Trans youth are at the center of a political battle over trans rights. In response, many LGBTQ+ youth are engaging in political action, including protest, campaigning, and public education and testimony. We investigate whether two social identity-based factors - community connectedness and identity importance - are associated with political action related to trans issues among transgender youth and cisgender LGBPQ youth. We conducted cross-sectional and longitudinal analyses of data from a longitudinal annual survey of transgender and cisgender youth in 2022 and 2023 (total N = 396). We found that youth who feel connected to those who share their trans or LGBPQ identity are more likely to engage in political action for trans rights. However, we found no evidence that those who feel their trans or LGBPQ identity is important to them are more likely to engage in political action. These findings contribute to the literature on social identity-based features that may motivate people to engage in political action.
Young people have recently been at the forefront of political action for trans rights (Fine et al., 2018; Yurcaba, 2021). In the US, over the last decade, there has been an exponential rise in anti-trans bills banning trans youth from accessing gender-affirming healthcare, participating in school sports, using public bathrooms, and more (Trans Legislation Tracker, 2024). Some Canadian provinces have also begun to introduce similar policies (Amnesty International, 2024; Ibrahim, 2023). Given the unprecedented nature of legal attacks on the rights of trans youth, a contemporary understanding of what motivates their political action is particularly important. In this paper, we examine the association of two social identity-related factors, community connectedness and identity importance, with political action for trans rights by trans and LGBPQ (Lesbian, Gay, Bisexual, Pansexual, and/or Queer) youth. We will use “trans” to refer to all people who identify as transgender, which includes trans boys and girls, nonbinary and genderqueer people, and those with other gender nonconforming identities (Davidson, 2007).
According to the expanded social identity model of collective action (Agostini & van Zomeren, 2021; van Zomeren et al., 2008, 2018), there are four interrelated contributors to collective action – social identity as a member of the disadvantaged group, an affective response to perceived injustice against members of one’s group, belief in the group’s efficacy to bring about social change, and moral conviction. As a first step in applying this model to the study of trans youth’s political action, we focus on the most foundational of these factors in the original model, social identity. Social identity is further broken down into a sense of connection with other members of one’s group (which we refer to as community connectedness) and the cognitive centrality of one’s group membership (which we refer to as identity importance). In our project, we examine community connectedness and identity importance both among trans youth - the youth being directly targeted by anti-trans legislation - and among cisgender LGBPQ youth, who may see themselves as acting in solidarity with members of a similarly marginalized identity.
Collective action is any action taken by individuals advocating for members of their disadvantaged group (Wright et al., 1990). These actions can be done by individuals alone or performed together as a group. Here, we use “political action” to refer to a subset of actions, such as educating others and political protest, that do not require coordinated group action.
Both community connectedness and identity importance may motivate trans and cis LGBPQ youth’s political action for trans rights (van Zomeren et al., 2008). For LGBTQ+ youth broadly, feeling connected to other members of one’s community can help connect their personal experiences with a larger collective struggle (DiFulvio, 2011; B. A. Robinson & Schmitz, 2021). This sense of community connectedness has also been linked to feelings of agency (Poteat et al., 2016), which is important for engagement in political action. Prior work shows that community connectedness is a strong predictor of trans adults’ civic engagement (Billard, 2022). To our knowledge, there is no existing research examining the relationship between community connectedness and political action in trans youth.
Identity importance is a related but distinct factor that may be associated with youth engagement in political action. People who believe their marginalized group membership is important to their self-concept tend to be more acutely aware of structural discrimination against their group and are more motivated to engage in behaviors that promote equity for their group (e.g., Dunn & Szymanski, 2018; Hope et al., 2019). Participatory action research with queer and trans youth finds that experiencing discrimination is associated with higher engagement in activism (Fine et al., 2018). It is an untested question whether, for trans and cis LGBPQ youth, the importance of their identity functions in similar ways to promote action for trans rights.
The Current Research
We analyzed responses from transgender youth and cisgender LGBPQ youth across the US and Canada enrolled in a longitudinal survey to investigate whether community connectedness (Analyses 1, 2, 3) and identity importance (Analysis 1) were associated with political action related to trans issues. As a secondary question, we explored whether the strength of the association between these social identity-related factors and political action for trans rights differed between transgender youth and cisgender LGBPQ youth. On the one hand, cisgender LGBPQ youth may see themselves as part of the broader LGBTQ+ umbrella, but as they are not directly impacted by anti-trans laws, their social identity may be less relevant to their political action for trans rights.
In this study, Analysis 1 is a cross-sectional analysis investigating the associations between community connectedness, identity importance, and participation in political action among transgender and cisgender LGBPQ youth in 2022. Analysis 2 is a replication of Analysis 1 with data from 2023, with two changes: (a) we did not study identity importance as a predictor of political action because it was not a significant predictor in Analysis 1, and (b) we included an additional measure of political action introduced in the 2023 survey as a dependent variable. Analysis 3 is a longitudinal analysis investigating whether community connectedness in 2022 predicted (a) political action in 2023, and (b) change in participation in political action from 2022 to 2023.
We predicted that both transgender youth and cisgender LGBPQ youth who reported higher levels of community connectedness and identity importance to their own respective communities/identities would also report higher levels of political action related to trans issues (Research Question 1). We also explored whether the relationship between those identity measures and political action would differ by participant group (Research Question 2).
Note that in the following Method and Results sections, we use the term “predictor” only to refer to the independent variables in our statistical models, and not as an indication of a causal relationship. The preregistration for Analysis 1 can be found at https://osf.io/sn4mj and for Analyses 2-3 at https://osf.io/svt9e respectively. While we had initially intended to use the first preregistration for all three sets of analyses reported here, we submitted a revised preregistration for Analyses 2 (2023 data) and 3 (2022 and 2023 data) after running Analysis 1 (2022 data) and before beginning the analyses for 2 and 3. The changes to Analysis 2 were made to improve model fit to the data; see the regression analyses below for details. In Analysis 3, besides no longer studying the association between identity importance and political action (as in Analysis 2), the updates also included an examination of how community connectedness is associated not only with political action in 2023, but also with change in political action from 2022 to 2023.
