Assessing Politicized Sexual Orientation Identity: Validating the Queer Consciousness Scale

ABSTRACT Building on psychological theories of motivation for collective action, we introduce a new individual difference measure of queer consciousness, defined as a politicized collective identity around sexual orientation. The Queer Consciousness Scale (QCS) consists of 12 items measuring five aspects of a politicized queer identity: sense of common fate, power discontent, system blame, collective orientation, and cognitive centrality. In four samples of adult women and men of varied sexual orientations, the QCS showed good test-retest and Cronbach’s reliability and excellent known-groups and predictive validity. Specifically, the QCS was positively correlated with identification as a member of the LGBTQ community, political liberalism, personal political salience, and LGBTQ activism and negatively correlated with right-wing authoritarianism and social dominance orientation. QCS mediated relationships between several individual difference variables and gay rights activism and can be used with both LGBTQ people and allies.

Little research has been conducted on the development of a politicized sexual orientation identity, though studies have focused on the connection between sexual and gender minority identification and political participation. For example, Renn and Ozaki (2010) found that college student leaders of LGBTQ organizations had strongly connected their leadership identities as activists to their queer identities. They found that for many LGBTQ students, "to be LGBT is to be [an] activist" (Renn & Ozaki, 2010, p. 22). However, Stürmer and Simon (2004) found that a strong identification with the gay rights movement (rather than a strong gay identity) predicted LGBTQ activism. Similarly, Van Dyke and Cress (2006) found that gays and lesbians were more likely to share a collective homosexual identity when they felt threatened by the anti-gay movement, in this case, around the HIV/AIDS epidemic. In heterosexuals, Wilkinson and Sagarin (2010) found that a strong identification with the gay community could lead to the development of a LGBTQ activist identity, providing evidence that identification as a sexual minority is not always a determinant of LGBTQ activism.

