Abstract
This study reports the development of an inventory to assess the perceived internalized homophobia of gay men in a collectivistic Chinese cultural context. The results of exploratory and confirmatory factor analyses using two samples suggested the viability and stability of a three-factor model: internalized heteronormativity (IHN), family-oriented identity (FOI), and socially oriented identity (SOI). The 11-item internalized homophobia inventory demonstrated good internal consistency and construct validity. Internalized homophobia was related positively to the extent of a sense of loneliness and negatively to self-evaluation and the discrepancy in self-identification as a gay man. In addition, the participants’ internalized SOI consistently predicted their coming out choices in their social surroundings, while their FOI predicted their decisions to enter into heterosexual marriages. The findings suggest that sexual self-prejudice was correlated with IHN, family values, and social norms. The present research demonstrates that a culturally sensitive scale is necessary to understand the cultural and family-oriented values that influence gay Chinese men’s everyday lives, self-constructs, and behavioral choices.
Even in relatively open Western cultures, social stigma and prejudice toward lesbians and gay men continue to exist; needless to say, discrimination and prejudicial attitudes toward lesbians and gay men are even more pervasive in the heterosexism-dominated culture of China, which denies, denigrates, and stigmatizes any nonheterosexual form of behavior, identity, relationship, or community (Neilands, Steward, & Choi, 2008; Zhou, 2006). Homosexual individuals internalize sociocultural discrimination and prejudices in their self-image (Allen & Oleson, 1999; Malyon, 1981). Not only is this internalized negative self-image reflected in the negative attitudes, identities, and beliefs with regard to themselves that these individuals hold, but it also deeply affects events in their everyday lives (Gencoz & Yuksel, 2006; Newcomb & Mustanski, 2010). Internalized homophobia is associated with many psychosocial experiences including low self-evaluation (Blais, Gervais, & Hebert, 2014), a sense of loneliness, homosexual identity formation (Rowen & Malcolm, 2002), the decision to enter into a heterosexual marriage (Higgins, 2002), and the disclosure of sexual orientation to others (Brown & Trevethan, 2010; Dew, Myers, & Wightman, 2006). Reducing internalized homophobia can increase homosexual individuals’ positive feelings about their sexual identity and healthy functioning (Dupras, 1994; Lock, 1998; Verduzco, 2014).
To understand their internalized homophobia and improve psychotherapeutic approaches to decrease their internalized homophobia, a culturally tailored scale to rate these individuals’ experiences of homophobia is necessary. Existing measures of internalized homophobia mainly focus on the dimensions of negative feelings about oneself as a homosexual individual (Currie, Cunningham, & Findlay, 2004; Mayfield, 2001; Ross & Rosser, 1996; Shidlo, 1994; Smolenski, Diamond, Ross, & Rosser, 2010; Theodore et al., 2013). A major criticism of current scales of internalized homophobia is that they measure only the negative cognitive, affective, and behavioral components of a gay person’s individually oriented identity (Nguyen et al., 2016). Hwang (2011) and Ren (2012) claim that individually oriented scales fail to consider the ideology of family and the ideology of community and human interconnectedness. Kam (2017) claims that homosexual individuals in East Asian cultures describe feelings of guilt and even panic at the thought of bringing shame to their parents and family and failing to continue the family line.
In neglecting to access culturally related internalized homophobia, current scales of internalized homophobia fail to measure an important aspect of the culturally related construct of homophobia. As these Western-based scales are focused on the cognitive, affective, and behavioral components of an individual’s self-identity, they insufficiently assess the internalized homophobia in those from socially oriented Chinese and other East Asian cultural backgrounds. There is a gap between the scales of internalized homophobia in Western cultures and the actual experiences of individuals from non-Western cultures. The need for modification is clear.
