Skip to content

Latest commit

 

History

History
130 lines (101 loc) · 4.74 KB

File metadata and controls

130 lines (101 loc) · 4.74 KB

Decision Dossier template

Use this template to turn a street-safety concern into a reviewable decision artifact. It is deliberately usable before the local-first Studio is built. Copy it into a dated working document, complete only the sections supported by your evidence, and retain the empty sections when evidence is missing.

This is a planning and communication template, not a substitute for local engineering judgment, legal review, a traffic study, or a causal evaluation.

1. Decision request

  • Dossier title:
  • Prepared by / community partner:
  • Jurisdiction and decision owner:
  • Corridor or area:
  • Decision requested:
  • Decision date or planning window:
  • Requested next action:

State one action that a named person or body can take. “Make this street safer” is a goal, not a decision request. “Fund a field audit before the 2027 capital plan” is a decision request.

2. Claim boundary

Choose the strongest statement the evidence supports. Delete stronger statements.

  • Tier 0 — coverage gap: Evidence is insufficient to rank this corridor.
  • Tier 1 — community signal: Repeated community reports warrant investigation.
  • Tier 2 — elevated rate: The observed report rate is elevated in the stated analysis window.
  • Tier 3 — triangulated priority: Independent evidence supports prioritizing this corridor for the stated next action.

Permitted claim:

This dossier does not establish: danger, fault, causation, the likely effect of a treatment, or that a treatment worked. Add any local limitations here. The generated dossier carries the same boundary: nearmiss dossier renders "does not establish danger, fault, causation" and a declared evidence tier into every corridor artifact, so the limit travels with the document rather than depending on whoever filled the template in.

3. Evidence snapshot

Evidence track Source and window What it shows Important limits
Community reports
Exposure / denominator
Official outcomes
Infrastructure / field evidence
Other independent source

Keep the tracks separate. Agreement across sources can support prioritization; disagreement is also a result and should remain visible.

4. Corridor finding

  • Analysis window:
  • Unit of analysis: block, corridor, intersection, or other stated unit
  • Report count (n):
  • Exposure denominator and tier: observed / modeled / proxy / unknown
  • Rate and 95% confidence interval:
  • Comparison set:
  • Spatial result, if used:
  • Dominant reported hazard pattern, if publishable:
  • Quality / sensitivity flags:

Explain the rate in one plain sentence. Do not call a raw count a rate or treat a wide confidence interval as a precise rank.

5. What is missing or uncertain

  • Coverage gaps:
  • Reporting-bias risks:
  • Exposure alignment, staleness, or representativeness:
  • Small-sample or privacy suppression:
  • Geographic-boundary sensitivity:
  • Data rights, consent, or publication constraints:

If a missing denominator prevents a rate, retain this section and use a Tier 1 claim. “Unknown” is a result, not a zero.

6. Recommended next action

Describe the next step that fits the evidence tier. Examples include a field audit, targeted count, design scoping, accessibility walk-through, pilot evaluation plan, or capital-program review. Do not infer a specific engineering treatment solely from observational reports.

  • Action:
  • Owner:
  • Decision forum:
  • Required partners or reviewers:
  • Budget / policy / design dependency:

7. Follow-up measurement plan

  • Baseline period:
  • Follow-up period:
  • Metric to freeze before follow-up:
  • Minimum evidence needed to interpret a change:
  • Comparison corridor or other counterfactual, if feasible:
  • How reporting, exposure, and method changes will be separated:
  • Commitment: publish an inconclusive or null result as prominently as a positive result.

8. Verification record

  • Dataset / dossier version:
  • Source declarations and licenses checked:
  • Analysis configuration and input hashes:
  • Privacy / small-cell review:
  • Conformance-verifier result:
  • External method or accessibility review:
  • Published artifact URL, if consented:

Never put raw report coordinates, exact timestamps, notes, reporter tokens, or other identifying data in this dossier.

Read before using