Method
Participants
Participants in the current study were recruited between 2013 and 2021 to participate in longitudinal studies of gender development. At the time of recruitment, participants belonged in one of four categories: (a) transgender boys and girls who had completed a binary social transition, (b) gender nonconforming children who had not completed a binary social transition, (c) cisgender siblings of transgender boys and girls, and of gender nonconforming children, and (d) unrelated cisgender children who were matched in age and gender to the transgender and gender nonconforming children. Over time, some participants have shifted between categories; e.g., a cisgender sibling who now identifies as transgender, or a transgender boy who now identifies as nonbinary. In general, participants whose gender does not align with their sex assigned at birth in this sample do explicitly identify as trans (Wittlin et al., 2024). The determination of participants’ gender identity and modality (trans vs. cis) for this study was done based on participant responses in the years included in our analyses (2022 and 2023). Participants were included in this sample as “transgender” if their response to “Are you currently transgender, nonbinary or gender diverse?” was “Yes”, and their response to “Which of these terms describe your gender identity? (select all that apply)” was “transgender” (either alone or in combination with other gender identity labels), and their response to “Are you transgender” was “Yes” (either alone or in combination with other identity labels). Participants were included in this sample as “cisgender LGBPQ” if their response to “Are you currently transgender, nonbinary or gender diverse?” was “No”, and their response to “Which of these terms describe your gender identity? (select all that apply)” was NOT “transgender” (regardless of other identity labels), and their response to “Are you transgender” was “No”; and their response to “Are you gay, lesbian, bisexual, pansexual or LGBPQ?” was “Yes”. For cross-sectional analyses, participants had to meet one of the above set of criteria (either for transgender or cisgender LGBPQ participants) in the year in which the analysis was being conducted. For longitudinal analyses, participants had to meet the same set of criteria in both 2022 and 2023 to be included in the analyses.
Although the sample in the current research overlaps with others in prior papers based on this dataset, the results of the measures and analyses used in this work have never been reported elsewhere.
Analysis 1
Youth between 12-21 years of age participated in the 2022 annual survey of gender development. For this study, participants had to be either transgender (of any sexual orientation, N = 175) or cisgender and LGBPQ (N = 81); see above for inclusion criteria in trans or cis LGBPQ categories. Eight participants were excluded because they did not respond to the questions of interest in this study, for a final sample of 248 participants (Mage = 14.75, SD = 1.99, see Table 1 for demographic information). 32 participants were part of a pair of siblings who took the survey (non-independence accounted for in robustness checks, see SM). The survey was conducted online using Qualtrics software.
| Variable | Analysis 1 (N = 248) | Analysis 2 (N = 306) | ||
| Trans (N = 171) | Cis LGBPQ (N = 77) | Trans (N = 211) | Cis LGBPQ (N = 95) | |
| Gender | n (n%) | n (n%) | n (n%) | n (n%) |
| Boy | 41 (24%) | 16 (20.8%) | 56 (26.5%) | 21 (22.1%) |
| Girl | 89 (52%) | 54 (70.1%) | 120 (56.9%) | 69 (72.6%) |
| Neither | 11 (6.4%) | 0 (0%) | 9 (4.3%) | 1 (1.1%) |
| Both | 6 (3.5%) | 2 (2.6%) | 3 (1.4%) | 1 (1.1%) |
| Changes | 14 (8.2%) | 0 (0%) | 14 (6.6%) | 0 (0%) |
| I don't know | 10 (5.8%) | 5 (6.5%) | 9 (4.3%) | 2 (2.1%) |
| Unknown | 0 (0%) | 0 (0%) | 0 (0%) | 1 (1.1%) |
| Race and ethnicity | ||||
| Asian, non-Hispanic | 7 (4%) | 1 (1.3%) | 5 (2.4%) | 0 (0%) |
| Black/African, non-Hispanic | 0 (0%) | 0 (0%) | 3 (1.4%) | 1 (1.1%) |
| Native American/Alaska Native, Hispanic | 0 (0%) | 1 (1.3%) | 0 (0%) | 0 (0%) |
| Multiracial, Hispanic | 3 (1.8%) | 1 (1.3%) | 2 (0.9%) | 1 (1.1%) |
| Multiracial, non-Hispanic | 21 (12.3%) | 12 (7%) | 30 (14.2%) | 18 (18.9%) |
| Multiracial, unknown ethnicity | 0 (0%) | 1 (1.3%) | 0 (0%) | 1 (1.1%) |
| White/European, Hispanic | 13 (7.6%) | 4 (5.2%) | 16 (7.6%) | 6 (6.3%) |
| White/European, non-Hispanic | 123 (71.9%) | 57 (74%) | 152 (72%) | 68 (71.6%) |
| Unknown race, Hispanic | 4 (2.3%) | 0 (0%) | 3 (1.4%) | 0 (0%) |
| Sexual orientation | ||||
| Gay | 10 (5.8%) | 5 (6.5%) | 13 (6.2%) | 5 (5.3%) |
| Lesbian | 12 (7%) | 5 (6.5%) | 21 (10%) | 8 (8.4%) |
| Bisexual | 23 (13.5%) | 36 (46.8%) | 29 (17%) | 40 (42.1%) |
| Straight | 47 (27.5%) | 1* (1.3%) | 63 (36.8%) | 2* (2.1%) |
| Asexual | 11 (6.4%) | 2 (2.6%) | 12 (5.7%) | 4 (4.2%) |
| Pansexual | 15 (8.8%) | 8 (10.4%) | 16 (7.6%) | 10 (10.5%) |
| I prefer a different term | 21 (12.3%) | 6 (7.8%) | 21 (10%) | 5 (5.3%) |
| Multiple | 29 (17%) | 14 (18.2%) | 33 (15.6%) | 19 (20%) |
| Unknown | 3 (1.8%) | 0 (0%) | 3 (1.4%) | 2 (2.1%) |
| Parent education | ||||
| High school diploma | 2 (1.2%) | 2 (2.6%) | 5 (2.4%) | 3 (3.2%) |
| Some college/Associate's Degree | 18 (10.5%) | 7 (9.1%) | 28 (13.3%) | 10 (10.5%) |
| College/Bachelor's Degree | 54 (31.6%) | 31 (40.3%) | 67 (31.8%) | 40 (42.1%) |
| Advanced Degree (MA, MD, PhD, etc.) | 95 (55.6%) | 36 (46.8%) | 108 (51.2%) | 42 (44.2%) |
| Other | 1 (0.6%) | 1 (1.3%) | 2 (0.9%) | 0 (0%) |
| Other + some other schooling also listed | 1 (0.6%) | 0 (0%) | 1 (0.5%) | 0 (0%) |
