We are delighted to announce that both Luis Ledesma and Di Shan successfully defended their Masters’ thesis in September.
Luis’ thesis was titled “Intercept estimation of semi-parametric joint models in the context of longitudinal data subject to irregular observations“; in it Luis showed how to extend previous multiplicative models for longitudinal outcomes and irregular observation times to estimate intercepts. Besides providing valuable information on prognosis, this also reduces computation time.
Di’s thesis was titled “Variable Categorization and Selection in the Context of Inverse-Intensity Weighting for Longitudinal data”; Di showed that adding variables associated with the outcome but not the intensity to the intensity model improves the precision of inverse-intensity weighted GEEs, while adding variables associated with the intensity but not the outcome to the intensity model inflates the variance of regression coefficients from inverse-intensity weighted GEE
