Template: templates/frame.question.md

# Question

## The Question

_In one sentence, ending in a question mark. If it takes a paragraph, you have a programme, not a question — say the question, then list the strands beneath it._

## Why It Matters

_What decision, practice, or belief changes depending on the answer. Who is currently acting on a guess._

## What Would Count As An Answer

_The standard of evidence you will hold yourself to, before you know what the data says. Stating it here is what stops it from drifting to meet the result._

## What This Programme Is Not Asking

_The adjacent question you are deliberately leaving alone, and why. This is where a research programme's scope is held._

Template: templates/frame.prior-work.md

# Prior Work

## What Is Settled

_The findings you are building on rather than retesting. Cite them; say how confident the field actually is, not how confident the abstracts sound._

## What Is Contested

_Where the literature disagrees, and what the disagreement turns on — a measure, a population, a definition._

## The Gap

_What nobody has asked, or has asked badly. Be specific about which part of the question is genuinely open; "more research is needed" is not a gap._

## Why Now

_What changed — a new instrument, a new dataset, a policy shift — that makes this answerable when it wasn't._

Template: templates/frame.method.md

# Method

## Approach

_The kind of evidence this programme produces — experimental, observational, qualitative, computational, mixed — and why that kind can answer the question._

## Standard of Evidence

_What you will treat as sufficient, decided in advance: effect sizes, saturation, replication, pre-registration, review. The bar, written before the results._

## Threats to Validity

_The ways this whole approach could mislead — selection, measurement, confounding, researcher degrees of freedom — and what you do about each._

## What You Will Share

_Data, code, materials, negative results. Where they go, and when. A programme that only publishes what worked is a programme that cannot be checked._

Template: templates/frame.ethics.md

# Ethics

## Who Is Affected

_Participants, their communities, and the people downstream of the finding — including those who never consented to anything but will live with the result._

## Consent & Governance

_What participants were told, what they agreed to, and what would fall outside it. The approving body, the approval, and its limits._

## Data Handling

_What is collected, how it is stored, who can see it, how long it lives, and what makes it identifiable again. Say what is sensitive, plainly._

## What Could Go Wrong

_Foreseeable harm from doing the work, and foreseeable harm from the finding being misread or misused. What you will do about each. A programme with an empty version of this section has not looked._

Template: templates/investigation.full.md

# {{title}}

## Intent

_The question this investigation answers, and why it is worth answering. One or two paragraphs, ending in an actual question mark. This is the stable part: if the question changes substantially, you have a different investigation, not an edit to this one._

### Assumptions

_What you are taking as given and have not tested — about the population, the instrument, the prior literature, the world. Nested under Intent because they frame it. Surface them so an untested assumption doesn't get read later as a finding._

## Open Questions

_What you don't yet know about how to run this — an undecided measure, an unresolved confound, a missing approval. Kept high so uncertainty is seen rather than buried. When one is settled it graduates into an Assumption, a Claim, or a Decision._

## Claims & Evidence

_What this investigation asserts, and what supports each assertion. Label each claim **C1:**, **C2:**, … (bold, unique within this investigation). Give each its evidence and its strength; a claim with no evidence yet is still a claim — mark it so. Push data, tables, and instrument detail into a sub-section a claim points to, not onto the line. Evolves freely._

- **C1:** <!-- the assertion --> <!-- evidence: … -->

## What Would Change Our Mind

_The falsifier: the specific result that would make you abandon or revise the claims above. Concrete enough that someone could go and look. An investigation with no falsifier is an opinion with citations, and this is the section this framework leans on hardest._

## Out of Scope

_What this investigation **cannot** tell you, and who will be tempted to think it can. Populations not sampled, mechanisms not tested, the causal claim the design does not support. The second-hardest-working section here: it is what keeps a finding from being over-read downstream._

## Dependencies

_What this needs before it can run: data, access, approvals, instruments, or another investigation's result. The `depends_on` property at the top is the investigation-only subset of this, as links. Reference other investigations as markdown links — never bare names._

## Method

_How the question is actually approached: design, sample, measures, analysis. Enough that a reader can judge whether the Claims follow. Detail belongs here rather than on a claim line; a full protocol can live in its own top-level doc this section points at._

## Notes & Decisions

_Two things under one header. **Notes**: threats to validity, things to chase, what surprised you. **Decisions**: an append-only log, one line each, with an optional but recommended date._

<!-- 2026-08-26: Dropped the second cohort, consent window closed. (Brenton) -->

Template: templates/investigation.note.md

# {{title}}

## Intent

_The question, and why it is worth answering. End it in a question mark. A paragraph is plenty here._

## Open Questions

_What you don't yet know about how to approach it. Honest holes beat invented certainty._

## Claims

_What you currently believe, labeled **C1:**, **C2:**, … with a word on what supports each — including "nothing yet." A note is allowed to be thin, but never vague about its own confidence._

- **C1:**

## Out of Scope

_What this could not tell you even if it went perfectly. Even in a note, this is the section worth filling — it is what stops a preliminary result from being quoted as a conclusion._