ADOLESCENT GIRLS' COGNITIVE APPRAISALS OF COPING RESPONSES TO SEXUAL HARASSMENT
The research was supported by grants to the first author from the University of California–Santa Cruz, Academic Senate and Social Sciences Division, and by grants to the second author from the University of California–Los Angeles, Center for the Study of Women. The students, teachers, and staff at the following sites are thanked for the help with the study: Expanding Your Horizons at University of California–Santa Cruz, Girl Scouts of Monterey Bay, River Trail Middle School, Los Angeles High School, North Hollywood High School, Rogers High School, and Westchester High School. Also, the authors thank Carly Friedman for helpful suggestions during survey construction, Agnieszka Spatzier for data coordination, and Bren Michelle Chasse and Nicole Nunez for data entry.
Abstract
Peer sexual harassment is a stressor for many girls in middle and high school. Prior research indicates that approach strategies (seeking support or confronting) are generally more effective than avoidance strategies in alleviating stress. However, the deployment of effective coping behaviors depends partly on how individuals evaluate different options (i.e., cognitive appraisal). The present study tested sociocultural (ethnicity, parents’ education), interpersonal (perceived support from peers, mother, and father), developmental (age, perspective taking), and individual (self-esteem, feminist self-identification) factors as predictors of girls’ cognitive appraisals of coping responses to sexual harassment. The sample comprised 304 girls (M age = 15.5 years, range = 14 to 18 years) from diverse socioeconomic and ethnic backgrounds who reported having previously experienced sexual harassment (e.g., unwanted sexual comments or actions). Cognitive appraisals of coping were based on the reported likelihood of confronting, seeking help, or using avoidance in response to sexual harassment. Regression analyses indicated that feminist identity, self-esteem, perspective taking, perceived support, and parents’ education were variously related to appraisals of different responses.
Several studies have documented the common occurrence of peer sexual harassment among youths in middle and high schools in North America and many other parts of the world (see Leaper & Robnett, 2011, for a review). Various forms of sexual harassment—including unwanted sexual comments, attention, or touching—are often directed at both genders; however, sexual harassment tends to be more distressing for girls than boys. For example, sexual harassment is more likely to be related to lowered esteem, adjustment difficulties, and academic declines among girls than boys (American Association of University Women [AAUW], 2001, 2011; Chiodo, Wolfe, Crossk, Hughes, & Jaffe, 2009; Goldstein, Malanchuk, Davis-Kean, & Eccles, 2007; Timmerman, 2005). These negative effects are not inevitable. The distress associated with sexual harassment can be mitigated through the use of effective coping strategies (e.g., Gruber & Fineran, 2007). Deploying an effective coping strategy, however, depends partly on how individuals evaluate different responses. These evaluations are known as cognitive appraisals (Lazarus & Folkman, 1984).
In the present study, we tested possible predictors of adolescent girls’ cognitive appraisals of coping strategies in response to sexual harassment. Whereas researchers have examined similar processes in adult samples (described later), there has been limited work examining coping responses to sexism during adolescence. This dearth of research is particularly important because adolescent sexual harassment is unique from adults’ experiences in several ways. First, it typically takes place within a school setting. Thus, negative social experiences are occurring in a setting intended for education. Negative emotions associated with perceived sexual harassment could impede the learning process and make girls feel uncomfortable at school. Second, adolescents are still developing attitudes about themselves and others, developing academically, and developing their attitudes about their bodies and sexuality. Identifying factors related to girls’ adaptive coping may prove helpful to parents, teachers, and school counselors who are concerned with improving youths’ resilience in the face of sexual harassment and other interpersonal stresses.
Stress and Coping Model
According to Lazarus and Folkman's (1984) model of stress and coping, effective coping with a stressful situation depends initially on the person's cognitive appraisal of the stressful event and subsequently on the type of behavioral strategy used. Cognitive appraisal involves recognizing the source of stress and evaluating the relative benefits and costs of different responses. In turn, cognitive appraisal of one's options may affect the kind of behavioral strategy that is deployed. In the present research, we examined possible predictors of girls’ cognitive appraisals of different coping responses to sexual harassment.
A general distinction is often made between approach (or engagement) and avoidance (or disengagement) coping strategies (Compas, Connor-Smith, Saltzman, Thomsen, & Wadsworth, 2001; Lazarus, 1999; Lazarus & Folkman, 1984). Approach strategies are oriented toward addressing the threat. These might include confronting the source of the stress (e.g., challenging someone about sexist behavior) or seeking social support (e.g., talking to someone about what happened). People may seek others to gain emotional reassurance, to clarify their understanding of the situation, or to get advice. In contrast, avoidance strategies are oriented away from the threat, such as downplaying or ignoring the event. In general, research indicates that approach strategies are more effective than avoidance strategies in reducing stress in adolescents (e.g., Compas et al., 2001) and adults (e.g., Lazarus, 1999).
Research has highlighted the relevance of the stress and coping model to women's experiences with sexism (e.g., Ayres, Friedman, & Leaper, 2009; Cortina & Wasti, 2005; Foster, 2009; Kaiser & Miller, 2004). For example, one study found that women's cognitive appraisals predicted their likelihood of confronting during a recent experience with sexism (Kaiser & Miller, 2004). Another study indicated that women's passive coping in response to gender discrimination was related to decreased well-being over time (Foster, 2009). To our knowledge, however, there has been no research examining adolescent girls’ cognitive appraisals of coping responses to sexual harassment. Given the pervasiveness of sexual harassment in school during adolescence (AAUW, 2011), research focusing on adolescent girls is warranted.
