Have an idea?

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

How to Compute RLH Values for LC?


I am wondering how to compute the RLH values for each respondent for a latent class analysis since the output of this method does not show them.
It would be great if someone could provide me a formula that I can use, because I haven't found a way to get these values yet.

Thank you so much!

Best regards!
asked Jul 21, 2017 by briniminii (490 points)

1 Answer

0 votes
For this you would need respondent-level utilities.  LC provides pseudo-respondent level utilities but I think it might be a stretch to compute RLH from them.  If you wanted to you could use the pseudo-respondent level utilities to simulate all of a respondent's choices, then identify the likelihood of each choice, the multiply those likelihoods together and take their nth root, where n is the number of choices a respondent makes.  

RLH is obviously NOT a common measure of model fit for LC, so I'm curious why you think you need it.
answered Jul 22, 2017 by Keith Chrzan Platinum Sawtooth Software, Inc. (85,075 points)