Two-wave panel models and causal analysis: A criminological example

This paper illustrates some general analytic procedures which may be helpful in assessing various assumptions underlying alternative models in the causal analysis of panel data. The role of two basic mechanisms of change, temporal inertia and entropic decay, is specifically examined. The models developed are evaluated using data on changes in the politicalization of 172 male felons over a six-month period following their incarceration in a maximum security institution in the southeast. Several issues and problems in drawing causal inferences from panel data are addressed.