CHAPTER I

INTRODUCTION

A. Background
In our daily life, we know that statistic is a study about how to make a
conclusion based on the incomplete data. In this case, incomplete data means a
sample data and data those are taken by certain methods from a big population
even great population. The existence of statistic in every side of our life is very
important for developing those sides, such us government, sport, farm, business
and economy, social, politics, law, education, medical, etc. All of them need
statistics, descriptively or inferentially.
The word Statistic is taken from Latin language, status that relate to a
country or a politic unity. Its mean that statistic has important role in service
administration of a country, such us the arranging of citizen information
accelerate the tax payment, and the citizen mobilization for the war forces power.
Statistic relate to the scientific methods for collect, organize, summarize,
analyze data, and it is include with making a valid conclusion, and also making a
rational decision based on the certain analytic.
For deepen our understanding in statistic comprehensively, someone must
master or know some part of mathematics, such us calculus, probability theory,
matrix algebra, and real analytic.
The probability theory is very useful to discuss about Poisson distribution and
Multinomial distribution. Both of that distribution relate with probability, because
both of them describe about the probability of an event in our daily life happened
on the certain time with the certain frequency.
Based on the background above, the writer feels interest to write a paper
about The Poisson and Multinomial Distribution.

1
B. Problem Statement
Problem statements of this paper are:
1. What is the definition of Poisson distribution?
2. What is the relation between Binomial and Poisson distributions?
3. What is the definition of Multinomial distribution?

C. Objective of Writing
The objectives of this paper are:
1. To increase our knowledge about mathematics, especially about Poisson and
multinomial distribution.
2. To give some theory about the relation between binomial and Poisson
distribution and also Poisson and normal distributions.

D. Significance
The significance that can be obtained by the readers from this paper is:
1. Understand the definition of Poisson distribution.
2. Understand about the relation between Binomial and Poisson distributions.
3. Understand about the definition of Multinomial distribution.

2
 CHAPTER II
DISCUSSION

A. The Definition of Poisson distribution

Poisson Distribution is a discrete distribution to expect

probability, that the certain output probability will appear X times
in a unit that has been standardized by average velocity of the
appearance of an event in every units is constant ().

Poisson Distribution is introduced by S.D Poisson in his book about the

applying of probability theory in a law indictment problem, criminal court
session, etc. This book is published on 1839, titled Recherches sur la probabilit
des jugments en matire criminelle et en matire civile.
Experiments yielding numerical values of a random variable X, the number of
outcomes occurring during a given time interval or in a specified region, are
called Poisson Experiments. The given time interval maybe of any length, such
as a minute, a day, a week, a month, or even a year. Hence a Poisson experiment
can generate observations for the random variable X representing the number of
telephone call per hour received by an office, the number of days school is closed
due to snow during the winter, or the number of postponed games due to rain
during a baseball season. The specified region could be a line segment, an area, a
volume, or perhaps a piece of material. In such instances X might represent the
number of field mice per acre, the number of bacteria in a given culture, or the
number of typing errors per page. A Poisson experiment is derived from the
Poisson process and possesses the following properties:
1. Properties of Poisson Process

3
a) The number of outcomes occurring in one time interval or specified
region is independent of the number that occurs in any other disjoint
time interval or region of space. In this way we say that the Poisson
process has no memory.
b) The probability that a single outcome will occur during a very short time
interval or in a small region and does not depend on the numbers of
outcomes occurring outside this time interval or region.
c) The probability that more than one outcome will occur in such a short
time interval or fall in such a small region is negligible.

The number X of outcomes occurring during a Poisson experiment is called

a Poisson random variable, and its probability distribution is called the
Poisson distribution. The mean number of outcomes is computed from
t , where t is the specific time, distance, area, or volume of

interest. Since its probabilities depend on , the rate of occurrence of

outcomes, we shall denote them by the symbol p ( x; t ) . The derivation of
the formula for p ( x; t ) , based on the three properties of a Poisson
process listed above, is beyond the scope of this book. The following
concept is used for computing Poisson Probabilities.

The probability distribution of the Poisson random variable X,

representing the number of outcomes occurring in a given time
interval or specified region denoted by t, is

Like the binomial distribution, the Poisson distribution is used for quality
control, quality assurance, and acceptance sampling. In addition, certain

4
 important continuous distributions used in reliability theory and queuing
theory depend on the Poisson process.

THEOREM
Both the mean and variance of the Poisson distribution are .

Proof:
Let t

e x
e x
e x 1
E( X ) x x
x 0 x! x 1 x! x 1 ( x 1)!

Since the summation in the last term above is the total probability of a
Poisson random variable with mean which can be easily seen by letting
y x 1, it equals to 1. Therefore, E ( X ) . To calculate the variance of
X, note that

e x
e x2
E[ X ( X 1)] x ( x 1) 2
x 0 x! x2 ( x 2 )!

