Hello All, I am not even sure if this is possible, but I have used this dataset for several publication -albeit with different constructs-and have not seen such low t-values. For example, I have a model with all the BFI personality traits as constructs related to an established construct measuring the intensity of Facebook use. All the loading and other measurement model statistics are spot on, but the t-statistics of the outer loading related to Neuroticism are all nonsignificant. Removing some of the items does not improve the situation. Interestingly, the outer loadings are all above 0.7 and the other constructs are good. The AVE, CA and CR are also good. What could be cause of this? I am using SmartPLS 3 and running native bootstrapping.
Ok. I seem to have resolved the issue by upgrading to the latest version of SmartPLS (from 3.2 to 3.2.9).
The t loads are not the routing coefficients of the SEM models?
I have the same problem high loadings CR and AVE but insignificance or low t values for outer model? Is there any recommendation or how did you fix it?
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running