Discrete models for survey data

Methods and models to analyze relationships between two or more categorical data presented in this dissertation are applied in four phases of survey data processing. These phases include collecting data, estimating averages and totals, analyzing relationships between variables, and making data available to third parties.

This research includes not only the analyzes of the relationship between the variables measured in the survey, but also the analysis of interviewer effects and other measurement errors, the analysis of the effects of different modes of data collection, estimation of the means of subpopulations with few available observations and assessing the risk of revealing the identity of one or more respondents.

Pannekoek, J. (1992). Discrete models for survey data. Dissertation, University of Amsterdam. (Only available in print)