Reader Comments

Post a new comment on this article

why not use real measures to look at improvement?

Posted by ellenli on 28 Dec 2010 at 02:23 GMT

At first I thought this was meant as a caricature of a study but since it seems to be "real" in some sense why not make it really real? That is, instead of using typical junk measures, have the subjects keep a symptom/medication diary and prior to actually doing the study define what you will consider to be improvement?

wrt the other comments about using fiber as a placebo: I wonder just how much was being used. 4 pills a day of what size?

Last, a 7 point scale should not be analysed using a t-test but even so does anyone really believe that a 1 point difference, between "no change" and "slight change" has any substantive meaning?

No competing interests declared.