Home / Article Critique / Gender Discrimination, Article Critique Example

Gender Discrimination, Article Critique Example

Pages: 33

Words: 9120

Acknowledgement: Part of this project was presented as a poster at the 2002 Conference on Human Development, Charlotte, NC. This research was supported by a grant to Christia Spears Brown from the Debra Beth Lobliner Fellowship. We are extremely grateful to the director, John Combs, and the site leaders of Extend-A-Care in Austin, TX, and to the students who participated in the project and their parents. We also thank Enrique Barroso, Jensen Sapido, Tiffany Seaman, Leah Lambert, and Allison Davis for help with data collection and Alex Brown for the illustrations.

Discrimination on the basis of group membership (e.g., gender, race, or religion) is an important social problem in the United States and throughout the world. In the year 2000, for example, the U.S. Department of Education’s Office for Civil Rights received approximately 5,000 complaints regarding instances of discrimination. The majority of these complaints (approximately 70%) were filed on behalf of elementary and secondary school children.Undoubtedly, many more instances of discrimination affecting children occur every year but go unreported. Although existing research has examined factors that affect adults’ perceptions of discrimination (e.g., Swim, Cohen, & Hyers, 1998), little research has examined children’s perceptions of discrimination. In this study, we examined children’s judgments about scenarios involving possible instances of gender discrimination. We were especially interested in whether children are sensitive to contextual information in making attributions to discrimination and whether individual and developmental differences among children are related to their perceptions of discrimination.

In addition to its obvious practical importance, understanding how and when children perceive discrimination is important for developmental theory.Developmentalists have argued that children’s perceptions of discrimination play an important role in shaping many developmental processes and outcomes, including identity development and academic achievement (Bowman & Howard, 1985; Spencer & Markstrom-Adams, 1990). That is, researchers have argued that understanding children’s perceptions of discrimination is imperative for outlining the normative development of children who are members of stigmatized groups and who may, therefore, be the targets of discriminatory actions (Garcia Coll et al., 1996). It seems equally important to understand perceptions of discrimination among children who are members of privileged groups and who may, therefore, witness and benefit from discriminatory actions.

What is discrimination? Fishbein (1996, p. 7) defined discrimination as “harmful actions towards others because of their membership in a particular group.” Discriminatory actions can range from mild (e.g., ignoring someone) to virulent (e.g., inflicting physical harm). As racial and gender biases have become less socially acceptable in this country, discriminatory actions have become increasingly subtle and ambiguous, requiring individuals to make attributions about the motivations of others on the basis of situational information (e.g., Dovidio & Gaertner, 1986; Swim, Aikin, Hall, & Hunter, 1995).It is especially important, therefore, to examine whether, and if so how, children use contextual information in making decisions about discriminatory behavior.

The earliest research examining children’s perceptions of discrimination centered on race and was conducted after the court-ordered racial desegregation of schools in the United States. For example, Radke and Sutherland (1972) asked European American children, “What are Negroes like?” They found that 12% of 11- to 12-year-olds, 49% of 13- to 14-year-olds, and 59% of 17- and 18-year-olds mentioned discrimination and that all respondents described discrimination as negative. Other studies reported that African American children viewed themselves as victims of racial bias.Rosenberg (1979) found that 51% of African American children who attended desegregated junior high schools, and 34% who attended predominantly African American schools, reported that they had experienced teasing or exclusion because of their race. Patchen (1982) found that many African Americans in newly desegregated high schools complained about the discriminatory actions of European American teachers. Those African American students who perceived discrimination had slightly lower grades and more negative attitudes toward European Americans than did other African American students.

More recently, researchers have investigated children’s broad conceptualizations of discrimination. Researchers in the Netherlands have reported, for example, that most children are knowledgeable about the definition of discrimination by the age of 10 (Verkuyten, Kinket, & van der Weilen, 1997).Verkuyten et al. (1997) reported that name-calling was the most frequently cited example of discrimination (cited by 67% of the children), followed by unequal sharing of goods and social exclusion (cited by 10% and 8% of children, respectively). Verkuyten and colleagues also found that children failed to consider actions as discriminatory if they considered the target to be responsible for the negative behavior or the perpetrator to have acted unintentionally.

Other researchers have examined children’s understanding of the factors that produce discrimination and prejudice. Quintana and Vera (1999), for example, examined 7- and 12-year-old Mexican American and African American children’s explanations for ethnic prejudice. They found that children’s understanding of prejudice becomes more complex with age. Seven-year-old children stated that prejudice occurs because of either (a) an individual’s perceptual preferences (e.g., “They don’t like their color”) or (b) an individual’s disliking of a literal, nonsocial aspect of a person’s ethnicity (e.g., “They may not like Mexico”). By 12 years of age, children state that prejudice occurs because of either (a) an isolated, idiosyncratic social action related to nonperceptual characteristics of ethnicity (e.g., “Their mom might tell them not to play with African Americans”) or (b) the pervasive experiential influences of our society (e.g., “If one [Mexican] did something, it’s like all the Mexicans in the world did everything bad”).

Although existing studies such as these indicate that children are cognizant of discrimination, especially racial discrimination, little work has systemically addressed the question of how and when children detect gender discrimination. In fact, we were unable to locate any work aimed specifically at examining children’s perceptions of gender discrimination. Research suggests, however, that children are highly likely to encounter discrimination based on gender. Gender discrimination may be most common—and overt—in situations involving peers. Children show high rates of gender bias in their peer preferences, interactions, and evaluations (Leaper, 1994; Maccoby, 1998; Nesdale & McLaughlin, 1987; Powlishta, 1995). For example, a recent study by Kowalski and Kanitkar (2003) documented high rates of gender bias and discrimination among kindergartners, including many instances in which children were prevented from, or ridiculed for, engaging in cross-sex-typed play by their peers. Research also suggests that children are likely to encounter gender discrimination at the hands of their parents (e.g., see Leaper, 2000; Ruble & Martin, 1998) and teachers (e.g., Good, Sikes, & Brophy, 1973; Jones & Wheatley, 1990). Adults’ discriminatory behaviors are more likely than those of children’s peers, however, to be subtle and covert (see Meece, 1987).

