Initially, we did not intend to post letters to the web site, but have eventually decided that all comments should be posted.

Monday, June 8, 1998

Letter: To the Ethics Committee

I am quite surprised that you would fill the ethical guidelines for statistical practice with so much extraneous dribble. Of the Preamble, section B. is fine but section C. Can be entirely removed. For instance, not being civil to fellow statisticians might make one a real jerk but it does not make one unethical as a statistician. The only thing worth keeping from section C. is part 2., but it should be applied to multiplicity problems you run into when scientists want to test everything to find something to publish. Published results should be informative about the number of tests that have been performed to give some idea about multiplicity concern and actual significance level. In the ethical guidelines section A., 1. and 2 are good but respecting the contributions and intellectual property of others has a lot more to due with being a good citizen by respecting copyright laws and patents than it does with statistics and thus could be removed. You might put at the beginning that members are expected to obey the laws of the United States in spirit and intent. That would cover most of this stuff and then you could stick to issues faced more by statisticians than professionals in general. Similar shortening could be applied to the rest.

In short, if it does not have anything to do with statistics as a discipline, then it should not be in there. It is obvious that this principle was left behind quickly in favor of someone's narrow idea of what it means to be a good and reputable person.

Mark Farmen, Ph.D.
Abbott Laboratories
Biostatistician

Dr. Gardenier's reply: Thank you for commenting. The civility issue is listed as a "good professional citizenship" issue, rather than an ethical one. Within section II, the Committee includes only items we feel pertain to ethics of statistical practice. Your opinion may differ. In response to the majority of commenters, those guidelines have been growing, not shrinking. For more information, see the more detailed comments of Philip Pichotta (also of Abbott Laboratories) and my replies on the web at http://www.tcnj.edu/~asaethic. This letter will also be posted so others can express similar or differing opinions.