An adaptation of the Child Occupations and Traits (COAT) scale (Liben & Bigler, 2002; 0.99) measures stereotypes re: gender and race intergroup categories. Participants respond to three types of items (e.g., how smart or not smart; how hardworking or lazy; or friendly or mean) and this measure was also modified to utilize a 6 point rating scale. There are two sub-categories, gender and race. The questions probe the degree to which children assign traits to specific groups. The version we utilized contained 12 items, 6 about race and 6 about gender and pilot testing revealed that this version had a high internal reliability for the full scale (0.919), as well as for the gender and the race subcomponents (0.859 and 0.849, respectively).

Scores ranged from a minimum of 1 to a maximum of 6. A higher score indicates more positive (better) trait attributions about peers.

This set of items is adapted from Liben & Bigler (2002). These items probe participant perceptions of the likelihood of group members obtaining high status occupations. Participants were asked "Which kid do you think will grow up to be a [pilot/doctor/lawyer/scientist]?" with one item for each profession. Participants selected which child of 8 possible choices they thought would become the profession. The 8 choices included a boy and a girl of each of the following racial groups: Asian, Black, Latine, White. The measure was made binary by whether participants selected an underrepresented minority (URM) (Black or Latine) in high status professions, or whether they selected a race group that is not underrepresented (White or Asian). The goal of the intervention was that more students in the treatment condition selected a URM student after the intervention program. Therefore in the fields below, the category Underrepresented Minority (URM) indicates the number of participants who select

The diversity of friendships will be assessed using an adaptation of the Friendship Questionnaire (Bierman & McCauley, 1987). The original measure has been widely utilized to assess the quality and quantity of children's peer interactions and has proven to be reliable (ranges from .72 to .82). The adaptation consisted of adding the intergroup categories of gender and race/ethnicity. This adapted version was also found to be internally reliable (0.659). Using gender, racial, and ethnic background, we will record how often they play with outgroup peers to determine the proportion of cross-gender, cross-race, and cross-ethnic playmate interactions.

Scores on the scale ranged from a minimum of 1 to a maximum of 6. Higher scores indicate a higher level of reported cross-group contact, which is the more desirable outcome.

This assessment consists of 5 separate items about how much the child wants to play with X, where X refers to pictures of girls, boys, and kids who are depicted as African American, European American, and Asian American). Response choices are provided on a 6 point rating scale ranging from 1 (really do not want to) to 6 (really want to). During pilot testing this measure was found to have a high internal consistency reliability (0.767). Using students' gender, ethnic, and racial background, we will record their desire to play with outgroup peers to determine the proportion of cross-gender, cross-race and cross-ethnic playmate preferences.

Scores on the scale ranged from a minimum of 1 to a maximum of 6. Higher scores indicate a higher desire for intergroup contact, the more desirable outcome.

This assessments consists of 8 items about whether the child experiences exclusion in a school context. Response choices are provided on a 5 point Likert-type rating scale ranging from "Never" to "Always." This measure has been validated in previous research (Killen et al., 2022).

Scores on the scale range from a minimum score of 1 and maximum score of 5. Higher scores indicate higher levels of personal experiences of exclusion, which is the less desirable outcome. Thus the goal of the intervention is to result in lower scores in the treatment group at posttest.

This assessment consists of 2 items repeated 4 times to reflect different race and gender combinations for a total of 8 items. This measure is adapted from Cooley et al., 2019. The items ask participants to evaluate exclusion in an intergroup and same group social context. Responses were on a 6-point Likert-type scale ranging from 1 (really not okay) to 6 (really okay).

Scores on this scale range from a minimum of 1 to a maximum of 6. A higher score on this scale indicates viewing exclusion as more acceptable and a lower score indicates viewing exclusion as less acceptable. The goal of the intervention is to promote lower scores, indicating participants are less accepting of exclusion.

This assessment consists of 5 items repeated 4 times to reflect different race and gender combinations for a total of 20 items. This measure is adapted from Cooley et al., 2019. The items ask participants to predict the likelihood of a child inviting someone to their birthday party, where the children they can invite may be of the same or a different race. Responses were on a 6-point Likert-type scale ranging from 1 (very unlikely) to 6 (very likely).

Scores on this scale range from a minimum of 1 to a maximum of 6. Higher scores indicate participants expect intergroup inclusion to be more likely, which is the more desirable outcome.

