I ran a Confirmatory Factor Analysis (CFA) in Amos and obtained a model with high covariance between the factors. James described this covariance in his video as alarming.

What can I do to tackle this problem? All my factor loadings are high and it seems to be a good model, except that I have high covariance between the factors.

Additionally, the model fit seems to be poor. All indices indicate that.

Somehow, I was really surprised, because I used the TAM model and the standard items. Therefore, content validity should be given and I expected that construct validity would be given as well.

  • $\begingroup$ High covariance between the factors indicates a violation against construct validity (discriminant validity). The questions for each factor are intercorrelated. Thus, construct validity is not given. $\endgroup$ – TottiMSc May 11 '14 at 17:44
  • $\begingroup$ Try a single factor, instead of two (or more) factors. If the factor correlation is 1.00, a single factor model is equivalent. $\endgroup$ – Jeremy Miles May 22 '14 at 18:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.