2 Discussion Posts Social Work
The Effects of General and Homophobic Victimization on Adolescents’ Psychosocial and Educational Concerns:
The Importance of Intersecting Identities and Parent Support
V. Paul Poteat, Ethan H. Mereish, and Craig D. DiGiovanni
Boston College
Brian W. Koenig K12 Associates, Middleton, Wisconsin
Many adolescents experience peer victimization, which often can be homophobic. Applying the minority stress model with attention to intersecting social identities, this study tested the effects of general and homophobic victimization on several educational outcomes through suicidality and school belonging among 15,923 adolescents in Grades 7 through 12 on account of their sexual orientation and race/ ethnicity. Parent support also was tested as a moderator of these effects. Homophobic victimization had different effects on suicidality across groups, indicating the importance of considering individuals’ multiple social identities. However, homophobic victimization had universal negative effects on school belonging for all groups. Nearly all indirect effects of general and homophobic victimization on reported grades, truancy, and importance of graduating were significant through suicidality and school belonging across groups. Parent support was most consistent in moderating the effects of general and homophobic victimization on suicidality for heterosexual White and racial/ethnic minority youth. In nearly all cases, it did not moderate the effects of general or homophobic victimization for lesbian, gay, bisexual, transgender, and questioning youth. Furthermore, in most cases, parent support did not moderate the effects of general or homophobic victimization on school belonging. Findings underscore the need for counseling psychologists to work with parents of all youth on ways to provide support to those who experience homophobic victimization. Furthermore, they highlight the need for counseling psychologists to be involved as social justice advocates in the passage and implementation of school policies that address homophobic bullying and other forms of bias-based bullying and harassment.
Keywords: bullying, sexual orientation, parent support, homophobia, minority stress
As counseling psychologists become more involved in schools (Kenny, Waldo, Warter, & Barton, 2002), the issue of bullying is highly relevant to our profession. Students who face bullying report multiple academic and mental health concerns (Juvonen, Nishina, & Graham, 2000; Schwartz, Gorman, Nakamoto, & Tob- lin, 2005). Indeed, mental health mediates the effects of victim- ization on academic outcomes such as grades (Schwartz et al., 2005). Furthermore, victimization is associated with perceptions of unsafe school climates (Chesir-Teran & Hughes, 2009). There is a growing focus on bullying that is prejudicial and biased (Russell, Sinclair, Poteat, & Koenig, in press). This can include many forms of bias (e.g., sexual orientation, disability, race). In this study, we focus on homophobic victimization. Students often are called
homophobic epithets when bullied or because of their actual or perceived lesbian, gay, bisexual, transgender, or questioning iden- tity (LGBTQ; Kosciw, Greytak, & Diaz, 2009; Poteat & Espelage, 2007; Swearer, Turner, Givens, & Pollack, 2008). Counseling psychologists are in a central position to address this issue, given their focus on discrimination and mental health disparities (Thompson & Neville, 1999) and on social justice and resilience (Arbona & Coleman, 2008).
The Minority Stress Model as a Framework for Studying Homophobic Victimization
The minority stress model (Meyer, 2003) offers a framework to understand homophobic victimization as a source of stress and why its effects may differ from general victimization and for different groups. Individuals from disadvantaged social positions experience unique stressors related to these positions, founded in oppressive structures, and can be distal (e.g., ho- mophobic victimization) or proximal (e.g., internalized ho- mophobia; Meyer, 2003). Victimization perceived as homopho- bic may have added effects beyond general victimization for LGBTQ youth because it further denigrates their identity and emphasizes their marginalized position. In line with these ar- guments, homophobic victimization is associated with mental health and academic concerns and risk behaviors for LGBTQ youth (D’Augelli, 2002; Russell et al., in press).
This article was published Online First August 22, 2011. V. Paul Poteat, Ethan H. Mereish, and Craig D. DiGiovanni, Department
of Counseling, Developmental, and Educational Psychology, Boston Col- lege; Brian W. Koenig, K12 Associates, Middleton, Wisconsin.
We extend our thanks to Lisa B. Spanierman for her helpful feedback on this manuscript. We are grateful to the schools that participated in the 2009 Dane County Youth Assessment and to the organizations that collaborated in and sponsored the development and completion of the project.
Correspondence concerning this article should be addressed to V. Paul Poteat, Department of Counseling, Developmental, and Educational Psy- chology, Boston College, Campion Hall 307, 140 Commonwealth Avenue, Chestnut Hill, MA 02467. E-mail: PoteatP@bc.edu
Journal of Counseling Psychology © 2011 American Psychological Association 2011, Vol. 58, No. 4, 597– 609 0022-0167/11/$12.00 DOI: 10.1037/a0025095
597
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
These basic concepts of minority stress have been supported empirically, yet nuanced questions remain. For instance, scholars have called for attention to mediators, moderators, and consider- ation of individuals’ multiple social identities (Meyer, 2010). These are issues of particular importance to counseling psycholo- gists and ones we seek to address in this study.
Homophobic Victimization Among LGBTQ and Heterosexual Youth
Multiple studies indicate that LGBTQ youth report greater vic- timization, distress, and poorer academic performance than het- erosexual youth (Berlan, Corliss, Field, Goodman, & Austin, 2010; Poteat, Aragon, Espelage, & Koenig, 2009). Although these dif- ferences reflect the high rates of discrimination faced by LGBTQ youth and their need for mental health resources, attention to how homophobic victimization predicts certain outcomes in similar or unique ways for heterosexual and LGBTQ youth could inform efforts to tailor interventions.
Several studies indicate that heterosexual youth experience ho- mophobic victimization, albeit at lower rates than LGBTQ youth. This research has focused on boys and their use of homophobic language to enforce gender norms, assert dominance, or bully others (Pascoe, 2007; Poteat & DiGiovanni, 2010; Poteat & Es- pelage, 2005). Girls also use and are called homophobic epithets in relation to bullying (Poteat & Espelage, 2005). For heterosexual boys and girls, being called homophobic epithets is associated with psychosocial concerns even when controlling for prior levels of psychosocial functioning (Poteat & Espelage, 2007). Also, hetero- sexual boys who are victimized and called homophobic epithets report greater distress than those victimized but not called these epithets (Swearer et al., 2008). These studies, however, have been among mostly White samples. In an ethnographic study, Pascoe (2007) noted that homophobic discourse was more prominent among White youth than youth of color. We seek to extend this literature by comparing these processes for LGBTQ and hetero- sexual youth, and among a more racially and ethnically diverse sample.
The effects of homophobic victimization may differ for hetero- sexual and sexual minority youth. The minority stress model (Meyer, 2003) suggests that homophobic victimization will have stronger associations with negative outcomes related to mental health, school belonging, or academic concerns for LGBTQ youth because they are members of the group on which this bias is predicated. To our knowledge, studies have not compared these processes for LGBTQ and heterosexual youth among racially and ethnically diverse samples.
Moving Toward Greater Complexity: The Intersection of Multiple Social Identities
Although studies have focused on LGBTQ adults of color (Huang et al., 2010), there remains a paucity of adolescent re- search. There is little attention to the effects of homophobic victimization across adolescents on account of sexual orientation and race. During this formative developmental period, these inter- sections likely shape how youth perceive and navigate their envi- ronment, and how they react to this victimization.
There are competing arguments and mixed empirical findings on how the intersection of identities relates to mental health. One line of empirical data supports a greater risk perspective, which suggests that LGBTQ persons of color show magnified risk com- pared with LGBTQ Whites because of discrimination on account of multiple marginalized positions (Balsam, Huang, Fieland, Si- moni, & Walters, 2004; Greene, 2000). Other findings are incon- sistent with this position and suggest that individuals from multiple minority backgrounds do not always report poorer health than those from fewer minority backgrounds (Kertzner, Meyer, Frost, & Stirratt, 2009). These perspectives argue that LGBTQ persons of color may have greater resilience than traditionally assumed (Bow- leg, Huang, Brooks, Black, & Burkholder, 2003; Meyer, 2010; Moradi, DeBlaere, & Huang, 2010). This may stem from their ability to draw on other cultural resources or to use coping skills learned from racial discrimination experiences. Although LGBTQ youth generally report greater mental health concerns than hetero- sexual youth, these differences are not always as evident for youth of color (Consolacion, Russell, & Sue, 2004). Support for this alternative perspective is limited. Thus, more research is needed to clarify these mixed findings.
