Correlated Preferences

Update on my last post: I have encountered a link to this xkcd in four places now.

1. Once in my feed. I subscribe to xkcd precisely because it consistently provides cartoons that appeal to the quantitative nerd like me.
   
2. One in an e-mail from a friend of mine in the UChicago economics department. He thought I would appreciate it because of my teaching
   
3. Another time in a post at Economists Do It With Models, who talks at length about publication bias.
   
4. Finally, I came across this post from Russ Roberts, who reminds us to count the number of tests.
   

Each post is a different perspective on the cartoon. The paper on taking the 'con' out of econometrics is a classic on economic research methodology. It was nice to be reminded of this.

Of note, in her post on the topic, Jodi Beggs casually links to this other wonderful xkcd on being someone's statistically significant other.


With data like that, he should probably reject the null hypothesis.