Given the lack of previous research, it is unclear whether children perceive particular events as discriminatory and, if they do, whether certain situational cues must be present before children attribute adults’ differential behavior toward boys and girls to gender discrimination. In this study, we examined children’s judgments about whether a student had been the victim of a teacher’s gender discrimination, varying the contextual information (e.g., male vs. female victim) that was present. Our specific predictions concerning the effects of situational and individual-difference factors on children’s perceptions of discrimination were informed by research within social psychology.

Social psychological research has found that both situational and individual-difference variables affect adults’ perceptions of discrimination (e.g., Feldman Barrett & Swim, 1998; Mazur & Percival, 1989; Swim et al., 1998). Research on situational variables indicates that individuals are more likely to attribute a negative outcome to discrimination (rather than poor performance) when situations are low, rather than high, in ambiguity (e.g., Dion, 1975; Feldman Barrett & Swim, 1998). For example, participants who have knowledge that an evaluator may be biased against a particular social group are more likely to make attributions to discrimination than are participants who have no such knowledge (e.g., Feldman Barrett & Swim, 1998). We reasoned that children are unlikely to be knowledgeable about teachers’ gender attitudes but are likely to have information about the past behavior of their classroom teachers (e.g., whether a teacher more often calls on boys than girls to perform specific tasks). Thus, in the present study, we manipulated situational information by providing children with information about the past behavior of an elementary school teacher (i.e., whether the individual had a history of gender fairness or gender bias). We predicted that children who were given situational information suggesting that discrimination was likely (i.e., a teacher with a history of bias toward one gender) would make more attributions to discrimination than would children who were either given no situational information or given situational information suggesting that discrimination was unlikely (i.e., a teacher with a history of gender fairness).

In addition to situational characteristics, certain individual, or personal, characteristics affect adults’ perceptions of discrimination. For example, individuals’ attitudes toward social groups predict the likelihood that stigmatized group members will label an event as discriminatory (e.g., Mazur & Percival, 1989;Swim & Cohen, 1996). Specifically, research has found that women who hold traditional attitudes about gender roles are less likely to label an event as sexist than are women who hold more egalitarian attitudes (Jensen & Gutek, 1982; Swim & Cohen, 1996). Similarly, children’s attitudes toward gender may predict their perceptions of discrimination. Children’s gender attitudes have been linked to a large number of cognitive processes, including decision making and memory (e.g., Liben & Signorella, 1980). In the present study, we examined children’s beliefs about whether gender should constrain individuals’ activities, a form of sex typing (see Liben & Bigler, 2002). We expected children’s sex-typed beliefs about activities to relate to their perceptions of discrimination involving schoolteachers’ selection of children to perform school-related activities and roles. Imagine, for example, a child who believes that only boys should play soccer. Such a child is likely to believe that a coach who selects only boys for a soccer team is acting in a fair or reasonable manner. In contrast, a child who believes that both boys and girls should play soccer is likely to believe that a coach who selects only boys is acting in a biased or unreasonable manner. Thus, children with more egalitarian gender attitudes were expected to perceive gender discrimination more often than were children with less egalitarian attitudes.

In addition to gender attitudes, we expected gender per se to influence children’s perceptions of discrimination in several ways, on the basis of the premise that elementary-school-age children show at least some awareness of the lower status of females relative to males (Levy, Sadovsky, & Troseth, 2000; Liben, Bigler, & Krogh, 2001). First, we expected the gender of the participant to influence responding. Research suggests that girls, perhaps because of their experience in a lower status group, view gender-based exclusion as more wrong than do boys (Killen, Lee-Kim, McGlothin, & Stangor, 2002; Killen & Stangor, 2001). We predicted, therefore, that girls would make more attributions to discrimination than would boys. Second, we expected both the gender of the authority figures (i.e., teachers) and the gender of the targets of discrimination (i.e., students) presented in the scenarios to affect children’s judgments. If children believe that other individuals view males as more important than females, children should view girls as more likely than boys to be the victims of gender discrimination (at the hands of both male and female teachers). Finally, although no previous research has addressed this issue, it is also possible that children will view men (because they are higher in status and thus the beneficiaries of gender bias) as more likely than women to engage in gender discriminatory behavior.

Although we drew on social psychological data to inform many of our predictions, children’s perceptions of discrimination are likely to be qualitatively different from those of adults. Just as children’s understanding of prejudice as a social phenomenon is affected by developing cognitive skills (Quintana, 2001), it is likely that children’s tendency to perceive gender discrimination is also affected by cognitive development. One cognitive skill that may relate to the perception of discrimination is the ability to simultaneously classify people along multiple dimensions (Inhelder & Piaget, 1964). Making an attribution to discrimination requires that the child be able to assess several characteristics of the target simultaneously. For example, to perceive discrimination, a child must be able to simultaneously understand that an individual is a good math student (and thus deserving of a good math grade) but also a girl (and thus belonging to a social group associated with poor math performance). A child who lacks the ability to characterize individuals as a member of both a contextual group (e.g., student) and a stigmatized social group (e.g., girl) is unlikely to perceive the target as the victim of discrimination.

Relatedly, children’s understanding of the hierarchical nature of categories may also affect their perceptions of discrimination. Young children who lack an understanding of the hierarchical nature of categories may fail to place individuals within social groups in a fully consistent and integrated manner. As a consequence, although they may be knowledgeable and even endorse negative stereotypes about a particular social group, they may fail to view these stereotypic beliefs as relevant to a particular member of that social group (see Spencer, 1985) and thus may not understand why others might treat that person in a stereotype-consistent manner. Thus, we predicted that children with more advanced classification skills would make a greater number of attributions to discrimination than would their peers with less advanced classification skills.

These questions outlined above were investigated in children between the ages of 5 and 10 years. It is during this age period that children acquire the cognitive skills hypothesized to be associated with adultlike perceptions of discrimination and also show considerable variability in their gender attitudes (e.g., Signorella, Bigler, & Liben, 1993). Although we expected cognitive skill level (rather than age per se) to affect responding, we also examined changes in children’s perceptions of discrimination across the elementary school years.

Method

Participants

Participants were 76 children (37 girls and 39 boys) divided into two age groups: 5–7-year-olds and 8–10-year-olds (see Table 1 for age means and standard deviations). The participants were primarily (89%) European American (68 European American, 3 Hispanic, 3 African American, 1 Asian, 1 African American/European American). Participants were recruited from the after-school programs of three elementary schools in a large southwestern city that serve families from diverse socioeconomic backgrounds. The sample was reflective of the schools’ racial and ethnic populations.

