I am comparing ACA and ACBC methods and it is not clear for me what calculations are made in each of them to select the atribute levels that pass to the next levels. I have read in ACA technical paper that utilities are estimated on the fly ( “every time the respondent completes a pairs question, the estimate of the respondent’s utilities is updated. Upadating the utilities improves the quality of subsequent pairs questions”). Does this mean that utilities are being calculated during the whole survey and that the attribute levels with lower utilities are the ones discarded in subsequent phases?
Regarding ACBC, from what I understood, there are no utilities being estimated in early phases of the survey, correct?
I am sorry if my question sounds really basic but it would be very useful for me to understand if the adaptive component of these two methods works in a different way or if they have a common ground.
P.S. - I am estimating final utilities using HB.
Thank you very much in advance for your help.