Readings
Why most published research findings are false
Most published research findings are false - but a little replication goes a long way
Podcasts and links
Biomedical Research: Believe It Or Not?
- Hilda Bastian’s blog post about the Ionnadis paper. Essential reading for Topic 3.
Hertz Podcast 54: Cuckoo Science
In this episode, James sits in the guest chair as Dan interviews him on his recent work find and exposing inconsistent results in the scientific literature.
This podcast is optional. LANGUAGE WARNING: James is very free with the swear words. If you’re offended by profanity, you need not listen. If you like profanity, you’re gonna love it!
Essay Topic - Are most published research findings false?
In 2005, John Ioannidis published a highly controversial paper suggesting that most published research findings in biomedical science were probably false, and that this could be shown with mathematical modelling. This was hotly disputed at the time, with many commentaries on the published article. Since then, however, both the biomedical and psychological literature has come under increasing scrutiny, and there have been some interesting campaigns focusing on error detection in published research - both systematic (such as StatCheck) and more individual (e.g. “PizzaGate”, the downfall of Professor Brian Wansink). Outline what you think may have led to this situation, and whether you think it is hopeless or could be addressed in the future. Starter readings are in the Topic 3 section.
Some questions to consider:
- What did you think of the Ioannidis paper, the published comment on the paper, and the arguments followed up in Hilda Bastian’s blog post?
- Can you find some more recent discussions of the paper?
- Is the future of science doomed, or can reforms help fix the problem?
- What do you think about error detection in science (and can you find some well-publicised exposes of bad science)?
Forum
What did you think of this paper, and the arguments followed up in Hilda Bastian’s blog post?
- Can you find some more recent discussions of the paper?
- Does this make you more determined to do good science, or disillusioned about the state of science, or both?
- If you listened to the Hertz Podcast, what do you think about error detection in science (and can you find some of James’ more spectacular exposes of bad science)?
Please post your original thoughts below!