Resilience in the Face of Victimization: General Parent Support as a Moderator
Counseling psychologists emphasize resilience as part of inter- vention and prevention (Arbona & Coleman, 2008). Although it remains critical to note disparities faced by LGBTQ youth, it is equally important to identify sources that promote resilience. Par- ent involvement and support are associated with mental health and academic achievement among heterosexual youth (O’Donnell, Schwab-Stone, & Muyeed, 2002; Wentzel, 1998). Emerging find- ings show that general parent support (i.e., not specific to victim- ization) moderates the effects of victimization for heterosexual youth on mental health (Davidson & Demaray, 2007; Stadler, Feifel, Rohrmann, Vermeiren, & Poustka, 2010). Research has not yet examined this for homophobic victimization.
Findings are inconclusive on parents as a source of resilience for LGBTQ youth. Many LGBTQ youth face or fear parent rejection because of their identity (Savin-Williams & Ream, 2003). Also, many rely more heavily on peers than parents (Muñoz-Plaza, Quinn, & Rounds, 2002). Thus, parent buffering effects may be weaker for these youth. This may further vary within the LGBTQ community. Because fewer LGBTQ youth of color disclose their sexual orientation to their parents than LGBTQ White youth (Grov, Bimbi, Nanı́n, & Parsons, 2006), they may hesitate to use their parents for support for this type of victimization.
Alternatively, parent effects may be comparable for heterosex- ual and LGBTQ youth. Parents vary in their approval of their child’s sexual minority identity (Willoughby, Malik, & Lindahl, 2006), and both heterosexual and LGBTQ youth who perceive parents as generally supportive may rely on them when they face victimization. Perceived family support, in general and specific to sexual orientation, is associated with positive LGBTQ mental health (Hershberger & D’Augelli, 1995; Sheets & Mohr, 2009). Also, LGBTQ youth of color may receive parent support related to racially biased victimization. This could generalize to moderate the effects of other forms of bias. This would be congruent with recent
598 POTEAT, MEREISH, DIGIOVANNI, AND KOENIG
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
arguments that sexual minorities of color demonstrate resilience in the face of multiple forms of adversity (Meyer, 2010).
The issue of parent support is critical for counseling psycholo- gists. If support does buffer the effects of homophobic victimiza- tion, counseling psychologists could consider parents as a source to promote resilience. If, however, they do not have this effect, counseling psychologists need to educate parents on how to sup- port their children who experience this victimization.
The Present Study
Despite the research on bullying, there are pronounced limita- tions in attention to homophobic victimization and its effects, factors that promote resilience, and how this applies across groups. Thus, we tested a model in which general and homophobic vic- timization, as well as parent support, predicted educational con- cerns (grades, truancy, graduation perceptions), through suicidality and school belonging. We then tested parent support as a moder- ator of these effects. We compared our models across groups on account of sexual orientation (heterosexual, LGBTQ) and race/ ethnicity (White, youth of color).
We hypothesized that both general and homophobic victimiza- tion would predict higher suicidality and lower school belonging across groups. However, we expected greater variability in ho- mophobic victimization effects. On the basis of findings among mostly White heterosexual and LGBTQ samples, we expected homophobic victimization would predict suicidality among these youth. On the basis of findings from Pascoe (2007) and those that suggest racial minorities can possess greater resilience (Moradi et al., 2010), we hypothesized that this effect would be less evident for youth of color. We hypothesized that homophobic victimization would predict lower school belonging for all youth, as these experi- ences foster hostile school climates (American Association of Uni- versity Women [AAUW], 2001; Chesir-Teran & Hughes, 2009).
We hypothesized that parent support would predict lower sui- cidality and higher school belonging. We expected these main effects to be comparable across groups because general parent support is critical to promote the overall health of all youth.
We believed that suicidality would predict lower grades, higher truancy, and lower reported importance of graduating, whereas school belonging would predict higher grades, lower truancy, and higher importance of graduating. Furthermore, we hypothesized that the indirect effects of general and homophobic victimization on these outcomes through suicidality and school belonging would be significant. Prior studies have identified the mediating role of mental health in relation to victimization and academic outcomes (Schwartz et al., 2005). Thus, we included suicidality as a mediator and indicator of mental health in our model because of its partic- ular focus in LGBTQ youth research (Coker, Austin, & Schuster, 2010) and because suicide remains the third leading cause of death for adolescents ages 15–24 (Centers for Disease Control and Prevention, 2006). We included school belonging because this would highlight the importance for counseling psychologists to provide not only individual counseling to students who are vic- timized but also the need to engage in broader efforts to change the school climate in an effort to address both the psychological and academic concerns of students. Prior work indicates the relation between victimization and perceptions of the school environment (Chesir-Teran & Hughes, 2009) and between school belonging and
grade point average (Murdock & Bolch, 2005). We extend this to test the mediating role of school belonging on multiple educational concerns of students who experience victimization.
We expected that general parent support moderating effects would be significant for heterosexual youth, but less consistently for LGBTQ youth because of the additional risks that LGBTQ youth could face in seeking parent support when victimized (e.g., unintentional disclosure of their sexual orientation or parent rejec- tion). Finally, we hypothesized that these moderating effects would be more consistent in relation to their attenuating effects on suicidality than school belonging, given that parents are more removed from the school context.
Method
Participants and procedure. Data were analyzed from the 2009 Dane County Youth Assessment (DCYA), a county-wide, multischool survey in Wisconsin (population approximately 490,000; 88.5% White, 4.7% African American, 4.6% Asian, 0.5% American Indian or Alaska Native, 1.7% identified with two or more groups, and 5.1% also identified as Hispanic; U.S. Census Bureau, 2010). The county ranges from small working farms to a large city. The middle and high school student population was 32,200 for the 2008 –2009 academic year, inclusive of public, private, and parochial schools. The DCYA is modeled on the Youth Risk Behavior Survey (YRBS; Centers for Disease Control and Prevention, 2009). It is a collaboration among middle and high schools and several organizations: Dane County Youth Commis- sion, United Way of Dane County, Dane County Human Services, and Public Health of Madison and Dane County.
All but two public schools in the county participated (n � 45) from late fall 2008 to early spring 2009. A total of 17,366 students in Grades 7–12 participated. Free or reduced-cost lunches ranged from 13% to 58% across schools. Because our comparisons focus on sexual orientation and race/ethnicity, data from students who did not respond to one or both items, or who were not identified as heterosexual or LGBTQ, were not analyzed.1 The final sample was 15,923 students, ages 10 –18 (M � 14.85, SD � 1.74); 94.2% were identified as heterosexual and 5.8% as LGBTQ; 76.4% identified as White, 6.5% as Black, 6.5% as bi/multiracial, 3.7% as Hispanic, 2.0% as non-Hmong Asian, 2.0% as Hmong, 1.1% as Native American, and 1.8% identified as “other.” There was an even balance of boys and girls (50% each) and grade levels.
The school districts approved a waiver of active parental con- sent and use of child assent. Students completed the anonymous electronic survey independently in school computer labs. Proctors monitored sessions to ensure confidentiality and answer questions. Students were given resources and contact information to access free mental health services, and were encouraged to use these services if they experienced any emotional discomfort in complet- ing the survey.
Measures. The DCYA assesses student health, attitudes, and behaviors. Some items comprise previously developed measures, whereas others are modeled from similar population-based youth surveys and that would have strong face validity. Traditionally, coun-
1 We tested for differences between students who were or were not included in our analyses and documented no significant differences based on our practical significance criteria (i.e., �p
2 � .01).
599GENERAL AND HOMOPHOBIC VICTIMIZATION
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
seling psychology research relies on preestablished multi-item scales to assess specific constructs, yet this is less practical or possible in comprehensive population-based surveys. However, in these cases, many of the items match those from other population-based surveys and preestablished measures, facilitating the comparison of our results to those from other population-based research.