| Household income | ||||
| <$25,000 per year | 9 (5.3%) | 4 (5.2%) | 9 (4.3%) | 2 (2.1%) |
| $25,000-50,000 per year | 18 (10.5%) | 4 (5.2%) | 22 (10.4%) | 4 (4.2%) |
| $50,001-75,000 per year | 30 (17.5%) | 9 (11.7%) | 40 (19%) | 13 (13.7%) |
| $75,001-125,000 per year | 59 (34.5%) | 23 (29.9%) | 70 (33.2%) | 33 (34.7%) |
| >$125,000 per year | 55 (32.2%) | 37 (48.1%) | 70 (33.2%) | 43 (45.3%) |
| Variable | Analysis 1 (N = 248) | Analysis 2 (N = 306) | ||
| Trans (N = 171) | Cis LGBPQ (N = 77) | Trans (N = 211) | Cis LGBPQ (N = 95) | |
| Gender | n (n%) | n (n%) | n (n%) | n (n%) |
| Boy | 41 (24%) | 16 (20.8%) | 56 (26.5%) | 21 (22.1%) |
| Girl | 89 (52%) | 54 (70.1%) | 120 (56.9%) | 69 (72.6%) |
| Neither | 11 (6.4%) | 0 (0%) | 9 (4.3%) | 1 (1.1%) |
| Both | 6 (3.5%) | 2 (2.6%) | 3 (1.4%) | 1 (1.1%) |
| Changes | 14 (8.2%) | 0 (0%) | 14 (6.6%) | 0 (0%) |
| I don't know | 10 (5.8%) | 5 (6.5%) | 9 (4.3%) | 2 (2.1%) |
| Unknown | 0 (0%) | 0 (0%) | 0 (0%) | 1 (1.1%) |
| Race and ethnicity | ||||
| Asian, non-Hispanic | 7 (4%) | 1 (1.3%) | 5 (2.4%) | 0 (0%) |
| Black/African, non-Hispanic | 0 (0%) | 0 (0%) | 3 (1.4%) | 1 (1.1%) |
| Native American/Alaska Native, Hispanic | 0 (0%) | 1 (1.3%) | 0 (0%) | 0 (0%) |
| Multiracial, Hispanic | 3 (1.8%) | 1 (1.3%) | 2 (0.9%) | 1 (1.1%) |
| Multiracial, non-Hispanic | 21 (12.3%) | 12 (7%) | 30 (14.2%) | 18 (18.9%) |
| Multiracial, unknown ethnicity | 0 (0%) | 1 (1.3%) | 0 (0%) | 1 (1.1%) |
| White/European, Hispanic | 13 (7.6%) | 4 (5.2%) | 16 (7.6%) | 6 (6.3%) |
| White/European, non-Hispanic | 123 (71.9%) | 57 (74%) | 152 (72%) | 68 (71.6%) |
| Unknown race, Hispanic | 4 (2.3%) | 0 (0%) | 3 (1.4%) | 0 (0%) |
| Sexual orientation | ||||
| Gay | 10 (5.8%) | 5 (6.5%) | 13 (6.2%) | 5 (5.3%) |
| Lesbian | 12 (7%) | 5 (6.5%) | 21 (10%) | 8 (8.4%) |
| Bisexual | 23 (13.5%) | 36 (46.8%) | 29 (17%) | 40 (42.1%) |
| Straight | 47 (27.5%) | 1* (1.3%) | 63 (36.8%) | 2* (2.1%) |
| Asexual | 11 (6.4%) | 2 (2.6%) | 12 (5.7%) | 4 (4.2%) |
| Pansexual | 15 (8.8%) | 8 (10.4%) | 16 (7.6%) | 10 (10.5%) |
| I prefer a different term | 21 (12.3%) | 6 (7.8%) | 21 (10%) | 5 (5.3%) |
| Multiple | 29 (17%) | 14 (18.2%) | 33 (15.6%) | 19 (20%) |
| Unknown | 3 (1.8%) | 0 (0%) | 3 (1.4%) | 2 (2.1%) |
| Parent education | ||||
| High school diploma | 2 (1.2%) | 2 (2.6%) | 5 (2.4%) | 3 (3.2%) |
| Some college/Associate's Degree | 18 (10.5%) | 7 (9.1%) | 28 (13.3%) | 10 (10.5%) |
| College/Bachelor's Degree | 54 (31.6%) | 31 (40.3%) | 67 (31.8%) | 40 (42.1%) |
| Advanced Degree (MA, MD, PhD, etc.) | 95 (55.6%) | 36 (46.8%) | 108 (51.2%) | 42 (44.2%) |
| Other | 1 (0.6%) | 1 (1.3%) | 2 (0.9%) | 0 (0%) |
| Other + some other schooling also listed | 1 (0.6%) | 0 (0%) | 1 (0.5%) | 0 (0%) |
| Household income | ||||
| <$25,000 per year | 9 (5.3%) | 4 (5.2%) | 9 (4.3%) | 2 (2.1%) |
| $25,000-50,000 per year | 18 (10.5%) | 4 (5.2%) | 22 (10.4%) | 4 (4.2%) |
| $50,001-75,000 per year | 30 (17.5%) | 9 (11.7%) | 40 (19%) | 13 (13.7%) |
| $75,001-125,000 per year | 59 (34.5%) | 23 (29.9%) | 70 (33.2%) | 33 (34.7%) |
| >$125,000 per year | 55 (32.2%) | 37 (48.1%) | 70 (33.2%) | 43 (45.3%) |
Note. Participants reported their gender identity in the survey on an item that had these six, and only these six options. Participant age is calculated based on their birthdate. Participant race and ethnicity, parent education, and household income reported in this study reflects the earliest available parent report of these variables. Percentage values represent the proportion of participants in each group (trans or cis LGBPQ) in each Analysis, not the entire sample.
* There was 1 participant in Analysis 1 and 2 participants in Analysis 2 who identified as “straight” in the measure of sexual orientation used to report demographic information, but who were included in the “cis LGBPQ” group in our sample. They were included in this group because they met the preregistered inclusion criteria for this group, which included responding “Yes” to a question that asked ““Are you gay, lesbian, bisexual, pansexual or LGBPQ?”.
Analysis 2
Youth between 12-22 years of age participated in the 2023 annual survey, from the same six recruitment groups mentioned above. As in Analysis 1, participants had to be either transgender (of any sexual orientation, N = 221) or cisgender and LGBPQ (N = 97). Twelve participants were excluded because they did not respond to the questions of interest in this study, for a final sample of 306 participants (Mage = 15.07, SD = 2.29, see Table 1 for demographic information). Of the participants in Analysis 2, 99 trans participants and 49 cis LGBPQ participants did not overlap with participants from Analysis 1. Results of analyses including only these participants are included in the SM, and largely do not differ from those presented below. Thirty-two participants were part of a pair of siblings who took the survey (non-independence accounted for in robustness checks, see SM). The survey was conducted online using Qualtrics software.
Analysis 3
This sample included the subset of participants from Analyses 1 and 2 who had completed the measures of interest in the annual survey in both 2022 and 2023. For this study, participants had to be either transgender (of any sexual orientation, N = 111) or cisgender and LGBPQ (N = 44) in both years, excluding 3 participants who had changed their identity between 2022 and 2023 for a final sample of 155 participants (Mage in 2022 = 14.68, SD = 2.03). Four participants were part of a pair of siblings who took the survey (non-independence accounted for in robustness checks, see SM).