Sample Sizes, Mean Ages (Years, Months), and Standard Deviations (Months) Across Experimental Conditions

Overview of Procedure

All measures were given to children individually by a same-gender experimenter. Children were first given tasks designed to assess classification ability.Children then were read six scenarios and, following each scenario, were asked about their attributions concerning the outcome depicted. Upon completion of the attribution questions, children’s gender attitudes were assessed. The gender attitudes measure was always given following the discrimination tasks (i.e., scenarios) because this measure explicitly identifies gender as the variable to be used in the task and thus, if given first, might have primed children to attend to gender in the discrimination task. Children were then debriefed and thanked for their participation.

Attributions to Discrimination

Materials

After completion of the classification skill measure, a same-gender experimenter read six scenarios to individual children. Each scenario presented a schoolteacher and a male and a female student in his or her class. A routine, gender-neutral classroom activity was introduced, and the teacher evaluated the male and female students. Scenarios depicted students drawing a picture for an art contest, writing an essay for an essay contest, competing in a race, trying out for a soccer team, vying to be class leader, and breaking a jar while playing (see the Appendix for complete texts).

For each scenario, the situational information regarding the likelihood of discrimination was manipulated as a within-subjects variable. Specifically, each child heard two scenarios each in which the situational information suggested that discrimination was (a) likely (e.g., “Mr. Franks almost always gives boys higher grades than girls on their stories.”), (b) unlikely (e.g., “Mr. Franks almost always grades boys’ and girls’ stories about the same.”), and (c) ambiguous (i.e., no information about the teacher’s past behavior was given). In addition, the gender of the authority figure was manipulated as a within-subject variable. Of the six scenarios, each child heard three scenarios in which the authority figure was a man and three scenarios in which the authority figure was a woman. Finally, the gender of the target of discrimination was manipulated as a between-subjects variable. Half of the participants heard scenarios in which girls (rather than boys) received the desirable outcome (e.g., received the best grade on an essay), and half of the children heard scenarios in which boys (rather than girls) received the desirable outcome. Thus, for half of the participants, discrimination favored girls, and for the other half of the participants, discrimination favored boys. The ns and ages associated with the cells are presented in Table 1. The order of the scenarios, the situational likelihood of each scenario, and the authority figure’s gender within each scenario were all counterbalanced.Colorful pictures accompanied each scenario and were shown to the children as the story was read. Children were urged to pay attention to each scenario and told that they would be asked several questions about each scenario once it had been read.

Procedure

After each scenario was read, children were first asked, in an open-ended format, why the two children experienced different outcomes (see theAppendix). An open-ended question was asked first so that children had the opportunity to make a spontaneous attribution, without any prompting or suggestions by the experimenter. Their answers were recorded. The total number of times children attributed the teacher’s behavior solely to the gender of the student was computed, and thus scores ranged from 0 to 6. Because some children may have believed discrimination to be a secondary (rather than primary) cause of the differential outcome, and because we were interested in children’s judgments about discrimination relative to other potential causes of the differential outcomes, we included a second measure of the perception of discrimination. Specifically, after each open-ended question, children were read a list of four possible reasons for the different child outcomes. The possible reasons were (a) lower ability of one student (e.g., Jacob got a better grade because his story was better), (b) lower effort by one student (e.g., Jacob got a better grade because he tried harder), (c) general unfairness of the competition (e.g., Jacob got a better grade because the assignment wasn’t fair), and (d) the students’ gender (e.g., Jacob got a better grade because he is a boy), with the order of presentation being counterbalanced (see the Appendix for the full set of experimenter-provided attributions). Children rated the veracity of each statement on a scale ranging from 0 (not at all true) to 4 (very true). For younger participants, the scale was accompanied by a graphic representation of responses (i.e., cups with increasing levels of liquid), and the experimenter gave several practice items to ensure that the scale was understood. Children’s mean ratings of the veracity of each attribution type were computed.

Individual and Developmental Difference Measures

Gender attitudes

Children’s gender attitudes were assessed using the Activity subscale of a sex-typing measure (the Children’s Occupation, Activity, and Trait-Attitude Measure [COAT-AM]) developed by Liben and Bigler (2002). Specifically, children were asked “who should” perform each of 25 activities (10 stereotypically female, 10 stereotypically male, and 5 neutral). Children responded using the options “only boys,” “only girls,” or “both boys and girls.” The number of nonstereotyped responses (i.e., “both boys and girls”) given to stereotyped activities was computed, and thus scores ranged from 0 to 20.

Classification skill

Classification skill was assessed using procedures adapted from Jones and Bigler (1996), following Inhelder and Piaget (1964), and included tasks designed to measure (a) consistent sorting, (b) reclassification, or re-sorting, (c) multiple classification, and (d) class inclusion skill. Specifically, children were first asked to sort a set of 12 pictures (men and women reading and talking on the phone) into two groups (consistent sorting task). Children were then asked to sort the cards along a second, new dimension (re-sorting task). Next, children were shown a 2 × 2 matrix and, following a demonstration sort by the experimenter (using bears and elephants that were gray or brown), were asked to sort their original set of cards into the appropriate cells of the matrix. Finally, children were asked a series of seven questions designed to tap their understanding of hierarchical relations among categories (e.g., when shown 2 gray bears and 4 gray elephants, children were asked, “Are there more bears or more animals?”). One point was given for each correct response, and thus classification scores could range from 0 to 10, with higher scores indicating more advanced skill.