This scale is adapted from Liben and Bigler (2002). The scale probes participant perceptions of the inclusivity levels of different groups. These scales were created consistent with the literature on gender stereotypes (Berenbaum, Martin, & Ruble, 2015; Mulvey, Rizzo, & Killen, 2016). A 5 point rating scale ranging from 1 (none) to 5 (all) indicates the proportion of children within a hypothetical target group whom participants believe fit into a categorization. Pilot testing revealed 0.800 for the 5-item measure of inclusivity levels of different groups.

Scores on this scale range from a minimum of 1 to a maximum of 5. Higher scores indicate more positive beliefs about peers' inclusive behavior, the more desirable outcome.

This scale is adapted from Liben and Bigler (2002). The scale probes participant perceptions of the math and science abilities of different groups. This scale was created consistent with the literature on gender stereotypes (Berenbaum, Martin, & Ruble, 2015; Mulvey, Rizzo, & Killen, 2016). A 5 point rating scale ranging from 1 (none) to 5 (all) indicates the proportion of children within a hypothetical target group whom participants believe fit into a categorization. Pilot testing revealed 0.796 for the 5-item measure of math and science abilities of different groups.

Scores ranged from a minimum of 1 and a maximum of 5. Higher scores indicate more positive beliefs about peers' math and science abilities, which is the more desirable outcome.

The control group follows a Business-As-Usual plan (no "alternative" program is implemented for the control group.) They complete a pretest and posttest in the same weeks as the experimental group.

Total of all reporting groups

The control group follows a Business-As-Usual plan (no "alternative" program is implemented for the control group.) They complete a pretest and posttest in the same weeks as the experimental group.

This is a five-item measure including items about whether other students in my class care about me and other students in my class like me for who I am. The measure has been found to produce reliability figures of .92 (Van Ryzin, Gravely, & Roseth, 2009) and our modified version utilizing a 6-point response scale in place of the original 5-point scale found similar reliability scores (0.896). The rating scale ranged from 1 (really not true) to 6 (really true). The average of all items creates a score of perceived peer personal support in which higher scores indicate higher levels of perceived peer support. Performance on the Van Ryzin et al. (2009) scale has been demonstrated to have an independent positive effect on engagement in learning and has been validated by connections to academic engagement (Collie, Martin, Papworth, and Gins, 2016).

The degree to which students perceive their teachers as personally supportive will be assessed using this 4-item measure. Items include statements such as my teacher cares about me and my teacher likes me for who I am. This measure was also modified from the original rating scale to a 6-point rating scale, which ranged from 1 (really not true) to 6 (really true). Van Ryzin et al. (2009) reported an internal consistency reliability of .91, and our pilot testing found a similar internal consistency reliability, 0.893.

A composite score was computed using the mean of all items in the scale. Scores ranged from a minimum of 1 to a maximum of 6. A higher score indicated higher teacher support, the more desirable outcome.

An adaptation of the Child Occupations and Traits (COAT) scale (Liben & Bigler, 2002; 0.99) measures stereotypes re: gender and race intergroup categories. Participants respond to three types of items (e.g., how smart or not smart; how hardworking or lazy; or friendly or mean) and this measure was also modified to utilize a 6 point rating scale. There are two sub-categories, gender and race. The questions probe the degree to which children assign traits to specific groups. The version we utilized contained 12 items, 6 about race and 6 about gender and pilot testing revealed that this version had a high internal reliability for the full scale (0.919), as well as for the gender and the race subcomponents (0.859 and 0.849, respectively).

Scores ranged from a minimum of 1 to a maximum of 6. A higher score indicates more positive (better) trait attributions about peers.

This set of items is adapted from Liben & Bigler (2002). These items probe participant perceptions of the likelihood of group members obtaining high status occupations. Participants were asked "Which kid do you think will grow up to be a [pilot/doctor/lawyer/scientist]?" with one item for each profession. Participants selected which child of 8 possible choices they thought would become the profession. The 8 choices included a boy and a girl of each of the following racial groups: Asian, Black, Latine, White. The measure was made binary by whether participants selected an underrepresented minority (URM) (Black or Latine) in high status professions, or whether they selected a race group that is not underrepresented (White or Asian). The goal of the intervention was that more students in the treatment condition selected a URM student after the intervention program. Therefore in the fields below, the category Underrepresented Minority (URM) indicates the number of participants who select

The diversity of friendships will be assessed using an adaptation of the Friendship Questionnaire (Bierman & McCauley, 1987). The original measure has been widely utilized to assess the quality and quantity of children's peer interactions and has proven to be reliable (ranges from .72 to .82). The adaptation consisted of adding the intergroup categories of gender and race/ethnicity. This adapted version was also found to be internally reliable (0.659). Using gender, racial, and ethnic background, we will record how often they play with outgroup peers to determine the proportion of cross-gender, cross-race, and cross-ethnic playmate interactions.

