Hello,
I am currently investigating some reflective research models with SmartPLS and had a few questions about checking the Construct reliability, validity and discriminant validity.
Can i consider Composite Reliability's result and leave Cronback's Aplha for Construct Reliability?
Can I consider Fornell-Larcker results and leave HTMT for Discriminant Validity?
Yes, you should report Dillon-Goldstein/composite reliability coefficient instead of Cronbach's Alpha. Yes, you can use the Fornell-Larcker criterion instead of HTMT criterion.
Dear Muhammad Junaid Khalid
Regarding your first part about the reliability measure that you want to use. Yes, you can use Composite Reliability rather than Cronbach’s alpha’s due to Cronbach’s alpha’s limitations: (As such, it may be used as a more conservative measure of internal consistency reliability)
1. Cronbach’s alpha assumes that all indicators are equally reliable (i.e., all the indicators have equal outer loadings on the construct).
2. Cronbach’s alpha is sensitive to the number of items in the scale and generally tends to underestimate the internal consistency reliability.
Thus, composite reliability is proposed to be used. This measure of reliability takes into account the different outer loadings of the indicator variables
Source:
Hair, J.F., M. Hult, G.T., Ringle, C.M., Sarstedt, M. (Eds.), 2017. A Primer on Partial Least Squares Structural Equation Modeling (PLS-SEM), Second edition. ed. Sage, Los Angeles.
Regarding the Discriminant Validity measures, there are mainly, three approaches to address this, which are the following:
1.The cross-loadings
2.The Fornell-Larcker criterion
3.Heterotrait-monotrait ratio (HTMT) of the correlations.
you can rely on the Fornell-Larcker criterion if you do not want to use the HTMT, however, I would also suggest that you ensure the cross-loading results are emphasising the Discriminant Validity.
can also include the effects of the categorical variables to evaluate if there are differences in the construction means
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Probability
- Probability Theory
- Reliability