The Danger of Using Statistics to Pass Judgment

Jenny Nguyen, Maroon-News Staff

By now, most Colgate students have seen the flyers around campus. “Colgate is hiring Alan Dershowitz, an accused child rapist, to speak on November 5th. Considering only 2.1 percent of rape accusations are false, should we be welcoming him to campus?” The flyer then goes on to encourage students to contact the organizers of the event on campus, presumably to protest Dershowitz’s presence, along with links to news articles on the allegation.

Setting aside the obvious issue of free speech (an issue on which Colgate just recently passed a resolution), the flyer is inviting students to consider Dershowitz’s guilt or innocence based on nothing more than the statistical likelihood (or unlikelihood) that his accuser was lying. According to the articles on the flyer, Dershowitz’s accuser had not filed a report or formal complaint; she only spoke out as part of a separate lawsuit in which he served as attorney for the defense. Both her lawyers and Dershowitz subsequently settled the defamation charges they brought against each other. The articles offered few details on the actual allegation, and even less evidence to support it. They shed no light on what Dershowitz may or may not have done. Students are left to judge for themselves whether a potential guest lecturer had committed sexual assault based on that number, 2.1 percent.

It should not come as a surprise that statistics can and have been misused to pass judgment on someone’s guilt or innocence. In 1998, a woman named Sally Clark was convicted for the alleged murder of her two children after an expert testified that the probability of both of her children dying of SIDS (sudden infant death syndrome) was so infinitesimal that their deaths must have been intentional. The statistical analysis was later found to be flawed, and Clark was exonerated by forensic evidence. Flawed statistics were also used in the murder trial against O.J. Simpson, when the defense argued that only one in 2500 domestic abuse victims are murdered and so, therefore it would be improbable that Simpson had both abused and murdered his wife. Your statistics professor could probably tell you numerous other unfortunate cases in which a person was judged to be guilty or innocent based on such fallacies.

The current estimates for false allegations of sexual assault vary widely, but let’s say that it is 2.1 percent. That means 2.1 percent of sexual assault allegations that have been investigated were found to be false. But we as a society should keep in mind that 100 percent of uninvestigated, unsubstantiated cases are not true until proven so. In our fervor to support survivors and combat the epidemic of sexual abuse by powerful men, some of us have lost track of this fundamental principle. As someone who grew up in a country where extrajudicial arrest and detention is a matter of course, I am appalled at this idea that we could pass judgment on a person’s guilt, and encourage others to do the same, solely because someone makes a claim that fits a statistic. It is unrealistic and morally untenable.

Point to the evidence. Point to the accused’s character and conduct. Point to the accuser, if you must. But saying someone committed a crime simply because it is statistically likely is a habit we can all do without.

Contact Jenny Nguyen at [email protected].