JACKKNIFE ALGORITHM FOR THE ESTIMATION OF LOGISTIC REGRESSION PARAMETERS

Abstract

This paper proposes an algorithm for the estimation of the parameters of logistic regression analysis using Jackknife. Jackknife delete-one and delete-d algorithm was used to provide estimates of logistic regression coefficient. The Jackknife standard deviation provides an estimate of variability of the standard deviation of sample and it is a good measure of precision. The method was illustrated with real life data; and the results obtained from the Jackknife samples was compared with the result from ordinary logistic regression using the maximum likelihood method and results obtained reveals that the values from the jackknife algorithm for the parameter estimation, standard deviation and confidence interval were so close to the result from ordinary logistic regression analysis, this provides a good approximation to the result which shows that there is no bias in the jackknife coefficients.

Keywords: Jackknife algorithm, Logistic regression, dichotomous variable, maximum likelihood