This chapter describes the procedures used to scale the International Civic and Citizenship Education Study (ICCS) questionnaire data (for students, teachers, and schools) and the indices based on them. In general, it is possible to distinguish two general types of indices derived from the ICCS questionnaires: 1. Simple indices constructed through arithmetical transformation or recoding, for example, ratios between teachers and students; and 2. Scale indices derived from scaling of items, a process typically achieved by using item response modeling of dichotomous or Likert-type items. The first part of this chapter lists the simple indices that were derived from the ICCS data and describes how they were created. The second part outlines the scaling procedures used in ICCS. The third and final part, lists the scaled indices, along with statistical information on item parameters, scale reliabilities, and the factor structure of related item sets. The cross-country validity of item dimensionality and constructs was assessed during the field trial stage of ICCS. At this time, data were used to assess the extent to which measurement models held across participating countries. Extensive use was made of both confirmatory factor analysis and item response modeling, a process that made it possible to examine cross-national measurement equivalence before conducting the final selection of main survey questionnaire items.
Available at: http://works.bepress.com/tim_friedman/7/