Demographic information. Students completed items on race/ethnicity, gender, age, grade, and sexual orientation. The sexual orientation item was, “Do you identify yourself as any of the following? (Check all that apply)” Response options were Gay, Lesbian, Bisexual, Transgender, Questioning my sexual orienta- tion, or None of the above. Because of the options available and their many possible combinations, criteria were established to identify students as LGBTQ or heterosexual in consultation with several scholars with expertise in the field of sexual minority youth research. First, students who only marked None of the above were considered heterosexual. Second, students with any combination of responses to Gay, Lesbian, Bisexual, Transgender, or Question- ing and who did not mark None of the above were considered LGBTQ. This provided a conservative estimate of LGBTQ stu- dents. Those who did not respond or did not meet these criteria were not included in the analyses. Because of the numerous combinations of responses, it was impractical to separate or tabu- late participants meaningfully into subgroups (i.e., gay, lesbian, bisexual, transgender only). Because transgender individuals com- monly are considered members of the sexual minority community and experience sexual prejudice (Grossman & D’Augelli, 2006), those who checked this option, solely or in combination with LGBQ, were included in this group. Meyer, Schwartz, and Frost (2008) have argued that, although recognizing differences within racial/ethnic minority communities, a commonality is their expe- rience of racial/ethnic discrimination, and that examining broader comparisons between Whites and racial/ethnic minorities is still consistent with testing models related to minority stress. The present models were tested on account of individuals’ dichoto- mized sexual orientation (heterosexual, LGBTQ) and race/ ethnicity (White, youth of color).2
General victimization. The four-item University of Illinois Victimization scale (Espelage & Holt, 2001) assessed self-reported
experiences of victimization in the last 30 days: (a) Other students called me names; (b) I got hit and pushed by other students; (c) Other students picked on me; and, (d) Other students made fun of me. Response options were Never, 1 or 2 times, 3 or 4 times, 5 or 6 times, or 7 or more times (scored 0 – 4). Items do not assess for perceived bias. The scale has been widely used among diverse adolescent samples and with high internal reliability estimates documented (� � 88; Espelage & Holt, 2001). Scores converge with sociometric peer nominations and are associated with greater mental health concerns (Espelage & Holt, 2001). Higher average mean scores reflect more frequent victimization. The internal reliability for the items in this study was � � .86 –.91 across groups.
Homophobic victimization. One item assessed how frequently students experienced homophobic victimization. The item asked, “In the past twelve months, how often have you been bullied, threatened, or harassed about being perceived as gay, lesbian, or bisexual?” Response options were Never, Rarely, Sometimes, Often, or Very often (scored 0 – 4). Although the time span was not the same as for general victimization, we were not interested in comparing these rates directly, but rather examining their association and accounting for experiences of one to examine the unique effects of the other. This single item, slightly modified, has been used in other youth surveys and is associated with lower perceived school safety and higher depression (e.g., Preventing School Harassment Survey [California Safe Schools Coalition, 2008]).
2 To assess the general appropriateness of testing these models for youth of color more broadly, we performed several analyses. First, we examined the correlations for each group (i.e., Black, Hispanic, non-Hmong Asian, Hmong, Native American, bi/multiracial, and “other”), and they were comparable. Second, we tested our models for heterosexual and LGBTQ youth of color in which we excluded youth who identified as bi/multiracial or “other,” partly because “other” could be considered a very broad response option, and partly because bi/multiracial youth could have distinct experiences related their identity and could include both those who identify as White and with a racial/ethnic minority group and those who identify with more than one minority group. The results were nearly identical to our inclusive models of all non-White-identifying youth.
Table 1 Basic Group Differences on Account of Sexual Orientation and Race/Ethnicity
Measure
Heterosexual LGBTQ
M (SD) M (SD)
White (N � 11,557)
R/E Minority (N � 3,440)
White (N � 606)
R/E Minority (N � 320)
General victimization 1.65 (2.77) 1.67 (2.76) 2.64 (3.58) 2.71 (4.03) H. victimization 0.18 (0.60) 0.22 (0.71) 0.94 (1.25) 0.87 (1.34) Parent support 2.53 (0.51) 2.43 (0.59) 2.20 (0.72) 2.06 (0.89) School belonging 2.09 (0.60) 2.01 (0.63) 1.83 (0.70) 1.77 (0.77) Suicidality 0.07 (0.26) 0.12 (0.38) 0.39 (0.63) 0.44 (0.72) Grades 5.70 (1.44) 5.07 (1.62) 5.04 (1.76) 4.78 (1.85) Truancy 0.16 (0.55) 0.31 (0.77) 0.42 (0.93) 0.69 (1.23) Importance of graduating 2.82 (0.47) 2.76 (0.57) 2.70 (0.62) 2.50 (0.85)
Note. All measures represent average scale or item scores. LGBTQ � lesbian, gay, bisexual, transgender, questioning; R/E Minority � racial or ethnic minority students; H. victimization � homophobic victimization.
600 POTEAT, MEREISH, DIGIOVANNI, AND KOENIG
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
General parent support. Three items assessed general parent support and involvement: (a) My parents love and support me; (b) My parents encourage me to do my best; and, (c) My parents have talked to me about my future plans. Response options ranged from 0 (strongly disagree) to 3 (strongly agree). These are similar to items in other general parent support scales (e.g., Schaefer, 1965), and higher scores indicate greater general parent support. The internal reliability for the items in this study was � � .74 –.85 across groups.
Suicidality. Self-reported suicidal ideation (“During the past thirty days, have you seriously thought about killing yourself?” Response options were No, Yes but rarely, Yes, some of the time, and Yes, all of the time [scored 0 –3]) and suicide attempts (“Dur- ing the past twelve months have you attempted to kill yourself?” Response options were No, Yes one time, or Yes more than one time [scored 0 –2]) were assessed. The items were strongly asso- ciated (r � .44 –.65 across groups). Higher scores reflect higher suicidality.
School belonging. The four-item Psychological Sense of School Membership scale (Bosworth, Espelage, & Simon, 1999) assessed school belonging (e.g., “I feel like I belong at this school”). Response options ranged from 0 (strongly disagree) to 3 (strongly agree). Higher scores reflect greater school belonging, are associated with lower anxiety and depression, and internal reliability has ranged from � � .63 to .68 (Bosworth et al., 1999;
Poteat & Espelage, 2005). The internal reliability for the items in this study was � � .79 –.85 across groups.
Educational concerns. Three items assessed educational concerns. First, students reported their average grades. Re- sponse options, modeled from the YRBS, were Mostly A’s, Half A’s and Half B’s, Mostly B’s, Half B’s and half C’s, Mostly C’s, Half C’s and Half D’s, Mostly D’s, or Mostly below D’s (scored 0 –7; higher scores � higher average grades). Second, truancy was assessed: “During the last four weeks, how many days of school have you missed because you skipped (absent without permission)?” Response options were None, 1–2 days, 3–5 days, 6 –10 days, or More than 10 days (scored 0 – 4; higher scores � greater truancy). The third item was, “It is important to me that I graduate from school” (ranging from 0 [strongly disagree] to 3 [strongly agree]; higher scores � greater re- ported importance to graduate). Self-reported grades are com- monly used in research on the association between victimiza- tion and academic outcomes, although the size of the association tends to be smaller than in studies with access to students’ academic records (Nakamoto & Schwartz, 2010). To assess the general validity of the grades and truancy data, descriptive data were presented for these responses to the schools, and school officials were asked to compare these rates with their own records. School officials confirmed that the data
Table 2a Correlations Among the Measures for Heterosexual Youth by Race/Ethnicity
Measure 1 2 3 4 5 6 7 8
1. G. victim. — .39�� �.13�� �.16�� .22�� �.05�� .14�� �.11��
2. H. victim. .38�� — �.13�� �.11�� .13�� �.07�� .16�� �.21��
3. P. support �.12�� �.10�� — .27�� �.22�� .20�� �.14�� .27��
4. S. belong. �.19�� �.11�� .38�� — �.14�� .20�� �.14�� .42��
5. Suicide .20�� .14�� �.23�� �.17�� — �.08�� .18�� �.18��
6. Grades �.10�� �.09�� .28�� .25�� �.13�� — �.24�� .20��
7. Truancy .09�� .10�� �.19�� �.17�� .16�� �.26�� — �.18��
8. Graduate �.10�� �.12�� .30�� .40�� �.18�� .26�� �.20�� —
Note. Correlations above the diagonal are for racial and ethnic minority youth; correlations below the diagonal are for White youth. G. victim. � general victimization; H. victim. � homophobic victimization; P. Support � parent support and involvement; S. belong. � sense of school belonging; Suicide � suicidality; Graduate � importance of graduating. �� p � .01.
