Received 23 June 2016; accepted 24 July 2016; published 27 July 2016
As minority populations in the United States continue to rise, multicultural education is a growing necessity in society. It is estimated that by the year 2050, the majority of the U.S. will be represented by populations previously considered to be minorities (Romeny, 2008)  . This continually developing increase in diversity only heightens the need to address the stereotypes and misrepresentation of minorities that currently plague our society. Immigrants are typically viewed as responsible for problems with the economy, “overpopulation, pollution, increased violence, depleted social resources (i.e., medical and educational), erosion of cultural values, and terrorism” (Yakushko, 2009, p. 37)  , which intensifies the perception that these individuals are “criminal, poor, violent, desperate to live in the United States, and noneducated” (Yakushko et al., 2008, p. 375)  . Even elementary school aged children have been found to hold negative attitudes towards immigrants, especially those who are undocumented and/or Mexican, and believe that immigrants residing in the U.S. illegally should be imprisoned, regardless of their understanding that most migrate to the U.S. for freedom, better economic opportunities, and safety from the devastating events occurring in their home country (Brown, 2011)  . This intolerance of foreigners may be related to limited exposure to quality multicultural education. The available literature has focused on the development of biased or stereotypical attitudes in children and racial influences, with few studies examining the potential to increase favorable attitudes towards individuals of different races and ethnicities. As a result, there is a need to expand upon prior research aimed at eliminating racial biases by teaching children about the differing values of other cultures.
Early research has indicated that negative views of race and physical differences begin at young ages (Porter, 1971)  . Children can detect differences in appearance and skin tones as early as ages 3 and 4 (Bernstein, Zimerman, Werner-Wilson, & Vosberg, 2000  ; Aboud, 1987  ; Porter, 1971  ), and stereotypical attitudes may develop based on these learned physical differences. In addition, children also develop racial attitudes during early socialization, at home and in school. Research suggests that a mother’s implicit racial attitudes tend to be associated with their children’s emergent racial attitudes (Castelli, Zogmaister, & Tomelleri, 2009)  . A child’s family also begins to influence their racial attitudes early in development as children’s early attitudes often align with parents’ views (Njoroge, Benton, Lewis, & Njoroge, 2009)  , further demonstrating the significance of familial factors in racial attitude development.
Racial attitudes may be influenced as early as preschool. Phinney and Rotheram (1987)  suggest that the process of reducing biased attitudes begins with increased cultural knowledge and awareness including, same group and outgroup awareness. Specifically, ethnic awareness involves knowledge about ethnic groups, as well as the difference between oneself and others. Ethnic awareness changes as the child develops cognitive abilities and is exposed to new information and experiences (Phinney & Rotherman, 1897)  . Kowalski (1998)  demonstrated that increasing preschoolers’ ethnic/cultural awareness may in fact increase their identification with their own group. However, this increased in-group identity resulting from cultural knowledge and awareness did not lead to outgroup biases. Patterson and Bigler (2006)  found that preschool children exposed to group separation and labeling by colors (e.g. red or blue) tended to develop outgroup biased attitudes. Additionally, a study by Aboud (1987)  found that the process of self-identification becomes apparent when an individual is learning about new cultures and increasing their cultural awareness. These studies suggest that interventions that increase cultural knowledge might also increase self-identity with a group.
In sum, there are limited interventions that have been studied specifically aimed at multicultural education and at reducing racial biases. The increased awareness from exposure to multicultural groups (Jordan & Hernandez-Reif, 2009)  provides a useful framework for this study. The current study expands upon previous classroom interventions and continues efforts in this area by focusing on the attitudes towards alternative races/ ethnicities. The purpose of this study is to determine whether a classroom intervention aimed at increasing multicultural awareness influences preschoolers’ attitudes towards children of other races and cultural backgrounds. It is hypothesized that the group receiving the multicultural intervention will demonstrate more favorable attitudes and decreased negative attitudes towards children of other cultures compared to children in the control condition. In addition, it is expected that participants will view other members of their own culture with similar attitudes or increased positive views as they become more aware of cultural groupings as indicated in previous research (Kowalski, 1998)  . These hypotheses, if supported, may offer further evidence in favor of education as a tool in eliminating cultural or racial bias.
Participants consisted of sixteen preschool children ages three-to five-years-old (9 male, 7 female, mean age = 4.0, SD = 0.365) who were enrolled in a university-affiliated preschool center. Parents reported child’s race/ ethnicity on the demographic form. Based on parent report, participants were white/Caucasian (n = 10), Hispanic/Latino (n = 4), and black/African American (n = 2). All participants were born in the United States and spoke English as their primary language. Regarding parent’s employment, 62.5% of mothers and 80% of fathers reported having full time employment. All of the mothers and a majority of fathers (87.5%) reported having a bachelor’s degree or higher. A majority of participants were from families of above average socioeconomic status (SES) with an average income category of $85,000 to 99,999 (see Table 1 for detailed demographic descriptions).