Again, letting y x 2 , the summation in the last term above is the total
probability of aPoisson random variable with mean . Hence we obtain
2 E ( X 2 ) [ E ( X )]2 E[ X ( X 1)] E ( X ) [ E ( X )]2 2 2 t

2. Nature of the Poisson Probability Function

Like so many discrete and continuous distributions, the form of the
distribution becomes more and more symmetric, even bell shaped as the mean
grows large.

B. Some Properties of the Poisson distribution

Some properties of the Poisson distribution are listed in the following table:

5
 Mean
Variance 2
Standard Deviation
Moment coefficient of skew ness 3 1
Moment coefficient of kurtosis 4 3 1

C. Relation between Binomial and Poisson Distributions

In the binomial distribution, if N is large while the probability p of occurrence
of an event is close to zero so that q (1 p ) is close to 1, the event is called a
rare event. In practice we shall consider an event as rare if the number of trials is
at least 50(N 50) while Np is less than 5. In such cases the binomial distribution
is very closely approximated by the Poisson distribution with Np.
Since there is a relation between the binomial and normal distributions, it
follows that there also relation between the Poisson and normal distributions. It
can in fact be shown that the Poisson distribution approaches a normal
distribution with standardized variable X as increases indefinitely.

D. The Definition of Multinomial Distribution

The binomial experiment becomes a multinomial experiment if we let each
trial have more than 2 possible outcomes. Hence the classification of a
manufacture product as being light, heavy, or acceptable and recording of
accidents at a certain intersection t the day of the week constitute multinomial
experiments. The drawing of a card from a deck with replacement is also a
multinomial experiment if the 4 suits are the outcomes of interest.
In general, if a given trial can result in any one of k possible outcomes E1, E2,
, Ek with probabilities p1, p2, pk, then the multinomial distribution will give the
probability that E1 occurs x1 times, E2 occurs x2 times, , and Ek occurs xk times
in n independent trials, where
x1 + x2 ++xk = n

6
We shall denote this joint probability distribution by
f(x1,x2,,xk: p1,p2,pk,n)
Clearly, p1+p2++pk =1, since the result of each trial must be one of the k
possible outcomes.

General Form for Multinomial Probabilities

To derive the general formula, we proceed as in the binomial case. Since the
trials are independent, any specified order yielding x1 c\outcomes for e1, x2 for
E2, .xk for Ek will occur with probability p1x1 p2x2pkxk. The total number of
orders yielding similar outcomes for the n trials is equal to the number of
partitions of n items into k groups with x1 in the first group; x2 in the second
group, ; and xk in the kth group.

This can be done in

n n!

x ,
1 2 x ,...., x k x 1 ! x 2 !...x k !

Ways. Since all the partitions are mutually exclusive and occur with equal
probability, we obtain the multinomial distribution by multiplying the probability
for a specified order by the total number of partitions.

If a given trial can result in the k outcomes E1, E2,Ek with probabilities p1,

p2,.., pk, then the probability distribution of the random variables x1, x2,, xk,

representing the number of occurrences for E1, E2, .,Ek in n independent trials

is
f(x1,x2,,xk;p1,p2,pk,n)= p1x1 p2x2pkxk

with

The multinomial distribution derives its name from the fact that the terms of
the multinomial expansion of (p1+p2++pk)n correspond to all the possible
7
values of f(x1,x2,,xk;p1,p2,pk,n.
E. Fitting Theoretical Distributions to Sample Frequency Distribution
When one has some indication of the distribution of a population by
probabilistic reasoning or otherwise, it is often possible to fit such theoretical
distribution (also called model or expected distributions) to frequency
distributions obtained from a sample of the population. The method used in
general consists of employing the men and standard deviation of the sample to
estimate the mean and sample of the population. In order to test the goodness of
fit of the theoretical distributions, use is made of the chi-square test. And in
attempting to determine whether a normal distribution represents a good fit for
given data, it is convenient to use normal curve graph paper or probability graph
paper as it is sometimes called.

F. Solved Problems
1. The Poisson distribution
1) Ten percent of the tools produced in a certain manufacturing process
turn out to be defective. The probability that in a sample of 10 tools
chosen at random, exactly two will be defective by using
(a) binomial distribution

8
 (b) the Poisson approximation to the binomial distribution
Solution:
Probability of a defective tool = p =0.1
(a) Pr {2 defective tools in 1} = 10C2(0.1)2(0.9)8 = 0.1937 or 0.19
(b) Np 10(0.1) 1
x e 1 e 1 e 1
2
1
Pr {2 defective tools in 1} 0.18
x! 2! 2 2e
In general the approximation is good if p 0.1 and Np 5

2) During a laboratory experiment the average number of radioactive

particles passing through a counter in I millisecond is 4. What is the
probability that 6 particles enter the counter in a given millisecond?
Solution:
Using the Poisson distribution with x 6 and t 4
e 4 4 6 6 5
p (6;4) p ( x;4) p ( x;4) 0.8893 0.7851 0.1042
6! x 0 x 0

