Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Say it with me:

Correlation ≠ Causality



I'll add to this. When A and B are correlated, there are only three possibilities regarding causality (which may work in combination):

1. A causes B

2. B causes A

3. C causes both A and B

If we apply this where A is majoring in computer science and B is high earnings, which answers are true? #2 isn't, obviously. #1 is plausible, since perhaps employers are impressed by a computer science degree. But I think #3 is the most significant, where C is a strong desire to be a software professional. Very few people complete a degree in computer science unless they really want to work with software.

We all know English majors, psych majors, and history majors who became high-earning software engineers. Those are the people who had C (the desire to work with software) but not A (computer science degree). C is the thing that caused them to have B.


Assuming that you have a sufficiently large sample set. :)

The key here, of course, is that we don't know what C is. It could be that high IQ causes high wealth, regardless of degree. High IQ could also be a prerequisite of completing a CS degree, so necessarily the two correlate.

My point is that you know what A is, and you know what B is, but without measurement, you can't know what C is.


Correct. However there are people who will say there is no correlation between degree and salary. This proves otherwise.


Who says that?


Who says that?

The original post has some quotes that say that, e.g.: "A degree in English is just as valuable as a degree in Biology — it teaches you critical thinking!"


That quote asserts a lack of causality, not a lack of causation.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: