Search found 1284 matches

by jmbecker
Wed Aug 05, 2015 7:59 am
Forum: Method and application
Topic: MGA Result Discrepancies
Replies: 12
Views: 13464

Re: MGA Result Discrepancies

1) You should also be careful, that the above 0.95 and below 0.05 rule only applies for the "PLS-MGA" results, but not for all results that you get from a Multi-Group-Analysis (MGA). For all other results that are provided by the MGA procedure (e.g., Bootstrapping Results, Parametric Test,...
by jmbecker
Mon Aug 03, 2015 8:54 am
Forum: Method and application
Topic: Composite relaibility
Replies: 9
Views: 16502

Re: Composite relaibility

Why should a high composite reliability be a problem? Usually you want a high composite reliability.
by jmbecker
Mon Aug 03, 2015 8:50 am
Forum: Method and application
Topic: Individual Co-Effecient of Determination Values R2
Replies: 9
Views: 8340

Re: Individual Co-Effecient of Determination Values R2

We provide the f² statistic, i.e., the effect size in terms of explained variance. It calculates as (R²inlcuded - R²excluded) / (1 - R²included) Whereas R²exluded means the R² of the dependent variable, when the focal variable is excluded from the model. This is the more common statistic in this con...
by jmbecker
Wed Jul 29, 2015 8:04 pm
Forum: Method and application
Topic: Mediator - multiple IV
Replies: 4
Views: 5299

Re: Mediator - multiple IV

Yes, you should not use Sobel, but only the T and P values from the bootstrapping for the indirect and direct effects. SmartPLS2, however, does not report the results for indirect effects, but only for direct and total effects. Thus, you have to use either SmartPLS3, which gives you all necessary re...
by jmbecker
Wed Jul 29, 2015 12:10 pm
Forum: Method and application
Topic: Mediation: negative or positive coefficients?
Replies: 5
Views: 8517

Re: Mediation: negative or positive coefficients?

I would also think that an absolute value is more what you want to model for the mediator ("Asymmetry in Fairness Perception"), because any deviation from 0 indicates an asymmetry (both to the positive and negative side). Maybe there is no effect of fairness asymmetry? You may also think a...
by jmbecker
Wed Jul 29, 2015 7:13 am
Forum: Method and application
Topic: Bootstrapping - seeing the bootstrap estimates
Replies: 4
Views: 3845

Re: Bootstrapping - seeing the bootstrap estimates

If you have the bootstrapping result and you click an Path Coefficients, for example, you have 4 different results tabs:
  • Mean, STDEV, T-Values, P-Values
  • Confidence Intervals
  • Confidence Intervals Bias Corrected
  • Samples
by jmbecker
Tue Jul 28, 2015 2:19 pm
Forum: Method and application
Topic: Mediator - multiple IV
Replies: 4
Views: 5299

Re: Mediator - multiple IV

You need to assess the indirect effects for significance. You do not need to delete your mediator. If only your indirect effect is significant: full mediation. If both your indirect and direct effect are significant: partial mediation. If only your direct effect is significant and not the indirect e...
by jmbecker
Tue Jul 28, 2015 2:04 pm
Forum: Method and application
Topic: 3-Way Interaction of Moderation
Replies: 15
Views: 25853

Re: 3-Way Interaction of Moderation

You do not want to standardize any of the product terms. So it would be just: Standardize(A) x Standardize(B) x Standardize(C) In SmartPLS3 we make sure that the product term LV is not standardized after multiplication. This, however, will only work for two-way interactions. If you do your caclulati...
by jmbecker
Tue Jul 28, 2015 1:56 pm
Forum: Method and application
Topic: Bootstrapping - seeing the bootstrap estimates
Replies: 4
Views: 3845

Re: Bootstrapping - seeing the bootstrap estimates

You will always find all estimates (path coeffcients, weights, etc.) for all n estimations in the "Samples" Dialog.
by jmbecker
Tue Jul 28, 2015 1:50 pm
Forum: Method and application
Topic: Multigroup analysis - group sizes
Replies: 1
Views: 2459

Re: Multigroup analysis - group sizes

They need not to have the same size, but it would be better if sample sizes are not very disimilar. There might be problems that results from unequal sample sizes across groups and confound your results. You find some further explanations in the book on PLS-SEM Hair, J.F., Hult, G.T.M., Ringle, C.M....
by jmbecker
Tue Jul 28, 2015 1:41 pm
Forum: SmartPLS 3 - Requested Features
Topic: PLSc loading in reflective model higher than 1
Replies: 3
Views: 9495

Re: PLSc loading in reflective model higher than 1

PLSc makes the assumption about appropriate factor models (reflective models). It takes these assumptions very seriously. It always produces strange results if the model is bad or assumptions are not met. Unlike CB-SEM, which does not give you results or bad fit, it will provide results, but these m...
by jmbecker
Fri Jul 10, 2015 6:12 am
Forum: Method and application
Topic: moderating effect (smartpls 3)
Replies: 3
Views: 3622

Re: moderating effect (smartpls 3)

No. A moderating effect in PLS has a normal p-value from bootstrapping. Your effect is only significant if p is low (e.g. <0.05), but not if it is high.
by jmbecker
Thu Jul 02, 2015 7:21 am
Forum: Method and application
Topic: IPMA negative importance
Replies: 6
Views: 6614

Re: IPMA negative importance

Generally, negative importance makes sense, because the importance is based on the measurement and structural weights. If any of them is negative, you may have negative importance and the predictor is negatively related to your outcome. (You have to judge yourself if this makes sense in your specifi...
by jmbecker
Fri Jun 26, 2015 9:25 am
Forum: Method and application
Topic: MGA results vs Groups bootstrapping
Replies: 3
Views: 3567

Re: MGA results vs Groups bootstrapping

Just put the first sentence into google and you will find references.
by jmbecker
Fri Jun 26, 2015 8:00 am
Forum: Method and application
Topic: Negative path expected -- but positive
Replies: 1
Views: 2561

Re: Negative path expected -- but positive

Do you observe negative weights? It could happen, that if weights are negative the construct is in fact reverse coded compared to the indicators. Hence, a positive relationship between constructs might occur. Setting the initial weights to be negative might circumvent this problem. It is not cheatin...