Approaches to the analysis of behavioral medicine data with few individuals but many repeated observations are discussed. Difficulties with common analytic strategies are noted and applications from econometrics using cross-sectional pooled time series designs are recommended. Additional complications due to outliers, missing data, measurement error, and aggregation bias are discussed.
No takes yet. Share an insight, caveat, or question.
Jaccard et al. (1993) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: