Display:
The quoted article says European Tribune - Too many lefties here?
left-handers did well enough in "expressive English", but were less competent than right-handers in "social/emotional skills, gross and fine motor skills and receptive English skills"
They omit to mention correlations with logical, numerical and spatial reasoning, which is a bit puzzling since they are part of standard psychometric evaluations (such as IQ)
research shows that left-handers are more likely to suffer from language disorders, autism, dyslexia, "schizotypal behaviour patterns", seizures and post-traumatic stress disorder
That's a bunch of other correlates to look for.

However, correlation is not transitive. That A is correlated with B and B correlated with C tells you very little about the correlation of A with C. (This is an exercise in spherical trigonometry :-)

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith

by Migeru (migeru at eurotrib dot com) on Tue Feb 24th, 2009 at 10:54:16 AM EST
[ Parent ]
mig:


However, correlation is not transitive. That A is correlated with B and B correlated with C tells you very little about the correlation of A with C. (This is an exercise in spherical trigonometry :-)

If A,B,C are three series of numbers of the same length arranged in N triplets, I see a geometrical explanation why correlation should be transitive (it is no trigonometric however).

First high correlation implies points are close to a plane around X axis (and not Y,Z axis), second high correlation implies they are close to a plane around Y axis (and not X, Z axis). The intersection of the two plane is a line unless they are one and the same plane. If it is a line, the third correlation is also pretty high. In order to have the two planes intersect at a very narrow angle (to degrade the 3rd correlation), you need very different orders of magnitude between the standard deviations of the first and second series...

ah OK I get it, this is were you have spherical trigonometry to express the angle in terms of the standard deviation. Forget it.

Pierre

by Pierre on Tue Feb 24th, 2009 at 11:09:47 AM EST
[ Parent ]
So if you impose additional constraints like that all std dev are of comparable magnitude, you can have some sort of "transitivity of correlation" in that restricted universe of observations.

Pierre
by Pierre on Tue Feb 24th, 2009 at 11:11:11 AM EST
[ Parent ]
You can have 50% percent correlation between A and B and between B and C and 50% anticorrelation between A and C.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Tue Feb 24th, 2009 at 11:13:18 AM EST
[ Parent ]
doh ! anticorrelation is just a special case of correlation (in the meaning that one is informative of the other)

Pierre
by Pierre on Tue Feb 24th, 2009 at 11:47:15 AM EST
[ Parent ]
Well, okay, as long as the correlation between A and each of B and C is less than 70%, B and C can be uncorrelated.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Tue Feb 24th, 2009 at 11:50:16 AM EST
[ Parent ]
Yes, you need strong correlations in the first two combinations and you will have less of it in the third, it is impossible to make the two planes orthogonal with true correlations in the first two couples.

Pierre
by Pierre on Tue Feb 24th, 2009 at 11:58:16 AM EST
[ Parent ]
do you happen to actually have the formula of the angle at hand ?

Pierre
by Pierre on Tue Feb 24th, 2009 at 11:59:14 AM EST
[ Parent ]
The correlation coefficient is the cosine of the angle. Is that what you mean? You then apply spherical trigonometry to the triangle. Assuming all three variables are on a plane, you get upper and lower bounds (from cosine of sum and cosine of difference formulas).

This all follows from considering that "covariance" is an "inner product". The Cauchy-Schwarz inequality gives you "covariance is less than the product of standard deviations".

Does that answer your question?

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith

by Migeru (migeru at eurotrib dot com) on Tue Feb 24th, 2009 at 12:10:16 PM EST
[ Parent ]
yes

Pierre
by Pierre on Tue Feb 24th, 2009 at 12:42:15 PM EST
[ Parent ]

Display:
Login
. Make a new account
. Reset password
Occasional Series