Experience sampling method studies in physical activity research: the relevance of causal reasoning

Background. The experience sampling method (ESM), also known as ecological momentary assessment, is gaining popularity in physical activity research. This method involves assessing participants’ behaviors and experiences repeatedly over time. One key advantage of ESM is its ability to temporally separate the dependent and independent variable of interest, reducing the risk of reverse causality. However, temporal separation alone is insufficient for establishing causality. This methodology paper illustrates the importance of the identification phase in drawing causal conclusions from ESM data. In the identification phase the causal effect of interest (or estimand) is specified and the assumptions under which a statistical association can be considered as causal are identified. Causal directed acyclic graphs (DAGs) visualize these assumptions by displaying the assumed causal structure between the independent variable, the dependent variable and other related variables. Methods. The Methods section demonstrates how to construct an estimand and develop a DAG for a specific research question in an ESM study. Results. In the Results section, we describe how confounding bias revealed by the DAG can be addressed by modifying ESM study design or data-analysis. Each step is illustrated with an example from the field of physical activity research. Conclusions. Implementing the identification phase before conducting data analysis enables ESM researchers to make more informed decisions, thereby enhancing the validity of causal inferences in studies aimed at answering causal questions.