On Friday 15 July 2005 07:11 pm, Andrew Lentvorski wrote: > John H. Robinson, IV wrote: > > http://sd.znet.com/~schester/fallbrook/home_prices/ > > Scroll down a bit, to find the Distribution Of Prices For Home Sales In > > Fallbrook In 1995, 1997 and January-June 1998 topic. > > > > I think ``Normal Distribution'' is far less common than assumed. The > > bell curve is an ideal, and like most ideals, it is very hard to > > achieve. > > John, > > I know you know better than this. Are you purposely choosing small > sample sizes to be pedantic? In addition, this distribution *by > definition* is asymmetric--a house price can't be below zero but has no > upper bound.
What Andrew said. Moreover you are looking at two distributions: one for most of us and another for people for whom price is not an issue. Cut off the tops and the distributions look remarkably like the Poisson distribution one would naively expect for this situation. Despite what you, John, think, the "Normal Distribution" is actually far more common than assumed since it results from all kinds of different and not obviously similar processes including those for which the individual random variates have other than normal distributions. The amazing and counter- intuitive thing about the central limit theorem is that no matter what the shape of the original distribution, the sampling distribution of the mean approaches a normal distribution. It is far easier to achieve than is commonly understood. Take a look at the Central Limit Theorem, e.g. http://mathworld.wolfram.com/CentralLimitTheorem.html http://en.wikipedia.org/wiki/Central_limit_theorem http://www.ruf.rice.edu/~lane/hyperstat/A14043.html boblq Math is a demanding mistress so if you love her be prepared to accept her wishes. > I am still looking for aggregate national statistics for SAT's. I > welcome any help. Good luck. I would love to see > -a -- [email protected] http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
