3 ms·
Actually, not even this is not true! The Central Limit Theorem applies only to samples which have finite mean and variance (see: http://en.wikipedia.org/wiki/C
by joe_the_user 17y ago
Actually, not even this is not true!
The Central Limit Theorem applies only to samples which have finite mean and variance (see: http://en.wikipedia.org/wiki/Central_limit_theorem http://en.wikipedia.org/wiki/Central_limit_theorem).
Take a distribution which has infinite variance or mean and you can wind-up instead with one of the fractal distributions which Mandlebrot studied.
see http://en.wikipedia.org/wiki/Stable_distribution http://en.wikipedia.org/wiki/Stable_distribution
Distributions with infinite variance or mean are more common than one might imagine. Some might argue the stock market would qualify.
- lutorm 17y agoGood point. Distributions with infinite variance are less common in physics, so that's normally not something we have to worry about.