Page 1 of 1

Significance of Path Coeffeceint

Posted: Thu May 26, 2011 8:56 pm
by k_ganju

I have formulated a simple model with 10 latent variables and one dependent variable and have the path coeffecients. I was wondering if it is correct to divide the path coeffecients by the AVE to get the t-statistic to see if there is a significant relationship between the variables? Or is there another method of determinign the t-statistic?

Also, is there an option to quickly search through the forums?

Kartik Ganju

Posted: Fri May 27, 2011 7:58 am
by dibakar
You need to do a boot strapping to get the T values. For boot strapping select Cases as number of observations made in your study and set sample size to 1000. Hope it helps.


Posted: Mon Jun 13, 2011 1:17 pm
by sunilsahadev
I conducted bootstrapping analysis and got the t-values. Do we have to look at the significance level by going through the t-distribution tables. If that is so then what is the degree of freedom. Also is there any option in smart PLS to get the significance level of t-values in the output itself.

Posted: Mon Jun 13, 2011 1:19 pm
by christian.nitzl
Pleas check following post for an answer:


Best regards,


Significance level

Posted: Mon Jun 13, 2011 1:38 pm
by sunilsahadev
I read this paper. It says that the degree of freedom of t-value is effectiveley n-1 where n is the number of bootstrapping groups. So if i conduct a bootstrapping with 500 iterations then is the degree of freedom for the t-statistic 499. Please correct me if I am wrong.

Posted: Mon Jun 13, 2011 3:03 pm
by christian.nitzl
You absolutly right! :-)