Table 1.

Characteristics of Studies Included in the Systematic Review.

First author, year	Country	N	Gender of sample	Age (years)	Population	Religious affiliation	Transphobia measure	Religiosity measure	Major findings	Quality Score
Acker, 2017	USA	600	22% male, 78% female	M = 22.00, SD = 5.85	University students majoring in social work, nursing, occupational therapy, or psychology.	Not reported.	9-item Transphobia Scale (Nagoshi et al., 2008).	Two-item religiosity scale (adapted from Gottfried & Polikoff, 2012): attendance at religious services in the past year, and importance religion. 85% of the sample reported moderate to high levels of religiosity.	Religiosity: Transphobia was positively correlated with religiosity (r = .23, p < .01). In MRA, religiosity was a significant predictor of transphobia (R2 = .03, p < .006).	18
Adams et al., 2016	USA	339	57.2% male, 42.8% female	Males: M = 19.34, SD = 1.22 Females: M = 18.81, SD = .94	University students enrolled in introductory psychology courses.	35% Catholic; 32%, Protestant or “other Christian”; 4%, Jewish; 3%, Mormon; 11%, “other”; 17%, atheist or agnostic.	9-item Transphobia Scale (Nagoshi et al., 2008).	Twenty-item Religious Fundamentalism (RF) Scale (Altemeyer & Hunsberger, 1992).	Religiosity (fundamentalism): Transphobia was positively correlated with RF for both women (r = .20, p < .05) and men (r = .15, p < .05). In a pathway analysis, RF and RWA (indicative of a general adherence to social conventions) were significantly correlated with all indices of discomfort with violations of gender and sexual heteronormativity and with sexual prejudice and transphobia for both men and women.	18
Ali, 2016	Canada	74	56.8 % male, 41.9 % female, 1.3% unspecified	Not reported.	Psychiatrists (faculty members) and psychiatry residents at university).	Not reported.	32-item Genderism and Transphobia Scale (Hill & Willoughby, 2005).	Assessed by asking “How much guidance does religion provide in your day-to-day living?” (“None at all,” “Some,” “Quite a bit,” “A great deal”).	Religiosity (religious guidance): Transphobia scores descriptively increased as reported levels of religious guidance increased (none: M = 55.3, SD = 14.6; some: M = 59.4, SD = 14.9; quite a bit: M = 62.7, SD = 24.3; a great deal: M = 68.0, SD = 17.7). However, the sample sizes were underpowered from a statistical perspective and, as such, the reliability of these between-group findings remains to be confirmed.	15
Cragun & Sumerau, 2015	USA	1,612	30.3% male, 69.7% female	M = 18.67, SD = 1.21	University students	27.4% “none” or non-religious, 33.4% Catholic, 30.1% Protestant, 3.4% Jewish and 5.7% other	Feeling thermometer (range: 0–100): positive values indicating relatively warmer feelings toward transgender people.	Participants self-identified their religious affiliation, then rated their religiosity on a 10-point scale. Participants were also asked their view of the Bible (response options included: “The Bible is the actual word of God and is to be taken literally, word for word”; “The Bible is the inspired word of God but not everything in it should be taken literally, word for word”; “The Bible is an ancient book of fables, legends, history, and moral precepts recorded by men”; “The Bible is not part of my religious tradition” and other.	Religiosity: Religiosity was correlated with negative attitudes toward transgender people at the bivariate level (r = −.156, p < .001), and predicted these attitudes in MRA after controlling for sex, race and sexual orientation (ß = −2.085, p < .001). Bible interpretation: Biblical literalists reported more negative attitudes toward transgender people than students who viewed the Bible as inspired or as myth (ps < .001), and students who viewed the Bible as inspired had more negative attitudes toward transgender people than did those who viewed the Bible as inspired (p < .05). Religious affiliation: Christians and those of other religions reported more negative attitudes toward transgender individuals than Jewish participants (ps < .05) and non-religious participants (ps < .01). In MRAs, identifying as Jewish was predictive of positive attitudes towards transgender people (ß = 10.95, p < .03), while identifying as Catholic (ß = 7.71, p < .01), Protestant (ß = 7.63, p < .05), or as “another” religion (ß = 15.87, p < .01) was predictive of negative attitudes.	18