Materials and Procedure
Participants included in Analyses 1 and 2 completed the below measures in 2022 and 2023 respectively. Participants included in Analysis 3 completed the independent variables in 2022 and the dependent variables in 2022 and 2023. These were completed as part of a larger battery of measures addressing other topics (e.g., mental health, gender categorization).
Independent Variables
Participants responded to single-item measures of community connectedness (Analyses 1-3) and identity importance (Analysis 1). These two items were part of a battery of items adapted from the Gender Minority Stress and Resilience measure (Testa et al., 2015) and differed in their focus on trans vs. LGBPQ identity for transgender and cisgender LGBPQ participants respectively (see Table 2 ). Responses were on a scale of 1 (not at all) to 7 (very much) with an additional option of not applicable. “Not applicable” was considered a non-answer for the purposes of this project.
| Construct | Item | Analyses |
| Community connectedness (IV) | Trans Ps: I feel connected to other people who are trans/gnc/nb | 1-3 |
| Cis LGBPQ Ps: I feel connected to other people who are LGBPQ | ||
| Identity importance (IV) | Trans Ps: The fact that I am trans/gnc/nb is an important part of how I see myself | 1 |
| Cis LGBPQ Ps: The fact that I am LGBPQ is an important part of how I see myself | ||
| Political action 1 (DV) | Have you ever engaged in any of the following actions? Check all that apply | 1-3 |
| - met with school officials about trans issues | ||
| - attended a school board meeting about trans issues | ||
| - spoken at a school board meeting about trans issues | ||
| - attended a legislative hearing about trans issues | ||
| - spoken/testified at a legislative hearing about trans issues | ||
| - sued a school, organization or government body about trans issues* | ||
| - attended a protest or rally about trans issues | ||
| - educated family or friends on trans issues | ||
| - spoken to the media or written an editorial on trans issues | ||
| Political action 2 (DV) | Are you engaged in any activism regarding transgender/nonbinary issues? (Y/N) | 2, 3 |
| Construct | Item | Analyses |
| Community connectedness (IV) | Trans Ps: I feel connected to other people who are trans/gnc/nb | 1-3 |
| Cis LGBPQ Ps: I feel connected to other people who are LGBPQ | ||
| Identity importance (IV) | Trans Ps: The fact that I am trans/gnc/nb is an important part of how I see myself | 1 |
| Cis LGBPQ Ps: The fact that I am LGBPQ is an important part of how I see myself | ||
| Political action 1 (DV) | Have you ever engaged in any of the following actions? Check all that apply | 1-3 |
| - met with school officials about trans issues | ||
| - attended a school board meeting about trans issues | ||
| - spoken at a school board meeting about trans issues | ||
| - attended a legislative hearing about trans issues | ||
| - spoken/testified at a legislative hearing about trans issues | ||
| - sued a school, organization or government body about trans issues* | ||
| - attended a protest or rally about trans issues | ||
| - educated family or friends on trans issues | ||
| - spoken to the media or written an editorial on trans issues | ||
| Political action 2 (DV) | Are you engaged in any activism regarding transgender/nonbinary issues? (Y/N) | 2, 3 |
aThis item varied slightly between 2022 and 2023. The 2022 version did not include “government body.” Results did not differ meaningfully when removing this item from analyses in Analysis 2.
Dependent Variables
Analyses 1 and 2. There were two measures of political action, which we coded in three ways total, resulting in three dependent variables. Although the items differed in wording and response type, we treated them as measures of the same construct: political action for trans rights.
In 2022 and 2023, participants responded to an item measuring their engagement in political action related to trans issues (see Table 2 ). This item was coded in two ways: (1) Political Action (how many): a count DV by summing the number of options (0-9) that participants selected, and (2) Political Action (any): a binary DV (0 = did not check any option, 1 = checked at least one option). See SM for more information about how this item was coded. In 2023, participants also responded to an item (Political Action (yes/no); used only in Analyses 2 and 3) which asked “Are you engaged in any activism regarding transgender/nonbinary issues?”. We coded this as a binary DV (0 = No, 1 = Yes).
Analysis 3. For the longitudinal analyses, we used three dependent variables. The first was the Political Action (yes/no) variable described above. The other two dependent variables were a measure of change in responses to the item measuring participants’ political action related to transgender issues in 2022 and 2023 (see Table 2 ). We coded these change variables as follows: (1) Action Change (how many): a count DV summing the additional options selected by participants in 2023, beyond the options they selected in 2022, and (2) Action Change (any): a binary DV (0 = did not check any additional options, 1 = checked at least one additional option in 2023 over 2022). We only measured positive change because participants were asked about their political action in general, not in the specific year in which they were taking the survey (e.g., if someone reported having engaged in 5 actions in 2022, that number could stay the same or increase in 2023, but could not decrease).
Results
Data and analysis scripts are available at https://osf.io/ywzs3/. All statistical analyses were conducted using R [4.3.1] software (R Core Team, 2019). Analyses were performed using the following packages: tidyverse (v2.0.0, Wickham et al., 2019), knitr (v1.44, Xie, 2023), corrplot (v0.92, Wei & Simko, 2021), broom (v1.0.5, D. Robinson et al., 2023), easystats (v0.7.0, Lüdecke et al., 2022), modelsummary (v1.4.3, Arel-Bundock, 2022), ggeffects (v1.3.4, Lüdecke, 2018), and patchwork (v1.1.3, Pedersen, 2023). All analyses reported below were preregistered unless indicated otherwise.
Analytic Approach
In each regression analysis described below, we used a single simultaneous regression model with all theoretically relevant predictors, in order to avoid Type I errors associated with multiple testing, and to be conservative about assigning significance to predictors (community connectedness and identity importance) with shared variance. Additionally, this allowed us to control for uneven group sizes, with the transgender sample being much larger than the cisgender LGBPQ sample, and prevent us from inappropriately generalizing patterns from the transgender group to the entire sample. Throughout the manuscript, we interpret the statistical effect of each predictor from these simultaneous regression models - i.e., controlling for the other effects in the models.
Note that for all variables of interest other than Political Action (yes/no), we had preregistered that participants who did not respond to that variable would be excluded from all analyses. For the sake of consistency, we decided to apply this rule to Political Action (yes/no), as well.