Possible Assets and Risks Related to Girls’ Cognitive Appraisals of Coping Responses to Sexual Harassment
Our study investigated possible assets and risks related to adolescent girls’ cognitive appraisals of coping responses to sexual harassment. Assets are factors that strengthen individuals’ abilities to cope effectively with stressors, whereas risks are factors that can undermine coping (Masten & Coatsworth, 1998). As reviewed next, possible assets or risks at multiple system levels were considered in relation to girls’ cognitive appraisals of coping. We focused on a sample of adolescent girls who reported having previously experienced sexual harassment. Our reasoning was that girls who recognized having experienced sexual harassment (compared with girls who did not) would have a more realistic expectation of how they would respond to these incidents, given they have already faced them.
Two theoretical models guided our conceptualization of the kinds of factors that might predict girls’ cognitive appraisals of coping in response to sexual harassment. First, we used Bronfenbrenner and Morris's (2006) ecological-systems model of development, which proposes that children's development transpires within the context of multiple intersecting systems. These begin with the individual child interacting in particular microsystems, such as the family and peer group. Microsystems are embedded in the larger macrosystem that includes the prevailing practices, values, and socioeconomic opportunities of the society. Cutting across all of the levels is the dimension of time, or the chronosystem, which includes the child's developmental age and the historical period in the society. Thus, in the present study, we took into account potential assets or risks related to the girls’ (1) sociocultural background (macrosystem), (2) interpersonal contexts (microsystems), (3) developmental level (chronosystem), and (4) individual characteristics.
Our conceptualization was further guided by Brown and Bigler's (2005) developmental model of children's perception of discrimination. According to the authors’ framework, children's perception and understanding of discrimination is influenced by a combination of situational variables (e.g., perceived social support), cognitive–developmental factors (e.g., perspective-taking ability), and individual variables (e.g., self-esteem). The Brown and Bigler model is pertinent to the present study, given we examined girls’ cognitive appraisals of responses to sexual harassment, which is one form of gender discrimination. Prior research with adults indicates that women's coping responses to sexism depend on a combination of interpersonal and individual factors (e.g., Ayres et al., 2009; Swim & Hyers, 1999). To our knowledge, however, there is no pertinent research examining age-related factors associated with coping in response to sexual harassment.
In summary, based on the previously reviewed models, we investigated four sets of assets/risks in relation to girls’ cognitive appraisals of coping responses to sexual harassment. These four components include the sociocultural context, interpersonal contexts, developmental factors, and individual characteristics. As reviewed in the following sections, each of these four levels was expected to contribute independently to girls’ responses to sexual harassment.
Sociocultural Context: Ethnicity and Parents’ Education
We considered girls’ ethnic background and their parents’ education as potentially relevant sociocultural predictors of their responses to sexism. As articulated in feminist standpoint theory (Basow & Rubin, 1999; Stewart & McDermott, 2004), ethnicity and gender often intersect in complicated ways. Conflicting hypotheses indicate that ethnic-minority girls either (a) can be more attuned to experiences of sexism because of their experiences with racism or ethnic prejudice (Kane, 2000; Wasti & Cortina, 2002) or (b) can be less aware of sexism because ethnicity is more central to their identity (Levin, Sinclair, Veniegas, & Taylor, 2002; Turner & Brown, 2007). By extension, responses to coping with sexual harassment may vary depending on individuals’ ethnic background (e.g., Buchanan & West, 2010; Wasti & Cortina, 2002). However, little research on sexual harassment during adolescence has considered girls’ ethnic backgrounds.
In the present study, half of our sample comprised girls from Latin American backgrounds. Traditional gender-role patterns may be more common in Latin American than in European American families (Raffaelli & Ontai, 2004; Villarruel, 1998). Further, many Latina girls and young women have been shown to endorse traditional gender roles—perhaps reflective of a cultural ideal of marianismo (traditional feminine virtues)—and to reject notions of feminism that may be viewed as disparaging boys and men (Manago, Brown, & Leaper, 2009; Robnett, Anderson, & Hunter, 2012). Thus, because they may tolerate or accept traditional gender roles, Latina adolescents may be less likely than European American girls to endorse active coping responses to sexual harassment.
Another sociocultural variable in our model was the education level of the girls’ parents. Researchers often use parents’ education as a proxy variable for socioeconomic status (SES), given the correlation between education level and income (e.g., Finkelstein, Kubzansky, Capitman, & Goodman, 2007). This factor is important to consider when testing for possible variations based on participants’ ethnic background to avoid confounding ethnicity and SES (Buchanan & West, 2010; Leaper & Valin, 1996). Moreover, high or low SES may be an asset or a risk, respectively, that affects coping. Repeated stressors may undermine one's sense of control and thereby lead to more avoidance coping. For instance, Finkelstein et al. (2007) found that those adolescents from lower SES families (based on parents’ education) experienced more stress in their daily lives and were more likely to report avoidance coping. A similar pattern may extend to girls’ coping with sexism. That is, girls from lower SES families may be less likely than girls from higher SES families to use approach coping (and more likely to use avoidance strategies) in response to sexual harassment.
Interpersonal Contexts: Perceived Social Support
The interpersonal context is the second component in our model. In particular, we considered perceived social support in relation to girls’ cognitive appraisals of sexual harassment coping. Social support can be an important asset in adolescents’ coping responses to stressful events (e.g., Frydenberg & Lewis, 1993). Also, perceived social support is positively related to girls’ likelihood of recognizing discrimination when it occurs (Brown & Bigler, 2005). By extension, perceived social support may provide social resources that increase the likelihood of favoring approach coping strategies (and decrease the likelihood of endorsing avoidance coping) in response to sexual harassment. In support of this prediction, prior research indicated that coping with sexism was more effective among women who perceived social support (Moradi & Funderburk, 2006).
