Hacker News new | ask | show | jobs
by scottlocklin 2229 days ago
I'd be surprised if those studies modeled the survivor bias. For example; this link was a standard reference:

http://www.phred.org/~alex/kenkifer/www.kenkifer.com/bikepag...

The author of it died in a bike crash.

1 comments

The studies looked at predicted life expectancy when you consider the various health benefits of cycling vs increased severity of road accidents and expose to pollution. They don’t need to model for survivor bias, they weren’t tracking individuals, but deriving specific population stats from larger general populations stats.

> The author of it died in a bike crash.

I’m not sure what your point is?

People die that’s not a surprise, the fact the author died in a bike crash doesn’t invalidate their work.

Doctors die of smoking, some scientists studying nutrition are fat. None of that changes the validity of their work.