Approval from the university’s Institutional Review Board was obtained prior to the commencement of the study. Two preschool classrooms were selected for the intended population due to children’s ability to recognize physical differences as young as 3 year of age (Bernstein et al., 2000  ; Aboud, 1987  ; Porter, 1971  ). Classroom teachers approached parents and provided a letter describing the study and a parent consent form. In addition, parents were provided a demographic form to complete and return prior to the study commencement. Each child, whose parents consented to participate, was individually asked for verbal assent. Out of 24 students whose parents provided parental consent, 17 parents (71%) returned consent forms and agreed to participate. Of the initial 17 students with parental consent, 16 (94%) verbally assented and were included in the study. Children who did not receive parental consent or provide verbal assent engaged in an alternate activity with their classroom teacher during the study’s lessons.
2.2.2. Assessment Procedure
Assessment prior to the intervention was administered by the primary investigator and a research assistant, both doctoral students in clinical psychology. Post-intervention assessments were conducted only by the graduate research assistant, who was blind to the intervention condition of participants. Assessors alternated approaching participants from each classroom to conduct assessments. The assessment was introduced to each child and they were offered a sticker for participation in the assessment activity. Each child was tested individually a week be-
Table 1. Demographic characteristics of the sample (n = 16).
Note: Categories not adding to 100% represent areas in which participants did not respond.
fore and a week after the intervention period. The assessors were trained to conduct the assessment, including practice administration and scoring. Training included a practice assessment with three children (ages 3 - 6). Assessments lasted between 15 to 20 minutes per child.
2.2.3. Intervention Procedure
Setting parameters restricted randomization of the individual participants, therefore each classroom was randomly assigned to lesson content that is either multicultural stories in the experimental condition or animal stories in the control condition. Each classroom contained eight students (16 total) who participated in the study. Groups did not differ on demographic variables. Each group participated in a 30-minute lesson, twice a week for 8 weeks. The structure of the lessons was consistent between the two classrooms. Storybook readings and open discussion comprised the lessons, followed by the time spent on interactive activities reflecting the ideas and concepts covered in the stories. The primary investigator and another doctoral student in clinical psychology, who did not conduct assessments, conducted the bi-weekly lessons. In the opening and closing weeks of the multicultural intervention, instructors discussed general multicultural themes, while specific cultural groups were discussed each week for six consecutive weeks, in the following order: African Americans, Asian Americans, Arab and Middle Eastern Americans, Latin Americans/Hispanics, and Indian and American Indians. The control classroom lessons were themed around different animals and animal characteristics (e.g. different types of animals, animal families, how animals eat and sleep) in order to provide the same format and number of sessions without presenting the experimental content (e.g. cultural information).
2.3.1. Demographic Questionnaire
The demographic questionnaire asked parents for background information about the family. Questions included child race/ethnicity, parent occupation and education, parent and child nationality, family religion, language spoken at home, and income.
2.3.2. Racial Attitudes Assessment
Variables derived from the assessment included: Same group bias: This variable measures the child’s attitudes towards the same race. Same group positive bias was calculated by counting the number of positive adjectives endorsed for photos of the same race, with a possible range from 0 to 8, and divided by 8.Same group negative bias was calculated by counting the number of negative adjectives endorsed for photos of the same race, with a possible range from 0 to 8, and divided by 8. Same group bias was measured by subtracting same group negative bias from same group positive bias. Scores range from −1 to 1, with higher values indicating more positive attitudes towards the same race.
Outgroup bias: This variable measures attitudes towards other races. Outgroup positive bias was measured by counting the number of positive adjectives endorsed for photos of children of any other race (besides the child’s own), with a possible range from 0 to 24, and divided by 24. Outgroup negative bias was measured by counting the number of negative adjectives endorsed for photos of children of any other race (besides the child’s own), with a possible range from 0 to 24, and divided by 24. Outgroup group bias was calculated by subtracting outgroup negative bias from outgroup positive bias. Scores range from −1 to 1, with higher values indicating more positive attitudes towards alternate races other than one’s own.
2.4. Data Analysis
Descriptive statistics and frequencies were used to examine demographic variables (e.g. ethnicity, primary language, parental employment and SES, as well as parental education). These can be found in Table 1. The mean and standard deviation of each variable of interest derived from the Racial Attitudes Assessment were examined for both groups. Results were analyzed using a one-way within-subjects ANOVA design to examine each variable of interest, including same group bias and outgroup bias, where time (pre and post intervention) was the within-subjects factor and intervention group (multicultural theme or control) was the between-subject factor.
Means and standard deviations for each variable of interest (same group bias, outgroup bias) for each condition measured pre-and post-intervention can be seen in Table 2.
3.1. Same Group Bias
The results for the ANOVA did not indicate a significant interaction effect for the intervention group over time, Wilks’ Λ = 0.749, F(1, 14) = 0.107, ns = 0.749, multivariate η2 = 0.008. Therefore, on average, same group bias did not significantly change due to the intervention.
3.2. Outgroup Bias
The results for the ANOVA indicated a significant intervention group over time interaction effect, Wilks’ Λ = 0.740, F(1, 14) = 4.913, p = 0.044, multivariate η2 = 0.260. Therefore, outgroup bias significantly changed due to the intervention. In other words, participants’ attitudes towards other races changed positively after receiving the multicultural education compared to the participants’ attitudes in the control group. Mean changes over time for both groups are represented in Figure 1.
Although previous literature has considered the impact of increased awareness and exposure on racial attitudes, few have done so with very young children. This study expanded research in this area by examining explicit attitudes with preschool children. Exposure to children of other cultures, especially in a favorable manner, would presumably increase positive attitudes towards others (Klefstad & Martinez, 2013)  . In this study, it was hypothesized that the group receiving the multicultural intervention would demonstrate more favorable attitudes and decreased negative attitudes towards children of other cultures. In order to minimize the effects of attention and the story arrangement of the lessons, a control group was conducted in the same manner with neutral animal content.
As expected, the results indicated significant changes in outgroup attitudes from pre- to post-intervention in the multicultural group when compared to attitudes in the control group. Children in the multicultural (i.e., experimental) group demonstrated improved outgroup bias as a result of the intervention when compared to children in the animal (i.e., control) group. The mean differences between groups were significant, however, it should be
Table 2. Means and standard deviations of each variable for each group, pre and post intervention.
Note: *Significant mean difference pre vs. post across groups; p < 0.05.
Figure 1. Means of each group (intervention and control) for outgroup bias variable compared across time (pre and post). Group means, −1 to 1, represent the overall outgroup bias, which was calculated by subtracting outgroup negative bias from outgroup positive bias. Higher scores indicate more positive attitudes towards alternate races other than one’s own.
noted that the multicultural group exhibited a lower overall mean at pre-test for outgroup attitudes. This difference at the initial assessment was likely due to responses of three participants in which each chosen adjective during the assessment was negative. The same pattern was not found in the control group between pre- and post-assessments. These findings replicate Bernstein et al. (2000)  using a similar methodology for increasing ethnic/racial awareness and improving racial attitudes. Similar to Bernstein et al. (2000)  , children in our study were exposed to multiple cultures through the use of stories and demonstrated improved attitudes after the intervention compared to children who did not receive the intervention. Thus, storytelling can be an effective method to introduce children to other cultures and expand their knowledge and tolerance toward children of races other than their own. However, we believe that assessing racial attitudes by asking children to assign negative or positive attributes to photos of children of racially diverse backgrounds provided a more robust method than that used by Bernstein et al. (2000)  , which relied on a simple assortment of photos and the strategy used to categorize the photos (e.g., by race, age, or gender) was decided upon by the researchers.
While these results suggest a beneficial effect for reducing racial bias, some limitations of the current study deserve mention. Given that the sample size was relatively small, these findings should be considered preliminary and require replication in a larger sample. In particular, replications of these findings are needed within larger samples to control for alternate variables, such as the participant’s prior exposure to other cultural groups (e.g., friends/relatives, parental attitudes, books in household, knowledge or tolerance of other groups). In this study, two intact classes were chosen for either the intervention or control group, which allowed the experimental and control conditions to follow particular curriculums according to the purposes of each condition. Future studies could randomly assign more classrooms to each condition in order to increase the applicability of random assignment. While the study consisted of 16 lessons, these provided children with limited exposure to each cultural group. The benefits of the study may be enhanced with increased exposure towards various cultural groups. Finally, the Racial Attitudes Assessment was adapted from previously used measures for the purposes of this study. The instrument’s psychometric properties should be examined in a larger sample where it can be validated for use with children of various ages and ethnicities.
