Questionnaire design limitation


My research has four stages Quan Quan Qual Qual.

For the first stage I used cronbachs alpha to measure construct validity- this was ok.

For the next stage I had to simplify my constructs for a younger age group and reduce the items. I could not used Cronbach's as only have two items per construct one which is negatively keyed, the other positive.

It's now coming to light ( in terms of the mush that is my brain) that some of these constructs aren't measuring what they are supposed to. I would say its potentially two out of 8.

Its 8 weeks until submission and I can get my head on as to how I will acknowledge this design flaw:(

Someone please help!


Well you clearly don't have time to actually correct the methodological flaw, so my temptation would be to include a paragraph in either the methods section or the 'limitations of study' section highlighting the conceptual ambiguity or lack of validity. Just be honest about it. So:

During the process of research it was discovered that ... . This is problematic in as much as ... . The consequences of this problem, then, is that ... . It could be hypothesised that, should the concepts be more valid, the results could ... . In future, this problem could be resolved by: i) ... ii) ...

So make sure you state the problem, what influences it could have your research and results, and perhaps even how you would correct it if you were to do it again. The key thing is clarity. We all make mistakes. Just be clear about what those are, what the effect on your research is, and how you would avoid it in the future.


Sorry for the late reply traffic I did take this advice on board. Tthings weren't as bad as they seem but I've written it up in my limitations any way:)