> If your baseline is a questionaire, then the questions you would think to ask are going to be informed by your preconceived notions of the issue, and you're forcing the raw data out there to conform to these formats you've set for it.

Generally, you'd run a bunch of interviews to develop a question bank, and then run a bunch of pilots to figure out which set of questions work best.

I do agree that there's a big problem in that you base your new measure on how it performs on the old measures, which is problematic if both measures are survey based.

This is a real problem in a bunch of psychology research.

That being said, the actual problem is the culture. You are required to spout out theory if you want to be taken seriously. I only put a theory chapter into my PhD because I was forced to. Like, I think theory can be useful, but it's putting the cart before the horse, given that we're mostly in the stamp collecting phase of psychology as a science.