Tuesday, January 23, 2007

Follow Your Passions at a Distance

Robin Hanson:

Don't research a topic because it is popular; expose yourself to many popular topics and you will fall in love with a few.  Fall in love with a dozen topics; too few and you'll linger too long after they fail.

[T]he best way to start researching is to take a good research paper and really understand it.  Then make a minor change, such as an assumption or data analysis technique.  Then make bigger changes. 

Another problem with passion is bias.  Especially in social science, people pick topics in order to convince the world of their one true answer.  But it is healthier to focus on questions, not answers.  By picking an answer before you've really studied a topic, not only are you more likely to be wrong, but more important, you could miss interesting new angles.

No comments: