Have an idea?

Visit Sawtooth Software Feedback to share your ideas on how we can improve our products.

In ACBC Utility Estimation, How is the BYO Task Coded Into Multiple Lines of Individual Data?

In running an ACBC, I am choosing to include the BYO task in the data used to estimate the utilities.  According to the Sawtooth documentation, there is a separate line of data created for as many attributes as there are (excluding price).  So, for example, if there are ten attributes with one of those being price, my understanding is that this would gets coded into 9 rows of data, each of which is the equivalent of one of the rows of data for a configuration that was selected in the tournament.  In this case, does this mean that there would be nine identical rows of data, each with the BYO configuration and each treated as if they had been chosen in a tournament?
asked Jul 11, 2018 by Bruce

1 Answer

0 votes
For each BYO question, we create a choice task with as many alternatives (concepts) as there are levels within the attribute in question.  

So, if a respondent sees 4 brands in the BYO question involving brand and chooses the 2nd brand as best, then this means four rows (alternatives) are added to the utility estimation matrix (one task with four rows).  The task communicates that four brands were shown, other attributes were missing (unless Summed Price is used, in which case Price would also be coded for the brands), and the respondent picked the 2nd brand.

So, if there are 8 attributes each with 4 levels included in the BYO, then this would add 8 x 4 = 32 rows of data to the utility estimation matrix.
answered Jul 11, 2018 by Bryan Orme Platinum Sawtooth Software, Inc. (176,515 points)