Yang’s (1997) Chinese social orientation theory implies that the self-concept of Chinese people is socially oriented. His Chinese social orientation model includes four coherent sets of closely related modalities: familial orientation (lineage prolongation, family face, family prosperity, family belonging), relationship orientation (relational formalism, relational interdependence, relational harmony, relational determinism), authoritarian orientation (subordination to authority, authority sensitization, psychological impotence before a leader or authority), and “other” orientation (worrying about the opinions of others, strong conformity with others, deep concern about social norms; Kim, Yang, & Hwang, 2006; Yang, 1997). Choi (2016) and Liu (2013) suggest that surrounding social pressures may also be a unique factor in gay Chinese men’s internalized homophobia, as these pressures involve perceived obligations such as filial piety, obedience to hierarchy, familial values, carrying on the family line, a social expectation to marry, and conformity to heteronormative traditions. To fit into these structures, more than 70% of well-educated homosexuals enter into a heterosexual marriage that conceals their sexual orientation and places their heterosexual partner in the dark (Chou, 1997; Zhang, Li, Hu, Liu, & Shi, 2000).
To the best of our knowledge, there has been no study that focuses on gay Chinese men’s socially oriented internalized homophobia and that adequately addresses their life contexts. In Chinese hierarchical social structures, these feelings of homophobia are expected to be closely associated with psychological problems. As homophobia related to social surroundings is such a critical concept for individuals in China, the existing framework for interpreting homophobic feelings must be enlarged to incorporate it. Consequently, developing an internalized homophobia scale that is specifically oriented toward Chinese communities and culture is urgently needed.
Purpose of Study and Hypothesis
As an initial step toward this aim, the present study was conducted to describe the development and psychometric evaluation of a scale designed to assess internalized homophobia in Chinese cultures. The first sample was mainly used to determine the scale items through exploratory factor analysis (EFA). The second sample was used to cross-validate the factor structure through confirmatory factor analysis (CFA). Based on the previous literature, the association between internalized homophobia and psychosocial problems was examined. We hypothesized that a Chinese version of an internalized homophobia scale is a valid scale for testing gay Chinese men’s internalized phobia experiences and that gay Chinese men’s internalized homophobia is associated with low self-evaluation, a sense of loneliness, a discrepancy in self-identification as homosexual, the choice to enter into a heterosexual marriage, and the decision to disclose sexual orientation to others.
Method
Item Generation and Verification
One psychiatrist with experience working with homosexual individuals and one professor with experience in theoretical psychology and scale development developed the 40 items reflecting experiences of heteronormativity and culturally related homophobia from existing scales and from a previously published qualitative study (Wagner, 1998; Ren, Howe, & Zhang, 2017). To ensure content validity and improve clarity, two psychologists, one psychiatrist, two leaders who work in community lesbian, gay, bisexual, and transgender (LGBT) centers, and five gay men reviewed the items. After the items were reviewed and those that were redundant, unclear, or irrelevant were removed, 35 items were left in the initial item pool. For all 35 items, a 5-point response scale ranging from 1 (strongly disagree) to 5 (strongly agree) was used to increase the response rate and response quality and to reduce respondents’ “frustration level” (Babakus & Mangold, 1992). Nine items were negatively keyed, and the rest were positively keyed.
Data Sources
Internet-based survey questionnaire: Through an anonymous Internet-based survey in China, a cross-sectional survey was conducted over 6 months from December 2016 to May 2017. Participants were recruited by using advertisements on gay chat room websites, WeChat groups, QQ groups, and gay social media. The participants completed an Internet-based survey questionnaire via the professional survey website Wenjuan Xing.
Paper-and-pen questionnaires: Our recruiters worked closely with community leaders in the two LGBT centers. After an introduction about the present study by the community leaders to the gay men in the LGBT centers, the recruiters gave the men a brief introduction about themselves and the study and asked whether the men were willing to participate. The recruited participants were verbally informed of the nature and purpose of the study, the survey procedures, the sensitive nature of the questions, the confidentiality parameters, risks and benefits, their rights as potential participants, and the freedom to cease participation at any time. Each participant was asked to complete a self-administered anonymous questionnaire voluntarily between December 2016 and May 2017. A total of 132 participants completed the questionnaires; 16 of them did not finish the whole questionnaires, and we used the remaining 116 participants’ data in the final analysis.
