Have an idea?

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

Is there any way to write chip allocation only for one attribute (such as brands)?

Hello all,

Is there any way in order to write chip allocation question only for one attribute (such as brands) and by allocating 11 chips for each 2 brands (paired comparison for each combination) properly via SSI script?

Thanks in advance
asked Sep 25, 2017 by seslem (140 points)
retagged Sep 25, 2017 by Walter Williams

1 Answer

0 votes
You can "trick" CBC into doing this.  CBC requires two attributes, but you can add a "blank" attribute that just paints blank spaces on the screen.  Then, during estimation, make sure to set the second blank attribute to "excluded" so that CBC/HB utility estimation routine only considers the brand attribute to evaluate the chip allocation.
answered Sep 25, 2017 by Bryan Orme Platinum Sawtooth Software, Inc. (152,955 points)
Hi Bryan..

Thanks for response..but I actually tried to do that but I am failed. I also made sure  to use "complete enumeration" or "shortcut" design methods when I did. I have only one attribute which is brands and also use each of them as a level.

But when I add a second dummy attribute I also have to add least two levels for the second attibute. How can I avoid the second attribute combinations when I run it. I just want to see C(6,2) on the screen In other words choosing 2 brands out of 6. Is that possible?
Yes, you add a second attribute with at least two levels.  The second attribute just paints blanks on the screen.  You can use blank graphics for the text of those levels, which just paint 1 pixel by 1 pixel blank graphics.  Or, you can use the HTML code for a blank space, which is &nbsp.   

If you use Complete Enumeration or Shortcut, this forces minimal overlap (no repeating of the same brand within the same question).  So, when the CBC question is shown on the screen, only the brand appears to the respondents.  The blank graphic or the blank spaces are shown in the second attribute position, but respondents of course do not see that second attribute and respondents only think they are being asked to allocate chips between the two brands you are showing on the screen.

Then, during analysis you use the "exclude" setting for the second attribute, so that only the first attribute is analyzed.