Hi - I am working on a complex model, and I wish to reduce the number of paths. Some of the paths are strong and some are weak, - it would make some kind of sense to not include some paths.
The question is, - how strong or weak can a path be for inclusion? 0,1? 0,2? Is there some kind of best practice, or reference to cite on this issue?
I realize there may be differences in between fields, - but in marketing - what is a too weak path to include? What is a path strong enough to include?
Very Best Regards!
:-)
Smallest Path coefficient to report?
Re: Smallest Path coefficient to report?
Hi,Mattias_O wrote:Hi - I am working on a complex model, and I wish to reduce the number of paths. Some of the paths are strong and some are weak, - it would make some kind of sense to not include some paths.
The question is, - how strong or weak can a path be for inclusion? 0,1? 0,2? Is there some kind of best practice, or reference to cite on this issue?
I realize there may be differences in between fields, - but in marketing - what is a too weak path to include? What is a path strong enough to include?
Very Best Regards!
:-)
I think if the path is significant by checking its t-value then no problem. For more details check Hair et al., 2013
MURAD ALI, Ph.D