Results

Overview

The primary question of interest was whether children perceived discrimination differentially as a function of participant characteristics (i.e., gender and age) and contextual information presented in the scenarios (i.e., likelihood of discrimination, gender of authority figure, and gender of target child).Preliminary analyses of responding were conducted to examine possible effects on responding of (a) the order of scenario presentation and (b) the individual scenarios. The order in which children heard the scenarios did not appear to affect responding. 1 Responding did, however, vary across scenarios such that children made more attributions to discrimination for the footrace scenario than for the other scenarios, probably because the information about the students’ performance was more objective and well-defined than in the other scenarios. The responding variable did not, however, interact with any other variable, and hence responding was pooled across scenarios. 2

Initial analyses also indicated that the gender of the authority figure did not affect responding alone or in combination with any other variables. To minimize complexity, we dropped this variable from subsequent analyses. Thus, analyses of variance (ANOVAs) included participant variables (i.e., gender and age) and gender of the target child as between-subjects variables and likelihood of discrimination as a within-subject variable. The sample was divided approximately evenly across conditions (see Table 1). Because the Mauchley sphericity test was significant (i.e., the homogeneity of variance assumption was violated), the Greenhouse-Geisser procedure was used to adjust the degrees of freedom for the F tests involving the within-subject factor. All post hoc analyses were conducted using Tukey’s honestly significant difference test. Below we describe analyses and findings for children’s open-ended responses to the scenarios, followed by analyses and findings for children’s ratings of the veracity of the experimenter-provided explanations for the scenario outcomes.

Attributions to Discrimination

Open-ended responses

Each open-ended response was first categorized as either an attribution to discrimination or not. Only those responses in which a child reported that the target child’s gender was the sole cause of the differential treatment by the teacher were classified as attributions to discrimination (e.g., “The teacher picked him because he was a boy”). 3 Simple labeling of the target’s gender (e.g., “She’s a girl”) was rare (2 instances), and these responses were not coded as an attribution to discrimination. Other examples of responses that were not classified as attributions to discrimination included those in which children endorsed a gender stereotype (e.g., “Girls are better at art”) or reported that some gender-unrelated characteristic of the target was responsible (e.g., “The teacher thought her picture was better”). The open-ended responses were coded by Christia Spears Brown, and a proportion of the responses (40%) was then independently coded by a research assistant. Cohen’s kappa coefficients for attributions to discrimination were.94,.96, and.96 for the likely, ambiguous, and unlikely discrimination scenarios, respectively.

Descriptive analyses indicated that when discrimination was likely, the majority of children (72%) made at least one attribution to discrimination. In contrast, when discrimination was either ambiguous or unlikely, only a minority of children made one or more attributions to discrimination (22% and 17%, respectively).

Children’s open-ended attributions to discrimination were examined using a 2 (target gender: boy or girl) × 2 (participant gender: boy or girl) × 2 (participant age group: 5–7 years or 8–10 years) × 3 (likelihood: likely, ambiguous, or unlikely) ANOVA in which the last factor was treated as a within-subject variable. The total possible number of attributionsto discrimination per situational information condition was 2. The main effect for situational likelihood was significant, F(2, 117) = 78.14, p <.01. Means are presented in Table 2. Post hoc analyses indicated that children made significantly more attributions to discrimination when the situational information suggested that discrimination was likely than when it was either ambiguous or unlikely (both at p <.01). Children’s attributions to unlikely and ambiguous discrimination did not differ from one another. On the basis of the partial eta squared statistic (Cohen, 1977), 4 situational likelihood accounted for 54% of the variance of attributions to discrimination (?_p² =.54). According to Cohen’s (1977) classifications of ?_p², this indicates a large effect size.

Means (and Standard Deviations) for Children’s Open-Ended Attributions to Discrimination

The interaction between situational likelihood and age was also significant, F(2, 117) = 5.91, p <.01. Analyses of simple effects indicated that both 5- to 7-year-old children, F(2, 67) = 16.63, p <.01, and 8- to 10-year-old children, F(2, 67) = 44.30, p <.01, based their attributions to discrimination on the situational likelihood of the discrimination. Both groups of children made more attributions to discrimination when it was likely than when it was unlikely or ambiguous (both at p <.01 based on post hoc analyses). However, examination of the respective F values indicates that this pattern of results was more pronounced in the older children than in the younger children. On the basis of the partial eta squared statistic, the Situational Likelihood × Age interaction accounted for 4% of the variance in attributions to discrimination, indicating a small effect size.

The main effect for target gender was also significant, F(1, 68) = 4.18, p <.05, with children making more attributions to discrimination when the target was female than when the target was male. The means are presented in Table 2. This effect was subsumed, however, by a significant interaction between target gender and participant gender, F(1, 64) = 4.25, p <.05. Analyses of simple effects indicated that girls made significantly more discrimination attributions for female targets than for male targets, F(1, 68) = 9.33, p <.01, whereas boys showed no differentiation between targets. On the basis of the partial eta squared statistic, the Target Gender × Participant Gender interaction accounted for 6% of the variance in attributions to discrimination, indicating a medium effect size.

Ratings of experimenter-provided explanations

To examine children’s ratings of the veracity of the possible reasons for the negative outcome, we examined children’s responses to the different types of attributions using a 2 (target gender: boy or girl) × 2 (participant gender: boy or girl) × 2 (participant age group: 5–7 years or 8–10 years) × 3 (likelihood: likely, ambiguous, or unlikely) × 4 (attribution type: ability, effort, unfairness, or discrimination) ANOVA in which the last two factors were treated as within-subject variables. The means are presented in Table 3. The main effect for attribution type was significant, F(2, 145) = 4.06, p <.05.Post hoc analyses indicated that, overall, children rated effort as a more accurate explanation for differential outcomes than either unfairness or discrimination (both at p <.05). Children’s ratings of the other attribution types did not differ from one another. On the basis of the partial eta squared statistic, attribution type accounted for 6% of the variance in children’s ratings, indicating a medium effect size.

Means (and Standard Deviations) for Children’s Ratings of the Experimenter-Provided Attributions

The interaction between attribution type and situational likelihood was also significant, F(5, 327) = 12.89, p <.01. Analyses of simple effects indicated that when children were told that the discrimination was likely, F(3, 62) = 6.56, p <.01, they rated discrimination as a more accurate explanation for the differential outcomes than either ability, effort, or unfairness (post hoc analyses indicated that discrimination differed from ability, effort, and unfairness at the p <.01 level). In contrast, when children were told that the discrimination was unlikely, F(3, 62) = 8.63, p <.01, they rated discrimination as a less accurate explanation for the differential outcome than ability, effort, and unfairness (post hoc analyses indicated that discrimination differed from ability, effort, and unfairness at the p <.01 level). When the situational information was ambiguous, F(3, 62) = 3.36, p<.05, children rated discrimination as a less accurate explanation for the differential outcome than ability and effort but not unfairness (post hoc analyses indicated that discrimination differed from ability and effort at the p <.01 level). On the basis of the partial eta squared statistic, the Attribution Type × Situational Likelihood interaction accounted for 17% of the variance of children’s ratings, indicating a large effect size.

