How to judge whether a path is strong or weak?

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
lp527965873
PLS Junior User
Posts: 1
Joined: Thu Jun 14, 2012 8:38 am
Real name and title:

How to judge whether a path is strong or weak?

Post by lp527965873 »

Hi all,
I have a IV, a DV, and a mediator.
Running Bootstraping, I get the Path Coefficients and corresponding T-Statistics. The values for the Path Coefficients are: 0.372(Mediator to DV), -0.338 (IV to Mediator), and-0.425(IV to DV). T values are: 4.081, 3.476, and 4.690. I have two questions:
(1) How to judge whether these paths are strong? Are there any references? The T values tell that they are significant.
(2) Who have published a formal paper using the SmartPLS? I want to know the wording when using the SmartPLS in a paper.

Thank you. Best regards.
hxin009
PLS Senior User
Posts: 25
Joined: Mon Aug 13, 2012 4:23 am
Real name and title:

Post by hxin009 »

Hi,
for question 1:
here is a video about testing mediation effects
http://www.youtube.com/watch?v=fvk39T0p ... ature=plcp

and here is the link about sobel test
http://www.danielsoper.com/statcalc3/calc.aspx?id=31

For question 2:
i guess you could search in Google scholar. Just type any research topics + smartpls. i think that will give u a list of articles using smartpls for their data analysis.

Hope this answers your question. Happy data analyzing~~

Best regards,
Catherine
hxin009
PLS Senior User
Posts: 25
Joined: Mon Aug 13, 2012 4:23 am
Real name and title:

Post by hxin009 »

Hi there,
just in case you cant access to the video.
The video talks about:
step 1: run PLS-algorithm without mediator
step 2: run PLS-algorithm with mediator
step 3: run bootstapping with mediator and in the default report, under"bootstrapping"-->"path coeffients(Mean, STDEV,T-value)", find the value of "Original Value" for IV-->Mediator as A(as in the sobel test from the link:http://www.danielsoper.com/statcalc3/calc.aspx?id=31), the value of "Original value" for Mediator -->DV as B; also "standard error" for IV--> Mediator as seA and as well as Mediator -->DV as seB.
Step 4: you paste the value to the boxes respectively (http://www.danielsoper.com/statcalc3/calc.aspx?id=31)
Step 5: then it calculates the significance of mediation effects for you.

There is a simple test of mediation proposed by Baron and Kenny. Full mediation occurs when the IV no longer has a significant effect when the mediator is included.
Partial mediation occurs when the IV still has a significant effect but when its effect is diminshed.

Baron, R.B. and Kenny D.A. "The moderator- Mediator Variable distinction in social Psychological research: conceptual, strategic, and statistical considerations" Journal of Personality and Social Psychology (51:6) 1986, pp 1173-1182

Best regards,
Catherine
Post Reply