Abstract
We discuss the motivation for integrating qualitative information in statistical models of complex, partially observable causal mechanisms and suggest ways to minimize the dangers posed by Braumoeller and Kirpichevsky. We stress the importance of linking our statistical estimators to underlying theories of the data generating process, qualitative or otherwise.
Original language | English (US) |
---|---|
Pages (from-to) | 280-291 |
Number of pages | 12 |
Journal | Political Analysis |
Volume | 13 |
Issue number | 3 |
DOIs | |
State | Published - 2005 |
ASJC Scopus subject areas
- Sociology and Political Science
- Political Science and International Relations