Hi, I am trying to compare the predictive ability of several Cox proportional hazards regression survival models. Is there a way to calculate Akaike information criterion (AIC) and BIC using SPSS 28 for Cox regression models?

From a statistician:

"I think we report the -2*log likelihoods and the tests for different blocks. They can be used to do model comparison. However, I don't think we directly report AIC or BIC, although they could be converted by using the log likelihoods."

Something like this:

AIC: −2 × log-likelihood + 2 × number of parameters
BIC: −2 × log-likelihood + log(Number of observations) × number of parameters