The main effect and the two-way interaction were, however, subsumed by the three-way interaction of attribution type, situational likelihood, and age, F(5, 327) = 2.41, p <.05. Analyses of simple effects indicated that when children were told that discrimination was likely (see Figure 1), 8 – to 10-year-old children rated discrimination as a more accurate explanation for the differential outcomes, F(3, 62) = 6.35, p <.01, than ability, effort, and unfairness (post hoc analyses indicated that older children’s ratings of discrimination differed from their ratings of ability, effort, and unfairness at the p<.01 level). Five- to seven-year-old children’s attributions did not differ from one another. When children were told that the discrimination was unlikely(see Figure 2), both younger children, F(3, 62) = 4.05, p <.01, and older children, F(3, 62) = 4.76, p <.01, rated discrimination as a less accurate explanation for the differential outcomes than ability, effort, and unfairness (post hoc analyses indicated that both older and younger children’s ratings of discrimination differed from their ratings of ability, effort, and unfairness at, at least, the p <.01 level). When the situational information wasambiguous (see Figure 3), younger children rated discrimination as a less accurate explanation for the differential outcomes, F(3, 62) = 2.97, p <.05, than ability and effort, but not unfairness (post hoc analyses indicated that younger children’s ratings of discrimination differed from their ratings of ability and effort at the p <.01 level). Older children’s attributions did not differ from one another. On the basis of the partial eta squared statistic, the three-way interaction accounted for 4% of the variance in children’s ratings, indicating a small effect size. Significant effects that did not involve attribution type (i.e., collapsed across reasons) were uninformative and thus are not described (these findings can be obtained from the authors upon request).

Figure 1. Mean ratings of the accuracy of the experimenter-provided attributions as a function of age when discrimination is likely

Figure 2. Mean ratings of the accuracy of the experimenter-provided attributions as a function of age when discrimination is unlikely

Figure 3. Mean ratings of the accuracy of the experimenter-provided attributions as a function of age when discrimination is ambiguous

Relations Between Responding to Open-Ended Questions and Rating Tasks

To assess whether children’s open-ended attributions to discrimination were related to their ratings of the experimenter-provided attributions, we computed the correlations that are reported in Table 4. As can be seen in the table, when discrimination was either likely or unlikely, children who made more open-ended attributions to discrimination also rated discrimination as a more accurate explanation for the differential outcome. When discrimination was ambiguous, children’s open-ended attributions to discrimination and ratings of discrimination were not significantly correlated with one another.Partial correlations, controlling for age, indicated similar, but somewhat weaker, relations.

Zero-Order (and Partial) Correlations Between Children’s Open-Ended Attributions and Ratings of the Experimenter-Provided Attributions by Situational Information

Individual and Developmental Difference Measures

Children’s gender attitudes

The mean score on the COAT-AM was 13.05 (SD = 6.15), indicating a moderate level of gender stereotyping. Children’s gender attitudes were unrelated to age and classification skill (rs =.16 and.13, respectively, ps >.05). To assess whether children’s gender attitudes were related to their open-ended attributions to discrimination, we conducted a hierarchical multiple regression analysis on children’s responses to the likely discrimination scenarios. (Too few children made attributions to discrimination for the unlikely and ambiguous scenarios to test the effects of individual-difference variables.) Age in months and the child’s gender (dummy coded) were entered into the model in the first block, followed by gender attitudes in the second block, and the Gender Attitudes × Age interaction in the third block. Age significantly predicted children’s attributions to discrimination, ? =.73,p <.01. Older children made a greater number of attributions to discrimination than did younger children. In addition, gender attitudes predicted children’s attributions to discrimination, ? = 1.05, p =.06, although at a level just missing conventional levels of significance. Children with more egalitarian attitudes made a greater number of attributions to discrimination than did children with less egalitarian attitudes. Neither the child’s gender nor the Gender Attitudes × Age interaction predicted attributions to discrimination.

An identical regression model was used to assess whether gender attitudes predicted children’s ratings of the experimenter-provided attributions when discrimination was likely. Results indicated that neither age, child’s gender, nor gender attitudes significantly predicted children’s ratings of the veracity of discrimination as the reason for the students’ outcomes. The Gender Attitudes × Age interaction was, however, marginally significant, ? = 1.32, p=.06. Inspection of the data indicated that among older (but not younger) children, those with more egalitarian attitudes perceived discrimination as a more accurate explanation for the outcomes than did those with less egalitarian attitudes.

Classification skill

The mean classification skill score was 6.73 (SD = 2.38). Children’s classification ability was significantly correlated with age (r =.52, p <.01). To assess whether classification ability predicted children’s open-ended attributions to discrimination (independently of age), we conducted hierarchical multiple regression analyses. Age in months was entered into the model in the first block, followed by the classification skill measure in the second block, and the Classification Skill × Age interaction in the third block. Although age (as before) significantly predicted children’s attributions to discrimination when discrimination was likely, ? =.36, p <.01, neither classification ability nor the interaction term accounted for a significant amount of additional variance.Regression analyses of children’s ratings of the experimenter-provided attributions using the same predictor variables indicated no significant findings.

Discussion

The purpose of the present study was to examine whether children are sensitive to contextual information in making attributions to discrimination and whether individual and developmental differences among children are related to their perceptions of gender discrimination. The findings suggest that children are sensitive to contextual information in making judgments about gender discrimination. As hypothesized, children made more open-ended attributions to discrimination (and rated it as a more accurate explanation for the differential outcomes) when situational information suggested that discrimination was likely than when situational information suggested that discrimination was unlikely or ambiguous. Specifically, when children were given information suggesting that the evaluator was biased (e.g., “Mr. Franks almost always gives boys higher grades than girls on their stories.”), they made more attributions to discrimination than when given information suggesting that the evaluator was unbiased or than when presented with no pertinent information. This finding is consistent with the adult literature, which has shown that adults who are told that an evaluator may be biased against a particular social group are more likely to make attributions to discrimination than are participants who have no such knowledge (e.g., Feldman Barrett & Swim, 1998; Swim et al., 1998).

As is the case among adults, children were no more likely to make an open-ended attribution to discrimination when the contextual information was ambiguous than when it suggested that discrimination was unlikely. In addition, when presented with ambiguous information, neither younger nor older children rated discrimination as a more accurate reason for the differential student outcomes than effort and ability. Thus, it appears that children (like adults) are reluctant to attribute outcomes to discrimination unless bias is very apparent. This finding is important because discriminatory actions (at least those performed by adults) are unlikely to be accompanied by unambiguous information (e.g., individuals rarely have knowledge about an evaluator’s past biased behavior). Thus, children are likely to attribute negative feedback from adults, including teachers, to their own performance (e.g., low ability or effort).

Some theorists have suggested that individuals’ reluctance to perceive discrimination in ambiguous situations is due to the psychological costs associated with making an attribution to discrimination, such as a perceived lack of control over outcomes and a diminished belief in a fair and just world (Major, Quinton, McCoy, & Schmader, 2000). Although failing to perceive discrimination in ambiguous situations may have benefits (e.g., maintaining one’s belief in a just world), the tendency to make attributions based on one’s ability or effort (i.e., internal attributions), instead of discrimination (i.e., external attribution), can have important negative consequences as well. For example, Weiner’s attribution theory (e.g., Weiner, 2000) suggests that attributing negative outcomes to internal causes can lead to reductions in motivation, self-esteem, and future expectancies for success. Indeed, social psychological research examining the targets of discrimination finds that perceiving discrimination when it is likely to have occurred is associated with enhanced motivation and self-esteem (e.g., Crocker, Voelkl, Testa, & Major, 1991; Dion, Earn, & Yee, 1978). Future research should examine whether children who are the targets of discrimination themselves make internal (e.g., ability) versus external (e.g., discrimination) attributions for negative outcomes and whether such attributions affect their academic self-concept and motivation.

Although children, at the group level, made use of situational information, there were important developmental trends in responding. Specifically, older children’s attributions to discrimination varied more dramatically as a function of situational information than did younger children’s attributions. Even in those cases in which evidence strongly suggested that discrimination had occurred (e.g., “Mr. Parks almost always picks boys…”), only 27% of 5- to 7-year-old children consistently made spontaneous attributions to discrimination. In addition to making a greater number of open-ended attributions to discrimination, older (but not younger) children rated discrimination as a more accurate explanation for the differential student outcomes than other possible reasons provided by the experimenter. Faced with strong evidence of discrimination, young children were just as likely to say that the target was at fault for producing the outcome (e.g., by failing to make a sufficient effort) as they were to claim that the teacher was discriminatory. Thus, it appears that although children as young as 5 are aware of discrimination as a social possibility, they make relatively poor use of situational cues and are unlikely to label even overtly biased behavior as gender discrimination.

Results also indicated that girls (but not boys) made more open-ended attributions to discrimination when the target of the discrimination was female than when the target was male. This finding is consistent with research examining children’s judgments about peer exclusion (Theimer, Killen, & Stangor, 2001). Specifically, girls rate peer exclusion of girls to be more negative than peer exclusion of boys. Several researchers have hypothesized that increased experiences with peer exclusion make girls more sensitive than boys to this issue (Killen & Stangor, 2001; Theimer et al., 2001). It seems possible, however, that girls’ experiences also lead them to develop a broad awareness of the lower social status of females relative to males. Girls’ awareness of the lower status of the female role may, in turn, lead them to develop a higher sensitivity for, and expectation of, discriminatory treatment of females than of males (though these findings may not generalize to all racial and ethnic groups; see Killen et al., 2002). For reasons that we were unable to intuit, this finding did not generalize to the children’s ratings of the veracity of discrimination when presented by the experimenter.

Although the gender of the target child affected girls’ open-ended attributions to discrimination, the gender of the evaluator did not affect children’s judgments about discrimination. Children rated male teachers and female teachers as equally likely to be discriminatory. This finding is inconsistent with adult social psychological research indicating that men are more likely to be considered prejudiced, and more likely to be labeled as discriminators, than women (Baron, Burgess, & Kao, 1991). To provide a better understanding of how and why children’s and adults’ attributions to discrimination are (or are not) affected by evaluators’ gender, future research should explore individuals’ understanding of the motivations that drive gender stereotyping and discrimination.

As hypothesized, children’s gender attitudes were related to their perceptions of discrimination, although the relations were not particularly strong.Children who endorsed more egalitarian gender attitudes were more likely to report that the target student’s gender was the cause of the teacher’s behavior than were children with less egalitarian attitudes. Though not significant, there was also a trend for children with more egalitarian attitudes, compared with children with less egalitarian attitudes, to rate the student’s gender as a more accurate reason for the teacher’s behavior than the student’s effort or ability. The stronger relation between attitudes and responding within the open-ended attribution than within the rating task may reflect the tendency of open-ended responses to tap the salience of gender inequality. That is, it seems possible that differential treatment of boys and girls is more salient among children with more rather than less egalitarian attitudes, leading them to more readily recognize and report examples of gender inequality. Furthermore, it is important to note that the scenarios used here concerned gender-neutral (rather than strongly sex-typed) activities and, as a result, may have led to a conservative test of the relation between sex typing and the perception of discrimination. That is, children who endorse egalitarian rather than traditional gender role beliefs should show more discrepant patterns of responding when the content of the scenarios is highly gender stereotyped than when it is gender neutral, a prediction that could be empirically tested in future research.

Contrary to our hypotheses, the developmental difference measure (i.e., classification skill) did not predict attributions to discrimination independent of age. Thus, although 8- to 10-year-old children made more context-appropriate attributions to discrimination than 5- to 7-year-old children, this increased sensitivity to contextual information did not appear to be related to children’s developing classification ability. It is possible that children’s perceptions of discrimination are moderated by the development of other types of cognitive skills (such as social-perspective taking) and/or the acquisition of various types of knowledge about gender. For example, children’s knowledge of past gender inequalities within and across societies may increase with age, perhaps as a function of schooling, and may contribute to their understanding and perceptions of discrimination.

