A hierarchical model for the analysis of intra-individual variability

Download PDF

When analyzing and reporting results of studies involving multiple participants, researchers are often faced with two choices: average data across participants to estimate group-level effects or treat each participant separately. Neither of these two options is entirely satisfactory: averaging data across participants ignores inter-individual differences and treating each participant as a separate entity ignores commonalities across participants. Hierarchical (i.e., multi-level) Bayesian models (HBMs) provide a principled solution to this conundrum. We show that, in most cases, estimates of performance or of treatment effects computed using HBMs are more precise and more accurate than estimates of performance obtained using traditional methods.

Leave a Reply

Your email address will not be published. Required fields are marked *