Governance · Measurement Instrument
JRS Codebook
The authoritative definition of the five components of decision defensibility: the conditions used to evaluate whether a record can explain why a consequential decision was made, before that decision is questioned long after the fact. This codebook is the measurement instrument: review and analysis reference these definitions, and changes are version-controlled.
Maturity note: all conditions are currently classified Experimental. JRS is in operational validation. Nothing in this codebook constitutes validated measurement, statistical findings, certification, or legal guidance.
Conformance. A record conforms when it satisfies all five conditions, assessed against the record as it stands by a reviewer with no access to the original participants and no institutional memory of the events. A record is non-conforming when one or more conditions are not met, and the review identifies the specific condition not satisfied. Conformance reflects documentation sufficiency, not the correctness, lawfulness, or merits of the underlying decision.
RC 1
Reconstructability
Definition
Whether the conclusion in a record can be reconstructed from the record itself: a future reviewer can trace the path from documented evidence to the conclusion reached, without relying on the author's recollection or added explanation.
Detection Criteria
The evidence cited, the steps from evidence to conclusion, and the conclusion are all present in the file and connected.
Examples
A finding that states the documents reviewed, what each showed, and how they led to the determination.
Non-Examples
A conclusion ("performance concern") with no traceable path from any documented evidence.
Inclusions
Records where a conclusion or determination is asserted.
Exclusions
Purely descriptive logs that assert no conclusion.
False-Positive Risks
Marking a record reconstructable because the reviewer already knows the context (institutional memory) rather than because the file shows it.
False-Negative Risks
Penalizing concise records whose reasoning is implicit but genuinely recoverable from cited evidence.
Severity Guidance
High when the conclusion drives an adverse action and cannot be reconstructed; lower when the conclusion is minor or provisional.
Research Notes / Version
v1.0 (2026-06-03). Derived from the public five-condition framework. Revisions tracked in the Evidence Ledger.
RC 2
Basis Identification
Definition
Whether the source of each characterization (observation, measurement, audit finding, reported incident) is visible and traceable, rather than implied or summarized without attribution.
Detection Criteria
Each material characterization can be tied to an identified source in the record.
Examples
"Late on 3/4, 3/11, 3/18 (timestamps in shared drive)."
Non-Examples
"Has a poor attitude" with no source for the characterization.
Inclusions
Evaluative or factual characterizations of conduct, performance, or events.
Exclusions
Direct quotations or attached primary documents that are self-identifying.
False-Positive Risks
Accepting a generic reference ("per HR records") as an identified source.
False-Negative Risks
Requiring citation for uncontested, self-evident facts.
Severity Guidance
High when an unsupported characterization is central to an adverse conclusion.
Research Notes / Version
v1.0 (2026-06-03).
RC 3
Chronology
Definition
Whether the sequence of events is followable from the record alone, including timing of prior interventions, escalation steps, and the period under review.
Detection Criteria
Dated events in order; the review period and prior steps are identifiable.
Examples
Each incident documented with its own date and outcome.
Non-Examples
"Repeatedly counseled over the past several months" with no dates.
Inclusions
Records describing conduct or events over time.
Exclusions
Single-point-in-time records with no sequence.
False-Positive Risks
Treating a narrative order as a dated chronology.
False-Negative Risks
Requiring exact dates where approximate periods are sufficient and stated.
Severity Guidance
High when timing is material to the conclusion (e.g., progressive discipline).
Research Notes / Version
v1.0 (2026-06-03).
RC 4
Decision-Process Traceability
Definition
Whether a future reviewer can determine how the conclusion was reached: who reviewed the matter, what criteria or threshold triggered it, and whether responsive or mitigating information was considered before finalization.
Detection Criteria
Reviewer, applicable criteria/threshold, and consideration of responses are documented.
Examples
"Manager met with employee 3/20; notes on file; mitigating explanation recorded."
Non-Examples
A determination with no record of who decided or what standard applied.
Inclusions
Records reflecting a decision, finding, or escalation.
Exclusions
Informational records with no decision.
False-Positive Risks
Accepting a signature block as evidence of a documented process.
False-Negative Risks
Requiring elaborate process for low-stakes routine decisions.
Severity Guidance
High when an adverse decision shows no consideration of responsive information.
Research Notes / Version
v1.0 (2026-06-03).
RC 5
Evidentiary Sufficiency
Definition
The aggregate condition: whether the record stands on its own as an evidentiary document, such that an independent reviewer with no prior knowledge could assess whether the conclusion is supported by the documented evidence.
Detection Criteria
RC 1 to 4 are met to the degree that the record needs no outside context to be evaluated.
Examples
A record a reviewer encountering it cold could fully assess.
Non-Examples
A record that only makes sense to someone who was present.
Inclusions
All retained records intended to support later review.
Exclusions
Working drafts not intended for the file.
False-Positive Risks
Rating sufficiency on familiarity rather than on the file's contents.
False-Negative Risks
Holding routine records to a litigation-grade standard.
Severity Guidance
Reflects the most severe unmet condition among RC 1 to 4.
Research Notes / Version
v1.0 (2026-06-03).
Condition Maturity Model
Experimental
Definition drafted; little or no reviewer-agreement evidence. All five conditions are here now.
Emerging
Some reviewer-agreement data collected; definition stabilizing.
Stable
Consistent reviewer agreement across records; definition unchanged over revisions.
Validated
Independent replication and benchmark agreement established. Not yet attained for any condition.
This page is the human-readable codebook. Condition definitions and revisions are recorded in the Public Evidence Ledger. Study methodology and status are in the Research & Validation registry.