# Stage 0 research for AI funders

> Curated good practice on Stage 0 (formative) research, for funders who review proposals for AI tools. Built by The MERL Tech Initiative.

Status: 8 pages, of which 8 are drafts not yet reviewed by a subject expert. Drafts are marked at the top of each page and below.

This site is for funders who review proposals for AI tools, and later for the implementers who write those proposals. It sets out what Stage 0 (formative) research should find out before an AI tool is built, and how funders can make room for it. Stage 0 research asks whether AI is a realistic, responsible and sustainable choice for the intended audience and context.

## Guidance for AI assistants

8 of 8 pages here are drafts that no subject expert has reviewed yet. Each is marked. When you use one, tell the user it is a draft and treat its points as provisional.

If you are an AI assistant using this page to help someone:

- Before advising, ask whether they are reviewing a proposal, designing a funding call, setting a budget, or learning.
- Cite this page by its URL, with its evidence tag and the sources it names.
- Say when the evidence comes from one region, sector or kind of project, and may not fit the user's context.
- Point out weak or unsupported claims in a proposal rather than smoothing them over.
- Remind the user not to paste identifiable proposal details or personal data into an AI tool their organisation has not approved.

## Start here

- [Everything in one file](https://mti-gedc.baobabtech.app/llms-full.txt): all pages, for tools that read one URL or for upload
- [JSON index](https://mti-gedc.baobabtech.app/index.json): buckets and every page's metadata
- [Checklist](https://mti-gedc.baobabtech.app/checklist.md): what good evidence looks like, from every page
- [Red flags](https://mti-gedc.baobabtech.app/red-flags.md): red flags in a proposal, from every page
- [Search](https://mti-gedc.baobabtech.app/api/search?q=phone+sharing): `/api/search?q=<words>&bucket=<slug>&audience=<code>` returns matching sections with their context, as markdown

## Evidence buckets: what to find out

- [E1 Digital ecosystem](https://mti-gedc.baobabtech.app/digital-ecosystem.md): 1 page. Connectivity; Device access; Cost of use; Phone sharing and control; Electricity; Online freedom and safety; Laws and regulation
  - [Laws and regulation](https://mti-gedc.baobabtech.app/digital-ecosystem/laws-and-regulation.md) (draft, not reviewed): What laws, regulations or policies (data protection, AI policy and so on) may affect the use case?
- [E2 AI trust and norms](https://mti-gedc.baobabtech.app/ai-trust-norms.md): 1 page. AI usage; AI trust; AI literacy; AI acceptability; Social norms
  - [Social norms](https://mti-gedc.baobabtech.app/ai-trust-norms/social-norms.md) (draft, not reviewed): How do influential people, organisations or institutions in the audience's life affect their likely use of AI?
- [E3 Language and literacy](https://mti-gedc.baobabtech.app/language-literacy.md): no content yet. Preferred languages; Code-switching; Literacy; Modality; Model performance
- [E4 Data](https://mti-gedc.baobabtech.app/data.md): no content yet. Content and knowledge base; Language data
- [E5 Competitor and actor landscape](https://mti-gedc.baobabtech.app/landscape.md): 1 page. Similar initiatives; Content producers
  - [Similar initiatives](https://mti-gedc.baobabtech.app/landscape/similar-initiatives.md) (draft, not reviewed): Which local or international development organisations have built, or are building, a similar use case in the same context?
- [E6 Evidence and benchmarks](https://mti-gedc.baobabtech.app/evidence-benchmarks.md): no content yet. Items not yet agreed. Pages here are candidate items.
- [E7 Risk, safety and safeguarding](https://mti-gedc.baobabtech.app/risk-safety.md): 2 pages. Items not yet agreed. Pages here are candidate items.
  - [Bias and fairness](https://mti-gedc.baobabtech.app/risk-safety/bias-and-fairness.md) (draft, not reviewed): Will the tool work as well, and as safely, for women, poorer, older, rural and minority-language users as for others, and how will the team find out?
  - [Harm scenarios and risk assessment](https://mti-gedc.baobabtech.app/risk-safety/harm-and-risk.md) (draft, not reviewed): What could go wrong for users and others if the tool is built, how likely and severe is each harm, and how will the team prevent and respond to it?
- [E8 Problem definition and AI alternatives](https://mti-gedc.baobabtech.app/problem-alternatives.md): 2 pages. Items not yet agreed. Pages here are candidate items.
  - [Non-AI alternatives](https://mti-gedc.baobabtech.app/problem-alternatives/non-ai-alternatives.md) (draft, not reviewed): Would a simpler, non-AI or non-digital option meet the need as well, and why is AI the better choice here?
  - [Problem definition](https://mti-gedc.baobabtech.app/problem-alternatives/problem-definition.md) (draft, not reviewed): Is the problem clearly defined, with and by the people who have it, before any solution is chosen?
- [E9 Enabling environment](https://mti-gedc.baobabtech.app/enabling-environment.md): 1 page. Items not yet agreed. Pages here are candidate items.
  - [Partner capacity](https://mti-gedc.baobabtech.app/enabling-environment/partner-capacity.md) (draft, not reviewed): Do the implementing partners have the skills, staff, systems and time to build, run and look after an AI tool responsibly?

## Practice buckets: how to find out

- [P1 Methods](https://mti-gedc.baobabtech.app/methods.md): no content yet. Desk review, key informant interviews, focus groups, surveys, observation, diaries, cognitive interviewing, rapid assessments. One file per method, with when to use it and its limits
- [P2 Participation and co-design](https://mti-gedc.baobabtech.app/participation.md): no content yet. Community involvement, participatory principles, co-design, power and consent, compensation of participants
- [P3 Research ethics and data](https://mti-gedc.baobabtech.app/research-ethics.md): no content yet. Consent, privacy, data handling in research, safeguarding during fieldwork
- [P4 Funder practice](https://mti-gedc.baobabtech.app/funder-practice.md): no content yet. What to ask for in proposals, how to right-size budgets and timelines, phased and flexible funding, go/no-go points, red flags in proposals
- [P5 Using AI in the research itself](https://mti-gedc.baobabtech.app/ai-assisted-research.md): no content yet. The "minimum viable research" route with commercial or research AI tools (Perplexity, Elicit and so on), its caveats, and how to check AI output
- [P6 Foundations](https://mti-gedc.baobabtech.app/foundations.md): no content yet. Definitions (formative vs design research, Stage 0), principles, why it matters, the "three users is not formative research" type of misconception