As with all research, there are limitations to the present study. Most significantly, we developed new measures of children’s perceptions of discrimination and used only one of many possible experimental designs. Additional research using alternative designs (e.g., sex-typed rather than neutral activities; between-subjects rather than within-subject manipulations of contextual information) will be important for ruling out possible methodological artifacts.In addition, the sample of children included in this study was largely racially and ethnically homogenous (i.e., the majority were European American). It will be important for future research to examine perceptions of discrimination among children of racial and ethnic minority groups.

As in any new area of research, there are a host of additional questions that need to be addressed. For example, we examined children’s perceptions of gender discrimination in situations that were hypothetical and involved other individuals (rather than the self) as the target of discrimination. Future research should explore at what age and under which circumstances children perceive themselves to be the target of gender discrimination. In addition, there are a number of additional parameters that might guide children’s perceptions of discrimination that need to be examined (e.g., the availability of objective criteria for performance judgments and comparison data). Nonetheless, this study represents an important first step in understanding the processes involved in children’s perceptions of discrimination. Continued research is likely to be useful for creating social and educational policies and for designing intervention strategies to prevent and remedy perceptions of discrimination among children.

Footnotes

1 We assessed whether exposure to a likely discrimination scenario primed children to perceive discrimination in a subsequent, ambiguous scenario.Specifically, we selected children who heard an ambiguous scenario as the second scenario (n = 36) and compared the open-ended responses to that ambiguous scenario of children who had heard a likely discrimination scenario as the initial scenario (n = 19) and children who had heard an unlikelydiscrimination scenario as the initial scenario (n = 17). There were no significant differences in responding (Ms = 0.26 and 0.24, respectively).

2 To test for scenario effects, we examined children’s open-ended attributions to discrimination, as well as their ratings of the veracity of the experimenter-provided attributions, for each scenario by conducting a 2 (age group) × 2 (gender of participant) × 2 (gender of target) × 6 (scenario) analysis of variance (ANOVA), with the last variable treated as a within-subject variable. Overall, children made more open-ended attributions to discrimination in the scenario involving the footrace (M = 0.45) than in the other scenarios (average M = 0.23). A similar pattern of findings occurred for the ratings of the accuracy of discrimination (footrace scenario, M = 3.03; other scenarios, M = 2.6).

3 Such responses indicate that gender was the basis on which the teacher determined the students’ outcomes, and as such, the teacher’s behavior constitutes a form of discrimination. It is possible, however, that children viewed the teacher’s behavior as justifiable (rather than unfair), perhaps based on a view of one gender’s superiority to the other gender. This possibility was explored further in the analyses of children’s rating of the experimenter-provided responses.

4 The partial eta squared statistic was selected as the measure of effect size, as opposed to the eta squared or omega squared statistic, because it allows for the comparison of effect sizes across the two ANOVAs (i.e., for the open-ended responses and experimenter-provided responses), which have a differing number of factors, and is most appropriate for repeated measures ANOVAs (Cohen, 1977).

References

Baron, R. S., Burgess, M. L., & Kao, C. F. (1991). Detecting and labeling prejudice: Do female perpetrators go undetected?Personality and Social Psychology Bulletin, 17, 115–123.

Bowman, P., & Howard, C. (1985). Race-related socialization, motivation, and academic achievement: A study of Black youths in three-generation families. Journal of the American Academy of Child Psychiatry, 24, 134–141.

Cohen, J. (1977). Statistical power analysis for the behavioral sciences. New York: Academic Press.

Crocker, J., Voelkl, K., Testa, M., & Major, B. (1991). Social stigma: The affective consequences of attributional ambiguity. Journal of Personality and Social Psychology, 60, 218–228.

Dion, K. L. (1975). Women’s reactions to discrimination from members of the same and opposite sex. Journal of Research in Personality, 9, 294–306.

Dion, K. L., Earn, B. M., & Yee, P. H. (1978). The experience of being a victim of prejudice: An experimental approach. International Journal of Psychology, 13, 197–214.

Dovidio, J. F., & Gaertner, S. L. (1986). Prejudice, discrimination, and racism: Historical trends and contemporary approaches. In J. F.Dovidio & S. L.Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 1–34). Orlando, FL: Academic Press.

Feldman Barrett, L., & Swim, J. K. (1998). Appraisals of prejudice and discrimination. In J.Swim & C.Stangor (Eds.), Prejudice: The target’s perspective(pp. 12–37). San Diego, CA: Academic Press.

Fishbein, H. D. (1996). Peer prejudice and discrimination: Evolutionary, cultural, and developmental dynamics. Boulder, CO: Westview Press.

Garcia Coll, C., Crnic, K., Lamberty, G., Wasik, B. H., Jenkins, R., Vazquez Garcia, H., & McAdoo, H. P. (1996). An integrative model for the study of developmental competencies in minority children. Child Development, 67, 1891–1914.

Good, T. L., Sikes, J. N., & Brophy, J. E. (1973). Effects of teacher sex and student sex on classroom interaction. Journal of Educational Psychology,65, 74–87.

Inhelder, B., & Piaget, J. (1964). The early growth of logic in the child. New York: Norton.

Jensen, L., & Gutek, B. A. (1982). Attributions and assignment of responsibility for sexual harassment. Journal of Social Issues, 38, 121–136.

Jones, L. C., & Bigler, R. S. (1996, March). Cognitive-developmental mechanisms in the revision of gender stereotypic beliefs.Poster presented at the biennial meeting of the Conference on Human Development, Birmingham, AL.

Jones, M. G., & Wheatley, J. (1990). Gender differences in teacher-student interactions in science classrooms. Journal of Research in Science Teaching,27, 861–874.

Killen, M., Lee-Kim, J., McGlothlin, H., & Stangor, C. (2002). How children and adolescents evaluate gender and racial exclusion. Monographs of the Society for Research in Child Development, 67( 4, Serial No. 271).

Killen, M., & Stangor, C. (2001). Children’s social reasoning about inclusion and exclusion in gender and race peer group contexts. Child Development,72, 174–186.

Kowalski, K., & Kanitkar, K. (2003, April). Ethnicity and gender in the kindergarten classroom: A naturalistic study.Poster presented at the meeting of the Society for Research in Child Development, Tampa, FL.

Leaper, C. (Ed.). (1994). Childhood gender segregation: Causes and consequences. (New Directions for Child Development No. 65). San Francisco:Jossey-Bass.

