Have an idea?

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

Combination of attribute importance weights

Dear all,

I have conducted a CBC HB analysis. I explored attributes including price, some technical attributes and some digital services.

My results indicate that price is most important followed by the technical attributes followed by the digital services. This is plausible.

However, I was wondering if I could interprete the following: When combining the digital service attributes, their combined importance is higher than a sole technical attribute`s importance.

This is my approach: I built the sum of importance weights of the digital service attributes at the respondent level and then averaged it across all respondents. This new combined average importance weight is higher than the one of a technical attribute.

Is this a valid approach?

My motivation is that I would like to conclude „don‘t focus on technical attributes and price only“.

Thank you so much for your reply and best regards
asked Nov 7, 2020 by Ida27 (295 points)

1 Answer

0 votes
Importances are additive, so yes, but you need to interpret the results carefully.  If there's heterogeneity in the DIRECTION of utilities for any of the attributes, your importance results may be at odds with any simulation results you run.  You're better off to make this kind of interpretation from simulations.
answered Nov 7, 2020 by Keith Chrzan Platinum Sawtooth Software, Inc. (102,700 points)