Politicized group identities
In mainstream social and personality psychology, researchers have found that a politicized group identity is key to understanding motivation for activism; simply being a member of a stigmatized group is not enough to spur individual action (Duncan, 1999(Duncan, , 2012Duncan & Stewart, 2007;Stürmer & Simon, 2004;van Zomeren, Postmes, & Spears, 2008). The social identity model of collective action (SIMCA; van Zomeren et al., 2008) identified identity, injustice, and efficacy as key elements comprising a politicized group identification. Duncan (2012) integrated individual difference variables into the SIMCA model to capture within-group differences in group consciousness and activism (see Figure 1). She and her colleagues (Duncan, 1999(Duncan, , 2012Duncan & Stewart, 2007) adapted popular American National Election Studies items (Miller, Kinder, Rosenstone, & the National Election Studies, 1999) to create a measure of feminist consciousness (or politicized gender identity) that was then used to mediate between individual differences in personality and life experiences and women's rights activism. This approach is useful in integrating individual difference approaches to motivation for activism with social identity approaches, as it helps to clarify which members of a particular social group will tend to become politically active and which will not. This measure of politicized group identity (or group consciousness) assesses five key elements of stratum consciousness as described by Gurin and colleagues (Gurin & Markus, 1989;Gurin, Miller, & Gurin, 1980) and reflect the identity, injustice, and efficacy components of SIMCA (van Zomeren et al., 2008). Group consciousness consists of (1) identification with the group, or a sense that what affects some members of one's group will have an impact on all members of the group (also called common fate); (2) power discontent, or the sense that one's group does not have access to an appropriate level of power and resources in society; (3) rejection of legitimacy, or system blame, or a sense that the lack of power experienced by one's group is unfair and related to systemic rather than individual factors; (4) collective orientation, or the sense that the best way to address unfair power differentials is by organizing together and working as a group; and (5) cognitive centrality, or the tendency to use one's group membership as a cognitive schema when processing information. According to this model, politicized group identification relies on the development of a shared belief that the lowstatus group experiences social injustice, and that the group should work as a collective to change their status in society. This model has been applied to understand class consciousness, race consciousness, age consciousness (Gurin et al., 1980), and, most commonly, feminist consciousness (Duncan, 2010;Gurin, 1985;Henderson-King & Stewart, 1999). People scoring high on these measures tend to be more politicized and more politically active for their group than people scoring low on these measures. Using a composite measure comprising all five elements described above, Duncan (1999Duncan ( , 2012 Duncan & Stewart, 2007) has focused on understanding the ways in which group consciousness mediates the relationships between individual difference variables and collective action. The current studies extend this research to queer consciousness and related activism. A few studies have examined politicized identities around LGBTQ movements; however, there does not currently exist an individual difference measure of queer consciousness that encompasses all five of these elements. To be sure, previous studies have looked at the role of identification, sense of belongingness, or common fate (Cimino, 2007;Harris, Battle, Pastrana, & Daniels, 2015). For example, Stürmer and Simon (2004) measured politicized gay identity by asking participants to rate their identification with a gay social movement organization and Simon, Pantaleo, and Mummendey (1995) asked participants to rate how similar and different they were from in-groups and out-groups.
In the current article, we have created an individual difference measure of queer consciousness based on Gurin et al.'s (1980) model of stratum consciousness that not only includes all five aspects of a politicized collective identity but also has the ability to measure the queer consciousness of allies to the LGBTQ movement. Researchers attempting to examine the politicization of sexual orientation identities need a reliable and valid measure of group consciousness that can be used with members of the LQBTQ community and allies alike. This is important because not all people politicize their sexual orientation, and it is the politicization of this identity that is most closely related to activism for LGBTQ rights and that mediates the relationships between individual difference variables and activism.
This new measure is especially relevant in today's quickly changing political climate. For example, by 2005, almost 2000 U.S. schools sponsored Gay-Straight Alliances (GSAs) to provide a forum and a safe place for those who identified as sexual minorities and their allies (Savin-Williams, 2005), and in 2015 the U.S. Supreme Court validated the right of gay people to marry. There are many questions about the LGBTQ community and allies that cannot be answered without a reliable and valid measure of group consciousness, such as the effects of GSAs and pro-gay media on attitudes and behaviors. Our measure can be used to learn more about the LGBTQ community, LGBTQ activism, pro-gay heterosexuals, anti-gay homosexuals and heterosexuals, and the political climate around LGBTQ rights.
Additionally, we chose to make the measure inclusive for those who identify as straight with the knowledge that sexuality can be fluid (Diamond, 2008) and that many young people are rejecting categorical labels based on sexual orientation (Savin-Williams, 2005). We also theorize that it is possible for people to be heterosexual but to have a political queer identity, evidenced by the existence of allies participating in GSAs.