Participants were eligible for the study if they were male, aged 16 years or older, gay, and residents in China. A screening question of whether you are a gay man (your romantic attraction, sexual attraction, or sexual behavior exclusively in regard to people of the same sex) was asked. If the participants answered yes, then they were invited to complete the scale. This study was approved by the Southwest Hospital research ethics committee. The data sources for this study consisted of gay men who were Chinese citizens (N = 312). The respondents lived in major metropolitan areas and were predominantly highly educated: 95.2% had an undergraduate education or above. The other baseline characteristics of the cohort are reported in Table 1.
Demographic Characteristics of the Participants.
Note. aDemographic characteristics of all participants (N = 312). bDemographic characteristics of the participants in the exploratory factor analysis (N = 196). cDemographic characteristics of the participants in the confirmatory factor analysis (N = 116).
Potential Correlates Examined
Based on the previous literature, internalized homophobia is associated with low self-evaluation, a sense of loneliness, a discrepancy in self-identification as homosexual, the choice to enter into a heterosexual marriage, and the decision to disclose sexual orientation to others. The association between gay Chinese men’s internalized homophobia and these psychosocial problems is examined.
Perceived sense of loneliness
One item tested the perceived sense of loneliness; the item was rated from 0 (completely normal) to 10 (very negative).
Coming out experiences
Four items represented whether (yes or no) the respondent was open about his sexual identity to his parents, relatives, friends, and colleagues and others in his work environment.
Self-evaluation
One item represented the degree of the respondent’s self-judgment. The item invited each respondent to evaluate himself based on an integration of all aspects of his life and to rate his evaluation from 0 (totally negative) to 10 (totally positive).
Discrepancy in self-identification as a gay man
This item reflects the discrepancy or conflict between personal sexual orientation and social sexual identity. Based on the suggestions and reflections by clinicians, researchers, and gay men, some gay men experience romantic attraction, sexual attraction, or sexual behavior exclusively in regard to people of the same sex. They still claim or think of themselves as bisexual men or as tending to be straight men, which reflects their internalized phobia regarding their own sexual identity. The present study added this item to test the discrepancy in the respondents’ self-identification as gay men on a quantitative level. The rating for the item ranged from 0 (heterosexual) to 5 (bisexual) to 10 (totally homosexual). Higher scores implied higher gay self-acceptance.
Relationship situation
The respondents reported their marital status: single, marriage of convenience (marriage to a lesbian), heterosexual marriage, cohabitating with a sexual partner, or divorced from a heterosexual marriage.
Statistical Analyses
Factor analysis and item trimming
The present research conducted EFA based on the items’ polychoric correlations using SPSS 18. Sample 1 utilized 196 completed on-line platform questionnaires, and the other baseline characteristics of the cohort are reported in Table 1. Of those who completed the questionnaires, 94.9% had undergraduate degrees or higher. EFA was conducted using 196 participants. A scree test and parallel analysis indicated three factors (Patil, Singh, Mishra, & Donavan, 2007). The 16-item three-factor solution was chosen using the following item selection criteria: (a) minimum loadings of .40 on at least one factor and cross-loading no greater than .30 on a second factor and (b) at least four clean items on each factor (Brown, 2014). After conducting the EFA, this study tested the factor structure found in the first sample (the tentative model) on the second sample group using CFA. The 16 items retained from the EFA process were subjected to a three-factor oblique model CFA with sample 2 using robust maximum likelihood estimation methods in Mplus. The baseline characteristics of the cohort are reported in Table 1. Three fit indices were used to evaluate the fit of the model to the data: (a) the comparative fit index (CFI; a value of .90 or greater suggests a reasonably good model fit), (b) the root mean square error of approximation (RMSEA; a value of .08 or lower suggests a reasonable error of approximation), and (c) the standardized root mean square residual (SRMR; a value of .10 or lower suggests an adequate model fit).
Associations with the hypothesized correlates
Items were summed within the scale dimensions to form subscale scores and whole scores in the entire group of 312 participants. The Pearson correlation coefficient was used to examine the correlations between continuous variables. For the binary variable, logistic regression was used to estimate the adjusted odds ratio (OR) to predict the relationship between the subscale scores and the binary variable. The results are reported in Table 6.
