JSSM · 2020
Demonstrates how adaptive testing methods can dramatically reduce the number of survey questions needed to measure personality traits without sacrificing measurement accuracy. The …
UAI · 2020
Proposes GPIRT, a Gaussian process extension of item response theory that relaxes parametric assumptions about how respondents relate to latent traits. The model provides more …
SAGE Handbook of Research Methods in Political Science and International Relations · 2020
An invited handbook chapter providing a comprehensive overview of Bayesian approaches to model selection, comparison, and averaging for political scientists. Covers practical …
PSRM · 2019
Compares Bayesian and maximum likelihood approaches to estimating treatment effects in bivariate probit models with instrumental variables. The analysis provides practical guidance …
PA · 2015
Develops a statistically principled approach to detecting electoral fraud using forensic analysis of vote return distributions. By incorporating prior knowledge about legitimate …
IJF · 2014
Addresses the challenge of calibrating ensemble forecasting models when training data are sparse, a common problem in social science prediction. The paper develops methods for …
PA · 2013
Introduces computerized adaptive testing (CAT) methods to public opinion surveys, allowing researchers to measure latent attitudes more efficiently. By tailoring question selection …
PS · 2013
A brief methodological contribution outlining principled approaches to combining data from multiple sources through ensemble and aggregation methods. Argues for a coherent Bayesian …
PA · 2012
Extends Bayesian model averaging to ensemble methods, showing how combining multiple models improves prediction in political science. The paper demonstrates practical gains in …
PS · 2012
Applies ensemble forecasting methods to predict the outcome of the 2012 US presidential election. Demonstrates the practical utility of combining multiple models, with the ensemble …