Performance measurement is credible only when the construct, numerator, denominator, exclusions, data source, calculation and intended use are explicit. This note defines a CHSPS measure lifecycle and the 12-measure CPDS core set, with worked calculation and validation examples.
1. Measurement should answer a defined question
A credential-performance measure should state what construct it is intended to reflect, who is measured, what data are required, how the result is calculated and how the result should — and should not — be interpreted. Numbers without a specification create false precision.
2. CHSPS measure lifecycle
Define the problem, construct and intended use.
Define population, numerator, denominator, exclusions, data and calculation.
Evaluate feasibility, reliability, validity and unintended effects.
Run a learning cycle before public comparison.
Review performance, version changes and retirement triggers.
3. Minimum measure specification
| Identity | ID, title, version, status, steward and program. |
|---|---|
| Construct | What the measure is intended to represent and why it matters. |
| Population | Target population, denominator and unit of analysis. |
| Measure focus | Numerator or outcome definition. |
| Exclusions / exceptions | Named conditions removed from or handled outside the denominator. |
| Data | Source systems, data elements, lineage and timing. |
| Calculation | Formula, unit, rounding and stratification. |
| Testing | Feasibility, reliability, validity and known limitations. |
| Use | Internal monitoring, public reporting, benchmarking or recognition eligibility. |
| Change control | Version history and transition rules. |
4. CPDS core measure set
| ID | Core measure | Type | Key interpretation caution |
|---|---|---|---|
| CPD-01 | Candidate Volume & Completion | Context/operations | Volume and completion require denominator and scoreability context. |
| CPD-02 | First-Time Pass Rate | Outcome/context | Higher is not automatically better; candidate population and standard difficulty matter. |
| CPD-03 | Score Reliability | Measurement quality | The appropriate reliability statistic depends on the assessment design. |
| CPD-04 | Standard-Setting Currency | Context | Currency depends on current support for the active standard and material-change history. |
| CPD-05 | Material Exam Security Incident Rate | Rare event | Severity, detection capability and response matter as much as the rate. |
| CPD-06 | Accommodation Timeliness & Delivery | Conformance/service | Interpret timeliness together with correct delivery of approved accommodations. |
| CPD-07 | Appeal Rate & Overturn Share | Diagnostic | Neither a low appeal rate nor a high overturn share is interpretable alone. |
| CPD-08 | Decision Timeliness | Service | Use a defined trigger and separate external delays where possible. |
| CPD-09 | Renewal & Lapse Profile | Context | Renewal patterns depend on cycle design, population and lapse rules. |
| CPD-10 | Complaint Rate & Resolution | Diagnostic/service | Complaint volume requires scope and denominator context; resolution time is separate. |
| CPD-11 | Verification Responsiveness | Service | Measure only complete verification requests against the published service standard. |
| CPD-12 | Submission Data Quality & Traceability | Data quality | Completeness does not replace source-record agreement and lineage testing. |
5. Worked example
Numerator: 1,024 candidates whose first scoreable attempt met the approved passing standard.
Denominator: 1,303 candidates with a first scoreable attempt during the measurement period.
Result: 1,024 ÷ 1,303 = 78.6%.
78.6% is not a CHSPS target. A pass rate should not be interpreted without the assessment standard, candidate population, eligibility rules and stability of the measure.
6. Data-validation example
An organization reports 1,805 scoreable administrations. A validator selects a source-record sample of 76 records and confirms agreement on 75. The worked-example source-record agreement is 98.7%. The single discrepancy must still be characterized — for example, a harmless date-format issue is different from a denominator omission that changes the reported result.
7. Public reporting maturity
CHSPS recommends separating early learning from public comparison. A measure can progress from concept → pilot → reporting-only → validated reporting → benchmark eligible → recognition eligible. Public rankings should wait until specifications, validation, denominators and comparison methods are sufficiently stable.
CHSPS references
Accreditation requirements are stated exclusively in the published, version-controlled CHSPS standards. This note explains method, risk and good practice; it creates no requirement and changes none.
- CHSPS Standards LibraryThe controlled Edition 2026.1 standards for all 16 active CHSPS program codes.
- CHSPS Standards InterpretationsFormal interpretations issued under document control when a requirement needs clarification.
- CHSPS Technical ReportsPublished CHSPS methods and analyses supporting evidence quality, assessment integrity and performance measurement.