Parents and peers are generally considered the most important sources of support in adolescents’ lives (Harter, 2003). Accordingly, we hypothesized that their perceived support would be positively related to girls’ reported likelihood of approach coping (and negatively related to avoidance coping) in response to sexual harassment. With regard to perceived parent support, we separately assessed girls’ perceptions of mother support and father support. Mothers and fathers tend to play different roles as sources of social support for adolescent girls. For example, studies find that girls tend to be more disclosing with mothers than fathers (e.g., Smetana, Metzger, Gettman, & Campione-Barr, 2006). Hence, perceived mother support may be stronger than perceived father support as an asset in girls’ coping with sexism.
Perceived support from friends may be especially important when choosing to confront sexual harassment in school settings. Prior studies indicate that peers are the most common perpetrators of sexism (e.g., AAUW, 2001; Leaper & Brown, 2008). Friends may be an important resource following peer sexual harassment. That is, girls may view confronting a classmate more positively if they believe their friends will back them up. Conversely, girls may be more likely to favor avoidance if they see themselves lacking peer support.
Developmental Factors: Age and Perspective Taking
Girls’ reported experiences with sexual harassment tend to increase with age during adolescence (e.g., Leaper & Brown, 2008; McMaster, Connolly, Pepler, & Craig, 2002). In addition, developmental changes in cognitive and socioemotional functioning may affect girls’ awareness of sexism (Brown & Bigler, 2004). Thus, girls may evaluate responses to sexism differently during the course of adolescence as they develop social skills and become more aware of sexism. These developmental changes may act as assets that facilitate coping (Masten & Coatsworth, 1998). Accordingly, we included age as a factor in our analyses.
Perspective taking was a second developmental factor in our model. Age-related increases in sociocognitive skill may be related to girls’ responses to sexual harassment. During adolescence, individuals become increasingly capable of taking a third-person perspective, whereby they can view events as others might look at them (Selman, 1989). This ability is believed to be a necessary (but not sufficient) condition for youth to recognize the unfairness inherent in discrimination (Brown & Bigler, 2005). In a somewhat similar manner, adaptive coping with stressful events depends on recognizing the source of the stressor. In support of this contention, individuals with higher levels of perspective taking are more likely to select constructive solutions to conflict (Selman, 1989). Therefore, we predicted that perspective taking would be an asset that is positively related to girls’ likelihood of favoring approach strategies (confronting or seeking help) rather than avoidance responses to sexism.
Individual Characteristics: Self-Esteem and Feminist Identity
Self-esteem and feminist identity are two individual factors included in the last component of our model. Prior research indicates that experiences with sexual harassment can undermine girls’ self-esteem (e.g., AAUW, 2001) and contribute to adjustment difficulties (e.g., DuBois, Burk-Braxton, Swenson, Tevendale, & Hardesty, 2002). In other studies, self-esteem predicted reported use of approach coping in response to (a) racial discrimination among adolescents (Phinney & Chavira, 1995) and (b) sexist events among women (Gervais, Hillard, & Vescio, 2010). In an analogous manner, we hypothesized that self-esteem would be related to girls’ evaluations of different coping strategies. Self-esteem may give individuals the confidence to confront a perpetrator of sexism. Conversely, low self-esteem may increase the likelihood that girls favor avoidance strategies in response to discrimination (Major & Schmader, 1998).
The other individual asset in our model was girls’ feminist self-identification. Adolescence may be an important period of feminist awareness and feminist identity development for many girls (Zucker & Stewart, 2007). Prior research indicates that adolescent girls were more likely to recognize gender discrimination when they reported having learned about feminism from teachers, parents, or other sources (Leaper & Brown, 2008). Relevant research is more limited on feminist identity or beliefs in relation to coping responses to sexism. Two relevant studies of undergraduate women found that feminist self-identification increased the likelihood that women positively appraised confronting sexist acts (Leaper & Arias, 2011) or reported previously confronting an actual sexist event (Ayres et al., 2009). Thus, to the extent that sexist acts violate their values, we hypothesized that girls with a feminist identity would be more likely to seek redress of the injustice through approach coping.
Summary
To our knowledge, no prior study has investigated potential assets and risks related to adolescent girls’ cognitive appraisals of coping responses to sexual harassment. We used the ecological-systems model and research on discrimination and coping to investigate four sets of predictors: sociocultural background (ethnicity and parents’ education), interpersonal contexts (perceived support from mother, father, and peers), developmental variables (age, perspective taking), and individual characteristics (self-esteem, feminist identity). Each component was expected to independently contribute to the prediction of girls’ cognitive appraisals of three coping strategies: confronting, seeking help, and avoidance.
Method
Participants
Participants were part of a larger sample from a study of adolescent girls’ identity and experiences with sexism (Leaper & Brown, 2008). In the present set of analyses, the sample comprised 304 girls ranging in age from 14 to 18 years (M = 15. 58, SD = 1.20) who reported having experienced sexual harassment during the past year (defined later). They were recruited from middle and high schools (Grades 8 through 12) and extracurricular programs (Expanding Your Horizons and a Girl Scouts camp) in southern California (77%), northern California (17%), and Georgia (6%). Self-reported ethnic backgrounds were 53% Latina, 19% White European American, 9% African American, 9% Asian American/Pacific Islander, and 12% other ethnic backgrounds. Self-reported family situations were 54% living with both a mother and a father, 3% living with a mother and a father in separate houses, 29% living with mostly or only a mother, 4% living with mostly or only a father, and 11% citing other arrangements.
