Internet-based intervention on maternal sleep and anxiety: a randomized clinical trial.

The underlying data for this study consist of anonymised, processed student-level records from two quasi-/randomized controlled experiments evaluating the U-Behavior teaching and learning method in undergraduate Science, Technology, Engineering, and Mathematics (STEM) courses. These datasets include retrieval practice activity (RPA) usage logs (timestamps of quiz attempts, scores, derived spacing and mixing behaviour scores), U-Behavior (UB)-rubric levels, total RPA attempts, exam performance (midterm/final exams), overall course grades and prior academic performance covariates [high school Grade Point Average (GPA) and/or university GPA]. All data were extracted from the university’s learning management system (LMS) and course records and then de-identified to comply with institutional review board approvals, FERPA regulations and participant consent. Raw individual timestamps and identifiable information are not shared. Instead, the deposited datasets provide the processed, aggregated and summarized variables used for all statistical analyses reported in the manuscript (e.g. spacing scores, mixing scores, UB-rubric levels, total attempts, exam scores, regressions, descriptive statistics and effect sizes). These allow independent verification of the key findings – significant increases in spacing and interleaving under U-Behavior, no overall condition differences in exam performance and predictive relationships between spacing quality and learning outcomes – and replication of the reported regressions and group comparisons. The datasets have been deposited in Dryad, in accordance with Access Microbiology’s Open Data policy.