Skip to content

Read results and export outcomes

Open Results to understand submitted judgments and assessed coverage. Read the population, selection method, rubric version, and completion state before comparing scores.

Read score and coverage together

Assessed coverage is the share of applicable criterion weight reviewers could judge from the evidence. N/A answers are excluded from the applicable amount. Insufficient-evidence answers reduce coverage rather than becoming passes or failures.

For example, two reviews can show similar scores while one has much less assessable evidence. The lower-coverage result supports a narrower conclusion. This example illustrates interpretation; it is not a published benchmark.

Open periods can show provisional results. Their totals may change as reviews are submitted or corrected. A period with no completed assessments does not establish either good or poor quality.

Synthetic results showing review outcomes, assessed coverage, and selection context.

Compare the assessment outcome with coverage and sample context. Example data is synthetic.

Compare like with like

  • Keep rubric versions visible. A revised standard can change scores without a change in service.
  • Separate random, targeted, coverage, and manually included selections.
  • Check sample shortfalls, exclusions, and incomplete work.
  • Use later-period comparisons with the same context in view.
  • Keep calibration and standalone ticket reviews separate from ordinary period totals.

A targeted sample highlights an area of interest; it should not be presented as a random estimate for the whole service.

Export completed results

Use the completed-results CSV action in Results. The export contains authorized completed outcomes, criterion/rubric and selection labels, and coverage. It excludes private notes, quotations, and AI text.

Current Jira issue visibility governs the numeric rows. This is different from the source checks for content-bearing records, which also consider cited comments. Inaccessible issue rows are excluded. If the app cannot establish a complete and consistent authorized export, it withholds the file rather than returning a misleading partial result.

Downloaded files cannot be recalled after permissions change. Store and share them under your organization's policies, and export needed outcomes before uninstalling or deleting app data.

Decide what to do next

Use Findings and coaching to connect a result to a practical action. Avoid drawing a broad conclusion from one score, a low-coverage assessment, or a deliberately targeted sample.