de Jong, 2015	USA	113	Not reported	M = 21.46, SD = 3.49	Faculty members teaching primarily in undergraduate social work programs.	73% no affiliation, 7% Evangelical Christian, 10% Mainline Protestant, 9% Roman Catholic, 1% “other”.	20-item Attitudes Toward Transgendered Individuals Scale (Walch, Ngamake, Francisco, Stitt, & Shingler, 2012).	Participants were asked whether their institution had a religious affiliation or was secular.	Religious affiliation (institution level): Faculty members from secular institutions (M = 94.31, SD = 6.19) had more positive attitudes towards transgender people than faculty members from religiously affiliated institutions (M = 89.75, SD = 10.90; t(33) = 2.10, p = .044).	19
Elischberger et al., 2018	USA and India	USA: 218; India: 217	USA: 45.8% male, 50.5% female, 3.7% unspecified; India: 59.4% male, 37.8% female, 2.8% unspecified	USA: M = 34.04, SD = 10.11; India: M = 32.63, SD = 9.79	Community	Not reported	7-item scale measuring disapproval of gender atypicality (self-developed) and the 32-item Genderism and Transphobia Scale (Hill & Willoughby, 2005).	2-item religiously motivated disapproval of gender nonconformity scale (self-developed).	Religion-based disapproval: Religious disapproval of gender nonconformity was predictive of disapproving attitudes towards transgender people for both U.S. participants (ß = .44, p < .001) and Indian participants (ß = .36, p < .001) when controlling for demographic variables.	17
Elischberger et al., 2016	USA	281	45.6% male, 54.1% female, 0.3% unspecified	M = 32.96, SD = 11.70	Community	38% not religious, 18% agnostic, 20% Protestant, 11% Catholic, 2% Jewish, 1% Muslim, < 1 % Hindu, < 1 % Buddhist, 7 % other religion.	8-item scale measuring disapproval of transgender youth (self-developed).	Participants indicated which religious group they belonged (Protestant, Catholic, Jewish, Muslim, Hindu, Buddhist, or “other”) or whether they considered themselves not religious or agnostic.	Religious affiliation: Religiously affiliated participants were less favorable (M = 4.22, SD = 2.88) towards transgender youth than non-religious participants (M = 2.18, SD = 1.74; p < .001). There were considerable differences between religions (e.g., Jewish > Muslim), but small sample sizes precluded exploring this. MRAs found that religious affiliation was a significant predictor of negative attitudes towards transgender youth for males (ß = 0.20, p < .01) and females (ß = 0.13, p < .05), after controlling for other demographic variables.	18
Fisher et al., 2017	Italy	310	29.7% cisgender men, 31.0% cisgender women, 20.3% transwomen, 19.0% transmen	M = 33.60, SD = 10.35.	Gender dysphoric individuals, health care providers, and community.	59.4 % agnostics/atheists, 37.1 % Christians (99% Catholics and 1% orthodox), 2.6 % Buddhists, 0.6 % Hindus, and 0.3 % Jews.	20-item Attitude Toward Transgendered Individuals Scale (Walch et al., 2012).	Italian version of the 12-item RF Scale (Altemeyer & Hunsberger, 1992). Religious belief was measured with a single item (I do [not] have a religious belief). Religious services attendance was measured on a 4-point Likert-type scale (I do not attend religious services; I attend only for the main festivities; I attend about once a month, I attend about once a week). Religious education was measured on a 4-point Likert-type scale (not at all; a little; enough; a lot).	Religiosity (fundamentalism): Positive attitudes toward transgender people were negatively correlated with RF (r = −0.41, p < .001). For the gender dysphoric population, internalized negative attitudes towards transgender people was negatively correlated with RF (r = −0.43, p < .001). Religiosity (service attendance): Positive attitudes towards transgender people were negatively associated with a higher attendance at religious services (r = −.32, p < .001). Religious education: No significant associations were found between attitudes and religious education.	19
Garelick et al., 2017	USA	287	38.7% male, 61.3% female	M = 19.5, SD not reported	University students	Not reported	9-item Transphobia Scale (Nagoshi et al., 2008).	12-item RF Scale (Altemeyer & Hunsberger, 1992).	Religiosity (fundamentalism): Transphobia was significantly and positively correlated with RF for women (r = .35, p < .001), but not for men (r = .14, p > .05). RF predicted transphobia in MRAs (ß =.17, p < .05).	15
Grigoropoulos & Kordoutis, 2015	Greece	238	39.9% male, 60.1% female	M = 22.0, S = 4.5	University students	Most participants (n = 229) were either Christian or Orthodox Christian. The remainder (n = 9) did not respond to this item.	32-item Genderism and Transphobia Scale (Hill & Willoughby, 2005).	2-item religiosity scale (self-developed): items related to self-identification as religious and frequency of attendance at religious services. Participants reported their religion (if any) with an open-ended measure.	Religiosity: Transphobia was positively correlated with self-identification as religious (r = .25, p < .01.) and frequency of attendance at religious services (r = .30, p < .01.) In MRAs, neither self-identification as strongly religious, or frequency of attendance at religious services, predicted attitudes towards transgender individuals.	17
Haupert, 2018 (unpublished manuscript)	USA	442	36% male, 63% female, 0.4% genderfluid, 0.2% non-binary, 0.2% do not know 0.2% choose not to answer	M = 19.6, SD = 1.94	University students	Not reported	9-item Transphobia Scale (Nagoshi et al., 2008). 32-item Genderism and Transphobia Scale (Hill & Willoughby, 2005).	Self-report if religious (and if so, self-reported religious affiliation).	Religiosity: Transphobia (measured by both scales) was positively correlated with self-identification as religious (rs = .19 & .26, respectively). Transphobia (measured by the Transphobia Scale) was higher for religious participants (M = 2.82) than non-religious participants (M = 2.40; t(400.93) = −3.78, p < .001). Transphobia (measured by the Genderism and Transphobia Scale) was higher in religious participants (M = 3.701) than non-religious participants (M = 3.161; t(429.93) = −5.4528, p < .001).	N/A
Kanamori et al., 2017	USA	483	44.3% male, 55.7% female	Not reported	Community	52.4% none, 47.6% Evangelical Christian	The 29-item Transgender Attitudes and Beliefs Scale (Kanamori et al., 2017), comprises 3 factors: interpersonal comfort; sex/gender beliefs (dichotomy vs continuum); and human value (transgender persons’ intrinsic human value).	Self-report of religious affiliation (and then only participants who identified as Evangelical Christian or non-religious were included).	Religious affiliation: A gender by religion ANOVA revealed a significant main effect for religion for all three attitude - Evangelical Christians scored lower than did non-religious people on all three factors: interpersonal comfort, F(1, 479) = 99.71, p <.001, d =.91; sex/gender beliefs, F(1, 479) = 164.38, p <.001, d = 1.17; human value, F(1, 479) = 21.21, p <.001, d =.42. No gender differences existed (ps > .115).	20
