Keenan, R. et al
Department of Psychology Chair Dr. Martha Willets apologized this week for sponsorship of what was billed as the department’s first annual p-hackathon, but was actually a bid by a professor to improve the p-value of his data.
“I was led to believe the “p” stood for programming,” said Dr. Willets. “I never imagined a faculty member would pressure students into mining his data until a statistically significant p-value could be found.”
While Willets’s email did not identify the faculty member, sources named behavioral psychologist Joshua Beecham, who will be considered for tenure next year.
One post-doc, who spoke on condition of anonymity, characterized the data from Beecham’s latest study, looking at mood disorders in urban environments, as a “shit-show.”
“You’ve seen blood-spattered walls in slasher movies? Our scatter plots looked worse. Not one key variable was related to another. That’s when he came up with the hackathon idea.”
Participants at the 36-hour event were provided with pizza, energy drinks and temporary “I am ≤ 0.05” tattoos while Beecham stood on the sidelines offering encouragement. “At one point he was yelling ‘Rerun that data like it’s a Simpsons episode,’” said grad student Angela Yost. Other p-hackathon participants confirmed that Beecham distributed t-shirts reading: “You call it stepwise regression, I call it first author JAMA.”
When contacted, Beecham’s only reply was a PDF of his newly accepted paper: “Wonderwall: The effect of apartment building facade colors on anhedonia in city dwellers.”
