This document is part of a series of chapters described in SO 011 759. This chapter reports the results of Monte Carlo simulations designed to analyze problems of using maximum likelihood estimation (MLE: see SO 011 767) in research models which combine longitudinal and dynamic behavior data in studies of change. Four complications--censoring of observations, small sample size, collinearity among causal variables, and model misspecification -- influence the quality of all ML estimates of parameters in a log-linear rate model when data consists of lengths of time between events. Separate sections of the document outline the model studied, formally present the method of maximum likelihood estimation, discuss previous findings on the quality of ML estimates of parameters of rate models, and describe the method and results of generating the Monte Carlo data. Results of the Monte Carlo experiments indicate that the quality of MLEs of a correctly specified log-linear rate model is generally good. For the correctly specified model, censoring seems to be the most serious impediment to correct inference, but can be compensated for by increasing the sample size. Findings also indicate that misspecification noticeably reduces the quality of estimates, and that sample size, censoring, and collinearity affect quality quite differently in the misspecified and correctly-specified models. (Author/KC)