Evidence, meet assessment.

Compare recorded scores with your own judgment. Select records, edit assessments, then apply them together.

Local session · no uploads

Selected records

of 3 records

Evidence mean

40.00

2 scored · 1 not scored

Assessment mean

70.00

2 scored · 1 not scored

Records

Select to include in the comparison.

Seed set

Evidence is fixed. Seed evidence is a supplied reference, not independently validated here. Imported evidence is an unverified claim. Your assessment is separate. Blank = Not scored; 0 = scored zero. Use 0–100 with up to 2 decimals.

All assessments applied

Means use selected, scored records only. Zero counts; missing scores do not. Means are rounded half-up to 2 decimals. Summaries always reflect committed assessments.

Bring records in. Take them out.

Import replaces the current set. Undo brings it back.

Paste a comparison.v1 JSON package

1–50 records. Every imported evidence score is an unverified imported claim.

Exports keep recorded evidence separate from your assessments. Neither is independently validated here. Unapplied drafts are flagged; committed scores are exported.

Format & export rules

Each record needs id, label, evidenceScore and assessmentScore. Scores must be JSON numbers or null. IDs use letters, digits, underscores or hyphens. selectedIds contains existing IDs, without duplicates.

Export contains committed scores, selection, summary and pending-draft flags. Assessments never become validated evidence. Reimporting even a seed export labels all evidence as imported claims; provenance metadata cannot authenticate it.

Session history

Seed loaded. All three records selected. Nothing has been independently validated.

0 undo steps · up to 50 kept