« Tantalizing unreliable new data on half doses and single doses | Main | The wish that won't quit »


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

Coleen B

Correct me if I am wrong but I think an assumption of random incidence would be wrong in these cases (perhaps a carry over from throw ones hands in the air epidemiological modeling) and stratification is needed. The only question is with which variables. We come across this frequently, with the main problem being the achievement of projectable sample size. Better to be partly right though....


CB: Yep, the whole enterprise of observational studies is to make adjustments and correct biases. Not adjusting at all is to allow the biases to fester. I think a good way of putting it is "we are confident the answer is partly right if our adjustments capture the key biases". Also, what did the researchers do to check their confirmation bias at the door?

The comments to this entry are closed.

Get new posts by email:
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.

See my Youtube and Flickr.


  • only in Big Data
Numbers Rule Your World:
Amazon - Barnes&Noble

Amazon - Barnes&Noble

Junk Charts Blog

Link to junkcharts

Graphics design by Amanda Lee

Next Events

Jan: 10 NYPL Data Science Careers Talk, New York, NY

Past Events

Aug: 15 NYPL Analytics Resume Review Workshop, New York, NY

Apr: 2 Data Visualization Seminar, Pasadena, CA

Mar: 30 ASA DataFest, New York, NY

See more here

Principal Analytics Prep

Link to Principal Analytics Prep