<b>nparLD</b>: An<i>R</i>Software Package for the Nonparametric Analysis of Longitudinal Data in Factorial Experiments

TLDR

Longitudinal data from factorial experiments arise in many fields, but their unknown distributions and outliers make parametric methods unreliable. The paper addresses this need by developing robust nonparametric methods for analyzing longitudinal factorial data. The authors present the R package nparLD, which offers user‑friendly implementation of state‑of‑the‑art rank‑based techniques for longitudinal factorial analysis. The package’s effectiveness is demonstrated through case studies in dentistry, biology, and medicine.

Abstract

Longitudinal data from factorial experiments frequently arise in various fields of study, ranging from medicine and biology to public policy and sociology. In most practical situations, the distribution of observed data is unknown and there may exist a number of atypical measurements and outliers. Hence, use of parametric and semi-parametric procedures that impose restrictive distributional assumptions on observed longitudinal samples becomes questionable. This, in turn, has led to a substantial demand for statistical procedures that enable us to accurately and reliably analyze longitudinal measurements in factorial experiments with minimal conditions on available data, and robust nonparametric methodology offering such a possibility becomes of particular practical importance. In this article, we introduce a new R package <b>nparLD</b> which provides statisticians and researchers from other disciplines an easy and user-friendly access to the most up-to-date robust rank-based methods for the analysis of longitudinal data in factorial settings. We illustrate the implemented procedures by case studies from dentistry, biology, and medicine.