Table 2b Correlations Among the Measures for LGBTQ Youth by Race/Ethnicity
Measure 1 2 3 4 5 6 7 8
1. G. victim. — .53�� �.32�� �.33�� .33�� �.30�� .37�� �.45��
2. H. victim. .41�� — �.32�� �.33�� .27�� �.31�� .35�� �.29��
3. P. support �.12�� �.22�� — .29�� �.36�� .44�� �.38�� .36��
4. S. belong. �.12�� �.20�� .33�� — �.26�� .31�� �.29�� .64��
5. Suicide .26�� .32�� �.29�� �.24�� — �.31�� .24�� �.26��
6. Grades �.00 �.07 .22�� .20�� �.18�� — �.43�� .44��
7. Truancy .10� .16�� �.17�� �.23�� .31�� �.31�� — �.46��
8. Graduate �.15�� �.19�� .25�� .43�� �.18�� .31�� �.28�� —
Note. Correlations above the diagonal are for racial and ethnic minority youth; correlations below the diagonal are for White youth. LGBTQ � lesbian, gay, bisexual, transgender, questioning; G. victim. � general victimization; H. victim. � homophobic victimization; P. support � parent support and involvement; S. belong. � sense of school belonging; Suicide � suicidality; Graduate � importance of graduating. � p � .05. �� p � .01.
601GENERAL AND HOMOPHOBIC VICTIMIZATION
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
were comparable to their records and that neither appeared underreported.
Results
Preliminary analyses. We used multiple imputation in LISREL 8.80 (Jöreskog & Sörbom, 2006) to impute missing values with plausible simulated values based on the actual data. This is preferred over listwise deletion or mean substitution at the variable level, which can introduce statistical bias (Schafer & Graham, 2002). We performed the procedure at the item level, as we used items and not variables as indicators of latent factors. Missing data ranged from 2.1% to 9.3% (M � 5.8%, SD � 2.6%). To determine whether missing data varied on account of gender, sexual orientation, or race/ethnicity, we conducted a multivar- iate analysis of variance (MANOVA) with these demographic factors as independent variables and the items as dependent variables (coded as missing or nonmissing). The large sample size increased the likelihood of identifying significant differ- ences if they existed, even with very small effects. Thus, for all preliminary analyses, we relied on effect sizes to indicate the practical significance of any statistically significant differences. We considered differences with effect sizes of � p
2 � .01 as negligible. There were no differences for these demographic factors, and correlations between missing data and grade level were trivial (r � .00 –.03, p � .05, for all except two parent support items, which were nonsignificant).
Descriptive data are in Table 1, and correlations appear in Table 2. We conducted MANOVAs to test for gender, sexual orientation, and race/ethnicity differences on general and homophobic victim-
ization, general parent support, school belonging, suicidality, grades, truancy, and importance of graduating. There was only a significant effect for sexual orientation (Wilks’s � � .92), F(8, 15908) � 181.09, p � .001, �p
2 � .08. Follow-up analyses of variance (ANOVAs) revealed significant
sexual orientation differences on all measures, although trivial in some cases: general victimization, F(1, 15915) � 100.91, p � .001, �p
2 � .01; homophobic victimization, F(1, 15915) � 817.01, p � .001, �p
2 � .05; general parent support, F(1, 15915) � 303.85, p � .001, �p
2 � .02; suicidality, F(1, 15915) � 698.56, p � .001, �p 2 � .04;
school belonging, F(1, 15915) � 111.86, p � .001, �p 2 � .01; grades,
F(1, 15915) � 83.33, p � .001, �p 2 � .01; truancy, F(1, 15915) �
211.10, p � .001, �p 2 � .01, and importance of graduating, F(1,
15915) � 119.50, p � .001, �p 2 � .01. LGBTQ youth reported higher
scores on general and homophobic victimization, suicidality, and truancy. Heterosexual youth reported higher scores on general parent support, school belonging, grades, and importance of graduating.
Testing the latent model across groups. We used structural equation modeling (SEM) to test the direct effects of general and homophobic victimization on suicidality and school be-
3 We also included grade level and gender as covariates and documented no differences in the results. For parsimony, we excluded them from our models. Because of the nested data (i.e., students within schools), we tested multilevel models to determine the variability across schools. Intraclass correlations (ICCs) ranged from .00 to .04 on our variables (i.e., between 0% and 4% of the total variance in scores was between schools). ICCs below .10 are considered negligible (Lee, 2000), precluding the need in this case to model the nested data in our analyses.
Table 3 Fit Indices for Measurement and Latent Structural Models Across Groups
Model
Fit indices
IFI CFI NNFI SRMR RMSEA [90% CI]
Heterosexual White students Measurement model .98 .98 .97 .026 .040 [.038, .041] Foundational model .98 .98 .97 .035 .043 [.042, .045] P-Support � G-Victimization .95 .95 .94 .061 .054 [.053, .055] P-Support � H-Victimization .96 .96 .95 .052 .048 [.046, .049]
Heterosexual R/E minority students Measurement model .98 .98 .97 .025 .041 [.038, .044] Foundational model .97 .97 .96 .035 .046 [.044, .049] P-Support � G-Victimization .95 .95 .94 .068 .055 [.052, .057] P-Support � H-Victimization .96 .96 .95 .068 .052 [.050, .054]
LGBTQ White students Measurement model .98 .98 .96 .040 .048 [.040, .056] Foundational model .97 .97 .97 .044 .047 [.040, .055] P-Support � G-Victimization .95 .95 .94 .072 .058 [.052, .064] P-Support � H-Victimization .95 .95 .94 .075 .059 [.053, .065]
LGBTQ R/E minority students Measurement model .98 .98 .97 .042 .058 [.047, .070] Foundational model .97 .97 .96 .069 .069 [.059, .080] P-Support � G-Victimization .95 .95 .94 .160 .080 [.072, .088] P-Support � H-Victimization .95 .95 .94 .140 .078 [.070, .086]
Note. IFI � incremental fit index; CFI � comparative fit index; NNFI � nonnormed fit index; SRMR � standardized root-mean-square residual; RMSEA � root-mean-square error of approximation; CI � confidence interval; Foundational model � model with no moderating effects; P-Support � G-Victimization � model with parent support as a moderator of the effects of general victimization on suicidality and school belonging; P-Support � H-Victimization � model with parent support as a moderator of the effects of homophobic victimization on suicidality and school belonging; LGBTQ � lesbian, gay, bisexual, transgender, questioning; R/E Minority � racial or ethnic minority students.
602 POTEAT, MEREISH, DIGIOVANNI, AND KOENIG
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
longing, and their indirect effects on educational concerns.3 We used several fit indices to assess whether each model was a good representation of the data: root-mean-square error of approximation (RMSEA), standardized root-mean-square resid- ual (SRMR), incremental fit index (IFI), comparative fit index (CFI), and nonnormed fit index (NNFI). IFI, CFI, and NNFI values of at least .90 indicate an adequate fit, and values of .95 or higher indicate a strong fit (Kline, 1998); RMSEA and SRMR values of .08 or below are recommended (Hu & Bentler, 1999). It is inadvisable to rely on the chi-square fit statistic for model fit or model invariance because of our very large sample size (Bollen, 1989; Chen, 2007; Cheung & Rensvold, 2002). Although we subscribe to the common practice of referring to certain variables as predictors that have an effect on other variables, because our data are cross-sectional we do not intend to infer causality from these data.
