Good evening everyone,
I have been working on a structural model with moderating variables and everything has turned out quite fine so far. I am now thinking about how to best report my results and had a look at several published PLS studies that report interaction effects in order to identify a suitable benchmark. Of course Chin's seminal paper on modelling interaction effects already contains some hints (http://disc-nt.cba.uh.edu/chin/icis96.pdf) and I found others that look quite decently, so there is no real problem.
However, what is quite striking is that compared to papers that use standard regression models to investigate interaction effects, non of the PLS papers had an exact visualization of the actual interaction effects under investigation, which seems to be absolutely essential when considering papers that use standard multiple regression models. It seems that there are a lot of excel-sheets on the web and elsewhere that easily allow to visualize interaction effects using the regression output, however, I was not able to figure out a way to transfer the SmartPLS output to such a template. So here are my questions:
- Does anyone have an idea how to visualize interaction effects in PLS models?
- Is there a way to utilize the PLS output so that visualization tools such as this one (http://www.victoria.ac.nz/psyc/staff/pa ... dgraph.php) might work?
I guess visualizing the interaction effects is tremendously important, since most referees at journals are familiar with multiple regressions and therefore most probably will expect such a visualization, especially if they are not familiar with SEM techniques such as PLS.
Thank you for any suggestion you might have, I really would appreciate your help regarding this issue.