---
title: "Andrew Heathcote"
type: "entity"
entity_kind: "person"
status: "hub"
canonical_name: "Andrew Heathcote"
aliases: []
first_seen: "2026-07-25T00:00:00.000Z"
writer_model: "claude-opus-4-8"
connects_to: ["averaging artifact","power law of practice","exponential law of practice","cognitive psychology (response-time modelling)","Scott Brown"]
---


Cognitive psychologist and quantitative modeller of learning and response time. In the vault he anchors the **averaging-artifact** cluster: with Scott Brown and D.J.K. Mewhort he authored *The Power Law Repealed* (Psychonomic Bulletin & Review, 2000), which showed that the celebrated power law of practice fits worse than an exponential in every *unaveraged* dataset and appears only after linear averaging over learners with heterogeneous rate parameters. He matters to this vault because that single finding turns a supposed universal law into a property of how data is aggregated — a reflex the vault then tracks recurring across fields and decades.

## References

- [[claim-heathcote-2000-power-law-of-practice-is-an-averaging-artifact]] — the core repeal: individuals speed up exponentially, only the group mean looks like a power law.
- [[claim-heathcote-2000-averaging-distortion-requires-rate-parameter-variability]] — the precise algebraic condition: distortion requires, and is proportional to, variability in individual rate parameters (after Myung, Kim & Pitt 1998).
- [[claim-averaging-artifact-across-practice-and-scaling-laws-is-analogous-symptom-not-same-mechanism]] — his mechanism vs. the 2026 neural-scaling-law critique: analogous symptom, distinct process.
- [[observation-averaging-heterogeneous-learners-manufactures-a-power-law-no-individual-obeys]] — the cross-domain bridge his 2000 result forms with Hu et al. (2026).