Hypothesis Tests for Covariance Analysis Models for Circular Data

Abstract

We introduce covariance analysis models for circular dependent variables. In addition, we have also developed hypothesis tests to evaluate the significance of the parameters used. One of the tests is an extension of the Watson–Williams test. The tests performances have been evaluated using simulation studies.

Keywords:

• Circular data
• Covariance analysis
• Directional data
• Hypothesis testing
• Watson–Williams test

Mathematics Subject Classification:

• Primary 62J99
• Secondary 62F03

Notes

*between parenthesis

(2) Non parallel model; (3) Parallel model with different intercepts; (1) Parallel model with the same intercept.