We first tested the measurement model. Each item was an indicator of its respective latent factor and was constrained to load only on that factor. The correlations among the factors were free to be estimated, but the measurement errors were not allowed to correlate. The model was a good fit across the four groups (see Table 3). We then tested its metric invariance across groups (Marsh, 1994). First, we ran the measurement model for all groups simultaneously and allowed the factor loadings to be estimated for each group with no constraints. The global fit indices were a baseline comparison for our next step, in which we constrained the loadings to be equal across groups. The unconstrained model was a good fit (IFI � .96; CFI � .96; NNFI � .95; RMSEA � .065, 90% CIs [.064, .066]). As expected, the con- strained model was a slightly poorer fit (IFI � .93; CFI � .93; NNFI � .92; RMSEA � .089, 90% CIs [.088, .090]). The changes in the fit indices were small and suggested that the measurement model could be considered to demonstrate metric invariance. We tested the invariance of factor variance, and this constraint led to no decrease in model fit from prior constraints for metric invariance. Factor loadings for the multi-item latent variables ranged from .56 to .86 (general victimization), from .55 to .84 (parent support), from .65 to .76 (school belonging), and were .62 and .82 for the two suicidality items.
We tested the mediation model (see Figures 1a– d without moder- ating effects), which was a good fit across all groups (see Table 3). We base our group comparisons on unstandardized path estimates (Aiken & West, 1991). As hypothesized, general victimization effects on suicidality and, with the exception of LGBTQ White youth, school belonging were significant and comparable across groups.4 Ho- mophobic victimization effects were more varied, as hypothesized. The effect on suicidality was nonsignificant for heterosexual and LGBTQ youth of color, but was significant for heterosexual White youth and LGBTQ White youth. Also as hypothesized, general parent support predicted lower suicidality and higher school belonging. As hypothesized, suicidality and school belonging predicted educational concerns on all three indices with similar effects across groups. Finally, the total indirect effects of general victimization, homophobic victimization, and general parent support on educational concerns through suicidality and school belonging were significant in nearly all cases (see Table 4).
Testing the moderating effects of general parent support. We tested two models in which general parent support moderated the effects of general and homophobic victimization on suicidality and school belonging. We followed the recommendations of Marsh, Wen, and Hau (2004) and created interaction parcels
among the observed variables of these factors to serve as indicators of the latent moderating effect. There were four parcels for the general victimization moderator and three parcels for the ho- mophobic victimization moderator. The models were a good fit for each group, with the exception of the SRMR for LGBTQ youth of color (see Table 3). As hypothesized, the general parent support moderating effect was evident for general victimization on suicid- ality with the exception of LGBTQ youth of color. In contrast, with the exception of LGBTQ youth of color, parent support did not buffer the effect of general victimization on school belonging. Finally, general parent support only buffered the effect of ho- mophobic victimization on suicidality for heterosexual Whites and youth of color, only weakly buffered its effect on school belonging for heterosexual youth of color, and was not significant for LGBTQ youth.
Discussion
Peer victimization has serious consequences, yet little attention is given to homophobic victimization and even less to factors that promote resilience among youth who experience it. Addressing these limitations, we documented how homophobic victimization predicted psychosocial and educational concerns on account of students’ sexual orientation and race or ethnicity. Furthermore, we
4 We did not test for significant differences in the size of each specific path coefficient for each group in comparison to the size of the coefficients for each of the other three groups. The number of comparisons yielded would ultimately detract from the focus of this study on the overarching patterns across groups. We did, however, test the overall invariance of path coefficients across groups. As expected, the fit indices displayed a poor fit for the mediation model (IFI � .89; CFI � .89; NNFI � .89; RMSEA � .18) and general victimization moderation model (IFI � .90; CFI � .90; NNFI � .90; RMSEA � .11), and was not positive definite for the homophobic victimization moderation model. Finally, we tested the overall invariance of path coefficients across groups on account of single social identities (i.e., models comparing White and racial/ethnic minority hetero- sexual youth; White and racial/ethnic minority LGBTQ youth; heterosex- ual and LGBTQ White youth; and heterosexual and LGBTQ youth of color). For the basic mediation model, fit indices for the invariance models were acceptable across these comparisons (IFI, CFI, and NNFI indices ranged from .94 to .97, RMSEA ranged from .05 to .07). For the general victimization moderation model, fit indices for the invariance models were poor to marginally acceptable across these comparisons (IFI, CFI, and NNFI indices ranged from .86 to .93, RMSEA ranged from .07 to .15). The models were not positive definite for the homophobic victimization mod- eration models. These results for the basic mediation model suggest that consideration of greater complexity in individuals’ social identities (i.e., multiple as opposed to single social identities) captures greater nuance and more distinct differences in the path coefficients across groups in our model. At the same time, they suggest caution in overinterpreting differ- ences in the size of significant path coefficients across groups; however, comparisons between significant versus nonsignificant path coefficients across groups remain warranted, and we focus on these distinctions in the discussion of our findings. These results for the moderation models provide support for differences in these effects across groups, even when consid- ering individuals’ single social identities in isolation as opposed to indi- viduals’ multiple social identities in combination.
603GENERAL AND HOMOPHOBIC VICTIMIZATION
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
Figure 1 (opposite).
604 POTEAT, MEREISH, DIGIOVANNI, AND KOENIG
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
identified important but limited buffering effects of general parent support.
Universal and distinct effects of general and homophobic victimization across groups. For the most part, general victim- ization had similar direct effects on suicidality and school belong- ing and indirect effects on educational concerns across groups. This is in line with other studies (Juvonen et al., 2000; Schwartz et al., 2005). Our findings are novel by noting the added effects of homophobic victimization even when accounting for overall vic- timization. Studies have documented associations between ho- mophobic victimization and mental health concerns for heterosex- ual and LGBTQ youth (D’Augelli, 2002; Poteat & Espelage, 2007; Swearer et al., 2008) but have not tested which groups are most affected. Homophobic victimization predicted suicidality for LG- BTQ and heterosexual White youth. The effect for LGBTQ White youth may have been evident over and above that of general victimization because this is a blatant denigration of their identity, a key contributor to minority stress (Meyer, 2003). The effect for heterosexual White youth is congruent with other associations between homophobic victimization and mental health concerns among mostly White heterosexual samples (Poteat & Espelage, 2007; Swearer et al., 2008) and reports that youth view this as especially stigmatizing (AAUW, 2001). This link to suicidality underscores the need for counseling psychologists to become part of school-based prevention programs (Kenny et al., 2002), espe- cially those that address the interplay between bullying and prej- udice for all students. Counseling psychologists should collaborate
on prevention and intervention efforts and address issues of prej- udice that underlie the victimization that many youth experience.
Homophobic victimization did not predict suicidality for het- erosexual and LGBTQ racial and ethnic minorities when account- ing for general victimization experiences, perhaps because ho- mophobic banter is more disparaging among White adolescents than racial and ethnic minority adolescents (Pascoe, 2007). We note, however, that these factors were correlated when not con- trolling for general victimization. There is strong evidence for the association between racial discrimination and mental health and academic concerns for heterosexual racial and ethnic minority youth (Fisher, Wallace, & Fenton, 2000; Flores, Tschann, Dimas, Pasch, & de Groat, 2010). Thus, future research should continue to consider homophobic and racially based victimization as predic- tors of psychosocial and educational outcomes. Specific to LG- BTQ youth of color, this finding highlights the need for attention to the intersection of social identities and how this adds complexity to the minority stress model (Meyer, 2010). In this case, our findings were more supportive of the resilience perspective (Bow- leg et al., 2003; Moradi et al., 2010). LGBTQ youth of color may have drawn on coping strategies learned from racial discrimina- tion. Use of adaptive identity processes (Pittinsky, Shih, & Am- bady, 1999) also could explain our pattern of findings: LGBTQ youth of color might reorient their identities to identify more strongly with their race or ethnicity than sexual orientation when they experience homophobic victimization.
