the Appendix.

Table 3.

Items comprising the final TM with factor loadings in standardized form and Cronbach’s alpha from Study 2.

Transnormativity measure α = 0.90	Factor loading (SE)
Children cannot be nonbinary	0.71 (0.03)
A trans person’s gender presentation and gender identity must match	0.87 (0.02)
Nonbinary identities are less valid than binary trans identities	0.94 (0.02)
Trans people were born in the wrong body	0.53 (0.03)
Nonbinary people cannot make up their mind about their gender identity	0.86 (0.02)
Trans people cannot be happy unless they medically transition	0.74 (0.02)
Binary trans people must conform to traditional gender stereotypes of gendered behavior (e.g., walking, gesturing, speaking patterns)	0.77 (0.02)
Trans people should have wanted to socially transition since they were very young	0.77 (0.02)
Nonbinary people should not want to participate in gendered activities (e.g., sports, men’s/women’s social clubs)	0.75 (0.03)
Trans people must find their body incongruent with their gender identity	0.80 (0.02)
Nonbinary people must use gender neutral pronouns (i.e., they/them, ze/hir/hirs, etc.)	0.82 (0.02)

Invariance

The results of the test of configural invariance indicated that the factor structures were the same across participant gender identity groups (i.e., trans men, trans women, and nonbinary individuals), χ² [132] = 174.35 p = .008, CFI = .998, RMSEA = 0.042 [90% CI: 0.023, 0.058]). Invariance at the metric level was not supported. While the model fit indices where in the acceptable range, χ² [152] = 373.96 p < .001, CFI = .989, RMSEA = 0.090 [90% CI: 0.079, 0.102]), there was a significant change in fit, Δχ2 [20] = 199.61 p < .001. In response item loading constraints were added one by one to see which item loadings could be set to equal and results indicated that only item “trans people were born in the wrong body” could be constrained indicating that the measure is metric noninvariant.

Descriptive statistics of the transnormativity measure

The average scores across for the total sample and by gender identity for the TM can be found in Table 4. One-way analysis of variance test was conducted to determine if there were significant differences between gender identity groups regarding scores of the TM. There was a significant overall effect of gender identity on TM scores, F(2, 539) = 29.90, p < .001, η² = 0.10. Bonferroni post-hoc comparisons indicated that trans men’s scores were significantly higher than trans women’s and nonbinary individuals’ scores and that trans women’s score were significantly higher than nonbinary individuals scores. All remaining tests were run with both the complete sample as well as by gender identity in response the significant differences in TM scores.

Table 4.

Correlations between the transnormativity measure and construct and predictive validity measures from Study 2.

			Correlations
TM	Mean	SD	Anxiety	Depression	TCBS
Total	26.51	9.92	−0.04	0.08	−0.39**
Nonbinary	22.68	5.89	−0.07	0.15*	−0.07
Women	26.60	9.02	−0.02	0.01	−0.37**
Men	30.40	12.38	−0.10	0.16*	−0.57**

Note. Transgender Community Belongingness Scale (TCBS). *p < .05, **p < .01

Note. Bolded values represent factor loadings > .40.

Construct and predictive validity

To test for construct validity, the finalized TM was correlated with TCBS (Barr et al., 2016). The TM was negatively related to the TCBS for the total sample, trans woman and trans men, but not for nonbinary individuals providing partial support for construct validity (see Table 4). To test for predictive validity, the TM was correlated with the subscales of the DASS-21 (Henry & Crawford, 2005). The TM was only significantly positively related to depression scores for trans men and nonbinary individuals (see Table 4). There were no significant relations between the TM and the anxiety subscale. As such, predictive validity was partially supported.

Study 3

In study 3, the test re-test reliability of the TM was assessed. Specifically, it was expected that TM would demonstrate stability over time and that participants’ responses between the administration of the TM described in study 2 and subsequent administration two weeks later would be consistent.

Study 3 method

Recruitment

Participants from study 2 who indicated interest in participating in a brief follow up survey were contacted via email two weeks after they completed the Study 2 survey. The email included a link to the survey as well as a reminder that the survey would not be compensated.

Participants

One hundred and seventy-nine participants initiated the survey; however, 14 participants did not pass the engagement check, one did not include the unique personal code, and two indicated the same gender identity as their assigned sex and were removed. When the included codes of the remaining 162 participants were compared to the provided codes in study 2, only 107 participants had matching codes. Participants identified as women (n = 38), men (n = 29), and nonbinary (n = 48) individuals, ranged in age from 18 to 61 (M = 27.28, SD = 8.65), and the majority were residing in the United States (71.0%). Regarding race participants were able to select multiple identities and 89.7% of the sample identified endorsed white and 22.4% of the sample endorsed a marginalized racial identity. Table 1 includes further demographic information including education and sexual identity.

Procedure and measures

Participants completed an online survey that began with an informed consent document followed by demographic questionnaire, the reduced items of the TM from study 1 with one engagement check item, a question instructing participants to enter a unique personal code, and concluded with a prompt which thanked participants and provided them an opportunity to leave suggestions to improve future studies.

