Groupcompare: A holistic approach to two sample data analysis
SoftwareX, vol.35, 2026 (SCI-Expanded, Scopus)
- Publication Type: Article / Article
- Volume: 35
- Publication Date: 2026
- Doi Number: 10.1016/j.softx.2026.102835
- Journal Name: SoftwareX
- Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Compendex, INSPEC, Directory of Open Access Journals
- Keywords: Bootstrap, Permutation test, R package, Two-sample comparison
- Dokuz Eylül University Affiliated: Yes
Abstract
The groupcompare R package provides a unified framework for two-sample statistical comparisons. It addresses critical limitations of traditional methods by integrating parametric, non-parametric, permutation, and bootstrap-based techniques within a single package. The package notably incorporates robust descriptive statistics such as Huber's M-estimator and the Harrell-Davis quantile estimator, enhancing reliability when dealing with non-normal data or outliers. Furthermore, groupcompare offers helpful bivariate plots and statistical inferences, simplifying interpretation and reporting.