Modelování panelových dat s dichotomickou závisle proměnnou: obecné principy a ilustrace v programu STATA

This article outlines several techniques for analyzing panel data with a dichotomous dependent variable. This presentation is inspired by the classic work of Paul Allison [1999]. An example analysis is presented where public attitudes toward restitution of church property in the Czech Republic is explored using panel data. Here the focus is on exploring changes in the intra-personal agenda of respondents on this specific issue. There are three main conclusions from this research: (1) media exposure and (2) the education level of the respondent increase the odds of the church restitution issue being mentioned by a respondent as being important, and (3) mention of the church restitution issue in a particular wave of the panel survey is negatively associated with mention of this issue in later waves of the panel study examined. These findings are discussed in terms of their methodological and substantive implications.