DMCA
Democracy as a latent variable (2008)
| Venue: | American Journal of Political Science |
| Citations: | 68 - 1 self |
BibTeX
@ARTICLE{Treier08democracyas,
author = {Shawn Treier and Simon Jackman},
title = {Democracy as a latent variable},
journal = {American Journal of Political Science},
year = {2008}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Measurement is critical to the social scientific enterprise. Many key concepts in social-scientific theories are not observed directly, and researchers rely on assumptions (tacitly or explicitly, via formal measurement models) to operationalize these concepts in empirical work. In this paper we apply formal, statistical measurement models to the Polity IV data, a set of country-level indicators of democracy. In so doing, wemake explicit the hitherto implicit assumptions underlying scales built using the Polity indicators. We apply two models: one in which democracy is operationalized a latent continuous variable, and another in which democracy is operationalized a latent class. We show how tobetter exploit the information in thePolity data set so as to produce amore reliable scale measure (or classification) of democracy. Our modeling approaches also let us assess the ‘‘noise’ ’ (measurement error) in our resulting measure of democracy. We show that this measurement error is considerable, and has substantive consequences when using a measure of democracy as an independent variable in cross-national statistical analysis. Our analysis suggests that skepticism as to the precision of the Polity democracy scale is well-founded, and that many researchers have been overly sanguine about the properties of the Polity democracy scale in applied statistical work. 1 Latent Variables Abound in Political Science Social and political theories often refer to constructs that can not be observed directly. Exam-ples include public opinion, socio-economic status, social capital, ideology, or democracy. Instead of observing these quantities, researchers may have indicators of these concepts,
Keyphrases
latent variable measurement error polity democracy scale social capital formal measurement model political science social resulting measure substantive consequence latent continuous variable polity iv data empirical work modeling approach applied statistical work country-level indicator latent class independent variable political theory many key concept statistical measurement model hitherto implicit assumption social scientific enterprise reliable scale measure thepolity data public opinion many researcher cross-national statistical analysis socio-economic status polity indicator latent variable abound social-scientific theory