Results
Factor Analysis and Item Trimming
The analysis resulted in a three-dimensional scale consisting of internalized heteronormativity (IHN), family-oriented identity (FOI), and socially oriented identity (SOI). EFA on the 196 samples provided eight eigenvalues greater than 1. The scree plot had a substantial drop from the third (1.575) to the fourth (1.310) eigenvalue and was relatively flat for the remaining eigenvalues. Sensitivity analysis using parallel analysis identified that the three largest eigenvalues based on the observed Pearson correlation matrix (10.89, 2.97, and 2.16) were greater than the means (1.89, 1.78, and 1.69) and 95th percentiles (1.97, 1.86, and 1.77) of the corresponding random eigenvalues.
Three and four EFAs were examined using robust maximum likelihood estimation. The study maintained the three-factor solution. The CFA was conducted based on the EFA in the field sample to check the model fit. The loadings ranged from .57 to .99 for the first factor (items 1 to 7), from .57 to .87 for the second factor (items 8 to 12), and from .52 to .72 for the third factor (items 13 to 16). The initial fit indices indicated an insufficient fit for one of the three indices: CFI = .82, TLI = .78, and RMSEA = 0.105 (90% CI [.0087-.123]); additionally, χ2 (120 df) = 825.08 was significant (p < .0001). Therefore, the scale was further refined by sequentially eliminating items with modification indices over 10, resulting in an 11-item three-factor scale. In this scale, the model fit was good: CFI = .93, TLI = .90, and RMSEA = .008 (90% CI [.0052, .0113]) ; χ2 (55 df) = 497.50 was significant (p < .0001).
The scale established from the sample of gay Chinese men thus included 11 items on three dimensions in the Chinese Internalized Homophobia Scale (CIHS). Each of the items representing the three factors and their factor loadings are presented in Table 2.
Items and Factor Loadings.
Reliability
The subscales exhibited high internal consistency. The alpha coefficients of the whole scale and the three subscales were .84 for the total and .72 for IHN subscale, .77 for the SOI subscale, and .70 for the FOI subscale.
Associations With the Hypothesized Correlates
This study tested the correlation between the CIHS and its subscale scores with the hypothesized correlates. The hypothesized correlates are shown in Table 3.
Homophobia Scale Associations With the Hypothesized Correlates.
Note. N = 312.
p < .05. **p < .01. ***p < .001.
Perceived sense of loneliness
As hypothesized, the whole scale (r = .127, p < .001) and the FOI (r = .216, p < .001) subscale were positively associated with a perceived sense of loneliness. A greater perceived sense of loneliness was associated with greater gay Chinese men’s internalized homophobia (r = .127, p < .001) and FOI (r = .216, p < .001).
Self-evaluation
Lower self-evaluation was associated with greater homophobia in the whole scale and all subscales. Lower self-evaluation was moderately associated with perceived FOI (r = −.246, p < .001).
Discrepancy in self-identification as a gay man
Higher scores implied higher self-acceptance. The scores measuring the extent of the discrepancy in self-identification as a gay man was negatively correlated with IHN (r = −.182, p = .001).
Coming out experiences
Tables 4 and 5 present the associations between the CIHS subscale scores and the coming out behavior correlates. This research examined the association between the score of Chinese internalized homophobia and the participants’ experiences coming out to major significant others, namely, parents, friends, relatives, and colleagues. The results implied that gay Chinese men with higher levels of social phobia are less likely to come out in their social surroundings. Those with higher levels of social phobia together with IHN are less likely to come out to friends and relatives.
Binary Logistic Regression Model to Predict Practices of Coming Out to Parents and Relatives.
Note. N = 312.
p < .05.
Binary Logistic Regression Model to Predict Practices of Coming Out to Friends and People in the Workplace.
Note. N = 312.
p < .001.
Binary Logistic Regression Model to Predict Marital Choices.
Note. N = 312.
p < .05.
