Balancing data needs in pedestrian dynamics experiments: crowd size, number of trials, and trial duration

Abstract

Determining the required data for pedestrian dynamics studies is essential but complex. Researchers need to decide on (1) the number of participants required per trial (i.e., the crowd size) and (2) the number of observations needed for robust results. Using illustrative examples from published pedestrian dynamics studies, we discuss estimating crowd size, number of trials, and trial duration. Critically, these considerations depend on the level of measurement, with different approaches needed for microscopic versus macroscopic measurements (the latter is the focus of this contribution). Clarity in the research questions, the unit of measurement, and the analysis before data is being collected is therefore critical.