Have an idea?

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

Importance of scaling using HB (instead of Maximum Simulated Likelihood)

Hello everyone,

I do not have a Sawtooth specific question, but rather a general question on model estimation using HB: I would like to know how important the scaling of variables is in model estimation using HB? More precisely, for model estimation using Maximum Simulated Likelihood and optimization, it is recommended to scale the variables so that the parameter estimators are in similar orders of magnitude. For example, instead of using a linear price attribute with the values 50$, 100$, 150$, it is recommended to use 1, 2, 3 and to rescale the coefficients after estimation. However, it did not find any information about the importance of scaling when using HB. I suspect that scaling is not as important with HB as it is with optimisation.

Do you have any information on that topic?
asked Mar 16 by Nico Bronze (800 points)

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.