Entering into marriage
This research examined the association between gay Chinese men’s internalized homophobia and the choice of entering into marriage (defined as still in a heterosexual marriage, divorced from a heterosexual marriage, or in a marriage of convenience). The result implied that FOI can predict the choice to enter into marriage.
Discussion
The CIHS developed in this study is a self-administered scale that examines gay Chinese men’s internalized homophobia with the aim of expanding the conceptualization of internalized homophobia beyond the parameters established by Western studies. This scale provides a total score that indicates the extent of gay men’s internalized homophobia. It also provides three subscales that can be used to identify specific sources of IHN, FOI, and SOI. IHN is the belief that homosexuality is abnormal. SOI is about how other people will think about me or my family. FOI refer to the ideology of filial piety, familial first, carrying on the family line, a social expectation to marry, bringing up children with the aim to guard against troubles in one’s late years, and conformity to heteronormative traditions.
The scoring is simple. Individual item responses are added together for groups of items and for all 11 items to obtain the subscale scores and the total score, respectively. To the best of our knowledge, this study is the first to develop a reliable and valid instrument for measuring the level of internalized homophobia among gay men in the Chinese context. Both EFA and CFA were performed, and the results indicated that the three-factor structure was appropriate.
The total scale and the subscales appear to be internally consistent. Preliminary evidence suggests that the scale is valid. The Chinese internalized homophobia scores were found to be positively correlated with a sense of loneliness and negatively correlated with lower self-evaluation and discrepancy in self-identification as a gay man. The strength of a gay Chinese man’s FOI predicted his choice of entry into heterosexual marriage, and his social identity–related homophobia could predict the likelihood of his having come out to his parents, relatives, friends, and colleagues.
The findings from the field testing of the homophobia scale and the results from other studies suggest that internalized homophobia is related to low self-evaluation (Harris, Cook, & Kashubeck-West, 2008; Lock, 1998). The present research identified that a higher correlation between FOI and SOI was related to low self-evaluation, implying that family and social values were intertwined with the participants’ own self-judgments. In particular, the current study discovered low-to-moderate correlations between an individual’s FOI and sense of loneliness. Not surprisingly, as gay men cannot fulfill traditional family responsibilities, many undoubtedly feel lonely and marginalized in mainstream Chinese family culture. Interestingly, those who were less likely to self-identify as gay were more likely to feel IHN. The current research suggests that the scores on discrepancy in gay self-identification were negatively correlated with IHN. This result reflects that it is difficult to negotiate the internal and social conflict as a gay man in the sociocultural context of compulsory heterosexuality. In the present study, marriage of any sort is associated with FOI-related homophobia. Thus, a higher level of family-oriented internalized homophobia in an individual was identified as being related to a greater chance of entering into marriage. As expected, entry into heterosexual marriage can be predicted according to the strength of a gay Chinese man’s FOI: Men who reported higher rated FOIs were more likely to have entered into heterosexual marriages than men with lower rated FOIs were. Family and societal pressures may be important factors in gay Chinese men’s self-concepts, thus contributing to their internalized homophobia (Ren et al., 2017).
In Chinese culture, homosexual individuals have not only internalized heteronormativity but also absorbed family and sociocultural values in a negative gay self-identity (Hu & Wang, 2013). Considering a person’s individual identity without considering his sociocultural narrative and background is insufficient for examining his internalized homophobia (Fivush, Habermas, Waters, & Zaman, 2011; Kim et al., 2006). The Chinese concept of self is always defined by relationship, which is termed a social-oriented tendency (Wang, Wei, Zhao, Chuang, & Li, 2015). In Chinese communities, social expectations are deeply intertwined and integrated into one’s self-concept, and the social pressures surrounding gay men may also become a unique factor in their internalized homophobia (Kim et al., 2006; Yang, 1997).
