Moderated consolidation (normalise + inter-judge moderation)

Q: Moderated consolidation (normalise + inter-judge moderation)

Moderated consolidation What it does. The default (AEA) method. Each judge's raw scores are converted to z-scores and rescaled to a configurable target mean and standard deviation (defaults 85 ± 5), so a consistently harsh or lenient judge no longer skews results. Judges are then moderated against the lead judge and averaged per entry. Small categories (fewer than 3 entries) fall back to a linear map of raw scores onto the target band. How to use it. Admin → Program → Scoring → Score Consolidation = "Moderated". Set the target mean/SD in the Moderation Mean / SD row. When to prefer it. Multiple judges of differing strictness scoring overlapping entries, the standard fair-comparison choice.

Question

Accepted Answer

Moderated consolidation

What it does. The default (AEA) method. Each judge's raw scores are converted to z-scores and rescaled to a configurable target mean and standard deviation (defaults 85 ± 5), so a consistently harsh or lenient judge no longer skews results. Judges are then moderated against the lead judge and averaged per entry. Small categories (fewer than 3 entries) fall back to a linear map of raw scores onto the target band.

How to use it. Admin → Program → Scoring → Score Consolidation = "Moderated". Set the target mean/SD in the Moderation Mean / SD row.

When to prefer it. Multiple judges of differing strictness scoring overlapping entries, the standard fair-comparison choice.