{"stable_id":"ae70ae8b90d164ae","slug":"nigeria-ai-tutoring","company":"Edo State schools / World Bank study","workflow_name":"Teacher-facilitated GPT-4 tutoring","function_code":"enrollment","pattern_codes":["creator_to_judge"],"changed_assumption":"A general model becomes a tutor only inside structured pedagogy.","evidence_strength":"verified","publication_tier":"showcase","freshness":"current","reviewed_at":"2026-08-23","updated_at":"2026-08-22","source_quality_summary":"2 independent; publication outcomes are verified.","caveat_summary":"World Bank Policy Research Working Paper, not formal peer review; 654 observations in the main outcome tables; short, supervised intervention among interested students.","freshness_caveat":null,"workflow_summary":{"before":[{"actor":"Public-school student","action":"Continues business-as-usual classes and study without the six-week after-school intervention.","actor_type":"control"},{"actor":"Classroom teacher","action":"Delivers ordinary curriculum and grades school assessments.","actor_type":"human"}],"hinge":"A general model becomes a tutor only inside structured pedagogy.","after":[{"actor":"Student pair using Microsoft Copilot/GPT-4","action":"Completes curriculum-aligned English, AI, and digital-skills activities in twelve 90-minute lab sessions.","actor_type":"system"},{"actor":"Teacher facilitator","action":"Introduces activities, monitors model use, and redirects shortcutting or unsuitable responses.","actor_type":"human"}],"decision_mode":"shared","decision_marker":"Human moves from creator to judge"},"before":[{"order":1,"actor":"Public-school student","action":"Continues business-as-usual classes and study without the six-week after-school intervention.","handoff_to":"Regular school assessment","control":"Randomized control condition; the paper does not claim a separate prior tutoring queue."},{"order":2,"actor":"Classroom teacher","action":"Delivers ordinary curriculum and grades school assessments.","handoff_to":"Student","control":"Human pedagogy and assessment authority."}],"after":[{"order":1,"actor":"Student pair using Microsoft Copilot/GPT-4","action":"Completes curriculum-aligned English, AI, and digital-skills activities in twelve 90-minute lab sessions.","handoff_to":"Facilitating teacher","control":"Carefully crafted prompts; pair work; intermittent connectivity/electricity are documented failure modes."},{"order":2,"actor":"Teacher facilitator","action":"Introduces activities, monitors model use, and redirects shortcutting or unsuitable responses.","handoff_to":"Pen-and-paper and school assessments","control":"Teacher retains pedagogy; outcomes measured independently after six weeks."}],"decision_rights":"Teachers own pedagogy and monitoring.","exception_path":"Internet, electricity or unsuitable output is handled by teacher.","removed_work":["Some one-to-one practice demand"],"outcomes":[{"metric":"Endline assessment","baseline":"Control business-as-usual","result":"0.310 SD weighted total score; 0.238 SD English skills (abstract reports 0.23; narrative rounds to 0.24)","period":"Six weeks; 12 sessions","scale":"Nine public schools","attribution_caveat":"World Bank Policy Research Working Paper, not formal peer review; 654 observations in the main outcome tables; short, supervised intervention among interested students.","evidence_label":"verified"}],"executive_lesson":"The model made supervised practice abundant; it did not replace the teacher.","anti_pattern":"Do not generalize to unsupervised access.","questions_for_leaders":["Where is the operating threshold set and who can override it?","What measured result would trigger rollback or retraining?","Which residual decisions must remain human-owned?"],"collections":[],"bundle_version":"1.0.0","bundle_fingerprint":"sha256:c23c6cc2b88153d008ea8fda928f632ce0011fc2d4c5036672a16e5d895bab93","canonical_url":"https://brianletort.ai/transformations/nigeria-ai-tutoring"}