symptoms recorded and presence of gluten in the diet.

Did they test for the presence of any one of a milllion or more other substances in the diet that could have been responsible for the symptoms?

And was the inverse of the "placebo effect" factored in?