Queer consciousness
For our measure, we chose the term queer consciousness to describe a politicized identity around sexual orientation, and we chose to use the acronym LGBTQ to describe the low-status sexual minority group. Categorizing this community has been the subject of much debate in both psychology and sociology, and we have used theories and research from both fields in our research. Stryker (2008) referred to the constantly changing acronym as "alphabet soup" (p. 21), and Savin-Williams (2005) documented that in the history of activist groups, the acronym used to group sexual (and gender) minorities changed from LGB (lesbian-gay-bisexual) to LGBTQ 2 (lesbian-gaybisexual-trans-queer-questioning). Within our measure, we used LGBTQ to categorize sexual minorities because it includes both sexual orientation and gender identification (for more information on the inclusion of trans in the LGBTQ rights movement, see Broad, 2002). We also specifically included the term queer because of its use academically (in queer theory) and its mainstream use as an umbrella term.
Though the word queer was once used pejoratively to describe sexual minorities, it has since been reappropriated by the community as a term that describes someone who opposes the gender binary and heteronormative culture (Renn, 2007). As society progresses and views on homosexuality change, sexual orientation has become a more fluid concept; rather than using definitive categories, people are beginning to accept that sexual orientation exists along a spectrum (Barker, Richards, & Bowes-Catton, 2009). For example, Savin-Williams (2005) noted a recent trend of teenagers who have turned away from binding terminology such as gay and lesbian, instead preferring not to label their sexuality. In some ways, this has simplified the process of identity development by emphasizing the fluidity of sexuality (Horowitz & Newcomb, 2001). At the same time, it has complicated traditional views of sexual orientation, making it more difficult to create categories on the basis of LGBTQ identification.
We labeled the politicized identity around the LGBTQ movement as queer consciousness because it provides theoretical consistency as well as some practical benefits. The term used to describe someone who is invested in the women's rights movement is feminist; however, there is no truly homologous term for the LGBTQ movement. Furthermore, in the same way that men can be feminists, allies can make important contributions to the movement. Ji, Du Bois, and Finnessy (2009) found that it is possible to increase ally identification through education, providing evidence for the idea that an identity connected to the LGBTQ movement can be developed and become politicized for heterosexuals. Also, we felt that using queer in the measure's name reflected the politicized nature of the identity we wanted to assess.
The term queer has distinctly political connotations. Not only is it a broad term with which to describe the community, encompassing any behavior that is not traditionally heterosexual (if there is such a thing), but queer is also one of the most modern and political terms in use today. Johnson (2007) argued that queering could be important for modern coalition politics because it is issue-based and focused on breaking down boundaries. Though the term queer is controversial and the queer movement itself can have different goals from the lesbian, gay, bisexual, trans, and numerous other identity politics movements, it is both the broadest umbrella term and carries the heaviest political connotation. Furthermore, we wanted to follow the current trend of "queering" so that the title of our measure remained salient to modern and future sexual identity politics. (For more information on the historical use of the term queer, see Johnson, 2007;Wahlert, 2012).
In developing the Queer Consciousness Scale (QCS), reliability and validity data were collected from four samples: two samples of college women, an ideologically varied general sample of both men and women recruited from Internet blogs, and a large general sample of women and men collected via Mechanical Turk. Because there are no existing measures of politicized sexual orientation identity, we concentrate our efforts on establishing Cronbach's alpha and testretest reliability and known groups and predictive validity. In the first study, the measure is introduced and validated with a sample of young women. Test-retest reliability in a second sample is described. In the third sample, the results of the first study are replicated and extended to a sample of men. In the fourth sample, the results are further extended to a large general sample of women and men, additional correlates are presented, and the ability of QCS to mediate the relationship between some individual difference variables and gay rights activism is tested.
Past research has shown that members of oppressed groups are more likely than members of nonoppressed groups to develop group consciousness and to become politically active (Duncan & Stewart, 2007;Gurin et al., 1980;Montgomery & Stewart, 2012). Therefore, we hypothesize that women and people identifying as gay, lesbian, bisexual, and queer will score higher on the QCS than men and people identifying as straight. In addition, LGBTQ people who have "come out" to their families, friends, and at work should score higher on queer consciousness than those who have not (Swank & Fahs, 2013a, 2013b. Previous research has established that participating in women-only groups and explicitly teaching women to question existing gender relations are related to increased feminist consciousness in women (Bargad & Hyde, 1991;Henderson-King & Stewart, 1999). We expect that there should be analogous relationships for LGBTQ people. Therefore, when compared to nonparticipants, QCS scores should be higher for people who have participated in LGBTQ organizations, attended Gay Pride parades, and taken college-level queer studies classes.
In terms of predictive validity, the QCS should be correlated with attitudes and behaviors related to open acceptance of the rights of gay people. One of the most obvious measures to test is self-identification along the Kinsey sexual orientation scale (Kinsey, Pomeroy, & Martin, 1948). Identifying along the non-straight sexual orientation spectrum should be related to higher scores on the QCS. Liberals tend to endorse progressive political and social views at higher rates than do conservatives (Jones, 2015); therefore, liberal political orientation should be related to higher scores on the QCS. Right-wing authoritarianism (RWA) and social dominance orientation (SDO) are associated with conservative beliefs about a variety of social issues, and authoritarians are known to be hostile to gay people (Peterson, Doty, & Winter, 1993). Therefore, QCS scores should be negatively correlated with RWA and SDO scores.
Personal political salience (PPS), or the tendency to attach personal meaning to political events, is related to feminist and race consciousness and activism (Duncan, 1999(Duncan, , 2005Duncan & Stewart, 2007). We expect that PPS should also be related to queer consciousness. In terms of behavior, feminist consciousness is related to participation in women's rights activism (Duncan, 1999;Duncan & Stewart, 2007). Similarly, the QCS should correlate positively with LGBTQ-related political activism.

