ok, first you have to do a pilot study, with a limited sample (some dozens of subjects), in order to investigate whether the questionnaire is clear to understand and acceptable. after, you repeate the questionnaire and you see if you have variations in the answers scores. Moreover, you administer other already validated questionnaires related to the subject. You see the correlation of the answers to your questionnaires with the answers of the other validated questionnaires. After some months you use a larger sample (also with control groups if necessary) and after 2-3 weeks repeat it. In this way you can calculate the cronbach alfa, using the other questionnaires you can do the factorial analysis
Respected Nicola Bragazzi, so kind enough with your answer. Actually, I developed a questionnaire to assess the knowledge, attitude and practice of asthma patients with the consultation of pulmonologists. I administered that questionnaire to 30 patients. Made few corrections based on their response and their capacity to understand questions. Now what should i do further. Thank u very much for your concern to reply
OK, select another sample of patients and after 2-3 weeks repeat the questionnaire. by the way which other questionnaires did you administer together with your questionnaire?
No other validated questionnaire is available sir. So could not administer any other. Should i repeat it on same patients or different patients sir. What data should i collect.
OK if other validated questionnaires are no available, point this as a limitation of your study and or study protocol.
Repeat on different patients. Besides your questionnaires, collect these data:
age, gender, job, income
asthma symptoms gravity (using clinical scores, such as http://www.clinmedres.org/content/2/3/155.full, or use the ASS, RAD scores, etc)
It would be important to correlate knowledge, attitude with the symptoms severity scores.
Sir actually first time i am using this research gate community. I didn't knew RG have great opportunity to discuss with genius like u etc. Thanks for your efforts sir. By the by, now i administered the questionnaire to 30 patients and incorporated and modified few things based on their comments. If i repeat the same questionnaire to other 30 patients
1. If all patients understood the questionnaire and no more corrections are recommended what should i do
2. If again few corrections are recommended by patients to make them clear what should i do.
Sorry to put my question arrows sir if i disturb u.
I dn t get u Dr Mary. What do u mean as first question. Please clarify me how to calculate cronbatch alpha
If you use SPSS you should follow this way: Analyze->Scale->Reliability Analysis. In appeared window you should move in right part all items in interest and click OK.
But you should remember that table of data consists of items, not scales of the test.
Normal value of Cronbach's alpha starts from 0,7.
And validity is also a question to your test. So you should think about tests to which your test will have positive or negative correlations. Also you can make contrast groups. Or find some measures of real activity.
But I don't understand what exactly is the purpose of your test, so I can't give an advise.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics
- Survey Design
- Questionnaire