Figure 1 (opposite). (a) Model results for heterosexual White youth. All coefficients are significant at p � .01, with the exception of dashed paths that were not significant. Coefficients for all main effects are reported from the mediation model without moderators. In nearly all cases, these coefficients were identical when the moderating effects were included. (b). Model results for heterosexual racial/ethnic minority youth. All coefficients are significant at p � .01 unless otherwise noted (coefficients with an asterisk denote p � .05). Dashed paths were not significant. (c) Model results for lesbian, gay, bisexual, transgender, and questioning (LGBTQ) White youth. All coefficients are significant at p � .01 unless otherwise noted (coefficients with an asterisk denote p � .05). Dashed paths were not significant. (d) Model results for LGBTQ racial/ethnic minority youth. All coefficients are significant at p � .01 unless otherwise noted (coefficients with an asterisk denote p � .05). Dashed paths were not significant. P-Support � Parent-Support.
Table 4 Total Indirect Effects Estimates Through Suicidality and School Belonging
Total indirect effects on each educational concern
Heterosexual LGBTQ
White Racial/ethnic minority White Racial/ethnic minority
Effect of general victimization on: Grades �.24�� [�.28, �.20] �.14�� [�.20, �.09] �.10 [�.21, �.01] �.35�� [�.59, �.16] Truancy .08�� [.07, .10] .10�� [.07, .14] .08� [.02, .18] .21�� [.09, .37] Importance of graduating �.10�� [�.12, �.09] �.10�� [�.13, �.07] �.03 [�.08, .02] �.21�� [�.31, �.07]
Effect of homophobic victimization on: Grades �.05�� [�.07, �.03] �.04�� [�.07, �.01] �.10�� [�.16, �.05] �.11 [�.20, �.01] Truancy .02�� [.01, .03] .02 [.00, .04] .08�� [.04, .13] .07 [.01, .13] Importance of graduating �.02�� [�.03, �.01] �.03� [�.05, �.01] �.04� [�.07, �.02] �.08 [�.14, �.01]
Effect of parent support on: Grades .67�� [.61, .74] .34�� [.27, .42] .46�� [.32, .66] .53� [.28, .94] Truancy �.21�� [�.24, �.18] �.17�� [�.22, �.13] �.31�� [�.43, �.22] �.31� [�.56, �.15] Importance of graduating .31�� [.28, .33] .24�� [.20, .28] .24�� [.17, .34] .27� [.12, .39]
Note. We used bootstrapping procedures conducted on 500 generated samples to test the significance of the indirect effects. The bias-corrected bootstrap confidence intervals are reported in this table. Values are unstandardized total indirect effects through the mediating variables of suicidality and school belonging. LGBTQ � lesbian, gay, bisexual, transgender, questioning. � p � .05. �� p � .01.
605GENERAL AND HOMOPHOBIC VICTIMIZATION
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
Counseling psychologists should consider more closely how LGBTQ youth of color process these experiences. This could integrate an intersectionality framework in which identity is not the additive sum of social positions; rather, it is multiple social identities influencing each other with synergistic effects (Warner, 2008). Knowledge of how these youth experience resilience is critical given the mental health and educational disparities on account of race and sexual orientation (Coker et al., 2010; Matti- son & Aber, 2007). Counseling psychologists may draw on strengths, resources, and coping strategies from youths’ experi- ences of racial discrimination in ways that could also buffer the effects of homophobic victimization.
Our models suggest that across sexual orientation and racial or ethnic background, youth who experience homophobic victimiza- tion feel a lower sense of school belonging, which is associated with more frequently skipping school, poorer academic perfor- mance, and feeling it is less important to graduate. These effects are in addition to those for general victimization and support findings that LGBTQ youth are marginalized at school (Kosciw et al., 2009) and that victimization is associated with academic con- cerns (Russell et al., in press). We further extend this to hetero- sexual youth, who may fear exclusion for being perceived as a sexual minority. Thus, homophobic victimization must be consid- ered in bullying research because failure to assess for bias in bullying masks its added consequences. Our findings accentuate the need for counseling psychologists to address homophobic victimization in a way that is inclusive of all students.
Support through adversity: Working with parents to foster youth resilience. Parent support attenuates the effects of vic- timization (Davidson & Demaray, 2007; Stadler et al., 2010), which we replicated for heterosexual youth in that it moderated the effect of victimization (general and homophobic) on suicidality. It did not, however, moderate the effect of homophobic victimization on suicidality for LGBTQ youth, and it only moderated the effect of general victimization on suicidality for LGBTQ White youth. In most cases, parent support did not moderate the effects of general or homophobic victimization on school belonging.
There are several potential reasons for the series of nonsignifi- cant moderating effects of general parent support on suicidality for LGBTQ youth. Many face or fear parent rejection (Savin-Williams & Ream, 2003). For self-protection, youth who are not out may not seek parent support for victimization, particularly when it is ho- mophobic, because they may risk exposure of their sexual orien- tation. Also, many rely on peers more than parents for support (Muñoz-Plaza et al., 2002). Notably, however, general parent support directly predicted lower suicidality and greater school belonging and indirectly predicted lower educational concerns. Similar effects have been documented in other studies (Hersh- berger & D’Augelli, 1995; Sheets & Mohr, 2009) and suggest that parents do promote the overall health of LGBTQ youth. When they are victimized, however, general parent support does not consis- tently attenuate these effects. General parent support may be an insufficient buffer of homophobic victimization effects; more spe- cific forms may be necessary (e.g., affirmation of sexual minority identity). Parents also may feel less equipped to provide support in these instances. Counseling psychologists cannot assume that all parents will automatically serve as a resource to promote resilience for victimized youth, even when youth perceive them as highly supportive in general. Counseling psychologists thus need to con-
sider how to include parents in intervention and social justice efforts (e.g., working with parents to provide specific affirming support to their children).
These moderating effects were more consistent for heterosexual youth, building on past findings (Davidson & Demaray, 2007; Stadler et al., 2010). Heterosexual youth may cope with victim- ization differently than LGBTQ youth. Because they do not iden- tify as LGBTQ, fear of rejection from parents may not be present as it is for LGBTQ youth, even when this victimization is ho- mophobic. In line with minority stress (Meyer, 2003), because heterosexual youth do not identify as LGBTQ, they may not internalize the victimization as a reflection of their identity and oppressive societal structures.
Several findings underscore the need to examine minority stress from an intersectionality framework. The moderating effect of general parent support on victimization (general and homophobic) and suicidality was most evident for heterosexual youth of color relative to other groups. These youth may draw on parent-taught coping strategies for racial discrimination. Also, they might rely on their parents for support during instances of racial discrimination and thus may rely on them for support in these instances as well. The effect of parent support on general victimization and suicid- ality may have been significant for LGBTQ White youth but not youth of color because fewer LGBTQ youth of color may have been out to their parents (Grov et al., 2006). These youth may have feared rejection if their sexual orientation was inadvertently dis- closed in the process. Yet, parent support moderated general victimization effects on school belonging for LGBTQ youth of color. In these cases, they may rely more on their parents, poten- tially due to barriers in schools (e.g., a lack of academic resources, racist school climates). Counseling psychologists should work with schools and organizations (e.g., Gay-Straight Alliances) to address any of these possible barriers for LGBTQ youth of color.
It is important to note the largely absent moderating effects of general parent support on victimization and school belonging. This may be possible because we measured general and not school- based parent support or involvement (e.g., parent–teacher confer- ence attendance). Also, some parents may feel they lack a voice in school systems, or may be unsure of the procedures to report bullying to administrators. Whereas parents may have a stronger buffering effect in relation to their child’s mental health than school belonging, teacher support may have a stronger buffering effect on school belonging, as they can actively shape their class- room climate.
Given the absent or weak moderating effects of parent support on homophobic victimization across groups, counseling psychol- ogists should work with parents of all youth on ways to provide support for those who experience homophobic victimization. This may include training on how to approach administrators who may hesitate to address issues around sexual orientation or homophobic victimization. Similarly, counseling psychologists and schools should work collaboratively on policy reform that strengthens antibullying programs and policies and their protection of margin- alized youth.
