Assessing Cross-National Equivalence of Measures of Xenophobia: Evidence from Probing in Web Surveys

When items eliciting attitudes or beliefs toward immigrants are used in cross-national research, equivalence of measurement across countries is a necessary prerequisite. Equivalence requires that respondents in all countries have at least ‘‘comparable’’ groups of immigrants in mind, although this might include different groups in terms of ethnicity, education, or legal status, for instance. The ideal situation that respondents have exactly the same groups of immigrants in mind is, obviously, com-pletely unrealistic in countries in which either the composition of immigrants is dif-ferent or when the coverage of immigration issues by the media and in national politics is qualitatively or quantitatively different. In this article, we will focus on four items regarding beliefs on immigrants taken from the International Social Survey Program (ISSP) 2003 questionnaire on ‘‘National Identity.’ ’ The items relate to whether immigrants increase crime rates, whether they are generally good for the economy, whether they take jobs away from native people, and whether they improve society by bringing in new ideas and cultures. Several things could happen when respondents answer these items: (1) Respondents