# Can I compare a negative AIC with a positive AIC?

The original model gives me an AIC of 75.43 and the Box-Cox transformed model gives me -189.2013. Does this me the Box-Cox transformed model is a much better model here?

• Did you include the geometric mean as described at en.wikipedia.org/wiki/Power_transform#Definition? Only then are the AIC-values compareable. – Jarle Tufto Dec 21 '16 at 15:06
• You need to use the complete formula in wikipedia, otherwise you're comparing the likelihood of a model specifying the distribution of mor with a model specifying the distribution of mor^0.1818. These are not compareable. Using the complete formula, you get comparaable likelihoods and you'll probably see a much smaller difference in log likelihoods and AICs. – Jarle Tufto Dec 21 '16 at 15:18
• Instead of mor^.18 you need to use (mor^.18-1)/(.18*exp(mean(log(mor)))^(.18-1)) as your response. – Jarle Tufto Dec 21 '16 at 15:52
• ? not sure what this last comment means ? BTW, negative vs positive is a red herring, e.g see emdbolker.wikidot.com/faq – Ben Bolker Dec 21 '16 at 16:19
• Some miracle happened...my AIC for all models are positive now!!!!!! – lusicat Dec 21 '16 at 19:17