Abstract

This paper proposes a modified extended median test that adjusts for the effects of any possible ties in the specific data. The proposed method is illustrated with data. A comparison with the extended median test suggests that the proposed method is at least as powerful as the extended median test.In this study, the proposed extended median test presents an example indicating that the proposed modified extended median test is more powerful than the ordinary extended median test.

Key words: Median test, tied observation, chi-square, P-value, χ² distribution.