The simulation uses latent sub-competency values z in [0,1]. The fifth sub-competency is labelled interpretation and validation. Reported sub-competency scores are r = 1 + 3z on a stylised 1-4 scale, and the reported composite score is M = sum_k r_k on a 5-20 scale. Figure1_data.csv through Figure4_data.csv reproduce the four manuscript figures. Figure4_data.csv reports the full-integration sensitivity analysis used in the manuscript, while Scenario_sensitivity_summary.csv applies the same perturbation procedure to all four scenarios. Simulation_parameters.csv and Scenario_profiles.csv provide the baseline parameterisation and scenario exposure profiles. OC_Coding_Rubric.csv gives an operational coding guide for assigning modelling-opportunity level O and cross-curricular connection level C in later empirical applications, and OC_Coding_Reliability_Checklist.csv outlines a recommended two-coder or expert-panel reliability pathway before empirical O/C values are used for calibration. Common_random_numbers_design.csv documents the shared virtual cohort, shared initial residuals and shared semester shocks used across scenarios. Random seed: 20260510; synthetic individuals: 1000; semesters: 8. Sensitivity values are absolute_sensitivity_points under predefined perturbation ranges and should not be read as dimension-free standardized effects.
