Re. common method bias (Harman's single factor' value is high)

Frequently asked questions about PLS path modeling.
Post Reply
vivian1020
PLS Junior User
Posts: 5
Joined: Thu Jan 16, 2020 1:42 pm
Real name and title: Ms. Shen Yan

Re. common method bias (Harman's single factor' value is high)

Post by vivian1020 » Thu Jan 16, 2020 1:48 pm

Dear Sir
hello! i meet a problem when doing SmartPLS.
For the common method bias, i use Harman's single factor and find the largest factor accounted for 53.982% of the variance. it is just over the 50%. However, in the Pearson Correlation, no case above 0.9. the inner VIF is over 3.3, but is below 5. i find an article says like that:

"Common Bias Method In addition, the presence of a common bias method was evaluated using a consistent PLS algorithm, with a factor weighting scheme and 300 maximum iterations. The results indicated that the VIF values of all constructs were below 3.3, indicating that the structural model is free of common method bias (Kock, 2015)."

so my inner VIF is below 5, can i use this reason?
so can i stii use these data? or what i can do? Thanks a lot!!!

jmbecker
SmartPLS Developer
Posts: 1129
Joined: Tue Mar 28, 2006 11:09 am
Real name and title: Dr. Jan-Michael Becker

Re: Re. common method bias (Harman's single factor' value is high)

Post by jmbecker » Wed Jan 22, 2020 2:57 pm

If you have a problem with common method bias there is not much you can do, but collect new (better) data.

Harman's single factor test is usually very insensitive to common method bias. It has been critizied in the literature for this problem. Thus, if it detects a bias there is likely a problem in your data.

I'm not very familiar with the Kock (2015) reference, but the VIF threshold likely comes from there. So you want to have a look at that reference and see what they suggest.
I am personally not sure whether this is a very useful test, but if it helps you make an argument in favor of your data, you might go that way. However, as a reviewer, I would be very skeptical.
Dr. Jan-Michael Becker, University of Cologne, SmartPLS Developer
Researchgate: https://www.researchgate.net/profile/Jan_Michael_Becker
GoogleScholar: http://scholar.google.de/citations?user ... AAAJ&hl=de

Post Reply