Hacker News new | ask | show | jobs
by dalore 2879 days ago
Correlation doesn't mean cause. Otherwise you would draw conclusions like divorce rates in Maine affected by margarine consumption.

https://tylervigen.com/view_correlation?id=1703

People consuming fruit could be more healthy individuals in general, exercising and avoiding other types of food which might have more effect than consuming fruit. So be careful about drawing conclusions from correlations or epidemiological studies as you have to consider other possibilities.

1 comments

But I didn’t imply causation? I literally gave the same “could be skewed due to healthier lifestyle choices” in my comment...
Hey, you are on HN. The opportunity to shine the `correlation isn't causation` was too strong. /s