Outer Weights > 1

[inertia1]

Hallo,

wir haben gerade mit einer Test-Auswertung unseres Modells begonnen (momentan sind wir noch bei n=15). Jetzt ist dabei allerdings das Problem aufgetreten, dass wir Outer Weights > 1 bekommen.

Woher kann das kommen? Kann uns da jemand weiterhelfen?

Vielen Dank und viele Grüße,

Christoph und André

[Diogenes]

Hi,

Firstlly, with 15 cases just higher correlations and weights will be significant
And second, Do you have some variables with positive and others with negative correlations in the same LV?
Do you use the standardized solution (or unstandardized)?

Best regards.

Bido

[inertia1]

Hello Bido,

thank you for your answer. Well, we have negative and positive correlations in the same LV. And we have used the unstandardized solution first, but after that tried the standardized one. It became better, but we still had the problem.

We are working on the number of cases (-;

Best regards,

Christoph

[viswadatta]

Though it is not wrong, we usually prefer to have loadings of one sign for LVs. This will give stable results. Try splitting the LV into two LVs

Regards

Vivek