Lewis et al., 2017	USA	1,020	Not reported.	Not reported.	National representative sample.	Not reported.	Feeling thermometer (range: 0–100): positive values indicating relatively warmer feelings toward transgender people.	Participants self-reported belonging to a denomination of Christianity (if any) or as being non-religious. Religiosity was measured by frequency of religious attendance on a 6-point scale from “never” to “more than once a week”.	Religiosity and affiliation: In MRAs, after controlling for demographic variables, religious attendance (ß = −1.878, p < .01) and any Christian affiliation (ß = −7.916, p < .01) were significant predictors of negative feelings toward transgender people. Non-affiliation did not predict attitudes.	16
Mao et al., 2018	USA	319	29.78% male, 70.22%	Not reported.	Self-identified cisgender heterosexual students	1% Buddist, 29% Catholic, 1.9% Muslim, 0.6% Hindu, 6.4% Jewish, 22.9% Protestant, 10.2% another religion, 28% no religion.	9-item Transphobia Scale (Nagoshi et al., 2008). 32-item Genderism and Transphobia Scale (Hill & Willoughby, 2005).	Religious affiliation was self-reported. 20-item RF scale (Altemeyer & Hunsberger, 1992).	Religiosity (fundamentalism): Transphobia (measured by both the Transphobia Scale and the Genderism and Transphobia Scale) was positively correlated with RF (rs = .407 & .475 respectively). Religious affiliation: Different religious groups reported varying levels of transphobia (F(4, 247) = 421.6, p < .001). Protestants > other religions > Catholics > no religion had higher levels of transphobia (all ps < .001).	16
Mbote et al., 2018	Kenya	212	85.8% male, 11.8% female; 2.4% did not report their gender.	Not reported.	Religious leaders from registered churches and mosques.	59.0% Protestant, 22.1% Muslim, and 18.9% Catholic.	Three-items (self-developed) assessing beliefs and attitudes towards gender nonconformity.	Self-report.	Religious Affiliation: No difference between the three religious groups on overall attitudes of transgender people (p > .05). Muslims > Catholics > Protestants felt that it was morally wrong for a man to present himself as a woman χ2(2) = 18.47, p < .05, and for a woman to present herself as a man χ 2(2) = 22.52, p < .01). These opinions were notably strong.	19
Nagoshi et al., 2018	USA	294	36.4% male, 63.6% female	Female: M = 19.9, SD = 3.9 Male: M = 21.0, SD = 3.9	University students	Not reported	9-item Transphobia scale (Nagoshi et al.2008).	20-item RF scale (Altemeyer & Hunsberger, 1992).	Religiosity (fundamentalism): RF was positively correlated with transphobia towards male-to-female targets for both men (r = .30, p < .05) and women (r = .37, p < .001), and female-to-male targets for both men (r = .33, p < .05) and women (r = .51, p < .001).	14
Nagoshi et al., 2008	USA	310	50.6% male, 49.4% female	Female: M = 19.45, SD = 3.28 Male: M = 19.47, SD = 1.76.	University students	35% Catholic, 32% Protestant or “other Christian”, 5% Jewish, 3% Mormon, 12% “other”, and 14% atheist or agnostic.	9-item Transphobia Scale (Nagoshi et al., 2008); items adapted from the Flexibility of Gender Aptitude Scale (Bornstein, 1998).	20-item RF Scale (Altemeyer & Hunsberger, 1992).	Religiosity (fundamentalism): Transphobia was positively correlated with RF for women (r = .54, p < .001), and men (r = .28, p < .001) at the bivariate level. After controlling for homophobia, transphobia was still positively correlated with RF for women (r = .34, p < .001), but not for men (r =-.01, p > .05).	14
Norton & Herek, 2013	USA	2,281	44% male, 56% female	M = 45.87, SD not reported.	Community (heterosexual adults).	Not reported.	Feeling thermometer (range: 0–100): positive values indicating relatively warmer feelings toward transgender people.	Religiosity item, asking how much guidance religion provides in participants day-to day living (“None at all,” “Some,” “Quite a bit,” “A great deal”).	Religiosity: Positive attitudes towards transgender people were negatively correlated with religiosity only for women (r = −.30, p <.001) and not for men (r = −.02, p >.5). In MRAs, religiosity predicted women’s transgender positive attitudes (9% unique variance) but not those of men (< 1% unique variance).	20