The sample was also varied in terms of SES based on parents’ education level. Of mothers, 56% had no more than a high school diploma, 31% had attended some college or had a bachelor's degree, and 13% had completed some graduate work or had an advanced degree. Of fathers, 56% had no more than a high school diploma, 26% had attended some college or had a bachelor's degree, and 15% had completed some graduate work or had an advanced degree.
Procedure
The study was described as a survey about “what it means to be a girl.” Consent was obtained from parents and participants. Volunteers completed several survey measures in their classroom or similar settings. They were instructed that their completion of the survey was optional and they could stop at any time. Questions were asked about participants’ demographic background, family, and peer life, as well as their self-concepts and views regarding school achievement, gender roles, and sexism. In general, girls did not indicate difficulty reading or completing the survey at any age. As explained later in the section describing our measure of sexual harassment, we limited our sample in the present study to girls who reported having experienced sexual harassment.
Measures
The numbers of respondents, descriptive statistics, and alpha coefficients for the internal reliability of each measure are indicated in Table 1. All of the measures had acceptable levels of reliability (i.e., minimum α = .70).
| 1. | 2. | 3. | 4. | 5. | 6. | 7. | 8. | 9. | 10. | 11. | 12. | 13. | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1. Latina | – | ||||||||||||
| 2. European American | – | – | |||||||||||
| 3. Parent Education | −.72*** | .46*** | – | ||||||||||
| 4. Mother Support | −.05 | .05 | .05 | – | |||||||||
| 5. Father Support | −.14* | .12† | .22*** | .25*** | – | ||||||||
| 6. Peer Support | −.13* | .08 | .13* | .33*** | .23*** | – | |||||||
| 7. Age | .29*** | −.22*** | −.32*** | .07 | −.10 | −.09 | – | ||||||
| 8. Perspective Taking | −.13* | .05 | .14* | .00 | .06 | .19** | −.03 | – | |||||
| 9. Esteem | −.18** | .11† | .23*** | .42*** | .21** | .43*** | −.01 | .12* | – | ||||
| 10. Feminist Identity | .11† | −.05 | −.07 | −.02 | .10 | .00 | .02 | .15** | .10† | – | |||
| 11. Confront | −.06 | .04 | .13* | .09 | .03 | .26*** | −.01 | .33*** | .20** | .12* | – | ||
| 12. Seek Help | .09 | −.05 | −.06 | .21** | .16* | .20*** | .03 | .22*** | .21*** | .24*** | .46*** | – | |
| 13. Avoidance | .05 | −.11* | −.03 | −.09 | .11 | −.05 | −.09 | −.05 | −.16** | −.08 | −.24*** | −.28*** | – |
| N | 304 | 304 | 302 | 254 | 242 | 304 | 304 | 303 | 304 | 303 | 304 | 304 | 304 |
| M | – | – | 3.36 | 2.97 | 2.62 | 3.11 | 15.58 | 3.17 | 3.02 | 2.49 | 3.05 | 2.98 | 2.71 |
| SD | – | – | 1.63 | .75 | .83 | .50 | 1.20 | .64 | .66 | .92 | .86 | .89 | .83 |
| α | – | – | .85 | .79 | .80 | .77 | – | .70 | .86 | .76 | .70 | .85 | .72 |
Note
- Latina (1 = Latina, 0 = other ethnic group) and European American (1 = White European American, 0 = other ethnic group) were dummy coded. Parents’ education was an ordinal measure ranging from 1 = elementary school to 7 = graduate degree. The remaining scale items were rated on a 4-point scale and then averaged for each measure. When appropriate, alpha coefficients of scale reliability are indicated.
- *p < .05. **p < .01. ***p < .001. †< .10.
Personal Experiences with Sexual Harassment
Items from Klonoff and Landrine's (1995) Schedule of Sexist Events were adapted for use with adolescents to assess girls’ experiences with sexual harassment. Participants were asked to rate whether they had experienced the following using a 4-point scale (1 = no, 2 = yes—once or twice, 3 = yes—a few times, 4 = yes—several times): (a) called you a nasty or demeaning name related to being a girl, (b) was teased about your appearance related to being a girl, (c) was told an embarrassing/mean joke about girls/women, (d) received inappropriate or unwanted romantic attention by a male, or (e) received unwanted physical contact by a male. These five items also reflect the most commonly reported forms of sexual harassment cited in a recent national survey of sexual harassment during adolescence (AAUW, 2011).
The ratings for the five items were averaged to create a composite measure of experiences with sexual harassment. For the present set of analyses, girls were selected if they had a mean rating of 2.0 (2 = yes—once or twice) or greater, which indicated they reported experiencing each form of sexual harassment at least once or twice on average. Out of an initial sample of 510 girls, 203 girls had mean ratings between 1.0 and 1.8, and 307 girls had mean ratings between 2.0 and 4.0. To focus on girls who were aware of having experienced sexual harassment, the latter group was selected for the present analyses. Of these 307 girls, 3 girls had missing data regarding reported coping responses (defined later); therefore, they were dropped from the analyses. Thus, the effective sample size for the present study was 304 (see Leaper & Brown, 2008, for findings regarding the incidence of sexual harassment with the entire sample).
Parents’ Education
Participants separately indicated their mothers’ and fathers’ highest level of education using the following ordinal scale: 1 = elementary school, 2 = some high school, 3 = high school graduate, 4 = some college, 5 = bachelor's degree, 6 = some graduate school, or 7 = graduate degree (master's, doctorate, medical, law, etc.). Education levels were provided for 302 mothers and 281 fathers in the sample of 304 girls, and all participants provided education information for at least one parent. When information about both mothers and fathers was available, their rankings were averaged. Otherwise, we used the value for the parent that was provided.
