
Coleman implored the state to administer a lie detector test before his death sentence was carried out, and the governor permittedĀ this unusual request. Coleman failed the test. His advocates, unwilling to accept the results of the test they had requested, remained so positive he was innocent that they worked for years after his death to have DNA testing of key evidence. That unusual request was also granted and the tests proved there was a one in 19 million chance that the semen found on the victimās body belonged to someone other than Coleman.
Confronted with overwhelming DNA evidence of guilt, Peter Neufeld of the Innocence Project said āToday we got just one answer, and one man cannot speak for the correctness of the verdicts in a thousand other cases.ā The same logic applies, of course, to the unusual case that Rachel Aviv just wrote about in The New Yorker.Ā But Aviv wants her readers to generalize broadly from one case, and to that end she relies entirely on a single study that has been widely discredited.

study by Shaw and PorterĀ that claimed in 2015 that āseventy per cent of people, when subjected to highly suggestive and repetitive interviews, would come to believe that they had committed a crime.ā What Aviv did not tell her readers is that this study is well known to academic psychologists as an outlier; it is also known for its incoherence. Brewin and Andrewsā 2016 meta-study demonstrates that the study is an extreme outlier. Pezdek and Blandon-Gitlinās 2016 analysis concluded that the same study was essentially incoherent, employing āan unorthodox rating systemā that rendered it āimpossible to know what the high prevalenceĀ rate actually refers to.ā
There are two possible explanations for Ms. Avivās over-reliance on this discredited study: (1) herĀ research was so limited she was unaware of these critiques, even though they were published online nine months before her article was published, or (2) she was aware of these critiques and decided not to trouble her readers with complicationsĀ that contradict her position. Either way, once again,Ā The New Yorker blew it on a story about false memory, as they also did hereĀ (with the same study), and hereĀ (without the benefit of any study at all).