Skip to main content

Applications for the autumn mentorship cohort are open until 30 September. Apply now

DSI
Template

Auditable Decision Record & Correction Log

A complete, publishable record of how you reached a decision, with an optional standardised public summary and a permanent correction log.

Type
Template
Difficulty
Intermediate
Time
45 min
Version
v1.0.0
Updated
8/9/2026
decisiontransparencyprocess

What this helps you do

A decision record is the artefact a delegator can actually audit. It shows the reasoning, the evidence relied on, the alternatives considered, the risks and uncertainties, and any conflicts disclosed at the time.

The template treats Yes, No and Abstain identically. Abstain is a full decision with a full rationale, not a blank. The optional public summary standardises how the decision is communicated, and nothing about it standardises what the decision should be. The correction log stays open after publication so the record can be amended without being rewritten.

When to use this

  • You have reached a decision and want a record a delegator can audit.
  • You are abstaining and want the reasoning documented as fully as a Yes or No.
  • Facts have changed after publication and you need to correct the record properly.

When NOT to use this

  • You have not finished your analysis; the record documents a decision, it does not produce one.
  • You want a short public note only; publish the summary block and link to the full record later.
  • You are recording a decision that is not your own.

What you'll need

  • Your completed intake brief and claim-evidence map
  • Your risk register and any budget work
  • Your current conflict disclosure
  • The action identifier and the voting deadline

Estimated completion time

45 minutes

Working time for one governance action, assuming the inputs listed here are already to hand.

How to use it

  1. 1Record the governance action and your decision first, with the date and time.
  2. 2Write the executive rationale in three sentences or fewer, before you elaborate.
  3. 3List the primary reasons in order of weight, and be honest about which one actually decided it.
  4. 4Cite the evidence you relied on, and separate what you verified from what you accepted on trust.
  5. 5Record the alternatives you considered, including the decision you did not take and why.
  6. 6Carry across the material risks from your register and the uncertainties still open at your cut-off.
  7. 7Disclose conflicts as they stood at the time of the decision, and link to your standing disclosure.
  8. 8State any conditions or assumptions your reasoning depends on.
  9. 9Complete the optional public summary if you publish a short form, and link it to the full record.
  10. 10Leave the correction log open. Add corrections as dated entries; never edit the original text.

The tool

Full template - v1.0.0

Decision record fields

  • Governance action and identifier
  • Decision: Yes, No, or Abstain
  • Date and time of decision
  • Executive rationale, three sentences or fewer
  • Primary reasons in order of weight
  • Evidence relied upon, with links
  • Alternatives considered and why they were not taken
  • Risks considered, carried from the risk register
  • Conflicts disclosed at the time
  • Material uncertainties open at the cut-off
  • Conditions and assumptions the reasoning depends on
  • Links to supporting work

Optional standardised public summary

  • Decision
  • TL;DR in one sentence
  • Primary rationale
  • Key evidence
  • Primary risk or uncertainty
  • Conflict disclosure statement
  • Link to the full analysis

Correction log

  • Date of the correction
  • What changed in the record
  • Why it changed, with the new evidence
  • Whether the correction would have changed the decision
  • Who made the correction
  • Original text preserved rather than overwritten

Neutrality rules for this template

  • Yes, No and Abstain use the same fields and the same depth
  • Abstain requires a rationale, not an apology
  • The template contains no recommended position and no preferred outcome
  • The public summary standardises the format of communication only
  • No field asks what anyone else decided
Fictional Example

Invented for illustration. It does not describe a real governance action, proposal, or organisation.

Fictional decision record: abstain

  • Action: Regional Developer Hubs Pilot (fictional)
  • Decision: Abstain
  • Date: 2026-08-11
  • Executive rationale: the programme is coherent, but the two claims that would justify the amount remain unevidenced at my cut-off. I could not reach a defensible position either way in the time available.
  • Primary reason: the 1,200 developers figure remained self-reported after a public request for cohort records
  • Evidence relied upon: proposal text, question tracker log, budget reconstruction showing a 13 percent unexplained variance
  • Alternatives considered: voting No on evidence grounds, rejected because absence of evidence is not evidence of overstatement
  • Risks considered: single technical lead, residual severity 12
  • Conflicts disclosed: none relevant to this action; standing disclosure linked
  • Material uncertainties: contingency base undefined; cohort methodology unpublished
  • Correction log: open, no entries

How to interpret the result

  • A record where the executive rationale does not match the primary reasons usually means the real reason went unwritten. Rewrite the rationale rather than the reasons.
  • A short evidence list next to a strongly worded rationale is a signal to re-read your own work.
  • Abstain records with full reasoning are read as diligence, not indecision, and they are auditable in exactly the same way.
  • A correction that would have changed the decision should say so plainly. That is the most credible entry a log can contain.

Limitations

  • The record documents reasoning; it cannot make the reasoning sound.
  • It captures conflicts as you understood them at the time, which is not the same as all conflicts that existed.
  • Publishing a record does not make it correct, and readers should treat it as a claim like any other.

Sources and methodology

Learn this method

This tool records the work. The Training Lab teaches how to do it.

  • communication-representation · Writing a Vote Rationale People Can Audit

    Writing a Vote Rationale People Can Audit

    Teaches how to record a decision and correct it well.

    Open lesson
  • Lab lesson planned

    A dedicated lesson on correction logs is planned.

Downloads

  • Markdown export: the complete tool, including metadata, instructions, the template, limitations, sources and version history.

No other file formats are published for this resource. We list a format only when the file exists.

Version and governance

Current version
v1.0.0
Published
8/9/2026
Last updated
8/9/2026
Next scheduled review
2/9/2027
Licence
CC BY 4.0
Editorial status
Not yet externally reviewed

No named reviewer is shown because no external review has been completed. Attribution appears only once a verified reviewer has signed off.

Changelog

  • v1.0.0 (2026-08-09) - First published edition.
Next in the workflow

Use these alongside it

The tools that usually come before or after this one.

WorksheetIntermediate60 min

Claim-Evidence Map & Citation Log

A row-per-claim map that separates what a proposal asserts from the evidence actually supplied for it, with a citation log you can publish.

Workflow
Verify Claims & Evidence
Formats
Markdown
Use ToolView Details
WorksheetIntermediate60 min

Delivery-Risk Register

A structured register of the risks that could prevent a funded proposal from delivering what it promised, scored on probability and impact.

Workflow
Assess Delivery & Technical Risk
Formats
Markdown
Use ToolView Details
Use This Tool