Have an idea?

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

attribute importance interpretation - combine average importances and compare?


I have six attributes in my CBC study and the importances (HB estimation) are:
A: 36 %
B: 21 %
C: 16 %
D: 13 %
E: 8 %
F: 6 %

Do I have to conclude that D-F are rather not very important to consumers? Or could I also combine say D & E and say something like D & E together are as important as attribute B?

asked Sep 16 by anonymous

1 Answer

0 votes
Importances are ratio level data, so they can be added and compared.  So . . .

. . . you could say that D-F are not as important as A or B, for example, but they are relative to A and B and are not absolute measures (e.g. they might be extremely important, but still less important than A and B).

. . . You could add D+E+F together (at the respondent level) and then take a mean of that sum and see how their mean compares to the means of A and B you show above.
answered Sep 16 by Keith Chrzan Platinum Sawtooth Software, Inc. (95,475 points)
Thank you! Very helpful!