DM-02 Hypothesis tree (hypothesis-driven decomposition, claim decomposition)
Splits a claim into the sub-claims that must all be true for it to be true and produces a tree of claims that goes into a claims register.
What it does
Splits a claim into the sub-claims that must all be true for it to be true. Artefact: a tree of claims; read leaf claims for which no evidence is possible.
When to use it
Use when a claim is too broad to test and you need to know which sub-claims need evidence.
When it does not fit
Leaves are vague, the parent can be true while every leaf is false, or the first hypothesis is kept after its top branch is refuted.
Result format
The result is a tree in which the main claim is split into the sub-claims that must all be true for it to be true, down to leaf claims. It lists the leaf claims for which no evidence is possible.
Requirements on the material
| Property | Requirement type | Requirement |
|---|---|---|
| Unit of analysis | Always applies | requires=claim |
| Appraises source credibility | Applies in some cases | no |
Not restricted on: Structure, Content type, Variable types, Time structure, How the sample was chosen, Level of detail available, Number of sources, Whether sources can be joined, Amount of material, Known data problems, Producer interest, Source dependence, Origin, Document genre, Language, Personal information, Security classification, Licence.
| Minimum material | A claim and knowledge of what evidence would support or refute each branch. |
|---|---|
| Additional sources needed | none |
| Not for (rules) | none |
| Not for (text) | Leaves are vague, the parent can be true while every leaf is false, or the first hypothesis is kept after its top branch is refuted. |
Choosing and ordering
| Kind of technique | Question technique, runs in step 5.1 |
|---|---|
| Phase tag | frame |
| Other phase tags | analyse |
| Kind of finding | gap; risk |
| Question answered | why; what |
| Strongest statement it supports | describe |
| Needs as input |
|
| Makes as output |
|
| Must exist first | None |
| Loop target | return to frame |
Library checks by stakes
Each stakes level includes the checks of the level below it. Checks that need a person apply at high stakes only, so the process leaves them out at lower stakes (decision 38). Repeatability: judgement dependent.
| Stakes | Check | Person involved | Counts as a different method |
|---|---|---|---|
| exploratory | spot check | optional | no |
| decision-informing | spot check | optional | no |
| decision-informing | assumption and limitation statement | none | no |
| high-stakes | spot check | optional | no |
| high-stakes | assumption and limitation statement | none | no |
| high-stakes | expert review | needed | no |
| high-stakes | independent review | needed | no |
Failure modes and limits
| Model suitability | human judgement required |
|---|---|
| Typical model failure | Writes leaf claims that restate the parent in other words, so that no evidence could separate them. |
| Error sources | input data; coder or rater judgement; language model |
| Certainty limits | Risk of bias because the split reflects the analyst view of what evidence is possible; the tree states what must be shown and does not show that it is true. |
Related techniques
- pairs with TC1-13 Analysis of competing hypotheses (ACH): Competing hypotheses scoring rates evidence against hypotheses once the tree has defined them
- pairs with TC1-16 Disconfirmation search (falsification, seeking disconfirming evidence): A search for disconfirming evidence can be run on each leaf claim
Evidence and tag basis
Unconfirmed. No source was opened; two note files describe it from general knowledge only.
Tag basis: input profile inference; selection inference; sequencing inference; robustness inference. All tag values are inference unless a source is named.