Variables formed by indicators of different scales

Frequently asked questions about PLS path modeling.
Post Reply
Francisco
PLS Junior User
Posts: 7
Joined: Fri Sep 15, 2017 8:03 am
Real name and title: Francisco Manuel

Variables formed by indicators of different scales

Post by Francisco » Wed Mar 14, 2018 9:02 am

Hello,
In my model I have variables, built by a series of indicators with different scales (numerical, categorical, ratios ...) Is the analysis procedure for the model the same as if we had all the indicators with the same scale?
Thank you.

jmbecker
SmartPLS Developer
Posts: 888
Joined: Tue Mar 28, 2006 11:09 am
Real name and title: Dr. Jan-Michael Becker

Re: Variables formed by indicators of different scales

Post by jmbecker » Wed Mar 14, 2018 2:18 pm

All data is usually standardized in the PLS algorithm. That alleviates some problems with different scales.

However:
There might be specific problems with different scales as you need at least quasi-metric data (i.e., scales that allow a metric interpretation such as a categorical Likert scale from 1-7).
If you have categorical scales that do not allow such interpretation you need to transform them first. For example, you could use a dummy coding scheme, etc.
Ratios might have specific distributional problems that in the extremes could also lead to some problems, but are usually quite robust in PLS.
Dr. Jan-Michael Becker, University of Cologne, SmartPLS Developer
Researchgate: https://www.researchgate.net/profile/Ja ... v=hdr_xprf
GoogleScholar: http://scholar.google.de/citations?user ... AAAJ&hl=de

Post Reply