ASA Ethics Case Study #2

The following article is a case study concerning the ethical use of statistics. You are invited to read it and join the discussion on it.  Please send your contributions in the form of an e-mail to the Ethics Committee Chair, specifying in the subject line of your message the case study you wish to contribute to.

 

Uncounted Data from the Scintillation Counter

The following case was borrowed from materials for the Third Intercollegiate Ethics Bowl (1997), prepared by the Center for the Study of Ethics in the Professions at Illinois Institute of Technology. Used with permission.

Armstrong is a first year graduate student, working in a molecular biology laboratory. She has great admiration for Hayes, who is just finishing his thesis work. He seems to have a golden touch in the laboratory. His experiments produce clean data, with scatter consistently less than or equal to theoretical predictions. Because his experiments seldom need to be repeated, Hayes has produced a thesis full of fascinating and demonstrably correct results. The laboratory has already followed up on several of these with success. One day Armstrong notices Hayes leaving the scintillation counter and can't help noticing he has 80 vials. This barely registers in her subconscious until later in the day he shows her his experimental results with 40 data points. When she asks about the missing points, he explains that it is standard practice to eliminate outliers from the analysis. He goes on to mention that the scintillation counter is a scientific instrument that frequently produces murky readings distorted by many different kinds of factors. The more Armstrong thinks about this, the more distraught she becomes. A week later she summons up her courage and tells her story to the professor in whose lab she and Hayes work. He seems uninterested and irritated. He hoped she had come to present him her experimental results, which she hasn't done for months.

What, should Armstrong do next, if anything, and why?

Additional or alternative scenarios might be offered. For example, would it make a difference if she noticed the discrepancy while reviewing a draft paper for publication? Would it make a difference whether she was to be a co-author on the paper? Would it make a difference if she were not a student, but a professor of statistics in a different department of the same university? Would it make a difference if she and Hayes were professional colleagues in a cancer research laboratory? Would it make a difference if she were Hayes' advisor or mentor?