Have an idea?

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

Inconsistency between Standalone and Lighthouse


I am running an HB analysis for my CBC data that I originally collected then exported using the Lighthouse studio to dual .csv files.

When I use the HB analysis in the Lighthouse studio (LH), I notice that the number of choices in each response category is different from the standalone program (SA) although the expanded takes totals are the same (I have only 2 concepts/task)

LH: concept 1 52.15% vs concept 2 47.85%; 3576 expanded tasks in total & 8.0 on average
SA: concept 1 45.78% vs concept 2 54.22%; 3576 expanded tasks in total & 8.0 on average

Note that I have a dual response none in the task, but I am excluding it from this analysis. I also have 2 fixed tasks that I excluded from the analysis.

What might be the possible reason for this discrimination? & which analysis I can rely on?

By the way, When I went into the Excel file for the responses & manually calculated the percentage of responses for choices of concept 1 overall respondents gave me the number consistent with the LH results. Also, when I tried the .cho file for the SA program, it gave me the same percentages as the LH (52.15% vs 47.85%; 3576 expanded tasks in total & 8.0 on average).

I wanted to use the .csv file because, in the Latent Class Analysis, I am adding an interaction term since I am analyzing the data using both LCA & HB for comparison requirements for my Project.

Then If I used the SA results, Is there a program to analyze the results to get comparisons between subgroups like the MaxDiff analyzer for the MaxDiff results? Can I import the SA results into the LH for this analysis?

asked May 18 by AMYN Bronze (2,980 points)
edited May 18 by AMYN
What I did over the past couple of days is that I manually modified the .cho file with the interactions I and included them as new variables in each concept. It was a very time-consuming effort but I believe this way I will have consistent results over both applications.

I hope to know what was the source of such inconsistency if possible.

Your solution to the original question

Please only use this to answer the original question. Otherwise please use comments.
Your name to display (optional):
Privacy: Your email address will only be used for sending these notifications.
Anti-spam verification:

To avoid this verification in future, please log in or register.