RESUMO
Researchers are sometimes interested in predicting a distal or external outcome (such as smoking cessation at follow-up) from the trajectory of an intensively recorded longitudinal variable (such as urge to smoke). This can be done in a semiparametric way via scalar-on-function regression. However, the resulting fitted coefficient regression function requires special care for correct interpretation, as it represents the joint relationship of time points to the outcome, rather than a marginal or cross-sectional relationship. We provide practical guidelines, based on experience with scientific applications, for helping practitioners interpret their results and illustrate these ideas using data from a smoking cessation study.
RESUMO
Intensive longitudinal data (ILD) have become increasingly common in the social and behavioral sciences; count variables, such as the number of daily smoked cigarettes, are frequently-used outcomes in many ILD studies. We demonstrate a generalized extension of growth mixture modeling (GMM) to Poisson-distributed ILD for identifying qualitatively distinct trajectories in the context of developmental heterogeneity in count data. Accounting for the Poisson outcome distribution is essential for correct model identification and estimation. In addition, setting up the model in a way that is conducive to ILD measures helps with data complexities - large data volume, missing observations, and differences in sampling frequency across individuals. We present technical details of model fitting, summarize an empirical example of patterns of smoking behavior change, and describe research questions the generalized GMM helps to address.