I have problem in assessing the discriminant validity for whole model. However, for exogenous factors the validity is met. Is it important to check the discriminant validity includes exogenous and endogenous latent factors?
Discriminant validity should be done for all the constructs under study.It is necessary to ensure the discreteness of the constructs whether exogenous or endogenous, irrespective of it.
Both...If the correlation between the construct is too high (e.g., >.90); that means they are the same thing.
What is it that you hope to learn through the discriminant validity?
c. Discriminant Validity. This validity indicates the measurement model of a construct is free from redundant items. AMOS could identify the items redundancy in the model through a discrepancy measure called Modification Indices (MI). High value of MI indicates the respective items are redundant. The researcher could delete one of the identified items and run the measurement model. The researcher could also constraint the redundant pair as “free parameter estimate”. Another requirement for discriminant validity is the correlation between exogenous constructs should not exceed 0.85. The correlation value exceeding 0.85 indicates the two exogenous constructs are redundant or having serious multicollinearity problem. Obtained from Zainudin Awang (2014). A Handbook on SEM for Academicians and Practitioners. MPWS Publication, Bangi Malaysia.
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis