How role-fit is calculated
The rollup follows a fixed chain. First, within each rubric criterion, the scores from all evaluators who scored that criterion are averaged. Next, the criterion averages combine according to the rubric's criterion weights to produce a task score. Task scores then roll up to the skills they measure.
Finally, each skill's contribution is weighted by its skill weight multiplied by its section weight, and the role-fit percentage comes from this weighted rollup. Because every step is deterministic, the same evidence always produces the same role-fit number.
Rubric scores override auto scores
Some tasks carry both an automatic score (for example, from a multiple-choice answer key) and a rubric score entered by evaluators. When both exist for the same task, the rubric score wins and is the one used in the rollup.
One related rule to keep in mind: an item with no answer key auto-scores as null, which means it needs manual review. A null is never treated as a zero, so unreviewed responses cannot silently drag a candidate's role-fit down.
Where you set the weights
Weights live in two places in the builder. Criterion weights are set in the rubric editor: each criterion carries a multiplier (written as ×N) that controls how much it counts within its task. Section weights are set on each section — open the section's edit form in the structure editor and set its Scoring weight. The Scoring & reporting panel then summarizes the weights read-only, alongside the counts of auto, rubric, and manual scoring across the assessment.
Weights are editable on both draft and live assessments. Because role-fit is computed from stored scores using the current weights, changing a weight on a live assessment re-aggregates role-fit for every candidate, including those who already submitted — raw scores never change, but the percentages shift. On blueprint-linked assessments, recorded decisions keep a role-fit snapshot from decision time. Closed and archived assessments are locked; duplicate to a new draft to change them.
- Criterion weights: rubric editor, per criterion (×N).
- Section weights: each section's edit form in the structure editor; summarized in Scoring & reporting.
- Weight changes on a live assessment re-aggregate role-fit for everyone; raw scores never change.
Where weights show up in reports
Candidates and reviewers can see that weighting was applied. In the candidate evidence report, tasks are grouped by section whenever the assessment has multiple sections or weights, each custom-weighted group carries a ×N weight badge, and a weighted-scoring note explains that scores were combined by weight rather than averaged flat. The audit pack also records section weights, rubric criterion weights, and skill mappings alongside a plain-language methodology.
At a glance
- Within a criterion, evaluator scores are averaged; criteria combine by rubric weights.
- Task scores roll up to skills; each skill contributes by skill weight times section weight.
- Role-fit % is the result of this weighted rollup.
- If a task has both a rubric score and an auto score, the rubric wins.
- Weights appear as badges in candidate evidence reports, with a weighted-scoring note.