Analyses 1 and 2
We report descriptive statistics for all variables of interest in Analyses 1-2 in Table 3 . We also report the distribution of participants who selected each individual political action in the item “Have you ever engaged in any of the following actions? Check all that apply” in Analyses 1 and 2 in Figure 1 .
| Variable | Analysis 1 (2022) | Analysis 2 (2023) | ||||||
| Trans | Cis LGBPQ | Trans | Cis LGBPQ | |||||
| M | SD | M | SD | M | SD | M | SD | |
| Community connectedness | 4.94 | 1.96 | 4.66 | 1.72 | 4.74 | 1.84 | 4.91 | 1.81 |
| Identity importance | 4.65 | 2.14 | 4.39 | 1.80 | ||||
| Political Action (how many) | 2.24 | 1.90 | 1.19 | 1.06 | 2.21 | 2.08 | 1.08 | 0.92 |
| Political Action (any) | 80% | 69% | 79% | 72% | ||||
| Political Action (yes/no) | 31% | 16% | ||||||
| Variable | Analysis 1 (2022) | Analysis 2 (2023) | ||||||
| Trans | Cis LGBPQ | Trans | Cis LGBPQ | |||||
| M | SD | M | SD | M | SD | M | SD | |
| Community connectedness | 4.94 | 1.96 | 4.66 | 1.72 | 4.74 | 1.84 | 4.91 | 1.81 |
| Identity importance | 4.65 | 2.14 | 4.39 | 1.80 | ||||
| Political Action (how many) | 2.24 | 1.90 | 1.19 | 1.06 | 2.21 | 2.08 | 1.08 | 0.92 |
| Political Action (any) | 80% | 69% | 79% | 72% | ||||
| Political Action (yes/no) | 31% | 16% | ||||||
Note. For Political Action (any) and Political Action (yes/no), we report % of participants who reported engaging in at least one action, and % of participants who responded “yes” to whether they had engaged in activism for trans issues, respectively.
Correlational Analyses
Table 4 displays correlations between community connectedness, identity importance, and political action in Analyses 1 (preregistered) and 2 (not preregistered; identity importance is excluded in these correlations). We conducted these analyses primarily to examine the extent to which our predictor variables were correlated.
| Variable | Analysis 1 (2022) | Analysis 2 (2023) | ||||||
| CC | ID | PA (how many) | PA (any) | CC | PA (how many) | PA (any) | PA (yes/no) | |
| Community connectedness (CC) | - | - | ||||||
| Identity importance (ID) | 0.48** | - | ||||||
| Political Action (how many) | 0.15* | 0.13* | - | 0.20** | - | |||
| Political Action (any) | 0.24** | 0.12 | 0.61** | - | 0.22** | 0.55** | - | |
| Political Action (yes/no) | 0.21** | 0.49** | 0.26** | - | ||||
| Variable | Analysis 1 (2022) | Analysis 2 (2023) | ||||||
| CC | ID | PA (how many) | PA (any) | CC | PA (how many) | PA (any) | PA (yes/no) | |
| Community connectedness (CC) | - | - | ||||||
| Identity importance (ID) | 0.48** | - | ||||||
| Political Action (how many) | 0.15* | 0.13* | - | 0.20** | - | |||
| Political Action (any) | 0.24** | 0.12 | 0.61** | - | 0.22** | 0.55** | - | |
| Political Action (yes/no) | 0.21** | 0.49** | 0.26** | - | ||||
Note. Asterisks indicate significant correlations.
*p < .05, **p < .001
Regression Analyses
We investigated whether community connectedness, identity importance, and participant group (trans vs. cis LGBPQ) were associated with political action related to trans issues (RQs 1 and 2). Because the dependent variable in this analysis is a count variable, we use Poisson regression. The regression model for Analysis 1 included community connectedness (continuous single-item measure), identity importance (continuous single-item measure), participant group (transgender or cisgender LGBPQ, contrast-coded such that transgender participants = 0.5, and cisgender LGBPQ participants = -0.5), the interaction of participant group and community connectedness, and the interaction of participant group and identity importance as predictors1. We used these to predict the number of different types of political actions participants undertook; i.e., Political Action (how many). In Analysis 2, we ran a nearly identical Poisson regression analysis (preregistered), but because identity importance was not a significant predictor of political action in Analysis 1, we no longer included identity importance or its interaction with participant group as predictors in the model.
Results of Analyses 1 and 2 are reported in Table 5 . In both 2022 (Analysis 1) and 2023 (Analysis 2), we found that, as predicted, higher levels of community connectedness were significantly associated with higher levels of political action related to trans issues (see Figure 2 ). Contrary to prediction, identity importance did not significantly predict higher levels of political action (Analysis 1). We also found a significant effect of participant group, such that trans youth were more likely to engage in higher levels of political action than cis LGBPQ youth. Neither of the interactions involving participant group were significant.
| Variable | Analysis 1 (2022) | Analysis 2 (2023) | ||||||||
| b | 95% CI | z | IRR | p | b | 95% CI | z | IRR | p | |
| (Intercept) | -0.02 | [-0.44, 0.37] | -0.12 | 0.98 | .907 | -0.22 | [-0.59, 0.12] | -1.20 | 0.80 | .230 |
| Community connectedness (CC) | 0.09 | [0.01, 0.16] | 2.22 | 1.09 | .026* | 0.13 | [0.07, 0.20] | 3.97 | 1.14 | <.001* |
| Participant group (PG) | 1.01 | [0.23, 1.85] | 2.45 | 2.74 | .014* | 0.87 | [0.18, 1.61] | 2.38 | 2.38 | .018* |
| PG x CC | -0.10 | [-0.25, 0.05] | -1.26 | 0.91 | .208 | -0.03 | [-0.16, 0.10] | -0.39 | 0.97 | .698 |
| Identity importance (ID) | 0.02 | [-0.05, 0.09] | 0.58 | 1.02 | .564 | |||||
| PG x ID | 0.02 | [-0.12, 0.15] | 0.25 | 1.02 | .800 | |||||
| Variable | Analysis 1 (2022) | Analysis 2 (2023) | ||||||||
| b | 95% CI | z | IRR | p | b | 95% CI | z | IRR | p | |
| (Intercept) | -0.02 | [-0.44, 0.37] | -0.12 | 0.98 | .907 | -0.22 | [-0.59, 0.12] | -1.20 | 0.80 | .230 |
| Community connectedness (CC) | 0.09 | [0.01, 0.16] | 2.22 | 1.09 | .026* | 0.13 | [0.07, 0.20] | 3.97 | 1.14 | <.001* |
| Participant group (PG) | 1.01 | [0.23, 1.85] | 2.45 | 2.74 | .014* | 0.87 | [0.18, 1.61] | 2.38 | 2.38 | .018* |
| PG x CC | -0.10 | [-0.25, 0.05] | -1.26 | 0.91 | .208 | -0.03 | [-0.16, 0.10] | -0.39 | 0.97 | .698 |
| Identity importance (ID) | 0.02 | [-0.05, 0.09] | 0.58 | 1.02 | .564 | |||||
| PG x ID | 0.02 | [-0.12, 0.15] | 0.25 | 1.02 | .800 | |||||
Note. Asterisks indicate significant effects.
