Have an idea?

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

ACBC Choice Tournament - Determining winning attributes

We have respondents going through an ACBC. At the end of the ACBC we show them the design winner from the Choice Tournament using the code below and ask them if they like the idea. This works well.

Question: How can I find the attribute levels the respondent saw (or the choice tournament winning levels)?  

In the data file I see variables titled: sys_ACBC_ACBC_dealWinnerXXX

However, the values in that variable don't seem to match either what is being shown on the screen, or the lists.

For example I initially assumed that "sys_ACBC_ACBC_dealWinner1" was the value selected for attribute 1. However, in the data file the variable has value of 13 and there are only 4 levels.  So doesn't match.

Second example "sys_ACBC_ACBC_dealWinner3" has a value of 2, but what is displayed on the screen is level 1 from that attribute.

Finally, there are five attributes in the conjoint, but no data in the data file for "sys_ACBC_ACBC_dealWinner4" nor "sys_ACBC_ACBC_dealWinner5".

<table border="0" cellpadding="0" cellspacing="0" width="100%">
  <td> <b>[% ListLabel(attributesList,1) %]</b>: </td>
  <td> [% ACBCWinnerLabel(ACBC, 1) %] </td>
  <td> <b>[% ListLabel(attributesList,2) %]</b>: </td>
  <td> [% ACBCWinnerLabel(ACBC, 2) %] </td>
  <td> <b>[% ListLabel(attributesList,3) %]</b>: </td>
  <td> [% ACBCWinnerLabel(ACBC, 3) %] </td>
  <td> <b>[% ListLabel(attributesList,4) %]</b>: </td>
  <td> [% ACBCWinnerLabel(ACBC, 4) %] </td>
  <td> <b>[% ListLabel(attributesList,5) %]</b>: </td>
  <td> [% ACBCWinnerLabel(ACBC, 5) %] </td>
asked Dec 19, 2022 by Nwiggin Bronze (1,930 points)

1 Answer

0 votes
I’m sure you resolved this a while ago, but just in case there were still questions: the easiest way to find each individual respondent’s Tournament Winner concept is to run a Counts analysis in the Analysis Manager. The Individual Results tab shows the Winning levels for each respondent in the final set of columns. I don't recommend trying to pull any of that selection information directly from the raw response data.
answered Jan 12 by Cameron Halversen Bronze Sawtooth Software, Inc. (625 points)