Strengths, limitations, and future directions. Our findings identify complex associations among victimization and risk out- comes and provide knowledge of how homophobic victimization may affect youth in similar and unique ways. Our attention to the intersection of sexual orientation and race/ethnicity adds nuance to
606 POTEAT, MEREISH, DIGIOVANNI, AND KOENIG
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
this. Many studies rely on mostly White LGBTQ samples. Our findings indicate that programs should be informed from an inter- sectional lens. Also, they provide a rigorous test of the added effects of homophobic bias by accounting for experiences of overall victimization. In addition, we focused on resilience, which is often overlooked in the victimization literature. Finally, the large sample size, multiple schools, and representativeness of LGBTQ participants strengthen the external validity of our findings and avoid many of the common limitations in LGBTQ youth research (e.g., small samples, snowball or other selective sampling proce- dures) that impact the validity of findings, especially for LGBTQ youth of color.
We also acknowledge several limitations. As with other population-based data sets, not all measures were preestablished assessments. Our items were similar to existing scales, yet multiple items within established assessments could better capture these constructs. Furthermore, studies should consider the cultural va- lidity of these indicators. Broadening the construct of parent sup- port to represent family or extended family may be more appro- priate for some racial or ethnic groups. Also, we could not consider the intersection of students’ sexual orientation and race/ethnicity on the basis of specific racial/ethnic minority groups. This would have yielded inadequate sample sizes for some groups. Qualitative studies could provide a particularly rich understanding of these experiences for underrepresented groups of youth (e.g., Native American youth). Additionally, our data were from a single time point, and longitudinal data are needed in this area. Finally, we assessed general parent support, which is a typical approach in the victimization literature. Nevertheless, specific types of parent support (e.g., toward their child’s sexual orientation) also should be examined.
Building on these findings, studies might identify other ways in which homophobic victimization and other forms of bias affect those in dominant and marginalized groups. Also, research should identify other sources of resilience among youth who are victim- ized. Furthermore, research should attend to other intersections of social identities in continual efforts to refine and tailor interven- tions for youth who experience invisibility due to their multiple minority group statuses. Finally, research should evaluate school- based prevention and intervention programs designed to counter homophobic bullying and victimization.
Counseling psychologists in schools are in a central position to address the serious issue of bullying. Their expertise and training in multicultural counseling places them at the forefront to address the neglected issue of homophobic victimization. Antibullying legislation and school policies continue to be debated on account of their inclusivity of protection for marginalized youth. Our findings strengthen the case for these policies to address homopho- bic bullying. With their emphasis on social justice, counseling psychologists must be involved in the creation, passage, and im- plementation of policies to ensure that schools are welcoming for all students.
References
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage.
American Association of University Women. (2001). Hostile hallways: Bullying, teasing, and sexual harassment in schools. Washington, DC: Author.
Arbona, C., & Coleman, N. (2008). Risk and resilience. In S. D. Brown & R. W. Lent (Eds.), Handbook of counseling psychology (4th ed., pp. 483– 499). Hoboken, NJ: John Wiley & Sons.
Balsam, K. F., Huang, B., Fieland, K. C., Simoni, J. M., & Walters, K. L. (2004). Culture, trauma, and wellness: A comparison of heterosexual and lesbian, gay, bisexual, and two-spirit Native Americans. Cultural Diversity and Ethnic Minority Psychology, 10, 287–301. doi:10.1037/ 1099-9809.10.3.287
Berlan, E. D., Corliss, H. L., Field, A. E., Goodman, E., & Austin, S. B. (2010). Sexual orientation and bullying among adolescents in the grow- ing up today study. Journal of Adolescent Health, 46, 366 –371. doi: 10.1016/j.jadohealth.2009.10.015
Bollen, K. A. (1989). Structural equations with latent variables. Oxford, England: John Wiley & Sons.
Bosworth, K., Espelage, D. L., & Simon, T. R. (1999). Factors associated with bullying behavior in middle school students. Journal of Early Adolescence, 19, 341–362. doi:10.1177/0272431699019003003
Bowleg, L., Huang, J., Brooks, K., Black, A., & Burkholder, G. (2003). Triple jeopardy and beyond: Multiple minority stress and resilience among Black lesbians. Journal of Lesbian Studies, 7, 87–108. doi: 10.1300/J155v07n04_06
California Safe Schools Coalition. (2008). 2008 Preventing School Ha- rassment Survey. Retrieved from http://www.casafeschools.org/ PSH_2008.pdf
Centers for Disease Control and Prevention. (2006). Welcome to WISQARS web-based injury statistics query and reporting system. Retrieved from http://www.cdc.gov/injury/wisqars/index.html
Centers for Disease Control and Prevention. (2009). Youth Risk Behavior Survey. Retrieved from http://www.cdc.gov/yrbss
Chen, F. F. (2007). Sensitivity to goodness of fit indexes to lack of measurement invariance. Structural Equation Modeling, 14, 464 –504. doi:10.1080/10705510701301834
Chesir-Teran, D., & Hughes, D. (2009). Heterosexism in high school and victimization among lesbian, gay, bisexual, and questioning youth. Jour- nal of Youth and Adolescence, 38, 963–975. doi:10.1007/s10964-008- 9364-x
Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Mod- eling, 9, 233–255. doi:10.1207/S15328007SEM0902_5
Coker, T. R., Austin, S. B., & Schuster, M. A. (2010). The health and health care of lesbian, gay, and bisexual adolescents. Annual Review of Public Health, 31, 457– 477. doi:10.1146/annurev.publhealth .012809.103636
Consolacion, T. B., Russell, S. T., & Sue, S. (2004). Sex, race/ethnicity, and romantic attractions: Multiple minority status adolescents and men- tal health. Cultural Diversity and Ethnic Minority Psychology, 10, 200 – 214. doi:10.1037/1099-9809.10.3.200
D’Augelli, A. R. (2002). Mental health problems among lesbian, gay, and bisexual youths ages 14 to 21. Clinical Child Psychology and Psychia- try, 7, 433– 456. doi:10.1177/1359104502007003010
Davidson, L. M., & Demaray, M. K. (2007). Social support as a moderator between victimization and internalizing-externalizing distress from bul- lying. School Psychology Review, 36, 383– 405.
Espelage, D. L., & Holt, M. K. (2001). Bullying and victimization during early adolescence: Peer influences and psychosocial correlates. Journal of Emotional Abuse, 2, 123–142. doi:10.1300/J135v02n02_08
Fisher, C. B., Wallace, S. A., & Fenton, R. E. (2000). Discrimination distress during adolescence. Journal of Youth and Adolescence, 29, 679 – 695. doi:10.1023/A:1026455906512
Flores, E., Tschann, J. M., Dimas, J. M., Pasch, L. A., & de Groat, C. L. (2010). Perceived racial/ethnic discrimination, posttraumatic stress symptoms, and health risk behaviors among Mexican American adoles- cents. Journal of Counseling Psychology, 57, 264 –273. doi:10.1037/ a0020026
607GENERAL AND HOMOPHOBIC VICTIMIZATION
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
Greene, B. (2000). African American lesbian and bisexual women. Journal of Social Issues, 56, 239 –249. doi:10.1111/0022-4537.00163
Grossman, A. H., & D’Augelli, A. R. (2006). Transgender youth: Invisible and vulnerable. Journal of Homosexuality, 51, 111–128. doi:10.1300/ J082v51n01_06
Grov, C., Bimbi, D. S., Nanı́n, J. E., & Parsons, J. T. (2006). Race, ethnicity, gender, and generational factors associated with the coming- out process among gay, lesbian, and bisexual individuals. Journal of Sex Research, 43, 115–121. doi:10.1080/00224490609552306
Hershberger, S. L., & D’Augelli, A. R. (1995). The impact of victimization on the mental health and suicidality of lesbian, gay, and bisexual youths. Developmental Psychology, 31, 65–74. doi:10.1037/0012-1649.31.1.65
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struc- tural Equation Modeling, 6, 1–55. doi:10.1080/10705519909540118
Huang, Y., Brewster, M. E., Moradi, B., Goodman, M. B., Wiseman, M. C., & Martin, A. (2010). Content analysis of literature about LGB people of color: 1998 –2007. The Counseling Psychologist, 38, 363–396. doi:10.1177/0011000009335255
Jöreskog, K., & Sörbom, D. (2006). LISREL 8.8. Lincolnwood, IL: Sci- entific Software International.
