Have an idea?

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

How I Can Estimate Willingness To Pay for CBC Partwort-Utility?

Hello, I am new on CBC...
And my question pretty simple (I think)
For example, I have 2 attributes: "Label" and "Price" and here are the levels:

Label: A, B,C,D
Price: $3, $7, 10

After gathering data and run a logit analysis, I got the effect (Part-wort utility) result here:

Label A: 2500
Label B: 2000
Label C:   500
Label D: -100

Price $3: 2000
Price $7:   700
Price $10: -200

From this data, how I estimate the WTP?? need help, please...
asked Jul 1, 2020 by yustian (280 points)

1 Answer

0 votes
You can check out this white paper:  https://sawtoothsoftware.com/support/technical-papers/general-conjoint-analysis/assessing-the-monetary-value-of-attribute-levels-with-conjoint-analysis-warnings-and-suggestions-2001.

Though there are common algebraic methods, those often over-estimate WTP, so we recommend the simulation-based method instead.
answered Jul 1, 2020 by Keith Chrzan Platinum Sawtooth Software, Inc. (111,275 points)
I've granted with SSI academic award, so I don't have access to market simulator. I think using common algebraic is enough for my paper.

But I got new problem, I've calculating WTP for 4 groups, but in the first group (largest size) I got very high WTP. According to the mentary.pdf paper from SSI, I calculate them with mean,, but the result still too high..

What can I do with this?
Unfortunately, this is the all -too-common result of using the algebraic method, which is why we advise against it.  If you must go the algebraic route, you may be better off with the median than the mean.