- In the post test, the experimental group indicates a significantly higher number and proportion of cross-race friendships than did the control groups. Black students named more White friends than White students named Black friends.
- The percentage of minority students in class was not significantly related to the number or proportion of cross-race friendships.
- However, the effect of multiracial learning groups was more pronounced in classes with high or low percentages of Black students than in racially balanced classrooms.
- One pair of classes was analyzed in nine month follow-up. The experimental students made cross-race friendship choices three times as often as the control group students.
- The proportion of cross-race friendships made by experimental group students was four times that of the control group students.
- The research supports multiracial learning teams as a strategy to improve cross-racial attraction, thereby improving race relations in desegregated environments.