Update To Partial Retraction Of Animal Value And Neuron Number
Read the original on Slate Star Codex →
Follows up on
↳ [PARTIALLY RETRACTED] Cortical Neuron Number Matches Intuitive Perceptions Of Moral Value Across Animals — Original research · Mar 2019
Summary
A model of good scientific self-correction. Scott's earlier (n=50) finding that people's moral valuation of animals scaled with cortical neuron count failed to replicate (commenter Tibbar's larger survey). David Moss (Rethink Priorities) resolved the discrepancy: Scott had offered a '99999 = basically infinity' option that ~5-10% used and Tibbar hadn't, and coding such N/A answers as 'infinity' vs discarding them reproduces either result. So the finding is fragile (a balance between equal-value voters and nine-spammers). Moral: explain your procedures well before replicating (and make sure the link to them works). This is the correction post itself, not a docked retraction.
Why this score
Quality 55 · Solid. A short, honest methodological-update post with a sound lesson (coding choices drive results; document procedures), but minor in scope -- resolving a small survey curio. Scored on merit as good self-correction (not docked-as-retracted, per the SSC-811 precedent). Mid-Solid.
Claude’s paradigm shift 30 · Slight. Slight. The 'explain your procedures' lesson is sound but familiar; a housekeeping update.
Real-world impact 2 · Minor. An honest methodological-update resolving why his animal-value/neuron finding failed to replicate (coding choices drive results); good self-correction modeling within the discourse, no material-world reach → RWI 2.