21¢‹ICOEuî•ñ‰ÈŠw‹Zpí—ªƒRƒAv


’´ƒƒoƒXƒg •sŠmŽÀ«‚̃‚ƒfƒ‹‰»ƒZƒ~ƒi[

“úŽžF2006”N 6ŒŽ13“ú(‰Î) 15Žž`16Žž40•ª
êŠFŒoÏŠw•”V“‚RŠK‘æ‚Q‹³Žº
u‰‰ŽÒFAtanu Biswas (Indian Statistical Institute)


ŠT—v

u‰‰ŽÒFAtanu Biswas (Indian Statistical Institute)

Title: Measures of association for nominal categorical variables

Abstract:
The present paper first reviews the existing measures of association for bivariate/multivariate nominal categorical variables and illustrates their drawback as measures of association. Some suitable measures of association are introduced in this context, which are flexible enough to find partial association between any two variables (for more than two variables at hand), eliminating the effect of the others. The methods are illustrated with some example.

˜A—æF‹î–Ø•¶•Ûi“àüF26941j


ŽÀs‘gDƒvƒƒWƒFƒNƒgŠT—vƒvƒƒWƒFƒNƒgÚ×ŽóÜ‚̏Љî‰ï‹cEŠÖ˜AsŽ–What's New—š—ð‚¨–â‚¢‡‚¹
iCj Copyright 2006 Information Science and Technology Strategic Core All Rights Reserved.