Note. Lines are model-based estimated marginal means and shading reflects 95% confidence intervals. Higher numbers on the X axis indicate higher levels of community connectedness.
Note. Lines are model-based estimated marginal means and shading reflects 95% confidence intervals. Higher numbers on the X axis indicate higher levels of community connectedness.
We then investigated whether community connectedness, identity importance, and participant group were associated with Political Action (any)—i.e., whether or not youth participated in political action at all. We ran logistic regression analyses to test this in Analyses 1 and 2, using the same predictors as above. Again, we did not include identity importance or its interaction with participant group as predictors in Analysis 2. Results of these analyses are reported in Table 6 . As predicted, in both 2022 (Analysis 1) and 2023 (Analysis 2), there was a significant positive effect of community connectedness on whether or not participants engaged in any political action. Contrary to expectation, identity importance did not significantly predict political action (Analysis 1). Unlike in the Poisson regression results reported for Political Action (how many), there was no significant effect of participant group. As in the results reported for Political Action (how many), the interactions involving participant group were not significant.
| Variable | Analysis 1 (2022) | Analysis 2 (2023) | ||||||||
| b | 95% CI | z | OR | p | b | 95% CI | z | OR | p | |
| (Intercept) | -0.19 | [-1.11, 0.73] | -0.40 | 0.83 | .688 | -0.20 | [-0.95, 0.56] | -0.51 | 0.82 | .608 |
| Community connectedness (CC) | 0.30 | [0.11, 0.51] | 2.99 | 1.35 | .003* | 0.29 | [0.14, 0.45] | 3.64 | 1.33 | < .001* |
| Participant group (PG) | 0.05 | [-1.79, 1.90] | 0.06 | 1.06 | .954 | 0.35 | [-1.15, 1.85] | 0.46 | 1.42 | .648 |
| PG x CC | -0.01 | [-0.42, 0.38] | -0.06 | 0.99 | .949 | 0.03 | [-0.28, 0.34] | 0.20 | 1.03 | .845 |
| Identity importance (ID) | -0.02 | [-0.21, 0.16] | -0.22 | 0.98 | .827 | |||||
| PG x ID | 0.13 | [-0.24, 0.52] | 0.70 | 1.14 | .482 | |||||
| Variable | Analysis 1 (2022) | Analysis 2 (2023) | ||||||||
| b | 95% CI | z | OR | p | b | 95% CI | z | OR | p | |
| (Intercept) | -0.19 | [-1.11, 0.73] | -0.40 | 0.83 | .688 | -0.20 | [-0.95, 0.56] | -0.51 | 0.82 | .608 |
| Community connectedness (CC) | 0.30 | [0.11, 0.51] | 2.99 | 1.35 | .003* | 0.29 | [0.14, 0.45] | 3.64 | 1.33 | < .001* |
| Participant group (PG) | 0.05 | [-1.79, 1.90] | 0.06 | 1.06 | .954 | 0.35 | [-1.15, 1.85] | 0.46 | 1.42 | .648 |
| PG x CC | -0.01 | [-0.42, 0.38] | -0.06 | 0.99 | .949 | 0.03 | [-0.28, 0.34] | 0.20 | 1.03 | .845 |
| Identity importance (ID) | -0.02 | [-0.21, 0.16] | -0.22 | 0.98 | .827 | |||||
| PG x ID | 0.13 | [-0.24, 0.52] | 0.70 | 1.14 | .482 | |||||
Note. Asterisks indicate significant effects.
Finally, in Analysis 2 (2023 data), we ran a logistic regression to test whether community connectedness and participant group were associated with Political Action (yes/no), the binary measure of political action newly introduced in 2023 (see Table 7 ). As predicted, we found a significant effect of community connectedness on whether or not participants engaged in political action. We saw no significant effect of participant group or interaction between participant group and community connectedness.
| Variable | b | 95% CI | z | OR | p |
| (Intercept) | -2.47 | [-3.59, -1.51] | -4.70 | 0.08 | < .001* |
| Community connectedness (CC) | 0.25 | [0.06, 0.44] | 2.56 | 1.28 | .011* |
| Participant group (PG) | -0.16 | [-2.10, 2.08] | -0.15 | 0.85 | .881 |
| PG x CC | 0.21 | [-0.19, 0.57] | 1.08 | 1.23 | .278 |
| Variable | b | 95% CI | z | OR | p |
| (Intercept) | -2.47 | [-3.59, -1.51] | -4.70 | 0.08 | < .001* |
| Community connectedness (CC) | 0.25 | [0.06, 0.44] | 2.56 | 1.28 | .011* |
| Participant group (PG) | -0.16 | [-2.10, 2.08] | -0.15 | 0.85 | .881 |
| PG x CC | 0.21 | [-0.19, 0.57] | 1.08 | 1.23 | .278 |
Note. Asterisks indicate significant effects.
Robustness Checks
Because of potential concerns about our analyses, we ran a series of preregistered robustness checks; see SM for details and results, which did not meaningfully differ from those reported here.
Interim Discussion
The more connected trans youth felt to other trans people, and the more connected cisgender LGBPQ youth felt to other LGBPQ people, the more likely they were to engage in political action related to trans issues. Trans youth engaged in more political actions than cis LGBPQ youth (controlling for community connectedness and identity importance). However, trans and cis LGBPQ youth did not differ in their likelihood of engaging in any political action. We hesitate to interpret our findings on the effect of participant group on political action without more conclusive evidence. In all analyses, we found no interaction between participant group and community connectedness, so we have no evidence that the association between community connectedness and trans-related political action is influenced by whether someone is transgender or cisgender LGBPQ. In Analysis 1, we also found no evidence that identity importance was uniquely associated with political action when controlling for community connectedness, nor did we find an interaction between identity importance and participant group.
