Have an idea?

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

Calculating importance scores for user-defined demographic groups

Hi Bryan,

I'm now being asked by my manager to create importance scores specific to subgroups which have different sub-populations of respondents than our latent class model.

I'm wondering if the best approach would be to take the individual respondent importances calculated using the HB model and then arithmetically average for the relevant sub-pops, or do a new importance model calculation from scratch using the respondent utilities in each sub-population.

Which is the most valid?

Thanks,
Alex
related to an answer for: Attribute Importances for Latent Classes
asked Sep 7 by alexrosenberg (180 points)

1 Answer

0 votes
Alex,

As you suggest, I would "take the individual respondent importances calculated using the HB model and then arithmetically average for the relevant sub-pops."
answered Sep 7 by Keith Chrzan Platinum Sawtooth Software, Inc. (107,050 points)
...