Participants and procedures
One hundred and twenty-three female students enrolled at a small women's college located in the northeastern United States were asked to complete an online survey using Surveymonkey.com for course credit or the chance to win a $50 Amazon.com gift card. Sixty-two percent of participants identified as White, 23% as Asian American, 10% as African American, 6% as Latina, and 1% as Native American. To evaluate test-retest reliability, a separate sample of 153 women recruited from three different psychology classes was administered the QCS in class on two occasions, approximately 1 month apart.

Measures
Queer consciousness scale As mentioned earlier, the QCS was created by adapting ANES items used to assess the four elements of Gurin's (1985) stratum consciousness, along with cognitive centrality (Gurin & Markus, 1989). The items were adapted with the following objectives in mind: (1) people of varied sexual orientations should be able to answer the questions; (2) the scale should include a balance of negatively worded as well as positively worded items; (3) each aspect of queer consciousness should be represented by at least two items, each worded in opposite directions; (4) the scale should be relatively short and quick to complete; and (5) the scale should include an item that explicitly asked participants to indicate their level of identification with the queer community.
Participants were asked to read 12 statements and rate how strongly they agreed or disagreed, using a 5-point scale where 1 = strongly disagree, 2 = somewhat disagree, 3 = neither agree nor disagree, 4 = somewhat agree, and 5 = strongly agree. The complete scale is included in Appendix A. Items 3, 4, 6, 7, and 10 were reverse scored. The mean of all items completed constituted the scale score.

Known groups validity
Participants were asked a series of questions about their membership in groups that might reasonably be expected to be populated with people who would score higher on queer consciousness than nonmembers. Participants were asked to describe their sexual orientation in an open-ended answer. Answers were coded as straight (63%) or queer (37%; including lesbian, gay, bisexual, and queer). In addition, queer participants were asked if they had come out to their families (63%), friends (89%), and at work (43%). Participants were asked whether they had been a member of an LGBTQ organization (35%), attended a Gay Pride parade (48%), or taken a collegelevel queer studies class (25%). We expected that people who had participated in queer-related activities and who had come out would score higher on queer consciousness than participants who had not.

Sexual orientation
We asked the participants to place themselves on the Kinsey Scale (Kinsey et al., 1948), a 7-point Likert scale ranging from exclusively heterosexual to exclusively homosexual. Answers were recoded so that higher numbers represented a straight sexual orientation. See Table 2 for alpha reliabilities for scales and means and standard deviations for all variables.

Liberalism-conservatism
Participants were asked to place themselves on a scale from 1 to 7, with a 1 indicating strong liberalism, a 7 indicating strong conservatism, and a 4 indicating a more moderate political attitude (Miller et al., 1999). The scale was reverse scored so that high scores indicated stronger liberalism. Altemeyer's (2006) Right-Wing Authoritarian scale consisted of 22 items, 10 of which were reverse scored. Participants were asked to indicate their feelings on an 8-point scale that progressed from very strongly disagree to very strongly agree. A sample item is "The established authorities generally turn out to be right about things, while the radicals and protestors are usually just 'loud mouths' showing off their ignorance."

Right-wing authoritarianism
Social dominance orientation Social Dominance Orientation (SDO; Pratto, Sidanius, & Levin, 2006) was measured by asking participants to indicate whether they agreed or disagreed with 16 statements (half of which were reverse-scored) on a 7-point Likert scale ranging from "Right now I feel strong agreement" to "Right now I feel strong disagreement." Items include statements such as "Some groups of people are simply inferior to other groups" and "We should do what we can to equalize conditions for different groups" (reverse scored).

