ResearchTrend.AI
  • Papers
  • Communities
  • Events
  • Blog
  • Pricing
Papers
Communities
Social Events
Terms and Conditions
Pricing
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 2008.11619
95
52
v1v2v3v4 (latest)

On the power of Chatterjee rank correlation

26 August 2020
Hongjian Shi
Mathias Drton
Fang Han
ArXiv (abs)PDFHTML
Abstract

Chatterjee (2021) introduced a simple new rank correlation coefficient that has attracted much recent attention. The coefficient has the unusual appeal that it not only estimates a population quantity first proposed by Dette et al. (2013) that is zero if and only if the underlying pair of random variables is independent, but also is asymptotically normal under independence. This paper compares Chatterjee's new correlation coefficient to three established rank correlations that also facilitate consistent tests of independence, namely, Hoeffding's DDD, Blum-Kiefer-Rosenblatt's RRR, and Bergsma-Dassios-Yanagimoto's τ∗\tau^*τ∗. We contrast their computational efficiency in light of recent advances, and investigate their power against local rotation and mixture alternatives. Our main results show that Chatterjee's coefficient is unfortunately rate sub-optimal compared to DDD, RRR, and τ∗\tau^*τ∗. The situation is more subtle for a related earlier estimator of Dette et al. (2013). These results favor DDD, RRR, and τ∗\tau^*τ∗ over Chatterjee's new correlation coefficient for the purpose of testing independence.

View on arXiv
Comments on this paper