Re: RE: Re: Arab growth rate

["Susan Garlinghouse" <suendavid@worldnet.att.net>]

> Could you explain the standard deviation to me?  Height between
individuals
> at each age group?
>
> K.


The real statisticians in the group will cringe at this very rough
definition, but here goes, using the data from this arab height project.

Let's say that at one year of age, the *average* horse in the study stood 54
inches at the withers, and the standard deviation was one inch (the standard
deviation value is derived through an algebraic formula).  This means that
about 67% of all the horses in the study stood the average height (54
inches), plus or minus the standard deviation (1 inch).  Thus, 67% of the
horses in the study were between 53 and 55 inches.  If you use two "standard
deviations of the mean", then that means that 98% of the horses in the study
stood the average height (still 54 inches), plus or minus *twice* the
standard deviation (or 2 inches).  Thus, 98% of the horses stood between 52
and 56 inches.

The average values can change, and standard deviations change, based on how
many subjects in your sample group, what you're measuring, etc etc etc.  The
percentages of 67% and 98% stay the same always, those are just part of the
definition of standard deviation.  Also, if you have a big standard
deviation, it may be telling you your results aren't very precise.  The
smaller your standard deviation, generally, the more accurate you are in
whatever it is you're hopefully trying to demonstrate with your study.
There are other statistical "tests" that tell you about accuracy and
conclusions and whatnot as well.  These are gross generalizations, but you
get the idea.

If this explanation has you hopelessly confused, I can try to give you a
better example.  The real stats people in the group can stop quailing in
horror now. <g>