Colleen A. Halliday^1*, Katherine A. Perkins², Claudia A. Salazar³, Carla Kmett Danielson⁴

¹Division of Global and Community Health, Department of Psychiatry & Behavioral Science, Medical University of South Carolina
²Yvonne and Schuyler Moore Child Development Research Center, College of Education, University of South Carolina
³Group on Diversity Affairs, College of Medicine Dean’s Office, Medical University of South Carolina  
⁴National Crime Victims and Treatment Center, Department of Psychiatry & Behavioral Science, Medical University of South Carolina

*Corresponding Author: Colleen A. Halliday, Division of Global and Community Health, Department of Psychiatry & Behavioral Science, Medical University of South Carolina

Abstract
African American adolescent girls have evidenced higher levels of disruptive behavior than girls from other ethnic groups. However, most research focused on understanding disparities in these outcomes has been conducted without consideration of gender or has focused exclusively on boys. Yet, prior research suggests that anger and aggression are less gender-typed in African American youth than they are among youth from other ethnic backgrounds. The purpose of this preliminary investigation was to examine the extent to which ethnic-specific gender schemas about anger mediated the relationship between ethnicity and girls’ disruptive behavior. Participants were 66 middle school girls (24.1 % African American, 46.3 % European American; M_age= 12.06). They completed measures of ethnic-specific gender schemas about anger, reactive and instrumental aggression, and classroom disruptive behavior. Results indicated that relative to girls from other ethnic groups, African American girls had higher levels of reactive aggression and classroom disruptive behavior, both of which are rooted in anger.  In contrast, no ethnic difference was found for instrumental aggression, which is not connected to anger.  Ethnic-specific gender schemas about anger at least partially accounted for ethnic differences in reactive aggression and classroom disruptive behavior.  Findings highlight the importance of examining gender schemas specific to ethnicity as factors in ethnic disparities in behavioral outcomes among adolescent girls.

Keywords: African American, Girls, Ethnic/Racial Differences, Disruptive Behavior, Aggression, Gender Schemas

Evidence suggests that aggression is less gender-typed in AA youth than it is among youth from other ethnic backgrounds [4]. According to gender schema theories [2,12], children develop schemas about their own gender that include beliefs about gender-appropriate behavior within their culture. Functionally, gender schemas are internalized gender stereotypes. Children view themselves, make decisions, and regulate their behaviors based on gender schemas. Because gender stereotypes vary over time and across ethnic groups, cultural context determines which behaviors are considered gender-typed. In dominant U.S. culture, displays of anger and physical aggression are generally stereotyped as the purview of boys and not girls [3]. However, AA adolescents have evidenced lower endorsement of conventionally held gender schemas of aggression than youth from other ethnic backgrounds [4]. Furthermore, AA girls and women are stereotyped as being hostile, argumentative, and aggressive [17]. As such, AA girls might be more likely than girls from other ethnic groups to have gender schemas that include anger and aggression as appropriate for girls.  What remains unknown is the extent that which ethnic-specific gender schemas about anger and aggression account for ethnic disparities in girls’ disruptive behavior.

The current study is a preliminary examination of the extent to which ethnic-specific gender schemas mediate the relationship between ethnicity and various forms of girls’ disruptive behavior. Middle school girls responded to questionnaires assessing reactive aggression, which is rooted in anger, and instrumental aggression, which is unrelated to anger [11]. They also completed self-report measures of disruptive classroom behavior and ethnic-specific gender schemas about anger. Established predictors of disruptive behavior, deviant peer affiliation, parental demandingness, and neighborhood income also were assessed to serve as covariates. We hypothesized that AA girls would evince more anger-related disruptive behavior than girls of other ethnic groups and that ethnic-specific gender schemas about anger would account for these differences. Specifically, we expected ethnic-specific gender schemas about anger to mediate the relationship between ethnicity and reactive aggression, but no such effects on instrumental aggression. Because anger at unfair treatment has been proposed as a factor underlying disruptive behavior at school more for AA girls than for girls of other ethnic groups [14], we also expected ethnic disparities in disruptive classroom behavior to be mediated by ethnic-specific gender schemas about anger.

