home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Loadstar 22
/
022.d81
/
t.math reflectio
< prev
next >
Wrap
Text File
|
2022-08-26
|
2KB
|
103 lines
The Central Limit Theorem
If samples of n numbers are
repeatedly chosen in a random fashion
from a given population with mean M
and variance
2
s
and if from each sample the mean is
formed (that is, the numbers in the
sample are added and divided by n),
then these means form a new
distribution.
There is a theorem in mathematical
statistics which states that the new
distribution of sample means will be
approximately normal (bell-shaped)
regardless of the shape of the
original distribution from which the
samples were drawn. Furthermore, the
distribution of sample means will have
the same mean, M, as the original
distribution, and have variance equal
to:
2
s /n.
As N (the sample size) increases, the
distribution will become increasingly
normal.
This result follows from a more
general theorem known as the central
limit theorem. Most of the results in
the field of inferential statistics
rest in some way on this theorem.
Our application will be to take the
uniform distribution generated by the
C-64's random number generator, take
averages of N random numbers and watch
the distribution of those averages as
they are plotted. Hopefully, instead
of looking uniform, this distribution
will look normal.
When you run the program,
choose the central limit theorem
option with, say, 25 subintervals,
averaging 10 random numbers, and 300
iterations. Then try other
combinations. Notice as you average
larger and larger samples, the
distribution becomes less spread out.
This is because the variance is being
reduced as the theorem claims it will.
---------< end of article >-----------