A recent John Oliver segment (~20 minutes) provides a terrific look at how scientific studies need to be evaluated. Here’s the link: John Oliver Scientific Studies
His main points:
- Scientists are under pressure to publish and sometimes publish a study with a title to grab interest
- P-hacking can be done to find statistical significance/correlation that is bogus
- Reports are often distorted by the media to generate a buzz. Smelling farts does not prevent cancer (see image below).
- Some reports extrapolate findings in animals to humans without any studies and without mentioning that these were animal experiments; in addition, most treatments on lab mice do not work for humans.
- Many media reports do not mention whether the study was industry-funded or the size of the study. Reports with 10 or 20 people are more likely to lead to false conclusions
Related posts:
- Understanding science | gutsandgrowth NPR explains why the new study may be wrong
- Alan Alda (aka Hawkeye Pierce) on Communicating Science …
- Marketing to Doctors -Informative Satire | gutsandgrowth Another commentary on a terrific John Oliver segment
- Why I have always liked Arthur Caplan… | gutsandgrowth
gutsandgrowth 