1step,rule,n_before,n_after,n_removed,rationale21,studies listed in data/studies.json with an extractable printed value,5,5,0,all five extracts contain the requested printed quantity32,harmonisable onto theta = ln(T_AI/T_noAI),5,4,1,S5 removed: DiD coefficient on GitHub output counts around a ban is not a task-completion-time AI vs no-AI comparison43,poolable (SE obtainable from printed quantities),4,2,2,S2 removed (CI only in unreproduced Figure 1; no SE/p printed); S3 removed (significance stars only)5