The simssd R package uses simulation to do sample size determination (SSD) and power computation for fixed effects in multilevel linear regression models. It has a focus on improving computational speed.
Project status
This repository contains the initial project skeleton. Development is currently on hold due to time constraints; the source code and documentation are placeholders as they have not yet been published.
The project is not abandoned and I intend to continue work on it when time permits, though there is no active development at the moment.
As described under Historical context, this package was originally intended as an implementation of my PhD research. Since the repository was first created, the underlying research has been extended, so any future development would reflect the methods described in that work.
Related publication:
Browne, W. J., Charlton, C. M. J., Price, T., Leckie, G., & Steele, F. (2025). Optimizing the Use of Simulation Methods in Multilevel Sample Size Calculations. Journal of Educational and Behavioral Statistics, 0(0). doi:10.3102/10769986251344939
Installation
Installation instructions will be provided at a later date. Meanwhile, see the Introduction to simssd.
Historical context
The predecessor to simssd was developed to support my PhD research, A faster simulation approach to sample size determination for random effect models, at the Centre for Multilevel Modelling (University of Bristol).
It extended ideas arising from the MLPowSim software written by William Browne and Mousa Golalizadeh.
Acknowledgements
I gratefully acknowledge funding provided for my PhD via UK Economic and Social Research Council (ESRC) grant number ES/H044094/1.
My thanks to the late Professor Jon Rasbash for getting the original project off the ground as well as Professor William Browne, Professor Fiona Steele, CBE, Professor Debora Price and the late Professor Harvey Goldstein for their invaluable guidance and support.
The MLPowSim manual by William Browne, Mousa Golalizadeh and Richard Parker contains a number of motivating examples.
The software design of simssd draws on ideas from:
Chalmers RP, and Adkins, MC (2020). Writing effective and reliable Monte Carlo simulations with the SimDesign package. The Quantitative Methods for Psychology, 16(4), 248–280. doi:10.20982/tqmp.16.4.p248.
In an ongoing way, tools provided by Hadley Wickham and his colleagues at Posit (formerly RStudio) enable me to develop much higher quality software in R than I otherwise would have been able to. Thank you Hadley & others at Posit 🙂
Last updated: 04 Aug 2026