Analysis 3
For our longitudinal analysis of data from 2022 and 2023, we report descriptive statistics for community connectedness and political action in Table 8 .
| Variable Type | Variable | Trans | Cis LGBPQ | ||
| Cross-sectional (values from responses in a single year) | Community connectedness (2022): Average feelings of community connectedness in 2022 | M = 5.07 | SD = 1.93 | M = 5.11 | SD = 1.57 |
| Political Action (yes/no) (2023): Percentage of participants who reported that they had engaged in political activism in 2023 | 38% | 16% | |||
| Longitudinal (variables measuring change from 2022 to 2023) | Action Change (how many): Average increase in number of actions reported from 2022 to 2023 | M = 0.74 | SD = 0.98 | M = 0.36 | SD = 0.84 |
| Action Change (any): Percentage of participants who reported at least one more action in 2023 than 2022 | 49% | 23% | |||
| Variable Type | Variable | Trans | Cis LGBPQ | ||
| Cross-sectional (values from responses in a single year) | Community connectedness (2022): Average feelings of community connectedness in 2022 | M = 5.07 | SD = 1.93 | M = 5.11 | SD = 1.57 |
| Political Action (yes/no) (2023): Percentage of participants who reported that they had engaged in political activism in 2023 | 38% | 16% | |||
| Longitudinal (variables measuring change from 2022 to 2023) | Action Change (how many): Average increase in number of actions reported from 2022 to 2023 | M = 0.74 | SD = 0.98 | M = 0.36 | SD = 0.84 |
| Action Change (any): Percentage of participants who reported at least one more action in 2023 than 2022 | 49% | 23% | |||
Note. Community connectedness was measured on a 7-pt. Likert scale (1 = not at all, 7 = very much). This measure was in relation to one’s own community (i.e., trans or cis LGBPQ for trans and cis LGBPQ participants respectively). Political Action (yes/no) was coded as 0 (no) or 1 (yes). Participants were asked about political action specific to trans issues. Action Change (how many) was measured as the sum of the number of political actions the participant reported in 2023 over the number in 2022, with the highest possible total being 9. Action Change (any) was coded as 0 (no change) or 1 (change in at least one action).
Regression Analyses
We first investigated whether community connectedness in 2022 predicted political action in 2023 for transgender and cisgender LGBPQ youth. We ran a logistic regression using community connectedness (continuous single-item measure), participant group (transgender or cisgender LGBPQ, contrast-coded such that transgender participants = 0.5, and cisgender LGBPQ participants = -0.5), and the interaction of participant group and community connectedness as predictors. We used these to predict the binary measure of Political Action (yes/no) newly introduced in 2023 (see Table 9 ). As predicted, we found that higher levels of community connectedness in 2022 significantly predicted higher likelihood that participants engaged in political action in 2023, but no significant effect of participant group or interaction between participant group and community connectedness.
| Variable | b | 95% CI | z | OR | p |
| (Intercept) | -3.34 | [-5.98, -1.51] | -3.01 | 0.04 | .003* |
| Community connectedness (CC) | 0.42 | [0.09, 0.85] | 2.22 | 1.52 | .026* |
| Participant group (PG) | 1.90 | [-1.81, 7.16] | 0.85 | 6.66 | .394 |
| PG x CC | -0.12 | [-0.98, 0.54] | -0.32 | 0.89 | .753 |
| Variable | b | 95% CI | z | OR | p |
| (Intercept) | -3.34 | [-5.98, -1.51] | -3.01 | 0.04 | .003* |
| Community connectedness (CC) | 0.42 | [0.09, 0.85] | 2.22 | 1.52 | .026* |
| Participant group (PG) | 1.90 | [-1.81, 7.16] | 0.85 | 6.66 | .394 |
| PG x CC | -0.12 | [-0.98, 0.54] | -0.32 | 0.89 | .753 |
Note. Asterisks indicate significant effects.
Next, we investigated whether community connectedness in 2022 predicted change in political action from 2022 to 2023 for transgender and cisgender LGBPQ youth. We ran a Poisson regression model using the same predictors as above to predict change in the number of political actions: Action Change (how many) (see Table 10 ). We found no significant effect of community connectedness, participant group, or interaction effect between participant group and community connectedness on change in the level of political action. We then ran a logistic regression model with the same predictors as above, using the measure of Action Change (any) as the dependent variable (see Table 11 ). Again, we found no significant effect of community connectedness, participant group, or interaction effect between participant group and community connectedness on whether or not there was change in political action. One possibility for why we found no significant results was the lack of change in political action over time (a concern we had preregistered), as demonstrated in the results reported in the Presence of Change Analyses below.
| Variable | b | 95% CI | z | IRR | p |
| (Intercept) | -0.38 | [-1.28, 0.35] | -0.92 | 0.68 | .356 |
| Community connectedness (CC) | -0.06 | [-0.21, 0.11] | -0.70 | 0.95 | .485 |
| Participant group (PG) | 0.12 | [-1.36, 1.90] | 0.14 | 1.13 | .886 |
| PG x CC | 0.12 | [-0.21, 0.42] | 0.74 | 1.13 | .457 |
| Variable | b | 95% CI | z | IRR | p |
| (Intercept) | -0.38 | [-1.28, 0.35] | -0.92 | 0.68 | .356 |
| Community connectedness (CC) | -0.06 | [-0.21, 0.11] | -0.70 | 0.95 | .485 |
| Participant group (PG) | 0.12 | [-1.36, 1.90] | 0.14 | 1.13 | .886 |
| PG x CC | 0.12 | [-0.21, 0.42] | 0.74 | 1.13 | .457 |
Note. Asterisks indicate significant effects.
| Variable | b | 95% CI | z | OR | p |
| (Intercept) | -0.55 | [-2.00, 0.70] | -0.82 | 0.58 | .414 |
| Community connectedness (CC) | -0.02 | [-0.26, 0.24] | -0.14 | 0.98 | .887 |
| Participant group (PG) | 1.28 | [-1.23, 4.18] | 0.95 | 3.58 | .342 |
| PG x CC | -0.02 | [-0.54, 0.46] | -0.08 | 0.98 | .933 |
| Variable | b | 95% CI | z | OR | p |
| (Intercept) | -0.55 | [-2.00, 0.70] | -0.82 | 0.58 | .414 |
| Community connectedness (CC) | -0.02 | [-0.26, 0.24] | -0.14 | 0.98 | .887 |
| Participant group (PG) | 1.28 | [-1.23, 4.18] | 0.95 | 3.58 | .342 |
| PG x CC | -0.02 | [-0.54, 0.46] | -0.08 | 0.98 | .933 |
Note. Asterisks indicate significant effects.
Presence of Change Analyses
For the two dependent variables that appeared in both the 2022 and 2023 data (Political Action (how many) and Political Action (any)), we preregistered an examination of the stability of the variable over time. We computed a Spearman’s rank correlation to assess the relationship between the Political Action (how many) variables in 2022 and 2023. There was a significant positive correlation between the two variables (r(153) = 0.64, p < .001). We conducted a McNemar’s chi-square test with continuity correction to test the difference between the Political Action (any) variables in 2022 and 2023 and found that the two were not statistically different (X2(1, N = 155) = 0.00, p > .999). These findings suggest that there was little change in either the Political Action (how many) or (any) measures between 2022 and 2023. Given this, the regression analyses measuring predictors of change in political action reported above should be interpreted with caution.
