The following is an excerpt of the intervention methodology. For more information, please see the full text of the article on the publisher's website or through your institution's library.

Sample [...] The sample focused on three cities—Antwerp itself, Mechelen, and Boom—with two to three times more people of Moroccan descent than the Belgian average. [...] A total of 1025 students in 63 classes participated, with on average 15.51 students per class and 65.13 students per school. Of these, 977 surveys were deemed valid. Across schools, 17% of the students had at least one parent born in Morocco. Exactly half of the students were male (50%). They were aged 13–21, with most aged 15 (45%) or 16 (35%). [...]

Procedure In spring 2016, the participating classes read book fragments together out loud during two course hours (100 min) of their Dutch class. [...] The class then read from one of two relatively easy-to-read adolescent novels by well-known Flemish author Dirk Bracke. Book A—read by 48% of the classes—is the ‘white-Belgian-majority’ novel Blinde woede (‘Blind rage’). It describes how 17-year-old Lena—a talented singer who is gradually going blind due to a hereditary disease—deals with school, love, and friendship issues. The book also has two male protagonists, but no ethnic-minority characters. [...] Book B—read by 52% of the classes—is the ‘Moroccan-Belgian’ novel Henna op je huid (‘Henna on your skin’). It tells the story of 15-year-old secondgeneration Moroccan girl Touria and her brothers, who negotiate between their family’s culture and encounters with Belgian-majority members. Although these encounters are mostly friendships among peers, some describe situations of discrimination. [...] The participating classes were thus first randomly distributed over the white-Belgian-majority and Moroccan-Belgian conditions, and as a second step matched so that classes in both conditions would be similar in their distribution of the class’ track, percentage of male students, and percentages of students from non-Western and—more specifically—Moroccan backgrounds. A ‘posttest only’ design was used: the students only completed an anonymous survey during the Dutch class following the experiment (usually one to three days later). [...]

Measures

Dependent variables Students’ attitude toward Moroccans and attitude toward Belgians were measured via two feeling thermometers with 0° implying very negative feelings and 100° very positive feelings. [...] In addition, students’ awareness of discrimination against Moroccan youngsters was measured with four five-point-scale items [...].

Independent variables As mentioned, students’ ethnicity was approached subjectively, allowing them to self-identify as Belgian, Moroccan, both or neither. [...] At the classroom level, the independent variables were the two types of interethnic contact. Vicarious interethnic contact through book reading was operationalized through the experimental manipulation use of a Moroccan-Belgian book. [...] The second type of interethnic contact, face-to-face contact, was operationalized with two continuous variables. Which of these two was used in the analyses depends on the hypothesis under scrutiny. Students’ face-to-face contact with Moroccan classmates was measured through the class Moroccan concentration, which is the class percentage of students with at least one parent born in Morocco. [...]

Control variables Three class-level controls were used [...]. These include teachers’ ethnic background [...] and age [...]. [...] Finally, the class’ track was also considered [...]. In addition, three student-level controls were used. Firstly, students’ sex [...]. Secondly, their reading skill was measured [...]. Finally, students’ parental cultural status was included [...].

Manipulation check A manipulation check was added to test that differences between the Moroccan-Belgian and white-Belgian-majority condition could not be ascribed to one book being more engaging than the other. [...]