Hi,
I've conducted a CBC with best/worst answers and I'm trying to figure out how the utility estimation works in this case. I've found out that the "best" answers are coded exactly as in standard CBC and the "worst" answers are coded with -1, but I still don't understand how the utility estimation itself works in this case or how it differs from traditional CBC.
Is there a paper or something that explains this issue?
Help would be greatly appreciated!
Thank you,
Sophia