Search found 899 matches
- Fri Sep 06, 2013 3:58 pm
- Forum: Method and application
- Topic: Total effect and number of latent variables?
- Replies: 1
- Views: 1750
Hi, I think that you are right, but the explanation should be done more in terms of mediators than paths. For example: Model 1) IV --> (V1) --> DV [R² = 0.20] Model 2) IV --> (V1 and V2) --> DV [R² = 0.30] Model 3) IV --> (V1, V2 and V3) --> DV [R² = 0.35] Probably the total effect of IV is bigger i...
- Fri Sep 06, 2013 3:26 pm
- Forum: Method and application
- Topic: Low AVE combined with high Cronbachs Alpha??
- Replies: 2
- Views: 2718
Hi, These results are uncommon, but if you are assessing 2nd order LV, the explanation is simple: SmartPLS uses the outer loadings or the repeated indicators instead the values between 2nd order LV and its 1st order LV (they are reported as path coefficients). See: http://forum.smartpls.com/viewtopi...
- Thu Sep 05, 2013 12:07 pm
- Forum: SmartPLS 3 - Small Talk Corner
- Topic: inter-item correlation
- Replies: 5
- Views: 7001
- Thu Sep 05, 2013 12:04 pm
- Forum: SmartPLS 3 - Small Talk Corner
- Topic: How can I delete my SmartPLS Account?
- Replies: 2
- Views: 4729
Hi,
if you want to delete your account, just send an e-mail to Christian Ringle.
c.ringle@smartpls.de
Best regards,
Bido
if you want to delete your account, just send an e-mail to Christian Ringle.
c.ringle@smartpls.de
Best regards,
Bido
- Thu Sep 05, 2013 11:52 am
- Forum: Method and application
- Topic: Always positive
- Replies: 2
- Views: 2437
Hi, You should look at outer loadings too. As the algorithm is “partial”, the measurement model is computed in one step and structural model in the next, until the convergence. For this reason, sometimes (rare) I found: - Positive path coefficient and all outer loadings of one LV negative - When I r...
- Thu Sep 05, 2013 11:34 am
- Forum: Method and application
- Topic: confirmatory factor analysis
- Replies: 1
- Views: 2400
Hi, CFA means that we do not have a structural model. All LV are connected on with each other, we use factor weighting scheme, and do not use the path coefficients, just the LV correlations and results of measurement model: Convergent validity: outer loading > 0.7 || AVE > 0.5 Discriminant validity:...
- Thu Sep 05, 2013 11:30 am
- Forum: Method and application
- Topic: Discriminant validity of 2nd Order Factor
- Replies: 5
- Views: 4851
Hi, If HOCs are formative, it does not make sense compute AVE and composite reliability. We do not expect that their LOCs will be correlated, for this reason they are used as formative. If HOCs were reflective, we should compute their AVE and CC based on their relations with LOCs (they are presented...
- Thu Sep 05, 2013 11:23 am
- Forum: Method and application
- Topic: different result in smartPLS and SPSS
- Replies: 2
- Views: 3389
Hi, the estimation of unstandardized LV in SmartPLS has a bug (see: http://forum.smartpls.com/viewtopic.php?t=812 ), for this reason, if you used these scores to compare SmartPLS with SPSS, the results are different. To have SPSS = SmartPLS: 1) Run your model in SmartPLS 2) Copy/paste the standardiz...
- Sun Jul 28, 2013 11:10 pm
- Forum: SmartPLS 2 - FAQ
- Topic: licence smartpls 2.0
- Replies: 1
- Views: 2671
- Tue Jul 23, 2013 10:48 pm
- Forum: Method and application
- Topic: Measuring model - structural model
- Replies: 4
- Views: 4094
- Fri Jul 19, 2013 11:42 am
- Forum: SmartPLS 2 - FAQ
- Topic: Errors in Bootstrapping
- Replies: 4
- Views: 4605
- Mon Jul 15, 2013 7:59 pm
- Forum: Method and application
- Topic: AVE of second order construct
- Replies: 2
- Views: 2225
- Fri Jul 12, 2013 8:09 pm
- Forum: Method and application
- Topic: Formula of beta weight
- Replies: 3
- Views: 3801
Hi,
It’s just regression, see one example here:
PPT = https://app.box.com/s/bd22560c4efd4a1eef50
XLS = https://app.box.com/s/6f69efce3761aebf884c
It is in Portuguese, sorry.
Best regards,
Bido
It’s just regression, see one example here:
PPT = https://app.box.com/s/bd22560c4efd4a1eef50
XLS = https://app.box.com/s/6f69efce3761aebf884c
It is in Portuguese, sorry.
Best regards,
Bido
- Fri Jul 12, 2013 5:18 pm
- Forum: Method and application
- Topic: Inter-manifest variable correlations?
- Replies: 1
- Views: 1551
Hi, Inter-construct = latent variable correlations (see the Report/Quality Criteria/Latent…) Intra-construct = inter-items correlations = use SPSS, R or other statistical software to compute these correlations = item-to-total correlations = use SPSS / analyze / scale / reliability Best regards, Bido
- Sun Jun 23, 2013 10:19 pm
- Forum: SmartPLS 3 - Small Talk Corner
- Topic: Calculation of Contribution to R2(%)
- Replies: 3
- Views: 5404
Hi, If you want to compute the contribution of each independent variable.. Example: y = b1*x1 + b2*x2 +b3*x3 Contribution of x1 = ry1 * b1 Contribution of x2 = ry2 * b2 Contribution of x3 = ry3 * b3 Were: ry1 = correlation between x1 and y. b1 = path coefficient Sum of these contributions = R² Best ...