Have an idea?

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

How to interpret the "Segment Membership" output from Latent Class analysis


I performed a Latent Class analysis in order to define a number of segments my respondents can be allocated to.

Now I decided to use three segment and want to enrich this data by demographic and sociodemographic data I also collected in the survey.
As I already read in other posts, this is not possible with Lighthouse. Therefore, I want to export the data to SPSS.

However, I do not really understand how to interpret the output "Segment Membership" for the Latent Class analysis. The columns say for instance "Prob_Membership Group 3/5", "Prob_Membership Group 4/5". But as I only want to include three segments in my final interpretation, how can I get a percentage saying e.g. that respondent X  has a probability of X% to be allocated to one of my three segments?

Thank you for your support!
asked Dec 13, 2019 by Stefanie

1 Answer

0 votes
If you included a solution for 3 groups in your estimation settings, then look for variables called "Prob_Membership Group 1/3," "Prob_Membership Group 3/3,"  and "Prob_Membership Group 3/3."   These contain your probabilities of membership in the three group solution.  The variables you mention are for the 5 group solution.

You may also want to look at the variable "3 Group membership" as it contains the modal segment for each respondent, i.e. the one segment for which the respondent has the highest probability of being a member.
answered Dec 13, 2019 by Keith Chrzan Platinum Sawtooth Software, Inc. (102,900 points)
Thank you! This helped a lot.