Coefficient of Correlation

[Anonymous]

Suppose that each time we sit for exams, for each of us there is a score of 0 to 10 of how good we "really" are (of course we will never know it). Call it x. Then there is of course, our official test result score y, of 0 to 10 also. Now, what would be your best guess for the correlation between x and y?
I pick 0.75

[1695814]

I pick .6

Get it? "6"?

cuteness aside, that's what I'd pick anyway.

[Wigmeister General]

0, 'cause I'm really, really bad.

[Mulan]

mine is .95. I always passed when I knew enough and failed when I didn't. I got high scores (not that we get scores on the upper exams in the CAS when we pass) when I knew a lot and low low low ones when I didn't have a clue.

One time I passed when I didn't think I had a clue and got a 7!! Very weird. Another time I got a 5 when I thought I'd probably passed.

[foghorn]

I always passed when I knew enough and failed when I didn't.
One time I passed when I didn't think I had a clue and got a 7!! Very weird. Another time I got a 5 when I thought I'd probably passed.

I always passed, except when I didn't.

[Wigmeister General]

Using computability theory lingo, my performance was a partially decidable question because I always failed when I knew I failed.

[Phil]

I always score within 3 standard deviations of my expected score x.