In correlation do we trust? Simpson’s paradox for Kendall’s rank coefficient

Denuit, Michel;Trufin, Julien;Zuyderhoff, Pierre
(2024) , 14 pages

Files

ISBA_DP_2024-26.pdf
  • Open Access
  • Adobe PDF
  • 658.91 KB
Download

Details

Authors
  • Denuit, MichelUCLouvain
    Author
  • Trufin, Julien
    Author
  • Zuyderhoff, Pierre
    Author
Abstract
This note revisits Simpson’s paradox in insurance and discusses confounding effects of hidden covariates on Kendall’s tau. As a result, observed correlation may vanish or even revert. This phenomenon is shown to have important consequences in insurance risk assessment and ratemaking.
Affiliations

Citations

  • APA
  • Chicago
  • FWB

Denuit, M., Trufin, J., & Zuyderhoff, P. (2024). In correlation do we trust? Simpson’s paradox for Kendall’s rank coefficient (LIDAM Discussion Paper ISBA 2024/26). https://hdl.handle.net/2078.5/279490