Perceived Social Support from Mothers, Fathers, and Peers
To assess perceived social support from mothers and fathers, scale items were adapted from Harter's (1988) Self-Perception Profile for Adolescents. Items were simplified by converting them to declarative statements to which respondents rated agreement on a 4-point scale (1 = disagree strongly, 2 = disagree somewhat, 3 = agree somewhat, and 4 = agree strongly). To measure perceived support from parents, participants were asked to rate their mothers and fathers on five items each (e.g., “My mother [father] doesn't really understand me” [reverse-scored item], “I feel very close and intimate with my mother [father]”). To measure perceived support from peers, we used eight items from DuBois, Felner, Brand, and Phillip's (1996) questionnaire (e.g., “I am as well liked by other kids as I want to be,” “I feel good about how much my friends like my ideas”). Both of the measures (Dubois et al., 1996; Harter, 1988) infer perceived social support through the individuals’ feelings of acceptance and closeness.
Self-Esteem
To assess self-esteem, eight items from the Rosenberg (1965) Self-Esteem Scale were used (e.g., “I feel that I have a number of good qualities,” “At times, I think that I am no good at all” [reverse-scored item]). The complete Rosenberg scale for adults includes 20 items. To shorten the form for use with adolescents, we cut the scale to eight items by deleting many of the similarly worded items; high internal reliability was attained (see Table 1). Respondents rated agreement on a 4-point scale (1 = disagree strongly, 2 = disagree somewhat, 3 = agree somewhat, and 4 = agree strongly).
Perspective Taking
We used the perspective taking scale from Davis's (1983) Interpersonal Reactivity Index (e.g., “I sometimes find it difficult to see things from the other person's point of view [reverse-scored item],” “I believe there are two sides to every question and try to look at both of them”). The scale includes five items that were rated for agreement on a 4-point scale (1 = disagree strongly to 4 = agree strongly).
Feminist Self-Identification
Many girls and women misunderstand what it means to be a feminist (Manago, Brown, & Leaper, 2009). Therefore, the following statement was presented: “As defined in the dictionary, feminism refers to the belief in equality for women and men. Feminists believe in equality and point to ways that society and certain individuals treat women and girls in unfair ways.…” Participants were subsequently asked to indicate if they personally identified as a feminist using two items from Szymanski (2004): “I personally consider myself a feminist” and “I describe myself as a feminist to others.” Each was rated for agreement on a 4-point scale (1 = disagree strongly to 4 = agree strongly). The two items were averaged to create a measure of feminist self-identification.
Cognitive Appraisals of Responses to Hypothetical Experiences with Sexual Harassment
The questions about experiences with sexual harassment (described previously) defined sexual harassment as unwanted sexual comments and behaviors. After completing those questions, the participants were asked to consider how they might respond to incidents of sexual harassment. The directions stated, “Suppose someone sexually harassed you. How likely do you believe you would do each of the following?” There were 10 responses to rate agreement on a 4-point scale (1 = disagree strongly to 4 = agree strongly). Items were derived from Lazarus and Folkman's (1984) Ways of Coping questionnaire for use with an adolescent sample. They included two items referring to confronting the perpetrator (“Tell the person the behavior was offensive,” “Tell the person that I was angry”), three items referring to seeking help from others (“Talk to someone to learn more about the situation,” “Talk to someone about how I was feeling,” “Ask someone that I respected for advice”), and three items referring to avoidance (“Not let it get to me,” “Try to forget the whole thing,” “Refuse to get too serious about it”).
Results
Preliminary Analyses
Descriptive statistics and bivariate correlations among the measures are presented in Table 1. Given that the measures were ordinal (for the rating scales) or dichotomous dummy-coded variables (for ethnicity), Spearman correlations were performed.
Associations with Coping Responses
As seen in Table 1, the two approach strategies—confronting and seeking help—were each positively correlated with perceived peer support, perspective taking, self-esteem, and feminist self-identification. In addition, confronting was positively correlated with parents’ education. Also, seeking help was positively correlated with perceived mother support and father support. Avoidance coping was negatively correlated with self-esteem, and it was less likely among White European American girls than other girls.
We additionally conducted a repeated measures analysis of variance to test whether girls varied in their reported likelihood of using the three coping strategies (see Table 1 for means). The effect for coping strategy was significant, F(2, 302) = 9.84, p < .001. Comparison tests indicated that avoiding was less likely than either confronting, t(303) = 4.40, p < .001, or seeking help, t(303) = 3.43, p < .001. There was no difference in likelihood between confronting and seeking help, t(303) = 1.34, p = .18.
Differences Associated with Ethnic Background
As seen in Table 1, there was one significant difference in reported coping based on ethnic background. Avoidance was less likely among White European American girls than ethnic-minority girls. Age was also correlated with girls’ ethnic background (see Table 1); that is, our sample of Latina girls was slightly older (M = 15.91 years, SD = 1.13 years) than the other girls (M = 15.21 years, SD = 1.18 years). We controlled for this difference by including age as well as ethnic background in our regression analyses.
Hierarchical Regression Analyses
A series of hierarchical regression analyses were performed with girls’ appraisals of the coping responses to sexual harassment. We separately tested predictors of three responses: confronting, seeking help, and avoidance. For each hierarchical regression model, four steps were tested. In Step 1, sociocultural background factors were entered: participants’ ethnicity and their parents’ education. Given the high proportion (53%) of girls from Latina backgrounds in the sample, we contrasted these girls with girls from other ethnic backgrounds (0 = non-Latina, 1 = Latina). In addition, we compared girls from all ethnic-minority backgrounds (i.e., Latina and other ethnic minority) and girls from White European American backgrounds (0 = ethnic minority, 1 = White European American). In the second step, we entered four interpersonal factors: perceived social supports from mother, father, and peers as well as reported sexual harassment. In Step 3, we entered two developmental factors: girls’ age (in years) and perspective-taking ability. In Step 4, two individual factors were entered: self-esteem and feminist self-identification. Any missing values for predictor variables were replaced with means.
