iFishIENCi (Jan 2023) — African catfish Hungary panel
https://zenodo.org/records/10842711 →benchmark:ifish-catfish-hungary-2023
Shape mix:
2
unimodal
Δ 0.50
4.10
()
→ 3.60
()
within-fixture gradient
Acceptance rubric
An SSR re-run acceptably reproduces this benchmark when it clears every check below on the same 2 paired questions the source publishes. The chip badges name the checks the round-38 → round-46 shape-mismatch finding argued are necessary but not sufficient in isolation.
CI overlap
2/2
Shape match
2/2
SSR mean inside CI · shape badge reproduces the source label.
What this benchmark discloses
Recruitment screener (verbatim)
Consumer/sensory panel recruited by HAKI (Research Institute for Fisheries and Aquaculture) in Hungary for the iFishIENCi project D3.6 catfish wave. Specific inclusion criteria not in the public spreadsheet; see project deliverable PDF at https://ifishienci.eu/wp-content/uploads/2024/01/iFishIENCi_D3.6.pdf.
Sample weighting & demographics
Unweighted hedonic. 9-point discrete scale per ASTM/ISO 11036. Single rating per product per panellist per attribute. Source spreadsheet aggregates per-attribute histograms only; raw per-panellist scores are not published. The paired distributions here are reconstructions from those histograms.
Fieldwork dates, country, language
Hungary (HAKI Research Institute for Fisheries and Aquaculture). Fieldwork month: 2023-01 (from filename _01_2023). Wave compares Control feed vs Algae-supplemented feed; no Insects or Bioactive arm.
Original question text & translations
What is your overall impression of this fish fillet? (rows 81-83 of the source workbook) — 9-point hedonic scale per ASTM/ISO 11036, single rating per panellist per product.
Likert / scale labelling
9-point hedonic per ASTM/ISO 11036 standard (Peryam–Pilgrim 1957). 1 = "dislike extremely". 5 = "neither like nor dislike". 9 = "like extremely". Intermediate anchors are conventional. Labels not embedded in the source workbook; standard sensory-science instrument cited.
FGI moderator guide & coding rules
Hedonic survey by consumer panel, not FGI. No moderator guide. The source spreadsheet uses a sample-per-sheet design with attribute-block questionnaires; the distributions here are reconstructed from those aggregated histograms rather than from raw per-panellist scores.
Data source & license
CC-BY 4.0. Attribution: Rost, A. et al., ttz Bremerhaven (2024), iFishIENCi D3.6 sensory analysis. https://zenodo.org/records/10842711
Anything not disclosed here was not available in the source dataset and is recorded as such.
SSR run — reproducibility keys
- template_version
- 3.0.0
- llm_seed
- 4209
- bootstrap_seed
- 8418
- llm_model
- deepseek-v4-flash
- embedding_model
- nvidia/nv-embed-v1
SSR score vs ground truth, by question
Each row holds one question. The teal mark and range are SSR. The sky mark and range are the source FGI/survey.
§
Overall, how much do you like this African catfish fillet? (Control — standard commercial feed)
SSR
—
Source
3.60 (3.18–4.02)
Source distribution
SD 0.98
skew -0.15
9 cells
unimodal
1.02.03.04.05.0
§
Overall, how much do you like this African catfish fillet? (Algae — microalgae inclusion in the feed)
SSR
—
Source
4.10 (3.82–4.37)
Source distribution
SD 0.65
skew -0.22
9 cells
unimodal
1.02.03.04.05.0
More on this shelf
Aquaculture sensory panels
9-pt hedonic on aquaculture species across feed systems (Control / Insects / Algae / Bioactive), iFishIENCi D3.6. The gradient is which feed system the panel actually prefers.