Changing Children’s Attitudes Through Story Time

In recent years, tensions have emerged in the south of England over the integration of refugees into mainstream British society. One way of easing this process may be ensuring that refugee children encounter a welcoming environment when they enter British schools. The present research tested an intervention that aimed to improve English schoolchildren’s attitudes towards refugees via an “extended contact effect.” This effect suggests that attitudes towards outgroup members can be improved through vicarious experiences of friendships between ingroup and outgroup members. In the current study, this was achieved by having children read and discuss stories featuring friendships between English and refugee children. While extended contact has been shown to be effective for adults and older children, little research has been done with children ages 5 to 11.

1. What They Did – Intervention Summary:

Researchers were interested in testing three different versions of the extended contact hypothesis. In one condition, children read stories that emphasized the individual qualities of the refugee characters, and then discussed these characters’ similarities and differences. The idea here was to encourage the children to think of refugees as unique individuals rather than homogenous members of an outgroup.

In another condition, the children read stories in which the refugee characters attended the children’s actual school, offering the English and refugee children a common identity. While the goal of the previous condition was to emphasize the uniqueness of each outgroup member, here the objective was to give outgroup members ingroup status by expanding the boundaries of the ingroup.

The third condition was similar to the common identity condition, except the intention here was to give all children a shared group with which to identify without minimizing their subgroup (English or refugee) identity. Thus, the stories read emphasized both the shared identity between the English and refugee children (as students attending the same school) and their differences.

1-2 weeks after the interventions, children were interviewed to assess their attitudes towards both ingroup and outgroup members. They were presented with a series of positive and negative traits (e.g. “hardworking) that they were asked to match with pictures of stick people representing different proportions (none, some, most, all) of group members, both English and refugee. The researchers were also interested in the children’s intended behavior. They presented them with 2 scenarios—one featuring an English child, and another with a refugee child—and asked the children to indicate how much they’d like to play with that child, have that child spend the night, etc.

2. What They Found – Results:

Children in all extended contact conditions expressed more positive attitudes towards refugee children than did children in the no-stories control group. As the researchers had predicted, attitudes were most positive in the dual identity condition. Children in the dual identity condition who did not have strong ties to their English identities also expressed more positive intended behaviors towards refugee children.

3. Who Was Studied – Sample:

253 White British children (116 boys, 137 girls) from 10 elementary schools, ages 5 to 11 years, from south-east England

4. Study Name:

Changing Children’s Intergroup Attitudes Toward Refugees: Testing Different Models of Extended Contact

5. Citation:

Cameron, L., Rutland, A., Brown, R., & Douch, R. (2006).  Changing Children’s Intergroup Attitudes Toward Refugees: Testing Different Models of Extended Contact. Child Development, 77, 1208-1219.

6. Link:

7. Intervention Categories:

Extended Contact

8. Sample Size:


9. Central Reported Statistic:

“outgroup attitude in the extended contact conditions was significantly higher than in control, t = 2.89, p < .01.”

“In low identifiers there was a significant main effect of condition, F(3,104) = 2.75, p < .05, MSE = 1.26. Planned contrasts revealed that control did not differ from the three extended contact variables (C1: t = 1.07, ns), but dual identity was, as before, significantly higher than the other two extended contact conditions (C2: t = 2.67, p < .01)”