EquiCut: A Web Tool for Determining Empirical Cut-Off Points Using an Entropy Measure

Read full text Summarise with VIRAT
Authors: Philippe Garvie, Jacques D. Marleau
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

(Background): The dichotomization of continuous or ordinal scales is a common practice in psychometric and survey research, but the justification for choosing a specific cut-off point is often weak, especially in the absence of an external criterion. (Tool Presentation): We introduce EquiCut, a free, open-source, and user-friendly web tool designed to address this challenge. (Methodology): Equi-Cut implements an information-theoretic method based on Shannon’s entropy. The entropy value for a score distribution is first computed, and an equivalent equitability index is obtained by normalizing it by the number of categories. The optimal cut-off is identified by finding the dichotomized distribution whose equitability index best matches that of the original multi-category distribution, thus preserving its structural properties. (Features): The tool requires no programming knowledge or software installation. Users simply upload a frequency data file, and the tool provides a comprehensive report. (Availability): The web tool and its full Python source code are freely available.

Fields of Study & Subjects

Psychometric Methodologies and Testing Statistical Methods and Applications Probability and Statistical Research

Publication details

Journal / source
PsyArXiv (OSF Preprints)
Publication date
25 March 2026
Language
EN
Document type
Preprint