Central Limit Theorem

Table of content
History
Introduction
Definition
Mean and Standard deviation
Probability density function
Applications

HISTORY
The actual term "central
limit theorem" (in
German: "zentraler
Grenzwertsatz") was first
used by George Plya in
1920 in the title of a
paper. Plya referred to
the theorem as "central"
due to its importance in
probability theory.
The first version of this
theorem was postulated
by the French-born
mathematician Abraham
de Moivre in a
remarkable article
published in 1733.
Laplace expanded De Moivre's finding by
approximating the binomial distribution with
the normal distribution. But as with De Moivre,
Laplace's finding received little attention in his
own time.
Russian mathematician
Aleksandr Lyapunov
Intro to CLT
We will begin by drawing samples
from a population that has a
uniform distribution.
The scores in the population
range from 1 to 5. The mean of
the population is 3.0 and the
standard deviation of the
population is 1.41 [=3.0;
=1.41].
Each level of the variable had the
same probability of occurring.
The population distribution is
presented left.
Six successive samples
Lets start small and draw 6 samples randomly from the uniform
population presented on the previous page.
Each sample has only 5 subjects (N=5). The central limit
theorem requires that all samples must have the same sample
size. The distribution of scores in each of these samples is
presented below.
The sample mean is shown in blue.

Distribution of sample means
25 additional sample means
Frequency Distribution

Let's see what happens when we randomly select 100
samples from the same population.
Compare and Contrast
Remember these sample means were taken from a uniform
distribution!
Compare and contrast this to the two previous frequency
distributions
We can see that this distribution of sample
means is even closer to a normal distribution
than the distributions calculated with 6 and 25
samples.
The mean of this distribution is quite close to
the population mean of 3
The standard deviation is smaller than the
population standard deviation of 1.41
Compare and Contrast
Infinite samples
Let's see what would happen if we were to
randomly select an "infinite" number of samples
from the same population.
The sampling distribution of the mean is
normally distributed.
The mean of the normal distribution is the same
as the population mean.
The standard deviation of the normal
distribution is equal to the population standard
deviation divided by the square root of the
sample size.
Summary so far
Increasing the sample size.
Increasing the sample size from 5 to 25
What do you notice
Increase n to 100
Central Limit Theorem
The central limit theorem states that when an
infinite number of successive random
samples are taken from a population, the
sampling distribution of the means of those
samples will become approximately normally
distributed with mean and standard deviation
/ N as the sample size (N) becomes larger,
irrespective of the shape of the population
distribution.
Sample Size
Conclusion of comparisons
With " infinite" numbers of successive random samples, the
sampling distributions all have a normal distribution with a
mean of 3.0, which is equal to the population mean (= 3.0).
As the sample sizes increase, the variability of each sampling
distribution decreases. The range of the sampling distribution
is smaller than the range of the original population.
The standard deviation of each sampling distribution is equal
to /N (where N is the size of the sample drawn from the population).
Taken together, these distributions suggest that the sample
mean provides a good estimate of and that errors in our
estimates (indicated by the variability of scores in the
distribution) decrease as the size of the samples we draw from
the population increases.
Mean and Standard deviation
Let x denote the mean of a random sample
of size n from a population having mean m
and standard deviation s. Let
m
x
= mean value of x and
s
x
= the standard deviation of x
then
sx = s/ n
m
x
= m




When the population distribution is normal
so is the distribution of x for any n.

For large n, the distribution of x is
approximately normal regardless of the
population distribution
Rule of thumb: n > 30 is large

Starting with a Poisson Distribution
Starting with a Normal Distribution
Probability density function
A function of a continuous random variable,
whose integral across an interval gives the
probability that the value of the variable lies
within the same interval.

for normal distribution curve
Where
F(x)= probability at a given point (x)
sigma = standard deviation of samples
u = mean of samples(population)


Integration of prob.
density function from
-infinity to x gives us
under area of the curve as shown
Application
This theorem says that a normal distribution
arises regardless of the initial distribution.
For e.g.
Our population has a skewed distribution,
which occurs when we examine things such as
incomes or peoples weights, a sampling
distribution for a sample with a sufficiently
large sample size will be normal. (of either
sample mean or sample proportions)
Starting with a Uniform distribution
Starting with a Poisson Distribution
Starting with a Normal Distribution
In Real Word

Involving hypothesis testing or
confidence intervals, make some
assumptions concerning the
population that the data was obtained
from.
x p(x) (x - m)
2
p(x)(x -m)
2

1 .45 1.21 .545
-1 .55 .81 .446
Variance(total)
.991
Suppose that a video game hub has a slot
machine such that player can play and
either double their bet or lose their bet. If
there is a 45% chance of winning . Then
find the probability that the player does not
lose any money ?
the mean is

1(.45) + (-1)(.55) = -(.1 ) = m
If a player throw 100 rupee into the slot machine
then player expect to average a loss of ten rupees
with a standard deviation of

first compute the z-score.
We have

0 - (-.1)
z = .0995 = 1.




P(z > 1.01) = 1 - P(z < 1.01)
= 1 - .8438 = .1562
There is about a 16% chance that
the player will not lose.
BIBLIOGRAPHY
http://www.wikipedia.org
http://www.Slideshare.net
Probability and Statistics




THANK YOU