Welcome to European Tribune. It's gone a bit quiet around here these days, but it's still going.
Display:
JakeS:
The person who made that graph should have his data analysis license revoked.

An explanation of why would be a positive.

by afew (afew(a in a circle)eurotrib_dot_com) on Wed Jan 26th, 2011 at 02:23:46 AM EST
[ Parent ]
It's what physicists call a "shotgun plot." I haven't checked, but it would greatly surprise me if you could find any meaningful sort of correlation in that data, nevermind enough to justify curve fitting. And even if you have a good theoretical reason to want to fit a curve to this sort of data, the amount of scatter means that if you start fitting anything noticeably more complicated than a straight line, you'll end up fitting to the noise rather than any signal that might or might not be there.

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Wed Jan 26th, 2011 at 02:44:34 AM EST
[ Parent ]
The fit could be to a straigth line in a log-log plot...

Keynesianism is intellectually hard, as evidenced by the inability of many trained economists to get it - Paul Krugman
by Migeru (migeru at eurotrib dot com) on Wed Jan 26th, 2011 at 04:27:11 AM EST
[ Parent ]
Unlikely. Look at how the points are distributed. A straight line fitted to a log-log plot would have been much flatter, unless you enforced some sort of theoretically derived intercept, which is a bad sort of thing to do unless you know some tricks for doing it without screwing things up.

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Wed Jan 26th, 2011 at 07:03:08 AM EST
[ Parent ]
I was thinking more of a circle...
by asdf on Wed Jan 26th, 2011 at 07:56:22 AM EST
[ Parent ]

Display:

Occasional Series