Riggs & Bartholomaeus, 2016	Australia	96	28% male, 72% female	M = 48.31, SD = 11.22	Mental health nurses	Not reported.	20-item Attitudes Towards Transgender Individuals Scale (Walch et al., 2012), and the 20-item Counselor Attitude Toward Transgender Scale (Rehbein, 2012).	Participants were asked their degree of religiosity (not at all, somewhat, quite, very).	Religiosity: Positive attitudes towards transgender people were negatively correlated with religiosity (r = −.330, p < .05).	17
Scandurra et al., 2017	Italy	438	30.8% male, 69.2% female	M = 32.21, SD = 5.51	University students (graduates studying teaching)	72.6% current religious faith (Catholic), 27.4% no current religious faith.	25-item Transphobia/Genderism Scale of the Genderism and Transphobia Scale (Hill & Willoughby, 2005) translated into Italian.	Participants were asked if they practiced a religious faith at the moment of the study and if they had a religious education (yes/no).	Religiosity (current practice): practicing religious participants (M = 2.56, SD = .88) reported higher transphobia scores than non-practicing participants (M = 2.02, SD = 0.79; t = 5.69, p < .001, d = .64). Religious education: Participants who received a religious education (M = 2.44, SD = .90) reported higher transphobia scores than those who did not (M = 2.08, SD = 0.77; t = 2.22, p < .05, d = .43). After controlling for demographic variables in MRAs, practicing religion was a predictor of transphobic attitudes (ß = −.12, p <.05), but religious education was not (ß = −.01, p >.05).	19
Skarsgard et al., 2014 (unpublished Conference Poster)	USA	318	38.68% male, 61.32% female	M = 38.0; SD = 15.10	61.64% university students, 38.36% community sample	Not reported	30-item Prejudice Towards Transsexual Women Scale (Winter et al. 2009).	Religiosity was rated on a likert scale.	Religiosity: Transphobia was correlated with self-reported religiosity for the student sample (r = −.33, p < .001), but not the community sample (r = .14, p > .05).	14
Solomon & Kurtz-Costes,2018 (Study 1)	USA	274	41.2% male, 58.8% female	M = 36.0; SD = 11.60	Community sample	48.5% Christians, 35.8% atheists, 15.7% Muslim, Jewish, Buddhist, and other non-Abrahamic religions.	20-item Attitudes Toward Transgendered Individuals Scale (Walch et al., 2012).	Religious affiliation was self-reported. Most participants identified as Atheist (n = 98) or Christian (n = 133), resulting in comparisons between these categories Religiosity was assessed by asking participants how important religion was to them on a 1 to 10 scale.	Religious affiliation: Christians (M = 3.50, SD = 0.99) reported more negative attitudes towards transgender people than Atheists (M = 4.31, SD = 0.70; F(1, 229) = 47.51, p < .001, d = .94). Religiosity: For Christians, positive attitudes towards transgender individuals were negatively correlated with religiosity (r = −.24, p = .006).	20
Solomon & Kurtz-Costes, 2018 (Study 2)	USA	450	46.7% male, 53.3% female	M = 39.50; SD = 12.70	Community sample	52.4% Christians, 23.8% atheists, 23.8% other religions.	20-item Attitudes Toward Transgendered Individuals Scale (Walch et al., 2012).	Religious affiliation was self-reported.	Religious affiliation: Christian participants (M = 3.41, SD = 1.01) reported more negative attitudes than Atheist participants (M = 4.13, SD = 0.71; F(1, 338) = 44.34, p < .001, d = .82).	20
