A meta-analysis estimating initial-final correlation in randomized clinical trials (RCTs) and a simulation study evaluating the performance of statistical methods in RCTs with repeated measures
Tipus de documentProjecte Final de Màster Oficial
Condicions d'accésAccés obert
This master thesis is split into two separate main parts. Part I is a meta-analysis of randomized clinical trials (RCTs) and Part II is a simulation study. First, a meta-analysis was performed through a literature search of RCTs of the years 2010 and 2013. The goal was to confirm results of a preliminary work which analysed RCTs from the years 2004 and 2007. Second, different data sets of RCTs with repeated measurements and a continuous outcome were simulated with varying sample sizes, effect sizes, residual variances, random slope variances and correlation structures. The goal was to compare statistical methods that can be applied to longitudinal data.