Time-varying effect modeling to address new questions in behavioral research: Examples in marijuana use.

Abstract

Time-varying effect modeling (TVEM), a statistical approach that enables researchers to estimate dynamic associations between variables across time, holds enormous potential to advance behavioral research. TVEM can address innovative questions about processes that unfold across different levels of time. We present a conceptual introduction to the approach and demonstrate four innovative ways to approach time in TVEM to advance research on the etiology of marijuana use. First, we examine changes in associations across historical time to understand how the link between marijuana use attitudes and marijuana use behavior has shifted from 1976 to present; gender differences in the relevance of attitudes diminished over time and were no longer significant after 2004. Second, we examine age-varying associations between heavy episodic drinking and marijuana use across developmental time and demonstrate that this dynamic association is substantially stronger during ages 14 to 16 compared to later ages. Third, we explore the complex association between age of onset of marijuana use and adult marijuana use to identify precise age ranges during which the onset of use is most risky, and demonstrate how this complex association is more salient for males. Finally, we examine changes in marijuana use as a function of time relative to the birth of first child and show how this transition is more crucial for females. All empirical examples in this methodological demonstration rely on existing data from cross-sectional or panel studies. We conclude with thoughts on future directions for the application and further development of TVEM in behavioral research.

Time-Varying Outcomes Associated With Maternal Age at First Birth

“…This statistical approach allowed for the examination of the influence of age as a continuous variable on the selected outcome variables. Allowing for the calculation of regression coefficients as a function of continuous time improves upon other statistical methods by enabling us to observe if and how outcomes vary with time (Lanza, Vasilenko & Russell, 2016). Moreover, TVEM does not use parametric constraints to determine the output, essentially allowing the data to "speak" in a more nuanced manner than other statistical methods (e.g., linear regression modeling).…”

Is Young Adulthood a Critical Period for Suicidal Behavior among Sexual Minorities? Results from a US National Sample

“…Second, we used weighted logistic regression to calculate the association between anti-LGB discrimination and recent suicidal behavior among SM adults and, separately, by biological sex and for SMs who did and did not identify as LGB. Third, we used logistic time-varying effect modeling (TVEM; Dziak et al 2017) to estimate age-varying prevalences of recent suicidal behavior and associations between anti-LGB discrimination and recent suicidal behavior (see Lanza et al 2016). With cross-sectional data, TVEM facilitates the identification of precise age ranges during which certain characteristics are associated with heightened risk for an outcome .…”

Shifting Age of Peak Binge Drinking Prevalence: Historical Changes in Normative Trajectories Among Young Adults Aged 18 to 30

“…These methods are very useful but impose parametric forms for all associations and typically assume that the observed associations are consistent over time (see also Patrick et al, 2017a). When seeking to determine whether the strength of a particular association changes across time, time-varying effect modeling (TVEM) provides an alternative modeling approach (Lanza et al, 2016;Li et al, 2015;Tan et al, 2012). TVEM models (described in greater detail below) allow for not only the behavior of interest (i.e., binge drinking) to change across time (i.e., age), but also the possible effects of covariates to change across time (e.g., gender over age), with no assumptions of parametric form for the observed changes.…”