1
$\begingroup$

I have a dependent construct (attitude) which has at least 2 dimensions in theory (affective & behavioral). The scale measure adopted for this research is found to have 2 or 3 dimensions in literature. Thus, on the theoretical level, this construct should be a second order factor when conducting CFA in AMOS.

But, the EFA of 782 cases data-set in SPSS shows that there is only one factor, not two, and this is what the sample presents.

So, in this case, when the results of EFA are different from those of theory, what should we do? Can we still treat this construct as a second (higher) order factor with two dimensions (sub-factors)? Or do we have to respect the results of the EFA and treat it as only one factor?

Improve this question
$\endgroup$
1
$\begingroup$

If you have a theory, then you should test it with CFA.

The problem with EFA is that there are an infinite number of solutions, all of which are (statistically) equally good when (talking about rotation).

There is no point testing a second order factor with only two indicators - that model is exactly equivalent to a two correlated factor model.

Improve this answer
$\endgroup$
3
  • $\begingroup$ Thanks a lot for your response. Actually the general theory of consumer attitudes say that this concept is multidimensional. Attitude here is a DV in an overall model. And, I couldn't figure out how to treat the results of EFA in this case. Should I ignore it completely, sir? and what should I say in my report about that? $\endgroup$
    – nisr
    Dec 5 '15 at 21:24
  • 1
    $\begingroup$ +1, but "infinite number of solutions" probably refers to the freedom of factor rotations, which does not change the number of factors, whereas this Q seems to be primarily about number of factors in EFA being less than expected. Perhaps you can expand your answer to comment on that? $\endgroup$
    – amoeba
    Dec 5 '15 at 21:38
  • $\begingroup$ @amoeba, yes, I should have been clear about that. $\endgroup$
    – Jeremy Miles
    Dec 5 '15 at 23:21

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.