Dropout is common in longitudinal studies and can depend on the unobserved outcomes, even after conditioning on observed data. When this is the case, dropout is considered missing not at random and nonignorable, requiring that the relationship be accounted for in the analyses. Frequently, the possibility of nonignorable dropout is disregarded and naïve methods that ignore the dropout mechanism are employed. This can lead to biased results and incorrect conclusions. Mixture models are a class of models that can be used to account for the relationship between a longitudinal outcome and dropout. Join us to learn about two mixture model methods that can be implemented using standard statistical software, the Conditional Linear Model and the Natural Spline Varying-coefficient model.
*This is a recorded webinar. A link to the Webinar will be sent via email once purchase information is confirmed.