Total Effects Analysis/Significance Level of Collect-Type HCM (Repeated Indicators Approach)

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
sgloeckner
PLS Junior User
Posts: 2
Joined: Tue Jan 22, 2019 8:30 am
Real name and title: Sebastian Glöckner

Total Effects Analysis/Significance Level of Collect-Type HCM (Repeated Indicators Approach)

Post by sgloeckner »

Hi everyone,

I wonder about the significance level of the effect between an antecedent latent variable and the HCM.
Following Hair et al. 2018 (https://goo.gl/b5oeuE; pp. 52-53) I have modeled the antecedent latent variable to have relationships with the HOC and all LOCs.
Figure 1 - Antecedent latent variable (Y5), HOC (Y4) and LOCs (Y1-Y3).png
Figure 1 - Antecedent latent variable (Y5), HOC (Y4) and LOCs (Y1-Y3).png (153.08 KiB) Viewed 12557 times
The path value I can calculate by building the sum of multiplying the path value of the antecedent latent variable and the the respective LOC and the path value of the respective LOC and the HOC plus once the direct effect of the antecedent latent variable on the HOC.

Is there also an equivalent formula to calculate the significance level of the total effect between the antecedent latent variable and the HOC and other typically reported parameters (e.g. STDEV)?

Could there also be a weighting regarding the T-values and STDEV?
Removing the LOCs and only using the HOC itself shows a significant relationship in my model. Ideally this could also be calculated in a sound way within the total model with the LOCs.

Thanks for your help.

Warm regards,
Sebastian
jmbecker
SmartPLS Developer
Posts: 1282
Joined: Tue Mar 28, 2006 11:09 am
Real name and title: Dr. Jan-Michael Becker

Re: Total Effects Analysis/Significance Level of Collect-Type HCM (Repeated Indicators Approach)

Post by jmbecker »

You are calculating the total effect of the antecedent to the HOC. You will find that also in the SmartPLS output and when doing bootstrapping you will also get the significance of the total effects.
Dr. Jan-Michael Becker, BI Norwegian Business School, SmartPLS Developer
Researchgate: https://www.researchgate.net/profile/Jan_Michael_Becker
GoogleScholar: http://scholar.google.de/citations?user ... AAAJ&hl=de
sgloeckner
PLS Junior User
Posts: 2
Joined: Tue Jan 22, 2019 8:30 am
Real name and title: Sebastian Glöckner

Re: Total Effects Analysis/Significance Level of Collect-Type HCM (Repeated Indicators Approach)

Post by sgloeckner »

My problem is that in this scenario p2, p3, p4 are highly signifikant (latent antecedent variable to LOCs), but p1 (latent antecedent variable to HOC)is not. If I remove the LOCs it is.

I would have seen this similar to the path strength, that I then can calculate by multiplying the paths.
I do not see how the bootstrapping algorithm is considering this scenario. Or am I totally off here?
jmbecker
SmartPLS Developer
Posts: 1282
Joined: Tue Mar 28, 2006 11:09 am
Real name and title: Dr. Jan-Michael Becker

Re: Total Effects Analysis/Significance Level of Collect-Type HCM (Repeated Indicators Approach)

Post by jmbecker »

Yes, the direct p1 effect (path coefficient) is usually not significant. That is because the LOC explain already all the variance in the HOC because they compose it. However, the total effect in SmartPLS is exactly, the indirect effect + direct effect so the effect p1_total that you describe.

You may also look at our simnal paper that suggest this approach: (Becker, J. M., Klein, K., & Wetzels, M. (2012). Hierarchical latent variable models in PLS-SEM: guidelines for using reflective-formative type models. Long Range Planning, 45(5-6), 359-394.)
Dr. Jan-Michael Becker, BI Norwegian Business School, SmartPLS Developer
Researchgate: https://www.researchgate.net/profile/Jan_Michael_Becker
GoogleScholar: http://scholar.google.de/citations?user ... AAAJ&hl=de
Post Reply