Your Cart

Your cart is empty

Home  >  Volume 21 (2012)

Likelihood Ratio Test as an Alternative to Chi-Square Test (Large Sample Approximation) by Amalare A.A, 2 Nurudeen T.S, and 3Adeniyi M.O Volume 21 (July, 2012), pp 199 – 206
Sale price: $5.00


In this paper , different methods of arriving at a reliable and authentic decision in hypothesis testing are discussed. The choice of the critical region, level of significance, testing equality of variances, maximum likelihood estimate and multinomial distribution are highlighted.  The test Statistic T of likelihood ratio test and its asymptotic approximation, –  T , are also carefully studied. Our study revealed that the proposed chi-square method  2∑▒0_i log(0_i/E_I ) as against the conventional chi-square ∑▒〖(〖0i-〗_(E_i ))〗^2/E_i  gives a better approximation of Likelihood ratio test ( -2logT▁((X) )) .

Keywords: Hypothesis testing, maximum likelihood estimate, likelihood ratio test, equality of variances, multinomial 

     distribution, restricted MLE