power contours

power contours visualizes statistical power across combinations of trial numbers and participant counts to guide experimental design decisions in psychology and neuroscience.


Key Features:

  • Statistical Power Visualization: Generates two-dimensional iso-power contour plots that map statistical power as a function of numbers of trials and participants.
  • Variance Analysis: Quantifies the influence of within-participant variance relative to between-participant variance and identifies when increasing trial numbers substantially affects power.
  • Application Across Paradigms: Validated using datasets from eight experimental paradigms, including reaction times, sensory thresholds, fMRI, MEG, and EEG.
  • Practical Implementation: Provides example code to estimate within- and between-participant variance for each method to inform study design.

Scientific Applications:

  • Study Design Optimization: Optimize numbers of trials and participants to achieve target statistical power while minimizing resource use in behavioral and neuroimaging experiments.
  • Modality-Specific Power Assessment: Assess detectability of effects in paradigms such as reaction times, sensory thresholds, fMRI, MEG, and EEG by accounting for trial-level and participant-level variance.
  • Reproducibility and Resource Allocation: Inform reproducible experimental configurations by balancing trial counts and sample sizes based on observed variance structure.

Methodology:

Computes two-dimensional iso-power contour plots from estimates of within- and between-participant variance derived from empirical datasets, with validation using datasets from eight paradigms including reaction times, sensory thresholds, fMRI, MEG, and EEG.

Topics

Details

Added:
1/18/2021
Last Updated:
11/24/2024

Operations

Publications

Baker DH, Vilidaite G, Lygo FA, Smith AK, Flack TR, Gouws AD, Andrews TJ. Power contours: Optimising sample size and precision in experimental psychology and human neuroscience.. Psychological Methods. 2021;26(3):295-314. doi:10.1037/met0000337. PMID:32673043. PMCID:PMC8329985.

PMID: 32673043
PMCID: PMC8329985
Funding: - Wellcome Trust: 105624 - U.K. Biotechnology and Biological Sciences Research Council: BB/H008217/1