Finalized trans and nonbinary coping measure

Participants were presented with the modified TM described in Study 2; however, analyses were conducted with the final items developed during Study 2. Results of tests of internal consistency for study 3 indicated excellent reliability (α = .91)

Data analyses

All data analysis took place in SPSS v26. Test-retest was assessed through an Interclass Correlation Coefficients (ICC) 2-way mixed-effects model (Portney & Watkins, 2009), with ICC values less than 0.50 indicating poor reliability, values between 0.50–0.75 indicating moderate reliability, values between 0.75–0.90 indicating good reliability, and values greater than 0.90 indicating excellent reliability (Koo & Li, 2016).

Study 3 results

Test re-test

To ensure that the TM demonstrated stability over time, ICC 2-way mixed-effects model were tested. The ICC results were .814 indicating good stability (Koo & Li, 2016).

Discussion

In the current study, we aimed to develop and validate a measure that could capture how TNB people have internalized and/or understand the attitudes and beliefs of cisgender and TNB communities regarding normed gender presentations, behaviors, histories, and identities of TNB people. The resulting measure from this process included an 11-item measure—the Transnormativity Measure (TM). To our knowledge, this is the first measure of its kind to capture transnormativity for TNB communities. This study confirmed that the TM is a valid and reliable measure to use with TNB populations and is capable of producing high quality data within clinical and research settings.

To date, studies focusing on transnormativity have sought to operationalize the experience (see Bradford et al., 2019; Bradford & Syed, 2019; Garrison, 2018; Johnson, 2016). To our knowledge, there has yet to be a way to capture these concepts for use in quantitative research or in clinical settings. Although we were able to identify six overarching constructs related to transnormativity in our review of the literature, the results of the EFA and CFA indicated that these domains do not hold up as distinct constructs quantitatively, but instead that an overall measure to capture attitudes and beliefs is a more robust way to capture the experience of transnormativity for TNB people.

Regarding specific validity components for the scale, there were several interesting findings. First, nonbinary TM scores were significantly lower than trans men’s and trans women’s scores. There are several possible explanations for this finding. Most nonbinary people will experience challenging the binary and confronting and transgressing narratives regularly/daily (McGuire et al., 2019), thus it is possible that they are inundated with more expansive ways and messages of how gender can be and that their answers to items reflect less adherence to transnormative attitudes and behaviors. Trans men’s scores were highest out of the three gender categories for this study. While both trans men and trans women are exposed to cisnormative and transnormative messages, there may be more pressure for trans women to not question the messages due to higher threats for physical safety (Boe et al., 2020). It is possible there is more room for trans men to question transnormative messages, given recent understanding of how manhood, brotherhood, and sexism are all being resisted and confronted in communities (see Phillips & Rogers, 2021).

In addition, when testing construct validity, results demonstrated that that the TM was moderately negatively related to trans community belongingness for the overall sample, meaning that when participants indicate that they acknowledge and observe higher levels of transnormativity, they also report feeling less connected to their trans communities. In their study regarding transnormativity, Bradford and Syed (2019) capture participants’ experiences, frustrations, and resistance to transnormativity. They note that connecting with other TNB people was considered a source of strength and positivity in the face of experiencing transnormativity. However, their study does not capture the nuances of the impact of belongness if transnormativity is occurring within community, which is likely what is being captured in this result for the current study.

We also tested the predictive validity of transnormativity and depression, anxiety and stress (as measured by the DASS-21; Henry & Crawford, 2005). Neither anxiety nor depression were correlated with the TM for the overall samples. However, depression was positively correlated with TM for trans men and nonbinary people, indicating higher reports of depression when also acknowledging and observing higher levels of transnormativity. Although all are impacted by cisnormativity and transnormativity in their own unique ways, it is most likely that depression for trans women is accounted for by factors other than transnormativity. For example, studies that focus on depression in trans women emphasize specific minority stress experiences (e.g., Klemmer et al., 2021; Nuttbrock et al., 2014); that may not account for or overlap with transnormativity experiences in the same way that they do for trans men and nonbinary people. A future area of research would be to understand if the TM relates to positive psychological well-being for TNB individuals. Specifically, it could be important for researchers to understand the potential relations between the TM and measures of positive TNB identification such as those measured by the Positive Identity for Transgender Identified Individuals (Riggle & Mohr, 2015).

Results demonstrate metric noninvariance in the sample, meaning that there is a latent construct of transnormativity and the 11-item measure is accurately measuring transnormativity for all groups; however, the interpretation of transnormativity will differ for trans men, trans women, and nonbinary people. This is not a surprising finding, given that gender socialization, norms, rules, and expectations are not uniform for gendered groups (see Tatum et al., 2020).

Limitations

The results from studies 1–3 provided support for the development, validation, and psychometric properties of the TM; however, it is not without limitations. First, this study utilized online recruitment of participants which can disproportionately sample White, educated, and middle-class participants (Christian et al., 2008). While the sample for Study 1 was racially diverse, the subsequent two studies were not as representative. Research has noted that providing compensation can positively incentivize racially marginalized individuals to participate in research (Williams et al., 2013), as compensation was only provided for Study 1, this may explain the noticeable differences in the racial compensation of studies. Future researchers should be mindful to incorporate compensation when conducting research with TNB participants to recruit racially diverse samples. Additionally, likely due to online recruitment, the majority of participants indicated having at least attended college which is not representative of TNB communities at large. The results should be interpreted in light of these limitations.

