Discrete Measurement, Continuous Time and Event History Modeling Public Deposited

Analytics

Downloadable Content

Download PDF
Request Version for Screen Reader
Last Modified
  • March 20, 2019
Creator
  • Desmarais, Bruce A.
    • Affiliation: College of Arts and Sciences, Department of Political Science
Abstract
  • Most even history models used in political science assume the time being analyzed is continuous. Discrete measurement causes this assumption to be violated. The violation of this assumption is shown to introduce non-trivial bias to parameter estimates. Analysis of discrete-measured data as interval-censored is shown to greatly reduce this bias. The empirical properties of the bias introduced by discrete measurement and the interval-censoring correction are explored through Monte-Carlo simulations and a replication of the analysis of civil war duration from (Fearon 2004). I also demonstrate that analyzing discrete-measured continuous-time data as interval-censored is a better approach than the discrete-time models proposed in (Box-Steffensmeier and Jones 2004). The conclusion of the analysis is that event-history analysis of continuous-time variables should always be implemented as interval-censored estimation.
Date of publication
DOI
Resource type
Rights statement
  • In Copyright
Advisor
  • Carsey, Thomas M.
Language
Access
  • Open access
Parents:

This work has no parents.

Items

Thumbnail Title Date Uploaded Visibility Actions
Discrete measurement, continuous time and event history modeling 2019-04-07 Public