BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:The perils of p-hacking and the promise of pre-analysis plans - Ja
 cob Stegenga (with Zoë Hitzig) (Department of History and Philosophy of S
 cience)
DTSTART:20171018T120000Z
DTEND:20171018T133000Z
UID:TALK85441@talks.cam.ac.uk
CONTACT:31287
DESCRIPTION:P-hacking involves the manipulation of data to find a statisti
 cally \nsignificant result. Many claim that p-hacking is a problem in scie
 nce\, \nespecially in the medical and social sciences\, while others deny 
 this. \nThe problem with p-hacking is usually articulated from a frequenti
 st \nperspective. In this paper we articulate the epistemic peril of \np-h
 acking using Bayesian confirmation theory and model selection theory\, \nw
 hich we then draw on to explain the arguments on both sides of the \ndebat
 e. This requires a novel understanding of Bayesianism\, since a \nstandard
  criticism of Bayesian confirmation theory is that it cannot \naccommodate
  the influence of biased methods. A methodological device \nwidely used to
  mitigate the peril of p-hacking is a pre-analysis plan. \nSome say that f
 ollowing a pre-analysis plan is epistemically meritorious \nwhile others d
 eny this\, and in practice pre-analysis plans are often \nviolated. We use
  the formal groundwork developed earlier in the paper to \nresolve this de
 bate\, offering a modest defence of the use of \npre-analysis plans. In th
 e longer run our ambition is to use this \napproach to make sense of scena
 rios in which scientists depart from \npre-analysis plans.
LOCATION:Seminar Room 2\, Department of History and Philosophy of Science
END:VEVENT
END:VCALENDAR