Given that many of the predictors were correlated with one another (see Table 1), collinearity was a potential concern in the regression analyses. Collinearity statistics were therefore performed to assess tolerance levels with each regression. Tolerance refers to the percent of variance associated with a particular predictor that cannot be accounted for by the other predictors. Tolerance values above .10 are generally considered adequate (Chen, Ender, Mitchell, & Wells, 2003). The minimum tolerance value across the three regressions was .46.
Confronting the Perpetrator
Table 2 summarizes the results from the hierarchical regression testing predictors of girls’ appraisals of confronting. Interpersonal factors (Step 2) and developmental factors (Step 3) were each significant when entered. In contrast, the other steps did not significantly add to the regression model. Thus, the third step was selected as the final model. It accounted for 18% of the variance, which is considered a large effect size (Cohen, 1988). The following predictors were significant in the final model. As hypothesized, girls with greater perspective taking were more likely to report that they would use confrontation. Also, as hypothesized, girls who perceived greater peer support were more likely to endorse confronting sexual harassers. Furthermore, parents’ education was positively and significantly associated with positive appraisals of confronting.
| Model 1 | Model 2 | Model 3 | Model 4 | |
|---|---|---|---|---|
| Type of Predictor | β | β | β | β |
| 1. Sociocultural | ||||
| Latina | .11 | .13 | .14 | .14 |
| White European American | .00 | .00 | .00 | .00 |
| Parents' Education | .20* | .19* | .17* | .16* |
| 2. Interpersonal | ||||
| Mother Support | .02 | .03 | .01 | |
| Father Support | −.06 | −.06 | −.07 | |
| Peer Support | .28*** | .23*** | .20** | |
| 3. Developmental | ||||
| Age | .01 | −.01 | ||
| Perspective Taking | .29*** | .28*** | ||
| 4. Individual | ||||
| Self-Esteem | .10 | |||
| Feminist Identity | .06 | |||
| Fmodel | 2.16† | 5.34*** | 8.00*** | 6.85*** |
| R2 | .02 | .10 | .18 | .19 |
| Fchange | 8.35*** | 14.54*** | 2.02 |
Note
- N = 304. Latina (1 = Latina, 0 = other ethnic group) and White European American (1 = White, 0 = other ethnic group) are dummy-coded variables.
- *p < .05. **p < .01. ***p < .001. † < .10.
Seeking Help
Table 3 summarizes the results of the regression testing predictors of girls’ appraisal of seeking help in response to sexual harassment. The interpersonal context (Step 2), developmental factors (Step 3), and individual characteristics (Step 4) were each significant when added to the regression model. Thus, the fourth step was selected as the final model. It accounted for 18% of the variance, which is considered a large effect size (Cohen, 1988). Several factors were significant in the final model. As predicted, girls who scored higher in perceived mother support, feminist self-identification, and perspective taking were more likely to endorse seeking help in response to sexual harassment. Also, there was a trend (p < .10) for girls with higher self-esteem to be more likely to report seeking help, which was hypothesized.
| Model 1 | Model 2 | Model 3 | Model 4 | |
|---|---|---|---|---|
| Type of Predictor | β | β | β | β |
| 1. Sociocultural | ||||
| Latina | .10 | .11 | .13 | .11 |
| White European American | .04 | .03 | .03 | .03 |
| Parents’ Education | .03 | .00 | −.03 | −.04 |
| 2. Interpersonal | ||||
| Mother Support | .18** | .19** | .16** | |
| Father Support | .08 | .08 | .06 | |
| Peer Support | .16** | .12* | .08 | |
| 3. Developmental | ||||
| Age | −.04 | −.04 | ||
| Perspective Taking | .22*** | .19** | ||
| 4. Individual | ||||
| Self-Esteem | .12† | |||
| Feminist Identity | .16** | |||
| Fmodel | .50 | 5.15*** | 6.12*** | 6.38*** |
| R2 | .01 | .09 | .14 | .18 |
| Fchange | 9.76*** | 8.25*** | 6.53*** |
Note
- N = 304. Latina (1 = Latina, 0 = other ethnic group) and White European American (1 = White, 0 = other ethnic group) are dummy-coded variables.
- *p < .05. **p < .01. ***p < .001. †p < .10.
Avoidance
The regression results associated with girls’ appraisal of avoidance are presented in Table 4. Interpersonal factors (Step 2) and individual factors (Step 4) were each significant when entered into the regression. Utilizing the fourth step as the final model, 7% of the variance was explained; this is considered a moderate effect size (Cohen, 1988). As hypothesized, self-esteem was negatively related to reported use of avoidance coping. Contrary to expectation, perceived father support was positively associated with the endorsement of avoidance coping.
| Model 1 | Model 2 | Model 3 | Model 4 | |
|---|---|---|---|---|
| Type of Predictor | β | β | β | β |
| 1. Sociocultural | ||||
| Latina | .02 | .01 | .02 | .02 |
| White European American | −.10 | −.10 | −.10 | −.10 |
| Parents’ Education | .03 | .01 | −.01 | .00 |
| 2. Interpersonal | ||||
| Mother Support | −.09 | −.09 | −.05 | |
| Father Support | .15* | .15* | .16** | |
| Peer Support | −.06 | −.06 | −.02 | |
| 3. Developmental | ||||
| Age | −.10 | −.09 | ||
| Perspective Taking | −.02 | .00 | ||
| 4. Individual | ||||
| Self-Esteem | −.14* | |||
| Feminist | −.08 | |||
| Fmodel | .97 | 1.90 | 1.78 | 2.14* |
| R2 | .01 | .04 | .05 | .07 |
| Fchange | 2.81* | 1.41 | 3.46* |
Note
- N = 304. Latina (1 = Latina, 0 = other ethnic group) and White European American (1 = White, 0 = other ethnic group) are dummy-coded variables.
- *p < .05. **p < .01.
Discussion
Prior research on resilience has noted that various factors may serve as assets or risks when individuals are coping with stress (see Masten & Coatsworth, 1998). In our analyses, we considered aspects of the sociocultural context (macrosystem), experiences in specific interpersonal contexts (microsystems), and the person's developmental and individual characteristics as potential assets or risks. As discussed below, the results support the independent influences of each level on adolescent girls’ cognitive appraisals of coping responses to sexual harassment.
Ethnic Background and SES
Girls’ ethnic background and SES (based on parents’ education) were the sociocultural factors included in our model. In the final regression models, ethnic background was not significantly related to cognitive appraisals of any of the coping strategies. However, parents’ education was positively related to girls’ likelihood of endorsing confronting perpetrators of sexual harassment. Girls from families with higher parental education and higher SES may have more resources that enhance their sense of self-control and help them cope with everyday stressors (Finkelstein et al., 2007). Further, research suggests that children from higher-SES families are more likely than those from lower-SES families to be assertive in response to perceived unfairness (Lareau, 2003). Thus, these girls may have been more likely to view confronting as having potentially positive outcomes than those girls who came from lower-SES families. Additional research is needed, however, to explore factors that may mediate the link between SES and girls’ cognitive appraisals of coping.
Interpersonal Factors
The interpersonal factors that we considered were girls’ perceived social supports from mothers, fathers, and peers. Researchers have highlighted the importance of parents’ and peers’ social support in adolescents’ responses to stressful events (e.g., Frydenberg & Lewis, 1993). Consistent with our predictions, perceived social support predicted the likelihood of all three forms of coping.
Perceived peer support was positively related to confronting and seeking help in response to sexual harassment. This finding is not surprising in light of prior reports that most sexual harassment and other forms of gender discrimination come from peers (AAUW, 2001; Leaper & Brown, 2008). Given the negative backlash that might follow a confrontational response, it may help girls to know that they can count on their friends to back them up. For example, support from classmates may mitigate concerns that confronting sexist behaviors will jeopardize one's popularity. In a similar manner, research points to the role of women's friendship circles when confronting sexual harassers in the workplace (e.g., Goldberg, 2001).
Both mother support and father support were positively associated with seeking help in the bivariate correlations (see Table 1); of these, however, only mother support appeared significant in the regression model. Perhaps due to many adolescent girls’ closer ties to their mothers than their fathers (e.g., Smetana et al., 2006), seeking help as a coping strategy was more strongly related to mother support than father support. In addition to feeling more open with their mothers, girls may have expected that their mothers were more likely than their fathers to have experienced sexual harassment.
One unexpected finding was that perceived father support was positively associated with girls’ expected use of avoidance in response to sexual harassment. In contrast, we had hypothesized a negative association. To our knowledge, there is no pertinent research to help interpret this result and therefore we can only speculate about its possible meaning. One hypothesis may be that some girls who are close to their fathers find it more dissonant to deal directly with sexual harassment—particularly if it comes from males. Of course, before taking this conjecture too seriously, it is necessary to see if the result can be replicated and explored more fully.
Developmental Factors
Perspective taking—but not age—proved to be a significant developmental predictor of girls’ cognitive appraisals of coping responses to sexual harassment. In particular, perspective taking was related to appraisals of confronting and seeking help. Perspective taking may be especially adaptive when appraising whether one should confront the perpetrator of sexual harassment, given the potential risks in confronting. In some instances, telling someone that their behavior was offensive may effectively address a problem with sexual harassment. In other cases, confronting a classmate for sexual harassment might lead to an angry reaction, physical harm, or social retribution (e.g., Shelton & Stewart, 2004; Swim & Hyers, 1999). Therefore, confrontation is not prudent in all situations. The ability to understand social situations through perspective taking may help girls realize that confronting can be helpful in some (but not all) situations. Consistent with this notion, prior research has indicated that adolescents with higher levels of perspective taking do better when negotiating interpersonal conflicts (Selman, 1989).
Perspective taking was also a predictor of girls’ cognitive appraisal of seeking help. Girls with higher perspective-taking ability may be more interested in seeking out others’ interpretations and guidance (Selman, 1989). They may be additionally concerned with getting validation from friends and family for their own experiences and viewpoint. Conversely, girls who generally do not consider other perspectives may be unmotivated to seek out advice from others following sexual harassment.
Individual Factors
Self-esteem and feminist self-identification are two individual assets that we hypothesized would predict girls’ likely coping responses to sexual harassment. First, we found that higher self-esteem was positively related to girls’ appraisal of seeking help. Although experiences with sexual harassment can undermine girls’ self-esteem (e.g., AAUW, 2001; Goldstein et al., 2007; Timmerman, 2005), approach coping may help to mitigate this effect. Thus, girls with high esteem may be more apt to view sexual harassment as undeserved and want to discuss it with someone. This proposal is consistent with prior research on adolescents’ coping with ethnic discrimination (Phinney & Chavira, 1995), which found youths with high self-esteem were more likely to seek discussion and were less likely to favor verbal retorts in response to discrimination.
