Publications
Below you can find the main publications pertaining to powerly
:
Constantin, M. A., Schuurman, N. K., & Vermunt, J. K. (2023). A General Monte Carlo Method for Sample Size Analysis in the Context of Network Models. Psychological Methods. https://doi.org/10.1037/met0000555
The network approach to psychology is an increasingly popular framework for studying interactions among variables. As the field matures and psychological network modeling becomes more prevalent, there is an increasing need to aid researchers with a network approach in mind that plan to collect data. In this paper, we introduce a general method for performing sample size analysis in the context of network models. The method takes the form of a three-step algorithm designed to find an optimal sample size value given a hypothesized network, an outcome measure (e.g., sensitivity), and a statistic of interest (e.g., power). It starts with a Monte Carlo simulation step for computing the outcome measure and the statistic at various sample sizes. It continues with a curve-fitting step for interpolating the statistic. The final step employs bootstrapping to account for the uncertainty around the interpolated curve. The method is implemented in the form of an R package called powerly, freely available on GitHub and CRAN.
Constantin, M. A., Schuurman, N. K., & Vermunt, J. K. (2025). A General Framework for Sample Size Analysis [Manuscript in preparation]. Methodology and Statistics, Tilburg University.
With the release of
powerly
v2.0.0
we go beyond just applying the method in Constantin et al. (2023) to network models. We provide a sophisticated system that can be intuitively extended to any statistical model and research question.
You may also be interested in:
Constantin, M. (2025).
boterview
: A Large Language Model Interview Toolkit for Social Science Experiments [Manuscript in preparation]. Methodology and Statistics, Tilburg University. https://boterview.devboterview
is aPython
package that enables social science researchers to easily deploy chatbot-based interviews with customizable protocols and randomized condition assignment. Complex study designs can be easily set up without programming knowledge beyond editing aTOML
(i.e., plain text) configuration file and providing interview content inMarkdown
files.Constantin, M. A. (2025).
parabar
: A Lightweight Framework for ParallelizingR
Code With Asynchronous and Decoration Semantics [Manuscript in preparation]. Methodology and Statistics, Tilburg University. https://parabar.mihaiconstantin.comparabar
is a package designed to provide a simple interface for executing tasks in parallel, while also providing functionality for tracking and displaying the progress of the tasks. This package is aimed at two audiences: (1) end-users who want to execute a task in parallel in an interactiveR
session and track the execution progress, and (2)R
package developers who want to useparabar
as a solution for parallel processing in their packages. Theparabar
can also be used in conjunction with the accompanyingdoParabar
R
package to enableforeach
semantics.