Volume 38 · Number 12
DECEMBER 2008

Statistics — What Can You Believe? If You Care About DSM-V, Read This

By Jan Fawcett, MD

First, let me say how proud I am that this last issue of Psychiatric Annals for 2008 is featuring a series of articles on behavioral statistics, guest edited by Robert Gibbons, PhD, a national resource. Dr. Gibbons has enlisted some of the best statistical minds in this country to write about modern issues in behavioral statistics. The material is well worth digesting. For research investigators, this issue should be a collector’s item. For clinicians who have been bombarded by conclusions in textbooks in their development, and are currently flooded with “new findings” from many sources, it is a key to informed skepticism — a much better position than no-nothing nihilism. Was that result you claim obtained from a last observation carried forward (LOCF) analysis? How did your handling of missing data affect your result? Much is not what it seems, and this series is a partial guide for the rabbit hole. My thanks to the authors for the informative challenge they have brought us.