37 France Germany Netherlands Religion Islam 177 Islam 123 Islam 234 Christianity 282 Christianity 287 Christianity 143 Non-religious 701 Non-religious 502 Non-religious 462 Other 39 Other 42 Other 64 To assess the migration background of the respondents, I asked them about the birthplaces of their parents. I had to ask this question first for sampling purposes. In order to avoid any potential ordering effects, I randomized the order in which the policy questions, propensity to vote questions (dependent variable Chapter 3) and experimental profiles were presented to the respondents. Additionally, I used a randomized wording approach to the policy questions to minimize acquiescence bias. For example, half of the respondents saw: “the government should raise the taxes for the rich” while the other half saw “the government should lower the taxes for the rich.” This approach also aligned with the policy positions expressed by the politicians in the experimental profiles. I used two kinds of conjoint experiments. First, I randomized the migration background, religion, and gender and subsequently asked which policy position respondents expect from the profile of the politicians I presented them (dependent variable Chapter 4 (van Oosten, 2022)). Second, I randomized the migration background, religion, gender, and policy positions of the politicians. For a complete list of the attributes and values used in the politician profiles, please see Table 2. After presenting one politician, I asked respondents whether they felt represented by the politician, trusted the politician, and felt they were competent to perform well on the job. After presenting two politicians, I asked the respondents to indicate which politician they were most likely to vote for (dependent variable Chapter 2, (van Oosten, 2023c)), see image 1. I repeated this process three times, resulting in six profile views per respondent. To measure religion, I included it as the final variable since it is typically considered an exogenous variable that is less likely to change based on preceding questions. The survey also interrogated participants’ sense of belonging to their respective countries and ethnic groups, as well as their perceptions of differences with various groups and preferences for interactions based on ethnicity and religion. Religious affiliation and denomination were explored, followed by participants’ views on intergroup interactions based on religion and the extent to which they prioritized their religion’s interests. Overall, the survey provided insight into participants’ historical political engagement, identity perceptions, past voting behavior, intergroup attitudes across countries, and its intersections. While planning these online survey experiments I did not know I would be collecting data during the outbreak of a pandemic. Fortunately, replications of online surveys Table 1. Continued.
RkJQdWJsaXNoZXIy MjY0ODMw