CH-13 Cognitive bias checks and debiasing (including consider-the-opposite)
Runs a conclusion through a list of known judgement biases, applies a counter-procedure, and produces a checklist of marked items where disconfirming evidence should be sought.
What it does
Runs a conclusion through a list of known judgement biases and applies a counter-procedure. Artefact: a checklist; marked items are places to seek disconfirming evidence.
When to use it
Use when a conclusion has been reached and you want to check which known biases may have shaped it.
When it does not fit
The check is run after the conclusion is fixed.
Result format
The result is a checklist with one row per bias, giving the bias, whether it is marked as present in the conclusion, and the counter-procedure applied. It ends with a list of the marked items, each paired with the disconfirming evidence to seek.
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 conclusion, its process and a bias list. |
|---|---|
| Additional sources needed | none |
| Not for (rules) | none |
| Not for (text) | The check is run after the conclusion is fixed. |
Choosing and ordering
| Kind of technique | Challenge technique, runs in step 6 |
|---|---|
| Phase tag | validate |
| Other phase tags | conclude |
| Kind of finding | risk |
| Question answered | what; why |
| Strongest statement it supports | describe |
| Needs as input |
|
| Makes as output |
|
| Must exist first | None |
| Loop target | proceed with limit stated |
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 | assumption and limitation statement | none | no |
| decision-informing | assumption and limitation statement | none | no |
| decision-informing | independent review | needed; left out below high stakes | no |
| high-stakes | assumption and limitation statement | none | no |
| high-stakes | independent review | needed | no |
| high-stakes | expert review | needed | no |
| high-stakes | second technique with different assumptions | none | conditional |
Failure modes and limits
| Model suitability | model alone |
|---|---|
| Typical model failure | Marks every bias as possible in general terms without linking any to a step in the actual reasoning. |
| Error sources | input data; coder or rater judgement; language model |
| Certainty limits | Only an abstract of a review was read and effect evidence was not confirmed; debiasing has barriers to real-world effectiveness, so certainty is lowered by indirectness. |
Related techniques
- pairs with TC1-16 Disconfirmation search (falsification, seeking disconfirming evidence): A search for disconfirming evidence follows each marked item
- pairs with CH-09 Structured self-critique: Structured self-critique covers the same ground with open questions
Evidence and tag basis
Unconfirmed. Only the abstract of a 2009 review was read; it discusses barriers to real-world effectiveness, and effect evidence was not confirmed. (Lilienfeld et al. (2009), debiasing, abstract)
Tag basis: input profile inference; selection inference; sequencing inference; robustness inference. All tag values are inference unless a source is named.