The results of the present study are consistent with Yang’s theoretical conception, which expand internalized homophobia beyond the individual to a comprehensive individual dimension and a dimension of socially oriented internalized homophobia, and they contribute to the existing internalized homophobia literature by assessing Chinese internalized homophobia. The socially oriented dimension of internalized homophobia implies that the components of the CIHS scale and the internalized homophobia of Chinese people are related to sociocultural values such as interdependence, harmony, and saving face. Such a socially oriented dimension has not been studied in other inventories of internalized homophobia and stands in stark contrast to their individualized conceptualizations, which primarily focus on identity constructs such as self-identity, personal emotions, cognitions, and behaviors.
The absence of this social and cultural factor or subscale in previous studies from North America and Europe suggests that simply understanding homophobia based on one’s personal emotions, cognitions, and behaviors is insufficient and that it is important to understand oneself within the Chinese cultural context, in which the social expectations of others are deeply intertwined and integrated into internalized homophobia. The relationships between a gay man’s SOI and his openness about his sexual orientation are likely to be consistent in regard to his perceived SOI and coming out. A man’s SOI may influence his level of openness. The intensity of a gay man’s social identity–related homophobia is deeply influenced by social norms, concerns for what other people think of him and his family, or sensitivity to the opinions of others, meaning that he will be less likely to come out to his social circles. This finding has important implications for clinical practice in China, as direct interventions can be focused on reducing not only the individual components but also the sociocultural and family-related components of internalized homophobia. The findings show that sexual self-prejudice is highly associated with IHN, family values, and social norms. This study suggests that this subscale may be able to predict the choice to enter into a heterosexual marriage and the likelihood of self-judgment, feelings of loneliness, and coming out behaviors. This domain is an important area for research and practice not only because addressing these factors is important for improving the collective social standing and well-being of sexual minorities but also because evidence suggests that mental health services in China should focus on these socioculturally related issues when serving gay men.
The implications of this research are threefold. First, it is clear that assessments of internalized homophobia must incorporate sociocultural dimensions. The scale developed is valid for assessing homophobia in gay Chinese men. Second, this research identifies a correlation between scores of Chinese internalized homophobia and psychosocial problems. The current research suggests that future possible interventions and public health services for gay men should address such relationships. In particular, given the cultural context of China, doing so may fill an important gap for this community and research area. Third, the findings suggest that addressing the needs of a culturally diverse population such as the global LGBT community requires that clinicians working with a given cultural group within that population possess knowledge and skill sets that are specific to that group’s context as well as the appropriate attitudes and beliefs.
Directions for Further Scale Evaluation and Improvement
In the present study, the respondents were predominantly young, well-educated, urban-dwelling gay men; the sample likely underrepresents gay men who are older, rural, poor, and isolated. Therefore, future research on the CIHS in gay men should also use other recruitment methods to reach these gay men who were not reached by the current study.
In terms of examining associations with external constructs, a limitation of this study is that many constructs were measured using single questions. Since single–item scales are less reliable than multiple-item scales, this method of measurement may have attenuated the estimated correlations. Within the limits of an online survey that covered multiple topics, such as the CIHS, the present study was unable to measure every construct with a multiple-item scale. The results from this study should therefore be supplemented with additional studies that use more rigorous measures. In identifying scale items, the present study included more items reflecting socially and culturally related phobias because participants in the prior qualitative study talked more about how they felt about themselves and the ways their situations were influenced by their social and family relationships and norms. Previous qualitative research and current quantitative research with this population may help further improve the CIHS and its subscales for use in Chinese cultural contexts.
To conclude, this study developed the CIHS, providing an important tool for the growing but still very limited research on the effects of sexual stigma in gay men. The CIHS, drawn from the international literature and tailored to incorporate the experiences and perceptions of gay men in a Chinese context, is a valid and reliable scale reflecting three components of these men’s internalized homophobia: IHN, FOI, and SOI. This study also revealed patterns in the associations between the CIHS and its subscales and external constructs in gay Chinese men that have theoretical relevance for the application and adaptation of research on internalized homophobia to a non-Western context.
Footnotes
Contributors
All authors contributed to the study conception and design. The first author Zhengjia REN wrote the first draft of the article. All authors contributed to the subsequent revision of the article and approved the final manuscript for publication.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
