On Power and Sample Size Calculation for Bayes Factors: A Tutorial on the BayesPower Package
Abstract
Power analyses and sample size justifications are increasingly encouraged and required by funding agencies across disciplines such as psychology and the biomedical sciences, highlighting the growing need for accessible sample size determination tools. The present article introduces BayesPower, an R package for conducting sample size determination, power calculation, and Bayes factor calculation in Bayesian hypothesis testing without simulation. The package supports a wide range of statistical tests commonly encountered in the social, behavioral, and biomedical sciences, including standardized mean differences, correlations, linear regression and analysis of variance, and tests for one and two proportions. In addition to the traditional comparison of a point-null hypothesis with a composite alternative hypothesis, BayesPower supports sample size planning for interval-hypothesis Bayes factors (i.e., equivalence testing). For most testing problems, sample size and power are returned almost instantly, enabling researchers to design informative and efficient studies rapidly. To reach a broader audience, the package provides a user-friendly Shiny application for researchers with little or no experience coding in R, in addition to command-line functions for more experienced users. The use of the software is illustrated through five empirical examples.
Related
- Introducing ResIN: An Applied Tutorial to the Response-Item-Networks Package for R
- Target Trial Emulation with the R Package TTE: A Tutorial and Methodological Guide
- How Wrong Can a Rank-Based Sample-Size Calculation Be? A Sharp Bound of 16/9 for Ordinal Outcomes
- Sample size calculation comparing two survival curves in the presence of nonproportional hazards utilizing the concept of Weighted Relative Time using two different Weibull distributions
- Support Vector Machines for Classification, Regression, and Outlier Detection: A Tutorial
BibTeX
@misc{wong2026power,
title = {{On Power and Sample Size Calculation for Bayes Factors: A Tutorial on the BayesPower Package}},
author = {Tsz Keung Wong and Samuel Pawel and Jorge Tendeiro},
howpublished = {PsyArXiv},
year = {2026},
month = sep,
doi = {10.31234/osf.io/pgdac_v4},
url = {https://osf.io/preprints/psyarxiv/pgdac_v4/},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - On Power and Sample Size Calculation for Bayes Factors: A Tutorial on the BayesPower Package
AU - Tsz Keung Wong
AU - Samuel Pawel
AU - Jorge Tendeiro
PY - 2026
DA - 2026/09/13
PB - PsyArXiv
DO - 10.31234/osf.io/pgdac_v4
UR - https://osf.io/preprints/psyarxiv/pgdac_v4/
AB - Power analyses and sample size justifications are increasingly encouraged and required by funding agencies across disciplines such as psychology and the biomedical sciences, highlighting the growing need for accessible sample size determination tools. The present article introduces BayesPower, an R package for conducting sample size determination, power calculation, and Bayes factor calculation in Bayesian hypothesis testing without simulation. The package supports a wide range of statistical tests commonly encountered in the social, behavioral, and biomedical sciences, including standardized mean differences, correlations, linear regression and analysis of variance, and tests for one and two proportions. In addition to the traditional comparison of a point-null hypothesis with a composite alternative hypothesis, BayesPower supports sample size planning for interval-hypothesis Bayes factors (i.e., equivalence testing). For most testing problems, sample size and power are returned almost instantly, enabling researchers to design informative and efficient studies rapidly. To reach a broader audience, the package provides a user-friendly Shiny application for researchers with little or no experience coding in R, in addition to command-line functions for more experienced users. The use of the software is illustrated through five empirical examples.
ER -