The Dynamic Measurement-Burst Model: A method for assessing psychological process features at multiple timescales.

Longitudinal research in psychology has traditionally been approached with aneither-or attitude: Either we are able to study momentary dynamics of psychological processesusing intensive longitudinal data (ILD), or we are able to study individual differences inlong-term change trajectories using panel data (PD). As a consequence, theoretical knowledgeof micro-timescale daily processes is disintegrated from knowledge about meso- andmacro-level (i,e, trait-level) knowledge regarding the stability and change in psychologicalconstructs. The foundational work by John Nesselroade proposes a solution for this problem:measurement burst data (MBD) contains repeated waves of ILD, allowing researchers toinvestigate how momentary processes in ILD change over long periods of time. In this paper,we propose a novel modeling strategy for these data, which unites two established frameworks:integrating the multilevel first-order autoregressive model (ML-AR(1)M) for ILD and therandom-intercept cross-lagged panel model (RI-CLPM) for PD, we propose a new modelingapproach, the dynamic measurement burst model (DMBM). We illustrate how this approachreveals nuanced patterns of individual variance in micro-level process features, bysimultaneously examining their meso-level dynamics across bursts, and their macro-levelbetween-person associations. In an empirical example we further demonstrate the feasibilityand substantive value of the DMBM for advancing the understanding of complex psychologicalprocesses. We discuss future directions for this approach and end with a broadened reflectionof the relative nature of temporal lenses in psychological research.