Robustness Checks
Because of concerns about dependency in the sibling data, we also replicated the regression analyses after randomly retaining only one participant from each set of siblings in Analysis 3, and report these results, which do not differ from the primary analyses, in SM.
Interim Discussion
Higher levels of community connectedness (but not participant group, or their interaction) in 2022 predicted youth engagement in political action in 2023. However, in analyses investigating whether community connectedness in 2022 and participant group predicted change in participation in political action from 2022 to 2023, we found no significant effect of either predictor, or their interaction.
General Discussion
We found cross-sectional and longitudinal evidence that higher levels of community connectedness were associated with higher likelihood of engagement in political action related to trans issues for both transgender youth and cisgender LGBPQ youth. This adds to previous research in LGBTQ+ adult populations showing that community connectedness is strongly associated with political action (Montagno et al., 2021; Montagno & Garrett-Walker, 2022). Feeling connected to others who share one’s identity, whether trans or LGBPQ, may provide motivation and opportunity to get involved in political action for trans rights. We found no evidence for the effect of community connectedness on change in political action for trans rights across time, but as mentioned above, more research on this question is necessary because our sample did not demonstrate substantial change in political action in the year-long span in which we collected this data.
Whether the nature of youth’s experience of community connectedness influences political action for trans rights remains an open question. In our study, participants were asked about the extent to which they felt connected to other members of their community. Would it matter whether they felt connected to other trans or LGBPQ individuals in their online communities, or in their schools and local environments? For trans adults, in-person community connectedness has a larger effect on civic engagement than online communication (Billard, 2022). However, given that transgender adolescents consider their online communities to be meaningful sources of support (Shah et al., 2024), online spaces might be especially important for them.
We found mixed evidence for the effect of identity on political action. While trans and cis LGBPQ youth did not differ in whether or not they engaged in any political action for trans rights, trans youth engaged in more political actions than cis LGBPQ youth, controlling for community connectedness (and identity importance in Analysis 1). We also found no evidence that the strength of the association between community connectedness and political action is moderated by participant identity. Overall, it is evident from our data that cisgender LGBPQ youth do participate in political action for trans rights, and that their feelings of connectedness to other LGBPQ people are positively associated with this action. This may suggest that it is important to create opportunities for cisgender youth to form connections with trans youth to encourage them to act to protect trans rights. Future research might also explore whether cisgender LGBPQ youth are seeing their engagement in political action for trans rights as an act of solidarity with members of a disadvantaged group (Agostini & van Zomeren, 2021), or if they consider trans youth to be members of their ingroup (i.e., LGBTQ+), in line with past work showing that people act in solidarity with members of marginalized groups when they identify with a higher-order group that shares some experiences of injustice (Subašić et al., 2011).
Finally, we had also predicted that the more important one’s trans or LGBPQ identity is to oneself, the more likely one would be to engage in political action for trans rights. In the current research, we found no evidence for the effect of identity importance on political action. However, we did find that the two independent variables of community connectedness and identity importance were highly positively correlated. This relationship has been demonstrated in prior work with LGBTQ people as well (Hinton et al., 2022).
Limitations
One limitation of this study is our relatively coarse and broad operationalizations of political action. In one measure, we asked if they had participated or not, without defining political action. In another, the options ranged considerably, from educating friends and family on trans issues, to attending protests and rallies, to testifying at legislative hearings, and these were treated equally in analyses. However, identity-related factors like community connectedness may have different associations with different types of actions, a topic for future research. Our broad operationalization of political action differs from some prior literature, which has taken a narrower definition of activism or collective action: e.g., participation in political protest (van Zomeren et al., 2008) or membership in activist groups (Montagno & Garrett-Walker, 2022). Additionally, our measures did not incorporate the frequency of engagement in these actions, which may also be associated with community connectedness.
A further limitation is that our samples are primarily White, and participants’ parents are largely upper class, highly educated, and supportive of their identities. To ensure generalizability, future research should test these questions with samples of trans and LGBPQ youth who have less supportive home environments and/or are marginalized along other dimensions of social identity as well.
A further limitation is that while we provided three tests of our research questions, they were not tests with fully independent samples. The participants in Analysis 3 were necessarily a subset of participants from 1 and 2, and there was also overlap in participants between the latter two analyses. Our results largely held when we ran the main regression analyses in Analysis 2 with only participants who had not also been included in Analysis 1 (see SM); however, it would be ideal to replicate our results with distinct samples.
Finally, the current research is observational and therefore not suited to determine causal relationships. Future work might investigate the potential bidirectionality of a relationship between community connectedness and political action among trans and LGBPQ youth. While community connectedness may motivate political action by situating one’s experiences within a collective struggle, participation in political action could help people find communities of like-minded others who share their identity (Conner et al., 2023; Szymanski et al., 2023). Community connectedness can also serve as a buffer against the negative impact of participation in political action (such as burnout and poor mental health) for youth in general (Conner et al., 2023). Future research could explore this potential “virtuous cycle” between community connectedness and political action, and their downstream consequences for wellbeing in trans youth.
Conclusion
In transgender youth and cisgender LGBPQ youth, community connectedness with other trans people and LGBPQ people respectively was associated with participation in political action for trans rights. This research contributes to the literature on the role of social identity-related factors in motivating people to participate in political action, focusing on a sample of trans and LGBPQ youth in the case of trans-related activism.
Author Contributions
Contributed to conception and design: MTYPT, NMG, NMW, KRO
Contributed to acquisition of data: NMG, KRO
Contributed to analysis and interpretation of data: MTYPT, NMG, NMW
Drafted and/or revised the article: MTYPT, NMG, NMW, KRO
Approved the submitted version for publication: MTYPT, NMG, NMW, KRO
Funding Information
Data reported in this publication was collected as part of a wave of data collection supported by the Eunice Kennedy Shriver National Institute Of Child Health & Human Development of the National Institutes of Health under Award Number R01HD092347. The content of this paper is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Ethics Statement
This work was approved by the Princeton IRB under protocol 12624.
Competing Interests
The authors declare no conflicts of interest.
Supplemental Material
Supplemental materials containing additional information about the measures and secondary data analyses are available for download from this article.
Data Accessibility Statement
Participant data and analysis scripts are available on this paper’s project page at https://osf.io/ywzs3/. The preregistration for Analysis 1 can be found at https://osf.io/sn4mj and for Analyses 2-3 at https://osf.io/svt9e respectively.
Footnotes
In Analysis 1, we had preregistered a linear regression to test this (see SM for results); however, this model violated the assumption of normality of residuals (see SM Figure 1). We instead report results of an exploratory Poisson regression analysis, with the same predictors as our preregistered analysis.
University of California Press