Children are often only exposed to the environments they grow up in and are likely unaware of the attitudes they may learn about other cultures. These attitudes are not necessarily a conscious decision, but may be a result of known traditions or implicit beliefs. Increased contact has resulted in more positive beliefs regarding other groups when children are given the opportunity to become familiar with other groups at a young age (Cameron et al., 2006  ; Bernstein et al., 2000  ; Phinney & Rotherman, 1897  ). Given that negative stereotypes regarding cultural minorities remain present in our society (Yoo & Pituc, 2013)  , educational efforts would benefit from an increased multicultural perspective. More work has to be done examining the best way to integrate this multicultural perspective into the classroom, especially on how to apply cultural education into the established curriculum. For example, barriers such as feasibility (e.g. costs, time) and inclusion into existing class- room material should be addressed. There have been recent efforts, in line with this study’s findings, to encourage incorporation of multicultural books into preschool curriculum and to enhance cultural awareness (Klefstad & Martinez, 2013)  . Overall, this study offers promising findings regarding the role that education can play in reducing negative racial biases in young children.
 Romney, P. (2008) Consulting for Diversity and Social Justice: Challenges and Rewards. Consulting Psychology Journal: Practice and Research, 60, 139-156.
 Yakushko, O. (2009) Xenophobia: Understanding the Roots and Consequences of Negative Attitudes toward Immigrants. The Counseling Psychologist, 37, 36-66.
 Yakushko, O., Backhaus, A., Watson, M., Ngaruiya, K. and Gonzalez, J. (2008) Career Development Concerns of Recent Immigrants and Refugees. Journal of Career Development, 34, 362-396.
 Brown, C.S. (2011) American Elementary School Children’s Attitudes about Immigrants, Immigration, and Being an American. Journal of Applied Developmental Psychology, 32, 109-117.
 Bernstein, J., Schindler-Zimmerman, T., Werner-Wilson, R. and Vosburg, J. (2000) Preschool Children’s Classification Skills and a Multicultural Education Intervention to Promote Acceptance of Ethnic Diversity. Journal of Research in Childhood Education, 14, 181-192.
 Aboud, F.E. and Ruble, D.N. (1987) Identity Constancy in Children: Developmental Processes and Implications. In: Honess, T. and Yardley, K., Eds., Self and Identity: Perspectives across the Lifespan, Routledge, New York, 95-107.
 Castelli, L., Zogmaister, C. and Tomelleri, S. (2009) The Transmission of Racial Attitudes within the Family. Developmental Psychology, 45, 586-591. http://dx.doi.org/10.1037/a0014619
 Njoroge, W., Benton, T., Lewis, M.L. and Njoroge, N.M. (2009) What Are Infants Learning about Race? A Look at a Sample of Infants from Multiple Racial Groups. Infant Mental Health Journal, 30, 549-567. http://dx.doi.org/10.1002/imhj.20228
 Smith, C.O., Levine, D.W., Smith, E.P., Dumas, J. and Prinz, R.J. (2009) A Developmental Perspective of the Relationship of Racial-Ethnic Identity to Self-Construct, Achievement, and Behavior in African American Children. Cultural Diversity and Ethnic Minority Psychology, 15, 145-157.
 Grøver Aukrust, V. and Rydland, V. (2009) “Does it Matter?” Talking about Ethnic Diversity in Preschool and First Grade Classrooms. Journal of Pragmatics, 41, 1538-1556.
 Kowalski, K. (1998) The Impact of Vicarious Exposure to Diversity on Preschooler’s Emerging Ethnic/Racial Attitudes. Early Child Development and Care, 146, 41-51.
 Patterson, M.M. and Bigler, R.S. (2006) Preschool Children’s Attention to Environmental Messages about Groups: Social Categorization and the Origins of Intergroup Bias. Child Development, 77, 847-860. http://dx.doi.org/10.1111/j.1467-8624.2006.00906.x
 Cameron, L. and Rutland, A. (2006) Extended Contact through Story Reading in School: Reducing Children’s Prejudice toward the Disabled. Journal of Social Issues, 62, 469-488.
 Jordan, P. and Hernandez-Reif, M. (2009) Reexamination of Young Children’s Racial Attitudes and Skin Tone Preferences. Journal of Black Psychology, 35, 388-403.
 Spencer, M. (1996) Racial Attitudes, Beliefs, and Stereotypes Measure. Center for Health, Achievement, Neighborhood, Growth, and Ethnic Studies (CHANGES) at the University of Pennsylvania, Philadelphia, PA.
 Cameron, L., Rutland, A., Brown, R. and Douch, R. (2006) Changing Children’s Intergroup Attitudes toward Refugees: Testing Different Models of Extended Contact. Child Development, 77, 1208-1219.
 Yoo, H.C. and Pituc, S.T. (2013) Assessments of Perceived Racial Stereotypes, Discrimination, and Racism. American Psychological Association, Washington DC. http://dx.doi.org/10.1037/14048-025