Date: Wednesday, Oct 19th, 3 PM

Location: Room 49 Hardin Hall North

Title: Modeling Repeated Measures Count Data Part I: How do autocorrelated count data arise?

Presented by: Martin Frenzel

ABSTRACT:

Models for analyzing repeated measures count data vary widely.  Comparison of these models is made difficult by the fact that many of the models are not fully parametric; they do not fully describe the probability process that gives rise to autocorrelated count data.    If we wish to compare methods for analyzing these data, we must first establish some reasonable probability processes that give rise to such data.  A mixture of observation driven and latent process approaches will be described and compared.