Personal political salience (PPS)
The PPS measure consisted of 31 political and social events rated on a 3point scale (1 = not at all personally meaningful, 2 = a little personally meaningful, 3 = very personally meaningful; Duncan, 1999Duncan, , 2005Duncan & Stewart, 2007). Participants were instructed to: "Please rate each of the following events for how personally meaningful it is (or was) to you (i.e., how much it affected your life or reflects your values and concerns)." The measure included contemporary as well as historical events (e.g., Obama presidency, Vietnam War). Overall PPS scores were computed by summing scores on all rated events and taking an item mean. In addition, we created a LGBTQ-events focused PPS score that consisted of mean ratings of the following events: Don't Ask, Don't Tell repeal, Stonewall, Matthew Sheppard's murder, Defense of Marriage Act (DOMA) repeal, same-sex marriage, lesbian-gay rights, transgender rights, and HIV/AIDS.

Activism
LGBTQ-related political activism was assessed by asking participants to indicate the type and level of their participation in three causes: LGBTQ rights (50%), Repeal of DOMA (26%), and HIV/AIDS activism (31%). Respondents marked whether or not they participated in up to six specific actions in support of each cause. Actions included signing petitions; attending meetings; writing, calling, or visiting a public official; contributing money; participating in organizations; and participating in rallies or demonstrations (Duncan, 1999;Duncan & Stewart, 2007). Scores could range from 0 to 6 actions for each of the activism variables.

Reliability analyses
The 12-item QCS showed adequate reliability. Cronbach's alpha for the first sample was .76. Alpha for the test-retest samples was .74 at time 1 and .79 at time 2. Over the course of approximately 1 month, the QCS showed strong test-retest reliability (r = .87, p < .001).

Validity analyses
Known groups validity Participants who identified as queer scored significantly higher on the QCS than participants who identified as straight (see Table 1). Queer participants who were out to their friends scored significantly higher on the QCS than participants who were not; however, there were no differences in QCS scores for participants based on whether they were out to their families or out at work. Participants who had been a member of a LGBTQ-related organization, attended a Gay Pride parade, or taken a college-level queer studies class scored significantly higher on the QCS than participants who had not.

Predictive validity
The QCS correlated strongly negatively with a Kinsey straight sexual orientation, right-wing authoritarianism, and social dominance orientation (see Table 2). The QCS correlated strongly positively with liberalism, overall PPS, and PPS focused on LGBTQ events. In terms of activist behaviors, the QCS correlated positively with LGBTQ rights and the repeal of the Defense of Marriage Act activism but was unrelated to HIV/AIDS activism.

Discussion
In this first study, the QCS showed good Cronbach's alpha and test-retest reliability and predictive validity in college women samples. In terms of known-group validity of the QCS, five out of seven groups that were expected to score higher on the QCS (self-identified queer people, queer participants who had come out to their friends-but not those who had come out to their families and at work-members of LGBTQ organizations, those who had attended a Gay Pride parade, and participants who had taken queer studies classes) scored higher on the QCS than nonmembers of these groups. In addition, the QCS was strongly related to liberalism and personal political salience and strongly negatively related to a straight sexual orientation, authoritarianism, and social dominance orientation. Finally, the QCS was correlated with two of out three types of LGBTQ-related political activism. HIV/AIDS activism was unrelated to scores on the QCS, possibly because HIV/AIDS is no longer seen as a primarily gay male-focused disease, and this was a college-student sample. We conducted two additional studies with mixed-gender samples. In the first of these studies (Study 2), we simply replicated the analyses conducted for Study 1.

