Have an idea?

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

Analysing ACBC data

I am very new with the software, so please excuse me for asking basic questions, if they are. I have been trying to test my survey before distributing it and I encounter a problem when I try to use the CBC/HB software.

My survey consist of a fictional BYO, (BYO just acts as a search interface in which the respondent chooses a few features of a product, it is not related to the attributes that I included for choice tasks), an ACBC question with 3 interrelated attributes that are shown to respondent by using Conditional Display and a number of Grid questions (Likert scale) that I want to include in my analysis as covariates, together with demographic questions (Select questions). I am testing the survey before distributing and when I try to incorporate it to the CBC/HB software it gives an error indicating that there is a fault on the "Heading''. I want to see the effect of these covariates on the attributes together with the main and interaction effects of the attributes in ACBC. Additionally, I want to calculate willingness-to-pay based on these and planning to further analyse the data on CCEA for possible segmentation.
Therefore, I wanted to ask if it is possible to use the ACBC data on CBC/HB software and is there anything I should do to avoid future problems on survey before distributing it? and also which approaches I can use in Sawtooth Software SSI or CBC/HB to be able to calculate willingness-to-pay together with main and interaction effects of both attributes and covariates.

Thank you very much in advance for your help, once more!

asked Dec 5, 2014 by anonymous

1 Answer

0 votes
You will find it easier to use the HB routine built right within the ACBC software (SSI Web) to conduct such analysis with covariates instead of exporting the ACBC data to a .CHO file and conducting the analysis in the separate CBC/HB standalone software.

The HB routine built right into the SSI Web interface allows you to implement covariates (but, be careful that using more than just a few covariates can often cause the time to compute to become multiple weeks, months, or years).

Also, please be aware that Willingness to Pay calculations are an extremely difficult topic to do well.  It is fraught with pitfalls and difficulties and we generally recommend avoiding trying to compute Willingness to Pay.
answered Dec 5, 2014 by Bryan Orme Platinum Sawtooth Software, Inc. (132,290 points)
WTP calculation ?