Hi everyone,
I am using smart PLS for one of my study. I have first order and second order constructs. In Smart PLS i got the AVE for first order constructs above 0.5, whereas the AVE for second order construct is less (0.30). is it acceptable ?if not do i need to report AVE for second order construct? support me with literature ( My Model is Reflective- Reflective model.
the other validity for both first order and second order are satisfying the cut off level ( say convergent, internal composite reliability etc.,)
Typically a value greater than .50 is desired for Convergent validity because it suggests that the majority of the variance in the indicator is due to the latent construct
have you considered that second order AVE is underestimate and for the same issue it's usually computed by hand...and check multicolinearity....
AVE is used to determine validity of scale and validity of scale is checked in first order CFA. Objective of second order CFA is different. It is used when a concept is multidimensional and contribution of each dimensions in concept is required to determined. I think AVE is not useful in second order CFA.
You can refer any research papers that has used second order CFA. The basic concept of first order CFA and second order CFA is different. You may refer the book Structure Equation Modeling with AMOS written by Barbara M Byrne.
I had a very similar problem in my thesis and I decided to drop the second-order construct. If AVE < 0.5 it means the variance explained by the latent variable is less than the measurement error. In other words, the first-order constructs do not have enough commonality to suggest a single, underlying latent factor.
PLS-SEM being distribution-free usually leads to a higher AVE value, especially for non-normal data. If AVE is only 0.3 using PLS-SEM, I would strongly suggest that you report that the hypothesized second-order factor did not emerge.
We can accept AVE < 0.5 if the composite reliability is > 0.6, because the convergent validity of the construct is still adequate (Fornell & Larcker, 1981)
Dear Raj Kumar
I agree with what Mohammad Suleiman Awwad is saying. Other than Fornell and Larcker, you can also cite Huang et al's., (2013) paper titled "An Empirical Analysis of the Antecedents and Performance Consequences of Using the Moodle Platform".
Regards.
- Badges
- Science topic
- Similar topics
- Thesis Research