Decision-Basis Disclosure¶
Document — instantiates Relevance-Substitution Detection and Correction
Requires decision-makers to state, on the record, which evidence — not which impressions — actually drove the call.
Decision-Basis Disclosure is a required written statement, produced at the point of decision, in which the decider names the specific evidence the call actually rests on. Its target is the gap between the reasons people give and the cues that moved them: by forcing a named, checkable basis into the record, it makes an impression-driven decision either declare its evidence or expose the absence of one. Crucially, it is built around the finding that disclosure by itself often under-corrects — people disclose and then keep leaning on the very cue they disclosed — so the mechanism pairs the statement with ongoing monitoring for residual influence. What makes it distinct is its forward-facing, on-the-record character: it captures the claimed basis at the decision and watches whether the disclosed evidence is really carrying the weight, rather than logging a numeric confidence change after the fact.
Example¶
A government procurement panel is awarding a multi-year IT contract. Two bids are close; one vendor gave a slick, warmly received presentation, the other a dry but detailed one. The Decision-Basis Disclosure form requires each panelist, before the award is finalized, to write which scored criteria and which pieces of evidence drove their ranking — total cost of ownership, references checked, security certifications, delivery timeline — and to explicitly exclude "presentation quality" and "rapport" as bases unless tied to a scored criterion. One panelist, writing it out, finds she cannot name an evidentiary basis for preferring the slick vendor beyond "they seemed more confident." That empty line is the point: the disclosure surfaces that her lean rests on an impression, not the criteria. The panel re-scores against documented evidence, and a monitor re-reads the final justifications a week later to check that the disclosed bases — not the lingering warmth of the presentation — are what the awarding memo actually cites.
How it works¶
- Demand a named basis. The decider must list the specific evidence items behind the call, in enough detail to be checked, not a vague gesture at "the whole picture."
- Exclude impressions unless grounded. Charged cues — rapport, confidence, prestige, presentation — are disallowed as bases unless they connect to a declared, relevant criterion.
- Audit the stated pathway. The disclosure is read against the decision to check that the named evidence could actually support it and that an undisclosed cue is not doing the real work.
- Monitor for residue. Because stating the basis does not guarantee acting on it, a follow-up check watches whether the disclosed evidence, rather than the excluded impression, is what the final rationale relies on.
Tuning parameters¶
- Basis specificity — how concretely each item must be named. High specificity makes hidden substitution visible but is burdensome and can invite tidy post-hoc rationalization; low specificity is fast but easy to fake.
- Impression-exclusion scope — how many charged cues are ruled out as standalone bases. Broad exclusion is protective but can suppress legitimately relevant soft signals; narrow exclusion leaves more loopholes.
- Audience of the record — private file, peers, or public. Wider audiences raise honesty pressure but also incentivize defensive, unfalsifiable prose.
- Monitoring intensity — whether the residual check is a one-time re-read or a standing audit. More monitoring catches persistent leakage but costs attention and can feel like surveillance.
When it helps, and when it misleads¶
Its strength is making the basis accountable: a decider who must name the evidence, and be watched for whether they acted on it, is far less free to be quietly swayed by an impression. It takes seriously the uncomfortable evidence that transparency is not self-executing — disclosing a conflict or a cue can even increase reliance on it, so a bare confession of "I was impressed by their confidence" corrects nothing on its own.[n1] Pairing the record with residual monitoring is what turns disclosure from theater into a control.
Its failure mode is manufacturing plausible-sounding bases: asked to justify a call they made on a hunch, people are fluent at reverse-engineering a respectable evidentiary story, so the document can certify a decision that was never actually made on those grounds. The classic misuse is exactly this — using the disclosure as a compliance ritual that launders impression-driven calls into evidence-cited ones. The guarding discipline is to keep the monitor independent of the decider, to compare the disclosed basis against the contemporaneous record rather than accepting it at face value, and to treat an unnameable basis as a stop signal, not a paperwork gap.
How it implements the components¶
decision_basis_disclosure_record— the document itself is the on-the-record statement of which evidence the decision rests on.update_pathway_audit— reading the stated basis against the decision audits whether the named evidence, and not a hidden cue, actually drove the update.residual_bias_monitor— the follow-up check watches for the well-known residue of influence that disclosure alone fails to remove.
It does not compute a before/after confidence delta or apply a correction_and_reweighting_rule to arrive at a revised number — that quantified reweighting after discounting the cue belongs to Reweighted Update Log, its document twin.
Related¶
- Instantiates: Relevance-Substitution Detection and Correction — provides the accountable, monitored statement of basis that forces a decision to declare its evidence.
- Sibling mechanisms: Affect–Evidence Split Prompt · Red Herring Filter Checklist · Analogy Mapping Table · Question–Evidence Matrix · Reweighted Update Log · Cue-Validity Audit · Proxy-Relevance Audit · Blinded or Masked Review
Editorial Notes¶
Form Classification¶
Form family: Rule, Policy & Commitment
Rationale: Decision-Basis Disclosure operates as a standing rule, threshold, contractual commitment, or policy constraint governing future conduct because it requires decision-makers to state, on the record, which evidence — not which impressions — actually drove the call.
Independent corroboration: The frozen evidence defines Decision-Basis Disclosure as 'Requires decision-makers to state, on the record, which evidence — not which impressions — actually drove the call', so its operative form is Rule, Policy & Commitment.
Review outcome: Independent reviewer agreement; medium confidence.
Origin Attribution¶
Primary origin: Law & Governance
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Administrative law is primary because it requires consequential decisions to place their findings, conclusions, and reasons or basis on the record for review. Public administration supplies the operating disclosure practice, while psychology supplies the residual-influence monitor because disclosure alone can license rather than eliminate bias.
Related originating lineages:
- Psychology — Bias and conflict-of-interest research supplies the warning that disclosure can leave or worsen cue influence, motivating follow-up monitoring.
- Public Administration & Policy — Administrative decision practice supplies written criteria-and-evidence statements for procurement, licensing, and public determinations.
Review resolution: Administrative law is primary because it requires consequential decisions to place their findings, conclusions, and reasons or basis on the record for review. Public administration supplies the operating disclosure practice, while psychology supplies the residual-influence monitor because disclosure alone can license rather than eliminate bias.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Researched adjudication after independent review; high confidence.
Sources consulted:
- 5 U.S.C. §557: Contents and record of agency decisions
- Cain, Loewenstein, and Moore (2005): The Dirt on Coming Clean
Notes¶
[n1] Research on disclosing conflicts of interest (Cain, Loewenstein, and Moore) found that disclosure can worsen the problem — advisors feel licensed to bias more, and recipients under-discount — which is why this mechanism attaches a residual-influence monitor rather than trusting the statement alone. ↩