Feminist self-identification additionally predicted girls’ positive appraisal of seeking social support in response to sexual harassment. In previous studies, girls’ perception of sexism was positively associated with their exposure to feminism (Leaper & Brown, 2008). Perhaps girls who embrace a feminist identity are more likely than other girls to recognize and disapprove of sexist events. Thus, feminist-identified girls may be more apt to consider these experiences as warranting discussion with friends, parents, teachers, or others in their social support network. Conversely, feminist-identified girls may be less apt to consider avoidance as an appropriate response to sexual harassment. Our findings lend support to proposals that feminist identification may help girls and women cope with gender discrimination (e.g., Ayres et al., 2009; Leaper & Arias, 2011). To our knowledge, no prior studies have tested whether feminist identity was related to adolescent girls’ coping response to sexism.
Limitations and Future Directions for Research
One limitation of the present research is that we assessed girls’ responses to hypothetical sexual harassment. Self-reported responses to hypothetical sexist events may not always reflect what would happen in real life (see Woodzicka & LaFrance, 2001). To increase the potential generalizability of the results, we focused on a sample of girls who had reported having previously experienced sexual harassment. We reasoned that considering how one might respond to sexual harassment would be less abstract for girls who had experienced sexual harassment compared with those who had not.
One alternative method would be to observe girls’ actual responses to sexual harassment from confederates in a laboratory setting (e.g., Swim & Hyers, 1999). However, this is potentially problematic to implement with an adolescent sample for ethical and practical reasons. Another strategy is to examine girls’ reported responses to past experiences with sexual harassment (e.g., Ayres et al., 2009; Ayres & Leaper, 2013; Kaiser & Miller, 2004). The corresponding cost of the latter approach, however, is that it limits one's sample to only those girls who can recount how they dealt with a prior experience with sexual harassment. Previous research with adolescents (Houston & Hwang, 1996) and adults (Crosby, 1984; Taylor, Wright, Moghaddam, & Lalonde, 1990) suggests that individuals tend to underestimate and downplay personal experiences with discrimination. In future studies, it might be revealing to collect self-reports of responses to both real-life and hypothetical incidents with sexual harassment.
Despite the limitations of our method, self-reported responses to hypothetical sexual harassment do reflect the girls’ cognitive appraisals of different coping responses. According to the stress and coping model (Lazarus, 1999; Lazarus & Folkman, 1984), cognitive appraisals guide the kinds of strategies that individuals are likely to deploy. That is, research based on this model suggests that girls are more likely to use approach coping strategies if they positively appraise (i.e., endorse) them. In support of this contention, Kaiser and Miller (2004) found that women's cognitive appraisals of confronting predicted their likelihood of confrontation responses to real-life experiences with sexism. Also, Foster (2009) observed that women's cognitive appraisals of coping with gender discrimination predicted their well-being 1 year later.
Another limitation of the present research is that we did not consider the potential roles of teachers and other school personnel in girls’ experiences with sexual harassment. As seen with classmates, teachers can be a hindrance or a help. On the one hand, teachers and other school staff can be sources of sexual harassment. Given the power asymmetries between students and school staff, students may have different cognitive appraisals of coping responses to sexual harassment from teachers and peers (e.g., Timmerman, 2003). On the other hand, teachers can be sources of support for students who experience sexual harassment. When teachers convey a message that sexual harassment is unacceptable, students may be more likely to approach them for help when sexual harassment occurs (Stone & Couch, 2004).
Finally, we note that we adapted some existing measures for use in the present study (see Method section). With some of the perceived support measures, we modified the response format to make them easier to read for our adolescent sample. With the self-esteem scale, we reduced the number of items given the length of the survey. In all cases, we attained satisfactory levels of internal reliability. Nonetheless, some caution may be warranted when comparing our results with those from other studies using these measures.
Conclusions
Our study highlights the combined and independent influences of sociocultural, interpersonal, developmental, and individual factors on girls’ cognitive appraisals of possible coping responses to sexual harassment. Compared with research with adults, relatively little research has examined adolescent girls’ possible coping responses to sexual harassment or other forms of sexism (see Leaper & Robnett, 2011). Yet, studies do indicate that repeated experiences with sexual harassment are associated with lowered socioemotional adjustment and decreased academic engagement (AAUW, 2011; Chiodo et al., 2009; Goldstein et al., 2007). Thus, further study in this area is highly warranted. By identifying assets and risks related to effective coping in response to sexual harassment, it may be possible to mitigate the impact of these experiences on youth. As Masten and Coatsworth (1998) observed, research on resilience “offers hope and guidance for those who seek to improve the odds of good developmental outcomes through policy and prevention” (p. 215).
An increasing number of schools in the United States and other countries have initiated programs to counter bullying and sexual harassment (Espelage & Holt, 2012). Based on our research, we can suggest a few components that might be included these programs. Two factors related to adaptive coping highlighted in our findings were perceived peer support and perspective taking. School counselors and teachers can potentially strengthen these in students through small-group, peer-facilitated discussions about the definition, emotional consequences, and possible responses to sexual harassment. These groups could become a context for experiencing peer support about sexual harassment. Also, role-playing exercises whereby individuals assume the roles of both perpetrator and victim can help to enhance perspective-taking skills (e.g., Gilbert, 1997). Part of these role-playing activities could involve practicing the use of approach coping strategies, such as confronting or seeking support, which can lead to later behavioral change (e.g., Lamb, Bigler, Liben, & Green, 2009). In these and other ways, girls (and boys) who are victims of sexual harassment can move from “learning powerlessness” to “learning agency” (Luft & Cairns, 1999, p. 112).
