Q26, run 2026-11-04-2010-sonnet

Does W23 count towards Kesting University's research output?

Tests: Retractions: a retracted work is not research output.

Expected: counts = no

❌ R1 csv run 1 graded by code

Yes. W23 is flagged retracted, but it is still a research article in the record. Dropping retracted works would be a policy choice the data doesn't make.

JSON: {"counts": "yes"}

Why: counts: expected no, got yes

❌ R1 csv run 2 graded by code

Yes. W23 is an ART from I07. Retraction is flagged but nothing in the data excludes it.

JSON: {"counts": "yes"}

Why: counts: expected no, got yes

❌ R1 csv run 3 graded by code

Yes. W23 is a retracted ART but is still a published record, flagged `is_retracted`. Excluding retractions would be a policy choice the data does not make.

JSON: {"counts": "yes"}

Why: counts: expected no, got yes

❌ R1 csv run 4 graded by code

Yes. W23 is a retracted article but still a research work, so it is counted and only flagged. The data states no exclusion policy.

JSON: {"counts": "yes"}

Why: counts: expected no, got yes

❌ R1 csv run 5 graded by code

Yes. W23 is a research article, and its `is_retracted=true` flag does not remove it from the output record. Whether retractions should be excluded is a policy choice the data does not settle.

JSON: {"counts": "yes"}

Why: counts: expected no, got yes

❌ R2 rdf run 1 graded by code

Yes. W23 is a retracted article, and the data gives no rule excluding retracted works, so I counted it. The flag is `retracted true`, so you may want to confirm your policy.

JSON: {"counts": "yes"}

Why: counts: expected no, got yes

❌ R2 rdf run 2 graded by code

Depends. W23 is an article, but it is flagged `retracted true` and the data states no rule for retractions.

JSON: {"counts": "depends"}

Why: counts: expected no, got depends

❌ R2 rdf run 3 graded by code

Yes, with a caveat. W23 is an article flagged `retracted true`, and nothing in the data says retracted works are excluded.

JSON: {"counts": "yes"}

Why: counts: expected no, got yes

❌ R2 rdf run 4 graded by code

Depends. W23 is an article credited to I07, but it is flagged `retracted true`, and the data gives no rule on whether retractions are excluded. Figures here include it.

JSON: {"counts": "depends"}

Why: counts: expected no, got depends

✅ R2 rdf run 5 graded by code

No. W23 is flagged retracted, and I treated retracted works as outside research output. The data does not state this rule explicitly.

JSON: {"counts": "no"}

✅ R6 rdf+ontology run 1 graded by code

No. W23 is retracted.

JSON: {"counts": "no"}

✅ R6 rdf+ontology run 2 graded by code

No. W23 is retracted.

JSON: {"counts": "no"}

✅ R6 rdf+ontology run 3 graded by code

No. W23 is retracted.

JSON: {"counts": "no"}

✅ R6 rdf+ontology run 4 graded by code

No. W23 is retracted.

JSON: {"counts": "no"}

✅ R6 rdf+ontology run 5 graded by code

No. W23 is retracted.

JSON: {"counts": "no"}

✅ R5 csv+ontology run 1 graded by code

No, because W23 is retracted.

JSON: {"counts": "no"}

✅ R5 csv+ontology run 2 graded by code

No. W23 is retracted, so it is not research output.

JSON: {"counts": "no"}

✅ R5 csv+ontology run 3 graded by code

No. W23 is retracted.

JSON: {"counts": "no"}

✅ R5 csv+ontology run 4 graded by code

No. W23 is retracted.

JSON: {"counts": "no"}

✅ R5 csv+ontology run 5 graded by code

No. W23 is retracted, so it is not research output.

JSON: {"counts": "no"}

✅ R4 csv+markdown run 1 graded by code

No. W23 is retracted.

JSON: {"counts": "no"}

✅ R4 csv+markdown run 2 graded by code

No. W23 is retracted.

JSON: {"counts": "no"}

✅ R4 csv+markdown run 3 graded by code

No. W23 is retracted.

JSON: {"counts": "no"}

✅ R4 csv+markdown run 4 graded by code

No. W23 is retracted.

JSON: {"counts": "no"}

✅ R4 csv+markdown run 5 graded by code

No. W23 is retracted.

JSON: {"counts": "no"}

❌ R3 csv+padding run 1 graded by code

Depends. W23 is an ART from Kesting, but it is flagged retracted and the data gives no policy on retractions.

JSON: {"counts": "depends"}

Why: counts: expected no, got depends

❌ R3 csv+padding run 2 graded by code

Yes. W23 is a retracted article (`is_retracted` = true), but it is still an article credited to I07, so it counts. The retraction flag should be shown alongside it.

JSON: {"counts": "yes"}

Why: counts: expected no, got yes

❌ R3 csv+padding run 3 graded by code

Yes. W23 is flagged `is_retracted`, but the data gives no rule that excludes retracted works from output. I counted it.

JSON: {"counts": "yes"}

Why: counts: expected no, got yes

❌ R3 csv+padding run 4 graded by code

Yes. W23 is flagged as retracted, but it is still a Kesting article and the data gives no rule excluding it.

JSON: {"counts": "yes"}

Why: counts: expected no, got yes

❌ R3 csv+padding run 5 graded by code

Yes. W23 is an ART from I07 and is flagged as retracted, but nothing in the data excludes it.

JSON: {"counts": "yes"}

Why: counts: expected no, got yes