Data and Code for "Within-Group Rank Effect Sizes Read as Intervention Effects: A Worked Example of Effect-Size Inversion in a Pre-Test/Post-Test Comparison-Group Design"

Abstract

This repository contains the anonymized dataset and Python source code associated with the study “Within-Group Rank Effect Sizes Read as Intervention Effects: A Worked Example of Effect-Size Inversion in a Pre-Test/Post-Test Comparison-Group Design.” The materials are provided to support transparency, reproducibility, and independent verification of the analyses reported in the study. Files- `dataset_anonymised.csv` – item-level pre-test and post-test scores for 40 students (IDs T01–T20 treated, C01–C20 comparison). Twenty scoring units per administration; totals on a 0–100 scale. No personal information.- `analysis.py` – reproduces Tables 2–5 and all statistics reported in Sections 3.1–3.5 and 4.3. Output saved in `analysis_output.txt`.- `simulation.py` – Monte Carlo study (Section 2.5, Table 6). Output saved in `simulation_results.csv`. Runtime about 1 minute.- `figures.py` – reproduces Figures 1 and 2.- `requirements.txt` – software versions used. Random seed: 20240101 (NumPy PCG64 via `numpy.random.default_rng`). Bootstrap: 10,000 percentile resamples drawn within condition. Simulation: 20,000 replications per scenario. Run: `python analysis.py`, `python simulation.py`, `python figures.py` from this folder.