2. The Multinomial distribution

1) A box contains 5 red balls, 4 white balls, 3 blue balls. A ball is selected
at random from the box, its color is noted, and then the ball is replaced.
Find the probabilities that out of 6 balls selected in this manner 3 are
red, 2 are white, and 1 is blue.
Solution:
5
Pt{red at any drawing} =
12
4
Pr{white at any drawing} =
12
3
Pr{blue at any drawing} =
12
Then 3 2 1
6! 5 4 3 625

3! 2!1! 12 12 12 5184

9
 Pr{3 are red, 2 are white, 1 is blue}

2) The complexity of arrivals and departures into an airport are such that
computer simulation is often used to model the ideal conditions. For a
certain airport containing three runways it is known that in the ideal
setting following are probabilities that the individual runways are
accessed by a randomly arriving commercial jet:

Runway 1 : p1 = 2
9

Runway 2 : p2 = 1
6

Runway 3 : p3 = 11
18
What is the probability that 6 randomly arriving airplanes are distributed
in the following fashion?
Runway 1 : 2 airplanes
Runway 2 : 1 airplane
Runway 3 : 3 airplanes
Solution:
Using the multinomial distribution, we have
2 1 3
2 1 11 6 2 1 11
f 2,1,3; , , ,6 0.1127
9 6 18 2,1,3 9 6 18

3. Fitting of Data by Theoretical distribution

No. of Accidents, X No. of Days, f
0 21
1 18
2 7
3 3
4 1

10
 The table above shows the number of days f in a 50 day period during which
X automobile accidents occurred in a city. Fit a Poisson distribution to the
data.
Solution:
Mean number of accidents is

fX 21 0 181 7 2 3 3 1 4 45 0.9
f 50 50

Then, according to the Poisson distribution,

Pr{X accidents}
0.9 Xe 0.9
X!
In table below are listed the probabilities for 0, 1, 2, 3, and accidents as
obtained from this Poisson distribution, as well as the expected or theoretical
number of days during which X accidents take place (obtained by multiplying
the respective probabilities by 50). For convenience of comparison, the fourth
column giving the actual number of days has been repeated.
No. of Expected Actual
Pr {X accidents}
Accidents, X No. of Days No. of Days
0 0.4066 20.33 or 20 21
1 0.3659 18.30 or 18 18
2 0.1647 8.24 or 8 7
3 0.0494 2.47 or 2 3
4 0.0111 0.56 or 1 1

CHAPTER III
CONCLUSION AND DISCUSSION

A. Conclusion
Based on the problem in the problem statement, the writer can conclude that:
1. Poisson Distribution is a discrete distribution to expect probability, that the
certain output probability will appear X times in a unit that has been

11
 standardized by average velocity of the appearance of an event in every
units is constant ( ).
The probability distribution of the Poisson random variable X, representing
the number of outcomes occurring in a given time interval or specified
region denoted by t, is
e t ( t ) x
p ( x; t ) , x 0,1,2,3,.....,
x!

2. In the binomial distribution, if N is large while the probability p of

occurrence of an event is close to zero so that q (1 p ) is close to 1, the
event is called a rare event. In practice we shall consider an event as rare if
the number of trials is at least 50(N 50) while Np is less than 5. In such
cases the binomial distribution is very closely approximated by the Poisson
distribution with Np.
Since there is a relation between the binomial and normal distributions, it
follows that there also relation between the Poisson and normal
distributions. It can in fact be shown that the Poisson distribution
approaches a normal distribution with standardized variable X as
increases indefinitely.

3. The binomial experiment becomes a multinomial experiment if we let each

trial have more than 2 possible outcomes. The formula of multinomial
distributions is
n
f(x1,x2,,xk;p1,p2,pk,n)= p1x1 p2x2pkxk
x1 , x 2 ,...., x k

B. Suggestion
In this paper, the writer wants to suggest in order that this paper can increase
our knowledge about mathematics especially for Poisson and Multinomial

12
 distribution. Besides, the writer hopes in order that this paper can help learning
and teaching process. But, as a common people, the writer recognize that there are
still so many mistakes in this paper, so the writer needs apology and some
suggestions from the readers for the perfection of this paper.

REFERENCES

Nugroho, Sigit. 2008. Dasar-Dasar Metode Statistika. Jakarta: Gramedia

Widiasarana Indonesia.
Spiegel,Murray R.1981.SchaumsOutline of Theory and Problems of Statistics in SI
units. Singapore: Mc Graw-Hill.Inc.
Tiro, Muhammad Arif et al. 2008. Pengantar Teori Peluang. Makassar: Andira
Publisher.
Walpole, E. Ronald et al. 2007. Probability and Statistics for Engineers and Scientist.
United State of America: Pearson Education.

13
14