TVEM represents an important statistical method for modeling non-linear, dynamic processes. However, TVEM models the average-level process for the focal effect. As addiction- and recovery-related processes demonstrate person-specific heterogeneity, we encourage researchers to model non-parametric and non-ergodic change to advance our understanding of how, when, and for whom change occurs.
Keywords: addiction; dynamic; heterogeneity; recovery; statistical modeling; time-varying effect modeling.