Participants and procedures
To oversample LGBTQ people and to validate the QCS on a sample of women and men with a wider range of ideological views and wider age range, we collected data in two ways. First, we used snowball sampling by posting a link and explanation of the survey onto the second and third authors' Facebook walls and asked others to do the same. To assure a balanced number of participants with different political views, we also posted the survey on a total of 15 online conservative and liberal blogs. The blogs were found using the Google search engine, and a link to the survey with a short description was posted. Participants were offered the chance to win a $50 gift certificate to Amazon.com in a raffle. Participants who did not answer at least two thirds of the QCS items were deleted from the sample. A total of 182 participants completed enough of the QCS to be included in the final sample. Seventy-six percent of participants identified as female, and 18% identified as male. The age of the online sample participants ranged from 16 to 60 (M = 27.78, SD = 11.20). For sexual orientation, 38% of the online participants self-identified as lesbian, gay, bisexual, or queer, and 50% of the sample identified as straight. In terms of ethnicity, 88% of participants identified as White, and 22% of the participants identified as people of color (Native American, Asian American, Black, or Latino). Thirty-one percent reported an annual family income under $40,000, 31% an income between $40,001 and $100,000, 28% an income between $100,001 and $200,000, and 11% an income over $200,000. Mean educational level was a 2-year college degree. Thirty percent had once been or were currently married or had made a life commitment to a partner, and 9% had been divorced.

Measures
Measures for Study 2 were identical to those used in Study 1. See Table 2 for alpha reliabilities for scales and means and standard deviations for all variables.

Results
The QCS showed good reliability with a Cronbach's alpha of .82. Consistent with the results of Study 1, participants who identified as queer scored significantly higher on the QCS than participants who identified as straight.
As hypothesized, women scored higher than men on the QCS (see Table 1).
Queer participants who were out to their families, friends, and at work scored significantly higher on the QCS than participants who were not. As in Study 1, participants who had been a member of a LGBTQ-related organization, attended a Gay Pride parade, or who had taken a college-level queer studies class scored significantly higher on the QCS than participants who had not. Consistent with the results of Study 1, the QCS correlated strongly negatively with a Kinsey straight sexual orientation, right-wing authoritarianism, and social dominance orientation (see Table 2). The QCS correlated strongly positively with liberalism, overall PPS, and PPS focused on LGBTQ events. In terms of activist behaviors, the QCS correlated positively with LGBTQ rights, repeal of the Defense of Marriage Act activism, and HIV/AIDS activism.

Discussion
Consistent with the results of the first study, the QCS showed good Cronbach's alpha reliability in a broader demographic sample. In terms of known-group validity of the QCS, all seven of the groups that were expected to score higher on the QCS (self-identified queer people; women; queer participants who had come out to their families, friends, and at work; members of LGBTQ organizations; those who had attended a Gay Pride parade; and participants who had taken queer studies classes) scored higher on the QCS than nonmembers of these groups. In addition, the QCS was strongly related to liberalism and personal political salience and strongly negatively related to a straight sexual orientation, authoritarianism, and social dominance orientation. In addition, the QCS was correlated with all three types of LGBTQ-related political activism.
We replicated these results in one additional, larger sample of men and women. We added two measures of political knowledge to expand our base of dependent variables. Although we expected no relationship between QCS scores and general political knowledge, we hypothesized positive correlations between the QCS and knowledge about LGBTQ political history, and that the QCS would mediate relationships between these variables and gay rights activism.

Participants and procedures
Participants consisted of 607 American adults recruited from the Web site Mechanical Turk, where people can complete short surveys for small payments (see Buhrmester, Kwang, & Gosling, 2011). Participants responded to a listing advertising a research study on attitudes about gender and sexual orientation. Participants were required to be at least 18 years old and American citizens. Sixty-three percent of participants identified as women, 37% as men, and .5% declined to identify their gender. Eighty percent of participants identified as White/Caucasian, 12% as Black/African American, 6% as Latino/Hispanic, 4% as Asian American, 3% as Native American, and 2% as "other ethnicity." Percentages sum to more than 100% because participants could check more than one ethnicity. Age ranged from 18 to 75, with a mean of 35.40 (SD = 12.56). Mean social class of origin was lower-middle to middle class, mean personal income was between $20,000 and $50,000, and mean education level was a 2-year college degree.

