"Why this sample size?" "But those aren't parametric data." "That's not an appropriate control group!" So it goes when editors meet to decide which studies to publish and which not. There is lots of wondering aloud, mixed with colorful editorial epithets, about why investigators conceived, carried out, and reported their project this way rather than that.
These deliberations turn on two points: Is the work valid and will the reader gain from having seen it? Such questions imply, rightly so I think, that the reporting of research is a collaborative effort between those who do it and those who read it. Only in the last decade has this been fully realized by the clinical epidemiologists who are in a position to do something about it. Many books, great and small (not a few of which have been reviewed in these pages), have appeared in an attempt to explain how valid