Second, the way in which racial identity was assessed may have not been affirming to participants identities. Specifically, there was a discrepancy between the number of participants who endorsed a biracial/multiracial identity and those who selected multiple racial identities. For example, no participants selected a biracial/multiracial identity in Study 3 yet 12.1% of the sample selected white and another identity. It is important for researchers to utilize current and accurate racial identity labels and acknowledge that while utilizing forced-choice options is more adaptable to quantitative research the categories themselves may lead to inaccuracies like those found in the current study (Ross et al., 2020). It tentatively appears that for TNB individuals the utilization of biracial/multiracial may not be as accurate and alternative identity choices should be explored.

Third, transnormativity is an evolving concept that is responsive to shifts in social norms. While we have captured transnormativity across five domains (i.e., emphasis on anatomical gender dysphoria, discovery narrative, medicalization, assimilation, victimization, and social interactions) there are likely to be additional domains that may become relevant. For example, an expectation to describe the affective aspects of TNB experience with negative valanced emotions could be an important missing link between transnormativity and psychological distress (Budge et al., 2021; Malatino, 2022). It is important for future researchers to continue to understand and integrate components of transnormativity and to revise, modify, and update the TM in response.

Future directions and clinical implications

The creation of the TM measure allows researchers and clinicians to better capture the extent to which TNB people are perceiving transnormativity. From a research perspective, researchers will be able to use this measure when they are conducting studies that focus on gender socialization, identity processes, gender norms, and community connectedness. Future researchers can use this measure to focus on often ignored components of the minority stress model—characteristics of minority identity—that include identity prominence, identity valence, and identity integration (see Meyer, 2003). Since transnormativity is exactly related to these characteristics, this measure can be used in studies that more compressively test the minority stress model. In addition, given the preliminary nature of the bivariate correlations with community belongingness and depression, future research can focus on if perceptions of transnormativity mediate social support and depression.

In addition to constructing additional research studies with the measure, there are several clinical contexts in which it might be useful for TNB people to complete this measure. First, in group therapy contexts when group members bring up how to navigate messages they receive about being “trans enough,” the TM could be useful tool for group members to take individually and then discuss as a group what their answers were. Having a normed measure that each group member can fill out may assist clients with feeling validated by knowing that these messages exist and allows for each group member to have a reference point for their responses. In addition to group therapy, providing this measure to individual clients can also have a similar effect, where the therapist and client can go through each item and discuss the client’s relationship to each message and process both cognitions and emotions related to the message. Although we did not test the therapeutic nature of the measure, this would be an important area for future research. For example, for clinical studies and intervention research, providing the TM as a baseline measure can offer intervention points for discussing the impact of specific types of transnormativity within a therapeutic context.

Appendix.

The transnormativity measure

Instructions: The items below are about various attitudes and beliefs that exist within transgender and nonbinary (TNB) communities about the ways trans people “should” behave or think about gender. In other words, these statements are about beliefs held by some trans people within TNB communities about how trans people “should” experience gender and who “counts” as trans. Prior research shows that these attitudes and beliefs include unwritten rules within TNB communities that are communicated in subtle and explicit ways that can impact well-being. Read each statement carefully and indicate how much you agree or disagree with each. Please remember, we are not asking about your personal experiences, but to what extent you may (dis)agree with these messages you have heard from or about other trans individuals. It is likely that you may agree with some statements and disagree with others.

Note: The term “trans” refers to all individuals whose gender identity differs from their assigned gender/sex, including transgender, nonbinary, and gender nonconforming individuals.

Scoring: Items are scored on a 6-point scale (“strongly disagree” = 1 to “strongly agree” = 6). A higher score therefore indicates greater endorsement of transnormativity. To obtain a score calculate the sum.

Items:

1	Children cannot be nonbinary
2	A trans person’s gender presentation and gender identity must match
3	Nonbinary identities are less valid than binary trans identities
4	Trans people were born in the wrong body
5	Nonbinary people cannot make up their mind about their gender identity
6	Trans people cannot be happy unless they medically transition
7	Binary trans people must conform to traditional gender stereotypes of gendered behavior (e.g., walking, gesturing, speaking patterns)
8	Trans people should have wanted to socially transition since they were very young
9	Nonbinary people should not want to participate in gendered activities (e.g., sports, men’s/women’s social clubs)
10	Trans people must find their body incongruent with their gender identity
11	Nonbinary people must use gender neutral pronouns (i.e., they/them, ze/hir/hirs, etc.)

Disclosure statement

No potential conflict of interest was reported by the authors.

Role of funding sources

This research received funding from the Baldwin Wisconsin Idea Endowment at the University of Wisconsin-Madison.

Conflicts of interest/competing interest

The authors have no conflicts of interest nor competing interests to report.

Availability of data and material

The authors have not deposited the data in a publicly accessible database.

Code availability

Not applicable.