Application of random-effects regression models in relapse research

摘要: This article describes and illustrates use of random-effects regression models (RRM) in relapse research. RRM are useful longitudinal analysis data since they allow for the presence missing data, time-varying or invariant covariates, subjects measured at different timepoints. Thus, can deal with "unbalanced" where a sample not all each every timepoint. Also, recent work has extended to handle dichotomous ordinal outcomes, which common Two examples presented from smoking cessation study illustrate using RRM. The first logistic model, examining changes status, treating status as an outcome. second example focuses on motivation scores prior following smoking. latter how be used examine predictors consequences relapse, occur any