evidoria

← Back to browse

Good practice Imported

STEPS — AI-Assisted Science Textbooks Scale to 101 Emmanuel Community Schools in DR Congo

Congo (the Democratic Republic of the) · Kinshasa · See the Congo (the Democratic Republic of the) profile · See the Kinshasa profile

Evidence: Descriptive / self-reported Top 17% 67/100 · Ask Evidence Copilot about this practice

GPE KIX's STEPS project used AI, reviewed by experts, to draft and translate 28 science textbooks for Benin, Cameroon and DR Congo. Partner Emmanuel Community Schools now scales the materials to all 101 DRC schools after 94% of teachers reported greater confidence.

28
Science textbooks and workbooks produced
94 %
Teachers reporting increased confidence teaching science
80 %
Teachers intending to continue using STEPS materials
101
Emmanuel Community Schools in DRC committed to scale-up
STEPS — AI-Assisted Science Textbooks Scale to 101 Emmanuel Community Schools in DR Congo

Details

Maturity
Scaling
Promoter
Emmanuel Community Schools (GPE KIX STEPS project)
Period
2021-2024
Keywords
primary education, STEM curriculum, teacher training, open educational resources

Context

Science teaching in many DR Congo primary schools is hampered by a shortage of curriculum-aligned, locally relevant textbooks and by limited teacher subject-matter confidence.

Objectives

Close the textbook gap using a combination of Open Educational Resources and generative AI, under GPE KIX's STEPS project (2021–2024), funded with IDRC and coordinated through the KIX Africa 21 Hub.

Activities

AI tools (ChatGPT and Anthropic models) were given detailed, curriculum-referenced prompts to draft initial lesson content, suggest culturally relevant examples, and translate materials from English to French; every AI-generated output was then reviewed and adapted by subject-matter experts and local educators before use in the classroom. The project produced 28 science textbooks and accompanying workbooks across Benin, Cameroon and DR Congo, and shifted teacher training from costly five-day external workshops to lower-cost, school-based sessions run by experienced local educators.

Results

In DR Congo, implementing partner Emmanuel Community Schools has committed to expanding use of the STEPS materials to all 101 of its schools. Across the three participating countries, 94% of teachers surveyed reported increased confidence teaching science, and 80% said they intended to keep using the STEPS materials after the project's external support ended. Benin's Ministry of Education initiated formal approval procedures for the material and Cameroon's Baptist Convention committed to systemwide rollout across its school network.

Conclusions

The reported gains are based on teacher self-report surveys rather than independent student assessment data, and student learning-outcome evidence has not yet been published, so the practice should be read as strong evidence of content-production efficiency and teacher buy-in rather than proven learning impact.

Implementation

Indicative cost
Medium (€50k–€500k) — Funded by GPE KIX with IDRC; cost efficiency driven by combining AI-assisted drafting/translation with lower-cost school-based (rather than external multi-day) teacher training.
Time to results
Medium (1–3 years) — Project ran 2021–2024; DRC partner has since committed to scaling materials to all 101 of its schools.
Staffing & skills
Subject-matter experts and local educators reviewing and adapting all AI-generated content before classroom use, Emmanuel Community Schools as DRC implementing partner, GPE KIX / IDRC / KIX Africa 21 Hub coordination

Conditions for success

  • Every AI-generated output reviewed and adapted by human subject-matter experts and local educators before use, not published raw
  • Shift from costly five-day external workshops to lower-cost, school-based teacher training sessions run by experienced local educators
  • Curriculum-referenced, detailed prompting used to keep AI drafts aligned to local curricula

Common failure modes

  • Reported gains (94% teacher confidence, 80% intended continued use) are based on teacher self-report surveys, not independent student assessment data; student learning-outcome evidence has not yet been published

Commonly funded by

Philanthropic / foundation funding

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful