I have a logistic regression model where failure observations are binary.y=cbind(failure, 1-failure). x= temperature. I have 24 observations in total, i.e. length(failure)=24. I wish to test if the coefficient of temperature is 0.

Model: glm(y~ temperature, family= binomial). From the anova output the change in deviance is 5.9441, which is the F statistic. I wonder if the distribution for the F test is F(1,1) or F(1,24*2-23)? I am not sure if the 2nd parameter n-g is 24-23 or $24*2-23$.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.