Skip to content
Sparqbox Docs

Why reviewers score independently

When two reviewers are scoring the same idea, neither one sees the other's score until both are done. The AI doesn't see the human scores either. This is intentional.

The point: prevent anchoring

Section titled “The point: prevent anchoring”

When the first score is visible to the second reviewer, the second reviewer scores against it instead of against the rubric. This is anchoring, and it's an unconscious effect, not a discipline problem. Independent scoring eliminates it.

Once every assigned reviewer has finished, the system unlocks the full scorecard view and you can see how each column landed. This is the right time to discuss differences.

What you see while scoring

Section titled “What you see while scoring”

While the idea is in_review and you haven't submitted yet:

  • Your own column is visible and editable.
  • Other reviewers' columns are hidden, even if those reviewers have already submitted.
  • The AI's column is hidden until you submit.

Once your scorecard is in, you can re-open the idea to see all columns side by side.

How reviewers are assigned

Section titled “How reviewers are assigned”

Reviewers are assigned automatically at submission: the category coordinators and the AI reviewer are added when the idea enters in_review. An admin can also adjust assignments after that point.

Admins can add or remove individual reviewers directly from the idea detail page, as long as the idea is still in submitted or in_review. Reviewers who have already completed their scorecard are protected and cannot be removed. The AI reviewer is always present and is not removable. Removing a reviewer who has started but not finished discards their in-progress scores.

The existing Reassign reviewers action in the idea header remains available for wholesale reassignment.

This means reviewer assignments are not final at submission. If the initial set is wrong - for example, the wrong coordinator was auto-assigned or a specialist needs to be added - an admin can correct it without reopening or rerouting the idea.

Comparing notes after submit

Section titled “Comparing notes after submit”

Differences between reviewers are useful, not embarrassing. A two-point gap on a single criterion almost always points at one of the same three things you'd find in a human-AI gap:

  • The criterion is ambiguous and the rubric needs sharpening.
  • The idea didn't say enough and a clarification would close the gap.
  • One reviewer has context the other doesn't.

Discuss the gap; don't average it away in your head.

What's next

Section titled “What's next”

Reading AI scores

Reading AI scores explains how to combine the AI's score with yours.