Measures
Ethnicity. Participants reported their ethnicity. Responses were dichotomized to reflect AA ethnicity (1 = AA, 0 = non-AA).
Aggression. Aggression was assessed with Little and colleagues’ (2003)  aggression questionnaire adapted to assess behavioral frequency. Two scores were computed: reactive aggression (α = .91; 12 items), which indexed angry aggressive defensive responses to peer transgressions, and instrumental aggression (α = .96; 12 items), which assessed aggressive behaviors enacted to achieve the desired outcome.
Disruptive classroom behavior.  Disruptive classroom behavior was assessed using the Patterns for Adaptive Learning Scales [13] disruptive behavior subscale (5 items; α = .80).
Ethnic-specific gender schemas about anger. The Stereotypic Roles for Black Women [17] was developed and validated for adult women to assess the adoption of four stereotypes of Black women. Townsend and colleagues (2010) simplified this instrument for use with Black adolescent girls. Participants in the current study completed the anger subscale, adapted to be applicable to girls of any ethnic group by replacing the words “Black girls” with “girls of my ethnic group” (9 items; α  = .87).
Parent demandingness. The demandingness subscale of the Parenting Styles Inventory-II [5] captured youth reports of parents’ setting and enforcing rules (5 items, α =.77).
Neighborhood income. The Federal Financial Institutions Examination Council's (2017) income classification for the census tract of each participant’s residence indexed neighborhood income.
Deviant peer affiliation. Ettekal and Ladd’s (2015) school conduct problems subscale and [6] friend antisocial behavior scale were combined to measure peer deviance (9 items; α = .94).

Table 1: Ethnic differences in variables (* p < .05, ** p < .01)

Reactive Aggression
The overall model for predicting reactive aggression was significant, F(5, 59) = 10.92, p < .001, R² = .693. The indirect effect of ethnicity on reactive aggression through ethnic-specific gender schemas about anger was significant at 95 % CI [.0420, .5168]. Thus, ethnic-specific gender schemas about anger mediated the relationship between ethnicity and reactive aggression. Results are presented in Table 2 and illustrated in Figure 1a.

Disruptive Classroom Behavior
The overall model for predicting disruptive classroom behavior was significant, F(5, 59) = 4.87, p = .0009 R² = .2921. The indirect effect of ethnicity on classroom disruptive behavior, through ethnic-specific gender schemas about anger was significant, 95 % CI [.0164, 0.1786]. Thus, ethnic-specific gender schemas about anger mediated the relationship between ethnicity and disruptive classroom behavior. Results are presented in Table 2 and illustrated in Figure 1b.

Figures (a, b): Path model depicting perceived ethnic-specific gender schemas as a mediator of the effect of ethnicity on (a) reactive aggression (b) disruptive classroom behavior, controlling for peer deviance, parent structure, and neighborhood income. Path values reflect standardized coefficients.

* p < .05, ** p < .01

Table 2: Regression results for classroom disruptive behavior and reactive aggression (* p < .05, ** p < .01)

Conclusion
Literature about gender schema for girls and women is based primarily on European American samples without consideration of ethnicity [3]. Because gender schemas are learned in cultural contexts, additional research is needed to the role that understanding ethnic-specific gender schemas play in disruptive behavior outcomes for African American girls. Research is also needed to improve the understanding of processes related to the internalization of negative ethnic stereotypes that are specific to one’s gender. More generally, this study highlights the need for future studies examining ethnic-specific gender factors in ethnic disparities in girls’ outcomes to inform efforts to reduce disparities in these outcomes.

Acknowledgments
This research was funded by grant K12HD055885 (PI: J. E. McGinty)  from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) and supported by the South Carolina Clinical & Translational Research Institute at the Medical University of South Carolina, through NIH grants UL1 RR029882 and UL1 TR000062. Preparation of this manuscript was supported in part by grant R01MD013812 (PI C. A. Halliday) and grant K24DA039783 (PI: C. K. Danielson). Views expressed in this article do not necessarily reflect those of the funding agency acknowledged.

Note: Supplemental analyses including only Black/AA and EA participants revealed a similar pattern of findings across all outcomes.