98 3 Methods Between March and June 2020, we conducted a survey, administered by survey agency Kantar Public, among 905 Dutch citizens, 597 of whom have a migration background in Turkey, Morocco and Surinam, of which 234 identify as Muslim. The survey agency knows the country of birth of the parents of the respondents from previous research, and oversampled respondents with a background in Turkey, Morocco or Surinam. Through government requests for research, Kantar had permission to gather data using the GBA (Gemeentelijke Basisadministratie, the municipal personal files of all inhabitants) as a sampling frame. It is very uncommon for researchers to have access to the GBA; usually, they have to rely on far less exhaustive sampling frames. After recruiting respondents for government research, Kantar is allowed to save the contact information of the respondents in their survey panel for future research. The survey agency invited respondents of their panel to join our survey through an email with a link to the survey, and there was an overall response rate of 54%. Participants received the equivalent of two euros for their participation, which they could use to buy small items in a gift shop. Even though approaching respondents via email might exclude some potential respondents, the unique reliance on the only possible complete sampling frame in the Netherlands (GBA) makes Kantar’s sampling superior to most other research. We asked all respondents about their ethnic and religious identification. For ethnic identification we asked: “In terms of my ethnic group, I consider myself to be…(max. 2 answers)”. We presented our respondents a list of 13 answer categories, including Turkish, Moroccan, Surinamese, Hindustani and Dutch (see Appendix 1 for the full list). The last questions of the survey were about religious identification. We asked: “Do you consider yourself as belonging to any particular religion or denomination?” If the respondent answered yes, we followed up with “Which one?” allowing respondents to answer “Christian, Muslim, Hindu, Buddhist, Jewish, Other, [specify]”17. Respondents were able to indicate that they identified with a max of two ethnic groups, of which one could be “Dutch,” and one religion. Table 1 shows the exact number of each group of respondents based on their migration backgrounds, and the percentage of which identified as Dutch, an ethnic minority group or belonging to a religion. For each ethnic group and religion respondents selected, the respondents then received a list of four statements with answers ranging from 0 (disagree) to 10 (agree), which together form an ethnic in-group favouritism scale (Bizumic et al., 2009). Respondents received this battery of four statements between zero and three times, depending on how many ethnic or religious groups they identified with. Table 2 lists these four statements measuring ethnic in-group favouritism. Respondents who did 17 European Social Survey (2016) ESS Source Questionnaire Round 8 2016/2017. 01(2016).
RkJQdWJsaXNoZXIy MjY0ODMw