If you read the underlying findings [1], there were two behavioral ethics studies with fabricated results. The first was conducted in 2012 and was supervised by Ariely.
I'm aware - in context, however, it appears *dsr was defending the use of 'he' used as a default without reading the article where the subject is clearly a 'she.' Whether another professor at Duke mentioned in the article supervised the study doesn't really matter.
[1] https://datacolada.org/109