Abstract
ObjectiveOftentimes, political scientists wish to measure how attitudes or behaviors have changed from one generation to the next. One obstacle to this analysis is that certain attitudinal and behavioral variables (e.g., political participation or policy preferences) correlate with the age of the respondent. In seeking to model both age‐ and cohort‐based changes in attitudes, researchers are left having to impose constraints within the generalized linear model framework to solve problems of underidentification.MethodsThis article uses a method that has been developed to work around this exact problem: the intrinsic estimator.ResultsBy using this approach to traditional age‐period‐cohort (APC) models, the IE is one solution to underidentified models of political change.ConclusionsThis article summarizes the identification problem in APC methodology and then illustrates the advantages of the IE to the traditionally used constrained generalized linear modeling (CGLIM) approach with specific applications to trends in American political attitudes regarding spending on space exploration.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have