Leaper, C. (2000). The social construction and socialization of gender during development. In P. H.Miller & E. K.Scholnick (Eds.), Toward a feminist developmentalpsychology (pp. 127–152). New York: Routledge.

Levy, G. D., Sadovsky, A. L., & Troseth, G. L. (2000). Aspects of young children’s perceptions of gender-typed occupations. Sex Roles, 42, 993–1006.

Liben, L. S., & Bigler, R. S. (2002). The developmental course of gender differentiation: Conceptualizing, measuring, and evaluating constructs and pathways. Monographs of the Society for Research in Child Development, 67( 2, Serial No. 269).

Liben, L. S., Bigler, R. S., & Krogh, H. R. (2001). Pink and blue collar jobs: Children’s judgments of job status and job aspirations in relation to sex of worker. Journal of Experimental Child Psychology, 79, 346–363.

Liben, L. S., & Signorella, M. L. (1980). Gender-related schemata and constructive memory in children. Child Development, 51, 11–18.

Maccoby, E. (1998). The two sexes: Growing up apart, coming together. Cambridge, MA: Belknap Press of Harvard University Press.

Major, B., Quinton, W. J., McCoy, S. K., & Schmader, T. (2000). Reducing prejudice: The target’s perspective. In S.Oskamp (Ed.), Reducing prejudice and discrimination (pp. 211–237). Mahwah, NJ: Erlbaum.

Mazur, D. B., & Percival, E. F. (1989). Ideology or experience? The relationship among perceptions, attitudes, and experiences of sexual harassment in university students. Sex Roles, 38, 167–185.

Meece, J. L. (1987). The influence of school experiences on the development of gender schemata. In L. S.Liben & M. L.Signorella (Eds.), Children’s gender schemata (New Directions for Child Development No. 38, pp. 89–105). San Francisco, CA: Jossey-Bass.

Nesdale, A. R., & McLaughlin, K. (1987). Effects of sex stereotypes on young children’s memories, predictions, and liking. British Journal of Developmental Psychology, 5, 231–241.

Patchen, M. (1982). Black-White contact in schools: Its social and academic effects. West Lafayette, IN: Purdue University Press.

Powlishta, K. (1995). Intergroup processes in childhood: Social categorization and sex role development. Developmental Psychology, 31, 781–788.

Quintana, S. M. (2001, April). Development of children’s understanding of ethnicity and race.Paper presented at the meeting of the Society for Research in Child Development, Minneapolis, MN.

Quintana, S. M., & Vera, E. M. (1999). Mexican American children’s ethnic identity, understanding of ethnic prejudice, and parental ethnic socialization.Hispanic Journal of Behavioral Sciences, 21, 387–404.

Radke, M. J., & Sutherland, J. (1972). Children’s concepts and attitudes about minority and majority American groups. In A. R.Brown (Ed.), Prejudice in children (pp. 138–156). Springfield, IL: Thomas Books.

Rosenberg, M. (1979). Conceiving the self. New York: Basic Books.

Ruble, D. N., & Martin, C. L. (1998). Gender development. In W.Damon (Series Ed.) & N.Eisenberg (Vol. Ed.), Handbook of child psychology: Vol. 3.Social,emotional,and personality development (5th ed., pp. 933–1016). New York: Wiley.

Signorella, M. L., Bigler, R. S., & Liben, L. S. (1993). Developmental differences in children’s gender schemata about others: A meta-analytic review.Developmental Review, 13, 147–183.

Spencer, M. B. (1985). Cultural cognition and social cognition as identity correlates of Black children’s personal-social development. In M. B.Spencer, G. K.Brookins, & W. R.Alen (Eds.), Beginnings: The social and affective development of Black children (pp. 215–230). Hillsdale, NJ: Erlbaum.

Spencer, M. B., & Markstrom-Adams, C. (1990). Identity processes among racial and ethnic minority children in America. Child Development, 61, 290–310.

Swim, J. K., Aikin, K. J., Hall, W. S., & Hunter, B. A. (1995). Sexism and racism: Old-fashioned and modern prejudices. Journal of Personality and Social Psychology, 68, 199–214.

Swim, J. K., & Cohen, L. L. (1996). Overt, covert, and subtle sexism: A comparison between the Attitudes Toward Women and Modern Sexism scales.Psychology of Women Quarterly, 21, 103–118.

Swim, J. K., Cohen, L. L., & Hyers, L. L. (1998). Experiencing everyday prejudice and discrimination. In J.Swim & C.Stangor (Eds.), Prejudice: The target’s perspective (pp. 37–61). San Diego, CA: Academic Press.

Theimer, C. E., Killen, M., & Stangor, C. (2001). Young children’s evaluations of exclusion in gender-stereotypic contexts. Developmental Psychology,37, 18–27.

U.S. Department of Education, Office for Civil Rights. (2000). Office for Civil Rights fiscal year 2000 annual report to Congress: Guaranteeing equal access to high-standards education. Washington, DC: Author.

Verkuyten, M., Kinket, B., & van der Weilen, C. (1997). Preadolescents’ understanding of ethnic discrimination. The Journal of Genetic Psychology, 158, 97–112.

Weiner, B. (2000). Intrapersonal and interpersonal theories of motivation from an attributional perspective. Educational Psychology Review, 12, 1–14.

APPENDIX

All scenarios depict the condition with the male perpetrator and the female target. The italicized sentences provide information about the likelihood of discrimination. When the scenarios involved ambiguous discrimination, the italicized sentences were omitted.

Scenario 1

Mr. Allen is an art teacher. Each year, Mr. Allen had an art contest. Mr. Allen wanted to pick someone with an original drawing. Mr. Allen almost always judges boys’ drawings to be better than girls’ drawings [or Mr. Allen usually judges boys’ and girls’ drawings to be equally good]. Michael and Jacqueline entered Mr. Allen’s annual art contest. They both consider themselves to be good artists. Michael loves to paint, and Jacqueline enjoys watercolors.Both Michael and Jacqueline worked equally hard for two weeks on their projects. Mr. Allen decided to give Michael 1st prize and Jacqueline 2rd prize.

Open-ended question:

Why did Michael win 1st prize and Jackie win 2nd prize?