Swank et al., 2013	USA	2,078	37.1% male, 62.9% female	M = 23.74, SD = 7.69	University students	34.8% no religious affiliation, 55.3% follow Christian traditions, 9.9% follow non-Christian traditions.	One item from the Genderism and Transphobia Scale (Hill & Willoughby, 2005): “If I found out that a friend was changing their sex, I could no longer be their friend.”	Participants were asked “How often do you attend religious services?” (0= never, 1 = very rarely, 2 = once a month, 3 = every other week, 4 = once a week, 5 = more than once a week).	Religiosity (attendance): Transphobia was correlated with attendance at religious services for straight (r(1168) = .187, p < 0.01), but not LGB students (r(366) = .063, p >.05).	16
Watjen & Mitchell, 2013	USA	114	All male	M = 21.2, SD = 0.41	University students	79% Christian; the rest identified as atheist, agnostic, or non-religious.	15 items derived (or modified) from the Genderism and Transphobia Scale (Hill & Willoughby, 2005).	Single item requesting religious affiliation.	Religious affiliation: Christians reported higher transphobia scores (M = 49.00, SD = 1.90) than non-religious participants (M = 33.70, SD = 3.70; z (Mann–Whitney U) = 3.39, p < .001).	15
Willoughby et al., 2010	Philippines	207	24.6% male 67.6% female, 7.7% self-identified as either both or neither male and female.	M = 18.50, SD not reported	Christian university students	75% Roman Catholic, 21% other Christian denominations	32-item Genderism and Transphobia Scale (Hill & Willoughby, 2005).	To measure religiosity, participants were asked (a) whether they were religious, and (b) how often they attended church or worship services in the past month.	Religiosity: Transphobia was correlated with self-reported religiosity (r = .25, p < .001) and worship frequency (r = .33, p < .001).	18
Worthen, 2014	USA	991	Not reported.	M = 21.58, SD = 3.36	University students	Not reported	32-item Genderism and Transphobia Scale (Hill & Willoughby, 2005).	Respondents were asked (a) if they religious, and (b) how often they attended church.	Religiosity: In MRAs, religiousness (ß = –1.76, p < .001) and church attendance (ß = –1.10, p < .001) were predictors of transphobia when controlling for various demographic factors.	19
Worthen, 2017a	USA, Italy and Spain	1311	Not reported	Oklahoma: M = 21.82, SD = 3.51 Texas: M = 20.24 SD = 1.90 Italy: M = 23.45, SD = 2.34 Spain: M = 23.53, SD = 2.82	University students	Not reported	Attitudes Toward LGBT People Scale (Worthen, 2012).	Respondents were asked how often they attend church on a scale ranging from 1 (never) to 5 (every week). Respondents were also asked if they believed that “The bible is the actual word of God and is to be taken literally, word for word” were coded as (1) for Biblical Literalism while others were coded as (0).	Religiosity: Church attendance was a significant predictor of negative attitudes in Oklahoma (ß = −.04, p < .001), but not in Texas (p > .05). Biblical literalism was a significant predictor of negative attitudes toward transgender people, when controlling for demographic factors, both in Oklahoma (ß = −.15, p < .001) and Texas (ß = −.20, p < .05). For Italians and Spaniards, neither church attendance nor biblical literalism were significant predictors of attitudes toward transgender people (when controlling for demographic factors; ps > .05).	20
Worthen, 2017b	USA	1,940	42% male, 58% female	Not reported	University students in the Bible Belt in the southern USA.	Not reported	Attitudes Toward LGBT People Scales (Worthen, 2012).	Religiosity scale included questions about general religiousness, church attendance, biblical literalism, and attitudes toward biblical passages about “sin” and “homosexuality” (Worthen, 2012).	Religiosity: In MRAs, after controlling for demographic factors, religiosity was a predictor of negative attitudes toward transgender men for both heterosexual (ß = −.29, p < .05) and LGB participants (ß = −.31, p < .05), and towards transgender women for both heterosexual (ß = −0.37, p < .05) and LGB people (ß = −0.28, p < .05).	19

Notes: RF = religious fundamentalism; RWA = right-wing authoritarianism; MRA = multiple regression analyses.