Have an idea?

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

How to analyze Discover Conjoint data in Lighthouse

Hi Discover Team

I downloaded some dummy CBC data from Discover to try to import and analyze it in Lighthouse (it's my first time considering to use it for a low budget project). My goal is to use the simulation capabilities incl sensitivity analyses therein.

The exported data format reminds me of ACBC as the design is apparently created on the fly. Not sure if and how it's feasible to feed this into Lighthouse...

Can you advise what is possible and what I need to consider?

asked May 14, 2018 by alex.wendland Bronze (2,485 points)
edited May 14, 2018 by alex.wendland

1 Answer

+1 vote
Best answer

To leverage the ease of data collection in Discover-CBC while taking advantage of the more capable Choice Simulator (that includes sensitivity analysis, etc.) then I'd take the following path:

1.  Use Discover-CBC's ability to export a .CSV file containing the design and respondent choices to submit to the standalone CBC/HB estimation tool.  Within Discover-CBC, select "More + Advanced Export + Export CBC Design & Choices".

2.  Use our CBC/HB standalone software to open that .CSV and estimate part-worth utilities.  This produces a .HBU file that contains individual-level utilities.

3.  Use the standalone Choice Simulator (the same market simulator as integrated within Lighthouse Studio) to import the .HBU file and create a Choice Simulator project.  You cannot import a .HBU directly into the Choice Simulator integrated within Lighthouse Studio.

Note: both standalone CBC/HB software and standalone Choice Simulator software may be downloaded from our website and they "look to" your Lighthouse Studio subscription licenses to verify you as a user.
answered May 14, 2018 by Bryan Orme Platinum Sawtooth Software, Inc. (187,915 points)
selected May 14, 2018 by alex.wendland
Perfect! Hadn't considered the standalone tools but will use these instead of Lighthouse.
Thank you!