Measures
Measures for Study 3 were identical to those used in Studies 1 and 2, although due to space constraints the SDO measure and several questions relating to known-group validity were eliminated. In addition, we did not ask participants to identify their sexual orientation categorically, although they did identify themselves along the Kinsey sexual orientation scale. Participants used sliders (assigned values of 0-9) instead of fixed numbers for continuously scaled measures.
In addition, two measures of political knowledge (one assessing general political knowledge and one assessing knowledge about LGBTQ political history) were created for this study as additional predictive validity measures. General political knowledge (Delli Carpini & Keeter, 1996) was assessed using six items taken from the 2012 American National Election Studies pre-and post-election questionnaires. Items covered basics about the U.S. political system (e.g., how many times someone can be elected U.S. president, how long a U.S. senator's term is), policies (e.g., what Medicare is) as well as knowledge about current political leaders (e.g., who the current Secretary-General of the United Nations is, which party holds a majority in the U.S. House, who the current U.S. Secretary of State is). See Appendix B for the exact wording of the items. Knowledge of LGBTQ political history was assessed with a 6-item quiz created for this study. Items covered famous historical events in the history of LGBTQ activism. See Appendix C for the exact wording of the items. Alpha reliability, means, and standard deviations for both scales are presented in Table 3. On average, participants answered about one half of the items correctly on both scales.
Finally, we tested the notion that the QCS would mediate relationships between individual difference variables and gay rights activism. Baron and Kenny (1986) specified three criteria for assessing mediating relationships: (1) the independent variable is significantly related to the dependent variable (individual difference variables are related to gay rights activism; Path C in Figure 1); (2) the mediator is significantly related to the independent and dependent variables (QCS is related to the individual difference variables and gay rights activism; Paths A and B); and (3) when the mediator is controlled for, the magnitude of the relationship between the independent and dependent variable is reduced (i.e., controlling for QCS scores, the relationship between the individual difference variables and gay rights activism-Path C-is reduced). In cases where Baron and Kenny's (1986) criteria for mediation were met, we tested whether the mediated effect was significantly different from 0 using a bootstrapping procedure (Shrout & Bolger, 2002).

Reliability and predictive validity
Similar to Studies 1 and 2, the QCS showed good reliability (Cronbach's alpha = .84). Women (M = 5.79, sd = 1.46) scored significantly higher on the QCS than did men (M = 5.28, sd = 1.48; t(602) = 4.15, p < .001). Consistent with the results of Studies 1 and 2, the QCS correlated negatively with a Kinsey straight sexual orientation and right-wing authoritarianism (see Table 3). The QCS correlated strongly positively with liberalism, overall PPS, and PPS focused on LGBTQ events. The QCS was unrelated to scores on general political knowledge; however, it was positively related to scores on the LGBTQ political history quiz. In terms of activist behaviors, the QCS correlated positively with LGBTQ rights and repeal of the Defense of Marriage Act activism. Table 4 shows the intercorrelation of independent variables included in the mediation analyses. The strongest (and moderate) relationship was between PPS and liberalism. The overall regression was significant, F(7, 595) = 52.74, p < .001, R2 = .38). Table 5 shows the results of the mediation analyses. After controlling for gender and education along all three paths, the relationship between three out of four of the individual difference variables and gay rights activism was mediated by queer consciousness. That is, the direct relationships between liberalism, sexual orientation, and personal political salience and gay rights activism (Path C) were significantly reduced when queer consciousness was added to the regression equations, and the indirect effects were significantly different from 0.

Discussion
Consistent with the results of the previous two studies, the QCS showed good Cronbach's alpha reliability in a large sample of the general population. The QCS was strongly related to liberalism and personal political salience and strongly negatively related to a straight sexual orientation and authoritarianism. Finally, the QCS was positively correlated with knowledge about LGBTQ political history and both types of LGBTQ-related political activism. Similar to other measures of politicized group identities, queer consciousness mediated the relationships between several individual difference variables and political activism. This shows that the QCS operates psychologically much in the same way as other types of politicized identities, providing motivation for collective action participation.

