On pg. 181 of Dobson, Intro to GLM, https://ocd.lcwu.edu.pk/cfiles/Statistics/EC/Stat-402/AnIntroductiontoGeneralizedLinearModelsbyAnnetteJ.DobsonAdrianGBarnettz-lib.org.pdf nominal logistic regression is described, quoting the text
logit(pij) = log(pij/pi1) = ....
But, I thought logit(pij) = log(pij/(1-pij)) note: I'm a novice
Question #1 - doesn't that seem a little peculiar?
Also, no likelihood or loglikelihood function in terms of pij/pi1 is given.
Question #2 - what is the likelihood function in terms of pij/pi1 for j = 1...J-1
Question #3 - loglikelihood, score functions, and info matrix?
Question #4 - is there are readable reference for this material?