Use response sets, verdicts and scoring options

Response sets are reusable answer scales containing labels and scoring metadata such as numeric value, colour and not-applicable behaviour. Verdicts provide consistent review outcomes for RFI evidence. They reduce repeated setup, but a shared scale is useful only when every option has a documented meaning and published responses remain bound to the release they answered.

Before you start

• Agree the order, meaning and score direction of the scale with evaluation owners.

• Check whether an existing response set already expresses the intended evidence before creating a near-duplicate.

Step by step

1. Review shared sets

Open Settings → Response sets and inspect the existing labels, colours, values and not-applicable options. Use the set whose meaning matches the question, not merely one with the right number of choices.

2. Create or revise a draft set

Name the set clearly and define each option. Add score, colour, flagged or correct-answer behaviour only where the evaluation uses it.

3. Configure not applicable

Use N/A when a criterion can genuinely be outside scope. Verify whether it is excluded from scoring rather than treating it as a zero by assumption.

4. Apply the set in the builder

Select the response set on eligible draft questions and preview both respondent labels and reviewer scoring. Explain the scale in question guidance when labels could be interpreted differently.

5. Maintain review verdicts

Use consistent RFI verdicts for evidence review, separate from the raw answer score. Choose a verdict only after reading the response and its supporting artefacts.

6. Version rather than reinterpret history

When a published meaning changes, update the draft and publish a new form version. Do not assume editing a shared set should recalculate historic responses.

What happens next

• Set scoring mode and weights after confirming that every scored option maps in the intended direction.

• Document when reviewers should use a verdict, score, comment or not-applicable response.

Troubleshooting

An edited set did not change an old response

This protects historic meaning. Published responses remain associated with their release; issue a new version for future respondents.

A set is unavailable in the builder

Check the question type, workspace and whether the set is active or valid. Some types use their own choices rather than a shared set.

N/A lowers the vendor score unexpectedly

Inspect the set’s not-applicable flag and the evaluation’s scoring behaviour. Correct the draft configuration and test it before publishing another release.

Related guides