Search found 1284 matches
- Thu Oct 15, 2015 9:15 am
- Forum: Method and application
- Topic: FIMIX algorithm not converging
- Replies: 3
- Views: 4687
Re: FIMIX algorithm not converging
You should not use those results from a run that did not converge correctly. The process sometimes does that and jumps between two local optima. Usually those results are conceptually inferior to other converged solutions.
- Wed Oct 14, 2015 10:07 am
- Forum: Method and application
- Topic: Index Values for Latent Variables
- Replies: 8
- Views: 98522
Re: Index Values for Latent Variables
I think we discussed this already somewhere else in the forum, but here are some hints: 1) Check Min. and Max. in the Ranges Dialog in the IPMA settings. Depending on which data set you use (Full Data(100) or Full Data) you will get wrong results when leaving the settings as they are. The Full Data(...
- Fri Oct 09, 2015 3:48 pm
- Forum: SmartPLS 3 - Requested Features
- Topic: When running FIMIX show if maximium iterations was hit
- Replies: 2
- Views: 6315
Re: When running FIMIX show if maximium iterations was hit
Thank you for the suggestion. It is a good idea to report some additional information about the other repetitions in FIMIX. I think we could easily report the lnL and the number of iterations of each intermediary FIMIX run in an additional table. Thereby, researchers can also investigate if the best...
- Fri Oct 09, 2015 9:06 am
- Forum: Method and application
- Topic: MICOM and reflective measurement models
- Replies: 2
- Views: 7139
Re: MICOM and reflective measurement models
In PLS path modeling all latent variables are represented as composites. In contrast to factor-based SEM (CBSEM/LISREL), where all latent variables are represented as common factors. If you have a reflective measurement model, you still have outer weights (see also the report output) that are calcul...
- Fri Oct 09, 2015 9:02 am
- Forum: Method and application
- Topic: How to check for measurement invariance?
- Replies: 12
- Views: 15317
Re: How to check for measurement invariance?
I would also add, that I think that it is questionable to test measurement invariance for interaction terms. They are not "real" latent variables, but only mathematical constructions to calculate the moderated regression analysis. Hence, I would advice only to check measurement invariance ...
- Wed Oct 07, 2015 3:22 pm
- Forum: Method and application
- Topic: How to check for measurement invariance?
- Replies: 12
- Views: 15317
Re: How to check for measurement invariance?
From the MICOM paper: "6. We test the null hypothesis that c equals one. If c is smaller than the 5%-quantile of the empirical distribution of cu, we must reject the hypothesis of compositional invariance, because the derivation from one is unlikely to stem from sampling variation." Hence,...
- Wed Oct 07, 2015 8:53 am
- Forum: Method and application
- Topic: Permutation test
- Replies: 3
- Views: 5174
Re: Permutation test
If you have a hypothesis that all effects of group A are larger/smaller than in group B, you can use one tailed test. If you don't have a specific hypothesis for the differences between your groups, use two tailed tests.
- Wed Oct 07, 2015 8:52 am
- Forum: SmartPLS 3 - Small Talk Corner
- Topic: work 2with original data
- Replies: 2
- Views: 6382
Re: work 2with original data
If you use the IPMA method, there are results for unstandardized data. I think that is what you are looking for.
- Tue Oct 06, 2015 1:17 pm
- Forum: Method and application
- Topic: How to check for measurement invariance?
- Replies: 12
- Views: 15317
Re: How to check for measurement invariance?
Please increase the decimal points. It seems that all your correlations are very close to 1. If there is no variation (all correlations are the same between permutations) the Permutation P-Value will be 0 (if you hide zero values, they are not shown). The zero value in this very special case might b...
- Tue Oct 06, 2015 1:00 pm
- Forum: Method and application
- Topic: Control Variable on SmartPLS 3.0
- Replies: 3
- Views: 6728
Re: Control Variable on SmartPLS 3.0
Yes, you are modeling it correctly. Just have an arrow from your control variable (e.g., age) to all constructs that you want age be controlled for.
- Tue Oct 06, 2015 12:57 pm
- Forum: Method and application
- Topic: Consistent PLS algorithm and outer loadings (form. cons.)
- Replies: 5
- Views: 7705
Re: Consistent PLS algorithm and outer loadings (form. cons.
This is an inherent problem of the consistent PLS algorithm. It sometimes produces inadmissable solutions.
Please read these related discussions:
viewtopic.php?f=5&t=3531
viewtopic.php?f=3&t=3766
Please read these related discussions:
viewtopic.php?f=5&t=3531
viewtopic.php?f=3&t=3766
- Thu Sep 24, 2015 7:13 am
- Forum: Method and application
- Topic: PC in groups not sig, but difference between groups sig/ MGA
- Replies: 4
- Views: 6446
Re: PC in groups not sig, but difference between groups sig/
There are several articles, if you search for it. This is one prominent example: Gelman, A., & Stern, H. (2006). The difference between “significant” and “not significant” is not itself statistically significant. The American Statistician, 60(4), 328-331.
- Wed Sep 23, 2015 3:12 pm
- Forum: Method and application
- Topic: How can i get unstandardized LV scores?
- Replies: 1
- Views: 3184
Re: How can i get unstandardized LV scores?
The unstandardized LV scores are now available from the IPMA results.
- Wed Sep 23, 2015 12:07 pm
- Forum: Method and application
- Topic: PC in groups not sig, but difference between groups sig/ MGA
- Replies: 4
- Views: 6446
Re: PC in groups not sig, but difference between groups sig/
It is definitely a strange result and should be further investigated. Generally, the difference between significant and insignificant must not be significant. This is, however, something different that your situation, but closely related. The difference test and the normal test are two different tes...
- Wed Sep 23, 2015 11:57 am
- Forum: Method and application
- Topic: Path Coeff. of total Group bigger than separated Group PathC
- Replies: 3
- Views: 4187
Re: Path Coeff. of total Group bigger than separated Group P
It is not surprising what you find. In brief, PLS estimates outer weights for the indicators of your constructs (which will be represented as weighted composites) and then estimates structural paths from these composites. There will be some iterations of outer weights and structural paths estimation...