Hyunkeun (Ryan) Cho
Department of Statistics
Western Michigan University
The varying coefficient model plays an important role in exploring the dynamic changes of regression coefficients over time. In this article, we employ a kernel smoothing method to capture time-dependent covariate effects for both discrete and continuous longitudinal data under the framework of generalized linear models. The proposed procedure takes into account the correlations among measurements within the same subject. This approach does not assume the exact knowledge of the true correlation structure nor estimate the parameters involved in the informative working correlation structure. The resulting estimator is more efficient than the one under an independent working correlation structure. Theoretical results show that the proposed method yields an asymptotically normal and consistent estimate. The methodology is illustrated through simulation studies and the analysis of a randomized controlled longitudinal data.
All statistics students are expected to attend.