Marginal Estimation of Parameter Driven Binomial Time Series Models

TL;DR

This paper develops asymptotic theory for parameter estimation in binomial time series models with latent dependence, showing that marginal likelihood methods yield consistent estimates even with autocorrelated latent processes.

Contribution

It provides theoretical proof that marginal likelihood estimation is consistent and asymptotically normal for binomial time series with latent dependence, despite bias in GLM estimates.

Findings

Marginal likelihood estimation reduces bias in parameter estimates.

Bias persists in binary data even with many trials, over 45% in long series.

Simulations confirm theoretical results and practical implications.

Abstract

This paper develops asymptotic theory for estimation of parameters in regression models for binomial response time series where serial dependence is present through a latent process. Use of generalized linear model (GLM) estimating equations leads to asymptotically biased estimates of regression coefficients for binomial responses. An alternative is to use marginal likelihood, in which the variance of the latent process but not the serial dependence is accounted for. In practice this is equivalent to using generalized linear mixed model estimation procedures treating the observations as independent with a random effect on the intercept term in the regression model. We prove this method leads to consistent and asymptotically normal estimates even if there is an autocorrelated latent process. Simulations suggest that the use of marginal likelihood can lead to GLM estimates result. This problem reduces rapidly with increasing number of binomial trials at each time point but, for binary data, the chance of it can remain over 45% even in very long time series. We provide a combination of theoretical and heuristic explanations for this phenomenon in terms of the properties of the regression component of the model and these can be used to guide application of the method in practice.