Page 2 of 2

Posted: Mon Jul 29, 2013 3:48 pm
by Katja G
Dear Christian,

Thanks for your reply! It worked very well. Did you mean A Primer on Partial Least Squares Structural Equation Modeling (PLS-SEM) when you talked about Hair et al. (2013)?

Greetings,

Katja

Posted: Mon Jul 29, 2013 6:43 pm
by nayabeyes
Katja G wrote:Dear Christian,

Thanks for your reply! It worked very well. Did you mean A Primer on Partial Least Squares Structural Equation Modeling (PLS-SEM) when you talked about Hair et al. (2013)?

Greetings,

Katja
yes it is

Negative path coefficient for indirect effect and its impact on the total effect

Posted: Fri May 10, 2019 7:41 am
by yokemay
Dear all,

The product of the two effects (direct and indirect) obtained from the Path coefficients (Mean, STDEV, T-Values) of the bootstrap results is negative, i.e., -0.2145. Meanwhile, the path coefficient of the direct effect is 0.3058. As a result, the total effect is 0.3058 - 0.2145 = -0.189.

I don't know how to interpret this result. Please advise. Thank you very much.

Best,
May

Re: Calculating T-Value for indirect effects: a*b/sd(ai*bi) ?

Posted: Wed May 15, 2019 1:29 pm
by jmbecker
That sounds like a typical example of competetive mediation: https://www.smartpls.com/documentation/ ... /mediation

There could be many reasons for such an effect. Most of them a subject specific theoretical reasons. So you need to consider based on your domain knowledge whether it makes sense from a substantive point of view.