Dear Abinash,

I suggest you the following interesting links and attached file

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3521886/

http://www.aijcrnet.com/journals/Vol_3_No_5_May_2013/8.pdf

https://www.academia.edu/513622/Development_of_a_questionnaire_on_knowledge_attitude_and_practice_about_immunization_among_Iraqi_parents

http://www.hindawi.com/journals/isrn/2014/403058/

http://www.sciencedirect.com/science/article/pii/S1319016415000122

With my best regards

Prof. Bachir ACHOUR

Hello,

It depends whether you are 'building' a survey based on already established questionnaires (then you should go for CFA and compare your data with past evidence, the same for internal consistency scores (reliability) or you are starting from scratch. 

Then, you need to go for EFA, it would be appropriate to test this new questionnaire across more than one sample (preferably drawn from general population) along with some well established measures to test for congruent validity. Internal consistency should be above .7 or not far below this level. 

Hope it helps a bit. I am sure other colleagues will add some more to this. Finally, some more details highlighted above would be appreciated, then advise could be better tailored to your needs.

Regards, Witold

 Here are links to two similar discussions here:

https://www.researchgate.net/post/How_can_the_reliability_and_validity_of_KAP_questionnaires_be_tested

https://www.researchgate.net/post/How_do_you_determine_the_validity_and_reliability_of_a_new_questionnaire