Juvonen, J., Nishina, A., & Graham, S. (2000). Peer harassment, psycho- logical adjustment, and school functioning in early adolescence. Journal of Educational Psychology, 92, 349 –359. doi:10.1037/0022- 0663.92.2.349
Kenny, M. E., Waldo, M., Warter, E. H., & Barton, C. (2002). School- linked prevention: Theory, science, and practice for enhancing the lives of children and youth. The Counseling Psychologist, 30, 726 –748. doi:10.1177/0011000002305004
Kertzner, R. M., Meyer, I. H., Frost, D. M., & Stirratt, M. J. (2009). Social and psychological well-being in lesbians, gay men, and bisexuals: The effects of race, gender, age, and sexual identity. American Journal of Orthopsychiatry, 79, 500 –510. doi:10.1037/a0016848
Kline, R. B. (1998). Principles and practice of structural equation mod- eling. New York, NY: Guilford Press.
Kosciw, J. G., Greytak, E. A., & Diaz, E. M. (2009). Who, what, where, when, and why: Demographic and ecological factors contributing to hostile school climates for lesbian, gay, bisexual, and transgender youth. Journal of Youth and Adolescence, 38, 976 –988. doi:10.1007/s10964- 009-9412-1
Lee, V. E. (2000). Using hierarchical linear modeling to study social contexts: The case of school effects. Educational Psychologist, 35, 125–141. doi:10.1207/S15326985EP3502_6
Marsh, H. W. (1994). Confirmatory factor analysis models of factorial invariance: A multifaceted approach. Structural Equation Modeling, 1, 5–34. doi:10.1080/10705519409539960
Marsh, H. W., Wen, Z., & Hau, K. (2004). Structural equation models of latent interactions: Evaluation of alternative estimation strategies and indicator construction. Psychological Methods, 9, 275–300. doi: 10.1037/1082-989X.9.3.275
Mattison, E., & Aber, M. S. (2007). Closing the achievement gap: The association of racial climate with achievement and behavioral outcomes. American Journal of Community Psychology, 40, 1–12. doi:10.1007/ s10464-007-9128-x
Meyer, I. H. (2003). Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: Conceptual issues and research evidence. Psychological Bulletin, 129, 674 – 697. doi:10.1037/0033- 2909.129.5.674
Meyer, I. H. (2010). Identity, stress, and resilience in lesbians, gay men, and bisexuals of color. The Counseling Psychologist, 38, 442– 454. doi:10.1177/0011000009351601
Meyer, I. H., Schwartz, S., & Frost, D. M. (2008). Social patterning of stress and coping: Does disadvantaged social statuses confer more stress
and fewer coping resources? Social Science and Medicine, 67, 368 –379. doi:10.1016/j.socscimed.2008.03.012
Moradi, B., DeBlaere, C., & Huang, Y. (2010). Centralizing the experi- ences of LGB people of color in counseling psychology. The Counseling Psychologist, 38, 322–330. doi:10.1177/0011000008330832
Muñoz-Plaza, C., Quinn, S. C., & Rounds, K. A. (2002). Lesbian, gay, bisexual and transgender students: Perceived social support in the high school environment. The High School Journal, 85, 52– 63. doi:10.1353/ hsj.2002.0011
Murdock, T. B., & Bolch, M. B. (2005). Risk and protective factors for poor school adjustment in lesbian, gay, and bisexual (LGB) high school youth: Variable and person-centered analyses. Psychology in the Schools, 42, 159 –172. doi:10.1002/pits.20054
Nakamoto, J., & Schwartz, D. (2010). Is peer victimization associated with academic achievement? A meta-analytic review. Social Development, 19, 221–242. doi:10.1111/j.1467-9507.2009.00539.x
O’Donnell, D. A., Schwab-Stone, M. E., & Muyeed, A. Z. (2002). Multi- dimensional resilience in urban children exposed to community vio- lence. Child Development, 73, 1265–1282. doi:10.1111/1467- 8624.00471
Pascoe, C. J. (2007). Dude, you’re a fag: Masculinity and sexuality in high school. Los Angeles: University of California Press.
Pittinsky, T. L., Shih, M., & Ambady, N. (1999). Identity adaptiveness: Affect across multiple identities. Journal of Social Issues, 55, 503–518. doi:10.1111/0022-4537.00130
Poteat, V. P., Aragon, S. R., Espelage, D. L., & Koenig, B. W. (2009). Psychosocial concerns of sexual minority youth: Complexity and cau- tion in group differences. Journal of Consulting and Clinical Psychol- ogy, 77, 196 –201. doi:10.1037/a0014158
Poteat, V. P., & DiGiovanni, C. D. (2010). When biased language use is associated with bullying and dominance: The moderating effect of prejudice. Journal of Youth and Adolescence, 39, 1123–1133. doi: 10.1007/s10964-010-9565-y
Poteat, V. P., & Espelage, D. L. (2005). Exploring the relation between bullying and homophobic verbal content: The Homophobic Content Agent Target (HCAT) scale. Violence and Victims, 20, 513–528. doi: 10.1891/vivi.2005.20.5.513
Poteat, V. P., & Espelage, D. L. (2007). Predicting psychosocial conse- quences of homophobic victimization in middle school students. Journal of Early Adolescence, 27, 175–191. doi:10.1177/0272431606294839
Russell, S. T., Sinclair, K. O., Poteat, V. P., & Koenig, B. W. (in press). Adolescent health and harassment based on discriminatory bias. Amer- ican Journal of Public Health.
Savin-Williams, R. C., & Ream, G. L. (2003). Sex variations in the disclosure to parents of same-sex attractions. Journal of Family Psy- chology, 17, 429 – 438. doi:10.1037/0893-3200.17.3.429
Schaefer, E. S. (1965). Children’s reports of parental behavior: An inven- tory. Child Development, 36, 413– 424. doi:10.2307/1126465
Schafer, J. L., & Graham, J. W. (2002). Missing data: Our view of the state of the art. Psychological Methods, 7, 147–177. doi:10.1037/1082- 989X.7.2.147
Schwartz, D., Gorman, A. H., Nakamoto, J., & Toblin, R. L. (2005). Victim- ization in the peer group and children’s academic functioning. Journal of Educational Psychology, 97, 425– 435. doi:10.1037/0022-0663.97.3.425
Sheets, R. L., & Mohr, J. J. (2009). Perceived social support from friends and family and psychosocial functioning in bisexual young adult college students. Journal of Counseling Psychology, 56, 152–163. doi:10.1037/ 0022-0167.56.1.152
Stadler, C., Feifel, J., Rohrmann, S., Vermeiren, R., & Poustka, F. (2010). Peer-victimization and mental health problems in adolescents: Are pa- rental and school support protective? Child Psychiatry and Human Development, 41, 371–386. doi:10.1007/s10578-010-0174-5
Swearer, S. M., Turner, R. K., Givens, J. E., & Pollack, W. S. (2008).
608 POTEAT, MEREISH, DIGIOVANNI, AND KOENIG
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
“You’re so gay!”: Do different forms of bullying matter for adolescent males? School Psychology Review, 37, 160 –173.
Thompson, C. E., & Neville, H. A. (1999). Racism, mental health, and mental health practice. The Counseling Psychologist, 27, 155–223. doi: 10.1177/0011000099272001
U.S. Census Bureau. (2010). State and county quickfacts. Retrieved from http://quickfacts.census.gov/qfd/states/55/55025.html
Warner, L. R. (2008). A best practices guide to intersectional approaches in psychological research. Sex Roles, 59, 454 – 463. doi:10.1007/s11199- 008-9504-5
Wentzel, K. R. (1998). Social relationships and motivation in middle
school: The role of parents, teachers, and peers. Journal of Educational Psychology, 90, 202–209. doi:10.1037/0022-0663.90.2.202
Willoughby, B. L. B., Malik, N. M., & Lindahl, K. M. (2006). Parental reactions to their sons’ sexual orientation disclosures: The role of family cohesion, adaptability, and parenting style. Psychology of Men & Mas- culinity, 7, 14 –26. doi:10.1037/1524-9220.7.1.14
Received October 20, 2010 Revision received June 24, 2011
Accepted June 29, 2011 �
609GENERAL AND HOMOPHOBIC VICTIMIZATION
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .