Library:Science Fictions
| Science Fictions: How Fraud, Bias, Negligence, and Hype Undermine the Search for Truth | |
|---|---|
| Author | Stuart Ritchie |
| First published | 2020 |
| ISBN | 978-1-250-22269-5 |
| Role | Analysis |
| Domains | Science & Environment Health & Medicine |
| Tactics | Statistical & Evidentiary Distortion Fabrication |
Science Fictions: How Fraud, Bias, Negligence, and Hype Undermine the Search for Truth is a 2020 book by Stuart Ritchie.
Summary
Stuart Ritchie, a psychologist then at King's College London, published this book in 2020. It is a clear-eyed account of the replication crisis: the discovery that many published scientific findings, especially in psychology and medicine, cannot be reproduced when other researchers repeat the work. Ritchie writes as a defender of science who believes it can be fixed.
The crisis
Ritchie explains how the crisis came to light. In 2011 a respected psychologist published a paper claiming evidence that people could sense the future, using standard statistical methods; the result was implausible, but the methods were common, which raised the question of what else they might be producing. In 2015 a large collaborative project attempted to replicate one hundred published psychology studies and succeeded in fewer than half. Similar problems emerged in cancer biology, economics and other fields.
Four failings
The book organizes the problem into four failings. The first is fraud: researchers who invent data, including a Dutch social psychologist who fabricated results for dozens of papers and a surgeon whose claims of successful artificial windpipe transplants concealed patient deaths. The second is bias: the tendency to publish only positive results and to analyse data in many ways until something reaches statistical significance, a practice known as p-hacking. The third is negligence: simple errors, such as a spreadsheet mistake in an influential economics paper on government debt. The fourth is hype: exaggerating findings in papers, press releases and media coverage.
The incentives
Ritchie argues that the root cause is the system's incentives. Scientists are rewarded for publishing often, in prestigious journals, with exciting results. Journals prefer novel findings to replications. These pressures push even honest researchers toward practices that produce unreliable results.
Fixes
The book ends with reforms already underway: registering studies and their methods before collecting data, sharing data openly, publishing results regardless of outcome and changing how scientists are evaluated.
Relevance to Bullshitopedia
This is a crucial text for editors because it shows that a peer-reviewed study is not automatically reliable. It is essential for the Statistical & Evidentiary Distortion tactic and a reminder that the site should check whether a cited study has been replicated or retracted.