Alpha and Omega
A cutoff of .70 is often used for reliabilities. McDonald’s Omega is a more general form of the Cronbach’s Alpha and more reliable (Hayes & Coutts, 2020). See the reference below for more information about the difference between these two.
IMPORTANT: to get correct reliabilities, you should reverse code items that are negatively framed.
AVE and SV
Looking at the Average Variance Extracted (AVE) and Shared Variance (SV) is another way to examine convergent and discriminant validity (Fornell & Larcker, 1981; Farell, 2010). The AVE is the average amount of variance that the latent construct is able to explain in the observed items. SV is the amount of variance that the latent factor shares with another latent factor.
- Convergent validity. The AVE should be equal or greater than .50 (Fornell & Larcker, 1981).
- Discriminant validity. The AVE for each latent construct should be greater than its SV with any other construct.
ICC
If you have indicated that your data is clustered, ConMET also displays the ICCs (Bliese, 1998).
- ICC1 represents the variance at the individual level that can be explained by group membership
- ICC2 represents the reliability of the group means.
These ICCs are calculated on estimated factor scores. You can also get the ICCs for an observed variable by putting them as a single-item factor in the model and then run the CFA.
References
Bliese, P. D. (1998). Group size, ICC values, and group-level correlations: A simulation. Organizational research methods, 1(4), 355-373.
Farrell, A. M. (2010). Insufficient discriminant validity: A comment on Bove, Pervan, Beatty, and Shiu (2009). Journal of Business Research, 63(3), 324-327.
Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39-50.
Hayes, A. F., & Coutts, J. J. (2020). Use omega rather than Cronbach’s alpha for estimating reliability. But…. Communication Methods and Measures, 14(1), 1-24.