Human cognitive neuroscience seminar

Speaker: Robert McIntosh (University of Edinburgh)

Title: The problem of power in single-case neuropsychology

Abstract: Researchers and clinicians in neuropsychology often compare individual patients against healthy control samples, to quantify evidence for cognitive-behavioural deficits and dissociations.  Statistical methods for these comparisons have been developed that control Type I (false positive) errors effectively. However, remark-ably little attention has been given to the power of these tests. In this practical primer, we describe, in minimally technical terms, the origins and limits of power for case-control comparisons. We argue that power calculations can play useful roles in single-case study design and interpretation, and we make suggestions for optimising power in practice.  As well as providing figures, tables, and tools for estimating the power of case-control comparisons, we hope to assist researchers in setting realistic expectations for what such tests can achieve in general.

Recommended papers:

1. McIntosh, R. D., & Rittmo, J. Ö. (2020). Power calculations in single case neuropsychology. https://psyarxiv.com/fxz49
2. McIntosh, R. D. (2018). Simple dissociations for higher-powered neuropsychology. Cortex, 103, 256-265.

Relevant points of discussion might include the following:

1. Does a lack of complete control over power mean that single-case studies are inherently unsuited to study pre-registration?
2. Does the chronically low power of case-control comparisons mean that such studies should generally be regarded as exploratory (that is, hypothesis-generating, rather than hypothesis-testing)?
3. Is publication bias (towards positive findings) particularly severe for single-case studies, and - if so - what implications does this have?
4. How should single-case studies address the issue of alpha correction, if deficits or dissociations have been examined in multiple patients?
5. Can the power constraints of single-case studies justify an alpha level (i.e. significance threshold) higher than the conventional .05?
6. Should we do away with the traditional distinction between classical and strong dissociations?