Aldri har så mange begavede mennesker oppnådd så lite
Av Ronny Klæboe Dato: 14.5.2009
Fra SEMNET—er det egentlig bortkasta tid å drive med strukturmodeller??
>>> Paul Barrett 05/13/09 7:30 AM >>>
I still find it one of the biggest mysteries in the world today - why so many clever and talented people continue working so slavishly with an analysis technique which yields results of so little value, practically or theoretically. In other sciences, investigators soon move on to more productive methods and thinking when a method appears to yield so little of benefit; in the social sciences, the opposite seems to be the order of the day. C’est la vie!
On Thu, May 14, 2009 at 9:43 AM, Ed Rigdon wrote:
Paul’s “why” is a very fair question. We all know that some applications of SEM are mimicry—doing what you have seeen others do. Mimicry explains why PLS path modeling is used so heavily in one field—information systems—but sparingly anywhere else. Add to those cases the applications which are due to an influential third party—dissertation advisor, journal editor, etc.—who commands the choice. Take away the ritualistic applications, and what percent of applications are due to a conscious, independent determination that SEM is the best method? Within that subset, what was the “why” that drove
the choice to use SEM? Did those researchers get what they wanted—did SEM deliver in those cases?
It’s a fair question. No boilerplate, please.