« Another day, another fake data scandal... This time, it's Twitter | Main | A/B Testing Primer and the DEED framework »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Ken

One of the things that behavioural researchers should start doing is to write protocols, and those would require registration for publishing in journals, the same as is done for medical research. Then it would be obvious how many outcomes they actually used, which hopefully would make them limit themselves to one. Then they would only be troubled by odd results, which could be checked by replication.

Kaiser

Ken: there is indeed a debate about pre-registering studies. The concern may be that there are sometimes things you discover during the studies that may be real effects that weren't anticipated. So there is pushback.

The comments to this entry are closed.

CONTACT KAISER



Link to Kaiser Fung Consulting Inquiry

Kaiser Fung. Business analytics and data visualization expert. Author and Speaker.
Visit my website. Follow my Twitter. See my articles at Daily Beast, 538, HBR, Wired.
Get new posts by email:

Search ...

  • only on this blog

MY BOOKS



Numbers Rule Your World:

Amazon - Barnes&Noble



Numbersense:

Amazon - Barnes&Noble

Junk Charts Blog



Link to junkcharts

Graphics design by Amanda Lee

Community