Stone-Geisser-Criteria

[AnjaGF]

Hello,

I would like to know, if there exists the possibility to validate the model by the Stone-Geisser-Criteria in Smart PLS? Which possibilities do I have to validate the Measurement and the Structure Model?

Thank you for your help!

Best,
Anja

[cringle]

Hi,

if you use the search function of this forum, you will find the following link that probably ansers your question on the Stone-Geisser-Criterion:

viewtopic.php?t=119&highlight=stonegeisser

For model evalutaion, check Tenenhaus et al. (viewtopic.php?t=21) and Chin (viewtopic.php?t=19).

Best
Christian

[tomcurl]

Dear all,

after preparing a review of the creiteria for evaluating the quality of the structural model, I have a question concerning communality and Q2.

Communality measures the quality of the measurement model for each block (tenenhaus et al. 2005, p. 173). It represents the sum of the correlations of a block of manifest indicator variables with a latent variable.

Stone-Geisser's Q2 is the relevance of the parameter estimators for a block j of manifest variables (Ringle 2004, p. 17). This is calculated using the sum of the squared errors for the parameter estimators (E), as well as the sum of the squared errors for the mean parameter estimators (O)(Ringle 2004, p. 17). In the following equation, D refers to the distance between two cases that are first left out and subsequently considered:

∑D ED
Qj²:= 1 – ________________
∑D OD

Just to be 100% sure: as far as I'm concerned, communality and Q2 are the same, aren't they? A quick comment from a PLS expert would be greatly appreciated.

Thanks in advance!

Best regards,

Thomas