Group comparison of "ranks" of path coefficients possible?

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
beautifulgarbage
PLS Junior User
Posts: 7
Joined: Fri Oct 10, 2014 2:42 pm
Real name and title: BeautifulGarbage

Group comparison of "ranks" of path coefficients possible?

Post by beautifulgarbage »

Dear PLS-community,

part of my PLS model contains four latent constructs (dimension 1, dimension 2 etc.), each having a path to another latent variable (my "overall judgment" construct). I am conducting a multigroup anaylisis and find significant differences between my groups regarding some of these four path coefficients. However, what I would like to show further is that the relative importance of the single dimensions for the formation of the "overall judgment" differs dramatically between the groups. To be more precise: within every group I can compare the path coefficients and conclude that e.g. "dimension 1 is the most important dimension for the formation of the overall judgment in group 1, but the least important one in group 2".

Is there any statistical test which I can use to compare the ranking of the dimensions between the groups in order to substantiate this finding?

Thank you very much in advance for your help!

Best regards
Post Reply