Skip to Main Content

03 May 2016 - 03 May 2016

What is the reproducibility crisis in science and what can we do about it?

  • Rhodes House
  • 5:30pm-7:30pm
Event Image

Professor Dorothy Bishop, FRS, FBA, FMedSci is a Wellcome Trust principal research fellow and professor of developmental neuropsychology at the University of Oxford. Her research focuses on developmental disorders, specifically those resulting in language impairments in children. She has been a pioneer in using twin studies to elucidate the genetic contributors of such disorders, and her work has challenged the paradigms underpinning the understanding of dyslexia, specific language impairment, and autism. Professor Bishop is also a prominent voice in the public understanding of her field and of science in general. She runs a popular and award-winning blog, BishopBlog, maintains a dynamic presence on Twitter, and recently organized a YouTube campaign to raise awareness of language learning impairments. Recently, she has been featured in a BBC Radio 4 project on discovering the source of this crisis, and methods of improving science integrity.

Professor Bishop began by introducing the reproducibility crisis in science and placing the recent interest in this phenomenon in a broader historical context. The roots of this problem are not new, and various methodological misuses of hypothesis generation and statistical analysis have been reported for decades. According to Professor Bishop, the renewed interest in the issue is due to an increasing number of studies thoroughly quantifying the problem and a greater outspokenness of the beneficiaries of technology, such as doctors, patients, and pharmaceutical companies. The specific sources of irreproducibility that Professor Bishop cited include researcher self-deception and bias, improper data analysis, and an inappropriate set of incentives in academia that discourage open collaboration and data sharing. Some of the solutions that she proposed include improvements in automatic data analysis methods, which would decrease the impacts of biases, improved education in statistical analysis, and a restructuring of scientific publishing and funding that would encourage research quality over newsworthiness and institutional prestige.

Key Note Speakers

Registration details

The event is now closed