I am running pilot test and the Cronbach's alpha is 0.71 for all of the items. However, when I come to analyze every section of the questionnaire individually, In found that it is too low. And I cannot delete all of these items. What can I do to improve it?
The quality of the items will determine the Cronbach alpha. Statistics reveal to us the positive results as well as the deficiencies in our work. If you want a questionnaire that your colleagues will want to use, the best way through is to replace the poor performing items/questions for more valid ones. Your stats are telling you that some of your individual items lack adequate validity, and ultimately will reflect a poor questionnaire. In one such study of mine, a 25-item questionnaire was originally reduced from over 400 items using factor analytical procedures. In closing, only valid items = valid questionnaires...and a good Cronbach alpha.
- Badges
- Science topic
- Similar topics
- Biological Science
- Histology
- Sectioning