PROSES INFERENSI PADA MODEL LOGIT

Rusgiyono, Agus

Media Statistika
Vol 1, No 2 (2008): Media Statistika

Let represent the response on a nominal random variable of Bernoulli distribution, with    ,  where is a parameter with unknown value. Problems of estimating used smallest square methods in  linier regression model  can overcome with used maximum likelihood method in  logistic regression.. Suppose  is maksimum likelihood estimstors of . In case can be obtained  from first condition, ln(Ln(p)) to be maximum  at point then be obtained and that is unbiased estimator because To be test hipothesis that , with a large sample size used fact that    Keywords : Estimator, unbiased estimator, test statistic