General discussion
The purpose of this study was to introduce and validate a measure of queer consciousness, or politicized sexual orientation identity, by establishing  .15*** .14*** none Note. Numbers in the first three columns are standardized ß coefficients controlling for gender and education. Unstandardized regression coefficients are available from the first author. Unstandardized indirect effects (bootstrap SE) are reported in the last column. Mediation was determined by running the SPSS "PROCESS" macro created by Andrew Hayes and available online at http://processmacro.org/ index.html. If paths A, B, and C were significant and if the 95% confidence interval for the estimates of the indirect effects did not include 0, then mediation was significant at p < .05. The relationship between QCS and gay rights activism ß = .26***, N = 602. *** p < .001. reliability and validity in several samples of adults. Overall, the QCS showed excellent test-retest reliability over the course of 1 month, and good Cronbach's alpha reliability. This new measure is short, easy to use, and tied to the existing and extensive psychological literature on politicized collective identities. It can be used with people of varied sexual orientations. Because there are no existing measures of queer consciousness based on the psychological literature on politicized collective identities, we concentrated our validation efforts on establishing known-groups and predictive validity. All groups that were hypothesized to score higher on queer consciousness (queer sexual orientation; women; gay people out to their families, friends, and work; members of queer organizations; those who had attended a Gay Pride parade; and those who had taken queer studies courses) scored higher on queer consciousness than nonmembers of those groups. In addition, QCS scores were positively related to identifying along the non-straight spectrum of the Kinsey scale, political liberalism, finding personal meaning in political events generally (PPS) and LGBTQ-related events specifically, and participating in LGBTQ-related activism. QCS scores were negatively related to right-wing authoritarianism and social dominance orientation. In the third study, the QCS was also positively related to LGBTQ political knowledge and unrelated to general political knowledge, as hypothesized.
Thus the QCS has shown good reliability and validity and should be used to understand the correlates of politicizing identity around sexual orientation, including political activism. One of its strengths is that it can be used with sexual minorities as well as allies. We believe that this new measure of queer consciousness can contribute to our understanding of LGBTQ psychology in a number of different ways. We would like to mention two specifically. First of all, this measure can be used to examine intragroup dynamics within the LGBTQ community. For those who identify as a sexual or gender minority, what personality characteristics and life experiences differentiate those who develop high queer consciousness from those who do not? We already know that, similar to women who develop higher levels of feminist consciousness, experiences with sexual orientation discrimination are related to higher levels of gay rights activism (Swank & Fahs, 2013a, 2013b. What factors are related to politicizing identities around sexual orientation, and what factors differentiate those who become politically active for LGBTQ causes from those who do not? Because we were able to show that the QCS mediates relationships between some of these variables and gay rights activism, researchers should be able to use this new measure in conjunction with other personality and life experience measures to develop a deeper understanding of individual factors motivating gay rights activism. Another burgeoning area of research is understanding LGBTQ ally identification and activism. Over the past few years, multiple studies have been published that seek to explain ally participation in the LGBTQ movement (Duhigg, Rostosky, Gray, & Wimsatt, 2010;Fingerhut, 2011;Goldstein & Davis, 2010;Russell, 2011;Swank & Fahs, 2011). Because our measure was specifically designed to include allies, it can be used in future studies to look at the variables that are related to higher levels of queer consciousness and activism in allies. Experimental studies could be conducted as well, wherein independent variables were manipulated and their effects on levels of queer consciousness examined.
The QCS fits squarely in the tradition of both the political science and social psychological investigations of politicized collective identities. Because the QCS items were developed from long-used items adapted from the American National Election Studies (ANES; Miller et al., 1999), they should seem familiar to social science researchers. In addition, the five elements of group consciousness assessed in the QCS overlap completely with the dominant social psychological models of politicized collective identities (Duncan, 2012;van Zomeren et al., 2008). Because this measure is short, easy to use, and consistent with the dominant theoretical models of group consciousness and collective action, it should allow researchers to incorporate an individual differences approach into their existing analyses of factors contributing to motivation for collective action in both non-straight and straight people. Integrating personality and social psychological approaches in research on this motivation will allow us to develop a more nuanced and complete understanding of queer activism.