rdharper

Hmmm... you are joking right? The x-axis is performance (of whatever) going from worst to best (left to right) of that parameter. The center of the distribution is where most will sit... be it calories eaten per day, IQ, income, time in the 50 meter freestyle... etc. Most of those parameters will be "normal distributions" because there are equal numbers on both sides of the mean forming a "bell shaped" curve. There are other "natural" distributions. Such as the Weibull distribution which measures failure rate of a product. Different shape.

So 3-sigma on the left of a normal distribution (such as racing skills) would be something you wouldn't want to watch.

Overlaying the distribution for the men and women would make the point more clearly (that's why graphs are valuable). I could dig up something to make the point if you are interested.

aviator

Sorry harper I was a little fried when I made the post I meant the right side of the curve vis a vie women. I corrected the original post to reflect that... LOL

So to put this in a way I can sort of grasp w/out it slipping out of my hands... when a teacher uses the "bell curve" and the class average is say 70% or a nice solid B before implimentation that nice B will actually drop to a C which means those who were still passing [C] before will now find them selves with a d or worse... failling in other words... which would argueably sque the class grade point average to the detriment of all concerned... except for the brainiacs who were aceing it anyway... right?
aviator

rdharper

Right. And it can work the other way... and probably often does. Depends on the "quality" of the class. I believe it used to be called "norming".

This can be against a national distribution, or state, or county, or against the professors own past class "standards".

There are theorems and formulas to make sure the distribution is really a "normal distribution." Sample size requirements etc. And ways of calculating probability of accuracy etc.

Its an extremely useful branch of math... used to very good purposes in modern manufacturing. But it also is misused constantly by so-called pollsters for spin purposes.

Anyway... you got it.

easttntoy

6 sigma is anything less than 6 standard deviations away from the mean is acceptable. .
3 sigma " " " " 3 " " " " " "
Atleast that's what I remember from Operations Managment.

Toyota uses 6 sigma quality control in their manufacturing.

4bytruckin

sorry it has to be said . trucks before _ucks man. u don't leave your truck out to dry like that

Kaydon

"Bros before hoes, trucks before both"