To the Editor.
—While I agree in principle with the criteria set forth in the Editorial by Drs Lundberg and Glass,1 I believe that there is an additional way to participate in research projects that can merit authorship. For projects that require large amounts of data management and computer programming, 1 or more of the programmers can meet authorship criteria. This does not apply to a programmer who just simply follows orders about what to do. I have participated in projects where the computer programmers were central to the research. These individuals participated in the regular research meetings and had a full understanding of the research. On a regular basis, comments and suggestions they made in response to programming requests would yield modifications to the analysis. In my experience, these individuals have been coauthors, and I fully believe that they should have been. I think that your criteria