Dear Colleagues and Scholars,
I have carried out a textbook analysis, based on which an interview with teachers was conducted. Then a five-point Likert scale questionnaire, consisting of 9 questions, was developed based on the interview. Now, for construct validity of the questionnaire, is it necessary to run a factor analysis? Or could the interview itself be considered as its validity?
Factorial analysis is necessary for the evidence of validity of the theoretical construct, a better evaluation would be based on a latent SEM analysis of the questionnaire.
Hello Hossain,
The answer depends on your intended use(s) of the scores from your questionnaire. If you believe that, as you generated the instrument, all nine (9) questions were worded so as to tap a single dimension (for example, the attitude of respondents towards smoking) and intend to use a summated (or averaged) score, it would make sense to confirm that you indeed have a single-factor model.
If you intend to use each item as an individual variable (not the best practice for assuring that scores are reliable), then factoring might still be useful
Dear Hossein,
I agree with David Morse here. Factor analyses are usually used for quantitative studies, as such the scales will be worded to give you a single dimension. On the other hand, if your intend to use the interview (which is qualitative in nature), factor analysis is possible by using Q methodology (yielding Q sorts) and Q factor analysis.
Dear Hossein,
When you are developing a questionnaire in quantitative type and it is based on reviewing a literature or a qualitative approach, you should do confirmatory factor analysis (CFA) to confirm that if your new designed constructs can identify themselves in your new questionnaire/model.
Good Luck
- Badges
- Science topic
- Similar topics
- Biological Science
- Anatomy
- Tissue