Text this: Balancing Rigor, Replication, and Relevance: A Case for Multiple-Cohort, Longitudinal Experiments