Scores on the scale ranged from a minimum of 1 to a maximum of 6. Higher scores indicate a higher level of reported cross-group contact, which is the more desirable outcome.

This assessment consists of 5 separate items about how much the child wants to play with X, where X refers to pictures of girls, boys, and kids who are depicted as African American, European American, and Asian American). Response choices are provided on a 6 point rating scale ranging from 1 (really do not want to) to 6 (really want to). During pilot testing this measure was found to have a high internal consistency reliability (0.767). Using students' gender, ethnic, and racial background, we will record their desire to play with outgroup peers to determine the proportion of cross-gender, cross-race and cross-ethnic playmate preferences.

Scores on the scale ranged from a minimum of 1 to a maximum of 6. Higher scores indicate a higher desire for intergroup contact, the more desirable outcome.

This assessments consists of 8 items about whether the child experiences exclusion in a school context. Response choices are provided on a 5 point Likert-type rating scale ranging from "Never" to "Always." This measure has been validated in previous research (Killen et al., 2022).

Scores on the scale range from a minimum score of 1 and maximum score of 5. Higher scores indicate higher levels of personal experiences of exclusion, which is the less desirable outcome. Thus the goal of the intervention is to result in lower scores in the treatment group at posttest.

This assessment consists of 2 items repeated 4 times to reflect different race and gender combinations for a total of 8 items. This measure is adapted from Cooley et al., 2019. The items ask participants to evaluate exclusion in an intergroup and same group social context. Responses were on a 6-point Likert-type scale ranging from 1 (really not okay) to 6 (really okay).

Scores on this scale range from a minimum of 1 to a maximum of 6. A higher score on this scale indicates viewing exclusion as more acceptable and a lower score indicates viewing exclusion as less acceptable. The goal of the intervention is to promote lower scores, indicating participants are less accepting of exclusion.

This assessment consists of 5 items repeated 4 times to reflect different race and gender combinations for a total of 20 items. This measure is adapted from Cooley et al., 2019. The items ask participants to predict the likelihood of a child inviting someone to their birthday party, where the children they can invite may be of the same or a different race. Responses were on a 6-point Likert-type scale ranging from 1 (very unlikely) to 6 (very likely).

Scores on this scale range from a minimum of 1 to a maximum of 6. Higher scores indicate participants expect intergroup inclusion to be more likely, which is the more desirable outcome.

This scale is adapted from Liben and Bigler (2002). The scale probes participant perceptions of the inclusivity levels of different groups. These scales were created consistent with the literature on gender stereotypes (Berenbaum, Martin, & Ruble, 2015; Mulvey, Rizzo, & Killen, 2016). A 5 point rating scale ranging from 1 (none) to 5 (all) indicates the proportion of children within a hypothetical target group whom participants believe fit into a categorization. Pilot testing revealed 0.800 for the 5-item measure of inclusivity levels of different groups.

Scores on this scale range from a minimum of 1 to a maximum of 5. Higher scores indicate more positive beliefs about peers' inclusive behavior, the more desirable outcome.

This scale is adapted from Liben and Bigler (2002). The scale probes participant perceptions of the math and science abilities of different groups. This scale was created consistent with the literature on gender stereotypes (Berenbaum, Martin, & Ruble, 2015; Mulvey, Rizzo, & Killen, 2016). A 5 point rating scale ranging from 1 (none) to 5 (all) indicates the proportion of children within a hypothetical target group whom participants believe fit into a categorization. Pilot testing revealed 0.796 for the 5-item measure of math and science abilities of different groups.

Scores ranged from a minimum of 1 and a maximum of 5. Higher scores indicate more positive beliefs about peers' math and science abilities, which is the more desirable outcome.

The control group follows a Business-As-Usual plan (no "alternative" program is implemented for the control group.) They complete a pretest and posttest in the same weeks as the experimental group.