“P-hacking” sounds like some sort of dark-web-cryptocurrency-type activity. But, in fact, it’s a shady practice from a different sphere: the pseudoscientific manipulation of a study to produce a desirable result. Coined by psychologists Uri Simonsohn, Joseph Simmons and Leif Nelson almost a decade ago, the “p” comes from “p-value,” a number given to research to show how likely the results could have come about by chance. A value lower than 0.05 (5%) allows researchers to declare their findings as “statistically significant, ” thus considerably increasing the likelihood of their work being published. Rather than setting out to prove a hypothesis, p-hacking means looking at the data and working backward to find out what hypothesis it could possibly prove.1 If a researcher is disappointed by the lack of findings from a particular study (on the benefit of eating breakfast for weight loss, for instance), they might This story is from Kinfolk Issue Forty-Three Buy Now Related Stories Arts & Culture Issue 45 Bad Idea: Paper Straws On the straw man of sustainability. Arts & Culture Issue 44 Bad Idea: Context Collapse Why misunderstandings multiply online. Arts & Culture Issue 42 Bad Idea: Year Wraps An algorithmic celebration of your most depressing digital data. Arts & Culture Issue 41 Bad Idea: Stereo Type The omnipresent embarrassment of “exotic” type. Arts & Culture Issue 38 Bad Idea: Gender Reveals It’s time to burst the (pink or blue) bubble of this trend. Arts & Culture Issue 37 Bad Idea: Lawns A green and pleasant death knell for diversity.
Arts & Culture Issue 42 Bad Idea: Year Wraps An algorithmic celebration of your most depressing digital data.
Arts & Culture Issue 38 Bad Idea: Gender Reveals It’s time to burst the (pink or blue) bubble of this trend.
