There is a fallacy about logical fallacies: using them to discredit non-logic based arguments. When one wants to assess the credibility of anecdotal facts, it i

There is a fallacy about logical fallacies: using them to discredit non-logic based arguments. When one wants to assess the credibility of anecdotal facts, it is entirely fair to ask whether a person is a reliable source for this information. This is why character witnesses are acceptable in court. We are not trying to prove a theorem or explain the results of an experiment. We are trying to use our intuition, for lack of better tools and information, to assess what we hear. So saying that this person can be considered unreliable as a source of information in this kind of analysis due to past behavior is valid. Objective data and detailed explanation should be provided to back up his argument logically in order for it to be acceptable, rather than taking Duncan's word for his analysis of the data he has presumably seen. In this case, the author's data-based assertion of non-correlation is much weaker than Douglas Duncan's. I think based on Duncan's background, it is fair to dismiss his quote entirely without further context and base conclusions only on the data presented. In addition, since the author presents an unreliable source, one may suspect the data is skewed to support that source, conciously or unconsciously.