evidoria

← Back to browse

Good practice Imported

Bertoua Pilot Tests a WhatsApp AI Chatbot for Cameroonian Teacher Training - With Real Trade-offs

Cameroon · Bertoua · See the Cameroon profile

Evidence: Quasi-experimental Top 56% 47/100 · Ask Evidence Copilot about this practice

A Saarland University-led field pilot with 47 primary-school teachers in Bertoua, Cameroon found a WhatsApp LLM chatbot rated modestly higher for usability and experience than an online-form baseline, while surfacing real barriers: French-speaker exclusion and data costs.

47 teachers
Teachers participating in the pilot (September 2025)
7.82 vs 7.33 mean score
Usability rating, chatbot vs. baseline (p=.046)
7.41 vs 7.03 mean score
Combined experience rating, chatbot vs. baseline (p=.044)
15–238 messages
Chatbot messages sent per teacher (interview subsample)
87.5 %
Teachers owning smartphones
77.1 %
Teachers using WhatsApp daily

Details

Maturity
Pilot
Promoter
Saarland University (with B2THEWORLD and Chatac.ai)
Period
September 2025
Keywords
teacher professional development, WhatsApp chatbot, generative AI, low-resource EdTech

Context

Researchers from Saarland University, with partners B2THEWORLD and Chatac.ai and funding from the Alexander von Humboldt Foundation, ran a one-week field pilot in September 2025 across six private primary schools in Bertoua, eastern Cameroon.

Objectives

Test whether a WhatsApp-based LLM chatbot for teacher professional development performs better than an online-form baseline, and surface practical barriers to low-resource EdTech delivery.

Activities

47 teachers (39 women, 8 men) used the chatbot or online-form condition; 14 teachers were later interviewed, with chatbot engagement ranging from 15 to 238 messages sent per teacher.

Results

Teachers rated the chatbot condition significantly higher on usability (mean 7.82 vs 7.33, p=.046) and combined experience (mean 7.41 vs 7.03, p=.044), both small-to-moderate effects; learnability differences were not statistically significant. Content was delivered in English only, excluding French-speaking teachers by the study's own account, and connectivity plus prepaid mobile-data costs were the primary access barriers, even though 87.5% of teachers owned smartphones and 77.1% used WhatsApp daily.

Conclusions

The authors frame the findings as early, small-sample lessons rather than proof of impact at scale.

Implementation
Indicative cost
Low (< €50k) — Academic research pilot scale; no infrastructure beyond existing teacher smartphones and WhatsApp, with prepaid mobile data cited as a recurring cost barrier for participants.
Time to results
Short (< 1 year) — One-week field pilot conducted in September 2025 across six private primary schools.
Staffing & skills
Saarland University researchers, B2THEWORLD and Chatac.ai (delivery partners), Funded by the Alexander von Humboldt Foundation

Conditions for success

  • High existing smartphone ownership (87.5%) and daily WhatsApp usage (77.1%) among participating teachers, lowering the adoption barrier for a WhatsApp-based tool

Common failure modes

  • English-only content excluded Francophone teachers, one of whom explicitly noted it excluded 'French (speaking) brothers and sisters'
  • Connectivity and prepaid mobile-data costs were the primary barriers to access
  • A small sample (n=47) across six private schools for one week limits generalisability

Commonly funded by

Philanthropic / foundation funding

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

Replication kit

Reusable artefacts from this practice — as published by their sources.

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.

Similar practices you may find useful