TREO-Based Team Formation: A Double Optimization Framework for Educational Research

This repository contains the R-based computational framework and example dataset for forming optimized student teams using the Team Role Evaluation Orientation (TREO) model. The tool is designed for educators and researchers who wish to compare different grouping strategies—specifically Balanced (maximizing complementarity) and Overlap (maximizing redundancy) conditions. The script performs a Double Optimization process by iterating through multiple threshold values ($T$) and group sizes ($N$) to find the configuration that maximizes the statistical separation ($S$) between team conditions. This ensures that experimental groups are robustly differentiated for subsequent performance analysis. Files Included Script.docx (or .R): The complete R code for data standardization, diversity diagnostics, team formation algorithms, and automated figure generation (SVG). Example_data.xlsx: A standardized template containing two parallels (Sheets A and B) with anonymized student data and TREO scores across six dimensions: Organizer, Doer, Challenger, Innovator, Team Builder, and Connector. Instructions for Users (README Context) 1. Data Preparation Your Example_data.xlsx must follow this structure to work with the script: Columns: N°, Sex, ID_student, and the six TREO roles (Organizer, Doer, Challenger, Innovator, Team_Builder, Connector). Anonymization: Ensure ID_student is a local reference (e.g., A001, A002) to protect student privacy. Sheets: The script expects two sheets named "A" and "B" representing different class parallels or cohorts. 2. Methodology Flow The algorithm follows a rigorous five-step process: Standardization: Calculation of Z-scores based on the grand mean of the combined parallels. Diversity Diagnosis: Calculation of the Effective Role Richness (R_ef) and the Role Diversity Index (IDR) based on Shannon Entropy to assess the sample's variety. Simulation Loop: Testing of thresholds (from 0.25 to 4.0) and group sizes (from 3 to 7). Optimization: Selecting the (T, N) pair that yields the highest Separation Score (S), defined as: S = (CI_{Balanced} - CI_{Overlap}) + (OI_{Overlap} - OI_{Balanced}) where CI is the Complementarity Index and OI is the Overlap Index. Output: Automated generation of high-quality SVG plots and CSV assignment lists.