Normal Distribution

(For M.B.A. I Semester)

by Prof. Nitin Karoulia

9033099006
 Normal Distribution
PROBABILITY DISTRIBUTIONS OF CONTINUOUS VARIABLE
The Binomial and Poisson distributions are distributions of
variables which are discrete i.e. the variable takes only certain
specified values or to be more specific, only integral values.
However, in real life, there are many situations where a
variable takes all possible values in its range. Such variables
are called continuous variables. Some of these are: life of an
individual or any item like electric bulb, height and weight of
a person, temperature, rainfall, distance, etc. For practical
purposes, however, these are measured in a discrete manner.
For example, life could be expressed in years, height in meters
and centimeters, rainfall in centimeters, etc. Even marks
obtained, in a question of, say 10 marks, is a continuous
variable but for convenience sake, the examiner rounds in full
 Normal Distribution
Marks, some examiners round up to ½ mark, and some even go
to the extent of rounding up to ¼ mark.
For continuous variable, the probability distribution is called
probability density function (abbreviated as p.d.f.) because it
is defined for every point in the range and not for certain
values like Binomial and Poisson distributions. It is
represented by f(x) where f(x) dx is the probability of a
variable taking a value from x to x + dx as shown by shaded
area in Fig. 7.10 below.
 Normal Distribution
The shaded area under the curve is the area of the rectangle
whose height is f(x) and base is dx, and is thus equal to f(x) dx.
Incidentally the probability density function is also known as
frequency function.
As mentioned in Chapter 5, Section 5.7, the probability that a
variable takes a value between a to b is given by calculating
the area under the curve which means integrating the
function from the lower value to the higher value.
b
P (a ≤ x ≤ b ) = ∫ f (x) dx
a
 Normal Distribution
Uniform/ Rectangular Distribution
This is the simplest type of distribution of a continuous variable.
Its mathematical form is
f(x) = 1 a≤x≤b (7.9)
(b – a)
and the graphical form is as shown below:
 Normal Distribution
It may be noted that the variable has a uniform value = 1/(b –a)
throughout its range from a to b.
Thus the probability that the variable takes any particular value is
the same throughout in its range. Because of this property, this
distribution is highly useful in simulation studies. Further, due
to its shape, it is also known as the Rectangular Distributions.
Mean = a + b
2
Variance = ( b – a) 2
12
If the values a and b are not known, and have to be estimated
from a sample of values x1, x2, x3,……xn, then their estimates
are found as follows:
 Normal Distribution
a=x–s√3

b = x + s √3
where, s is the standard deviation of the sample.
If a = 0 , and b = 1, the uniform distribution is called a standard
uniform distribution.
One of the most important application of Uniform distribution
is in generating random numbers.
 Normal Distribution
Normal Distribution
The normal law of error stands out in the mankind as one of
the broadest generalizations of natural philosophy. It serves
as the guiding instrument in researches in the physical and
social sciences and in medicine, agriculture and engineering.
It is an indispensable tool for the analysis and the
interpretation of the basic data obtained by observation and
experience. - W.J. Youden

Some of the situations where normal distribution is applicable

are:
• Life of items subjected to wear and tear like bulbs, batteries,
currency notes, tyres, etc.
 Normal Distribution
• Length and Diameter of certain manufactured products like
pipes, screws and discs.
• Breaking strength of textile thread, bursting/tensile strength
of paper and plastic bags, metallic wire, etc.
• Weekly sales of an item in a store.
• Daily rate of return for a stock or market index like SENSEX
(BSE).
• Height and Weight of children at birth.
• Aggregate marks obtained by students in an examination.
• Yield of a fertilizer used in different plots of a crop.
• Filling of a liquid item by an automatic machine in a container.
 Normal Distribution
Some other situations where normal distribution is applicable
are indicated in the Section 7.5.2.
The most common pattern of distribution of a continuous
variable, found in nature, is of the following type:
 Normal Distribution
Maximum number of men will be around the average height, and as one goes
to the left or right, the number of persons will keep on reducing. In fact,
even other human characteristics like weight, intelligence, etc. follow this
pattern or distribution.
Similarly, if a number of persons were asked to estimate the length of a line of
fixed length, say 20 cms, the maximum number will estimate the line in
the interval from 19.5 to 20.5 cms, the number of person estimating the
length between 18.5 to 19.5 & 20.5 to 21.5, 17.5 to 18.5 & 21.5 to 22.5,
16.5 to 17.5 & 22.5 to 23.5 cms, will go reducing as depicted in the
following diagram.
 Normal Distribution
The same type of diagram will be obtained, if one measures the
errors committed by persons while estimating the length of
the line. For this reason, sometimes it is referred to as
distribution of errors. It could appear as follows:
 Normal Distribution
This type of distribution was discovered by Karl Gauss. For this
reason, it is called Gaussian Distribution. But, it is found so
often in real life that it is called Normal Distribution – the
name which is commonly used. If the curve is drawn in a
continuous manner, it would looks as follows.
 Normal Distribution
Over a period of time, mathematicians tried to describe this type
of curve by a mathematical function, and ultimately found
that the following function describes it best:

f(x) = ( 1 / σ √2π ) e (-1/2) (( x – m) /σ )2 – ө < x < +ө

where, m is the mean of the variable x, and σ is the standard
deviation (s.d.) of x. It is a bell shaped curve symmetrical with
respect to mean (m). As can be seen, the function, called
probability density functions of the normal distribution,
depends on two values, viz. m and σ. These are also referred
as the two parameters of the normal distribution (Many
authors use the Greek symbol ‘µ’ for mean; we prefer to use
‘m’ for sake of simplicity).
 Normal Distribution
•   curve has maximum value at x = m, and tapers off either side
The
but never touches the horizontal line. Theoretically, the curve
on left side goes up to – , and on the right side, it goes up to
+ . However, as much as 99.73% of the area under the curve
lies between m-3σ and m+3σ, and only 0.27% of the area lies
beyond these points. One knows too well that, in real life, no
variable can extend up to – or + , but then it is a negligible
price to pay for this unrealistic assumption for the sake of
amenability to mathematical analysis and interpretations
found to be useful in real life.
 Normal Distribution
Illustration 7.5
For illustration, it was noted in a school, that the height of the
students of Class X was normally distributed with mean as 160
cms and s.d. as 3 cms. The curve of this distribution of height
would be as follows.
 Normal Distribution
•  may be observed that almost all of the students have height
It
between 151 to 169 cms but, however, the curve extends up to
+ on the right side and even beyond 0 cms going up to – which
is absurd. But then this is tolerated as mentioned earlier, for the
sake of mathematical amenability and making useful
interpretations. This would be better appreciated later in the
chapter when some problems from real live situations are
solved.
Standard Normal Distribution The normal distribution with mean
(m) = 0, and s.d. (σ) = 1, is called the standard normal
distribution. The random variable that follows this distribution
is denoted by z. If a variable x follows normal distribution with
mean m and s.d. σ, the variables z defined as
z=x–m
σ
 Normal Distribution
•   standard normal distribution with mean 0 and s.d. as 1. This
Has
id also referred as z-score. The shape of this distribution is as
follows:

It may be noted that the curve has the same shape as the
normal distribution, and has maximum at z = 0, and extends
from – or + like normal distribution.
 Normal Distribution
Use of Normal Distribution for Solving Real Life Problems The use
of the normal distribution for solving problems in real life is
explained in the illustration given below.
Illustration 7.6
If the mean and standard deviation of a variable which is normally
distributed are given e.g. suppose, the life of bulbs, manufactured
by a company, has mean as 1600hrs and s.d. as 30 hrs; several
types of questions relating to the variable (life in the above case)
can be answered. These are as follows:
(i) What is the probability that the life of a bulb selected at
random will be less than or equal to a certain value less than
the mean, say 1550 hrs?
or
What is the proportion or percentage of bulbs having life less than
or equal to a certain value less than the mean, say 1550 hrs?
 Normal Distribution

The evaluation of the probability requires integrating the

probability distribution of the normal distribution from –
infinity to 1550 i.e.
 Normal Distribution
It is not possible to find the value of this integral, and one has to resort to
numerical integration. If, for eh convenience of the users, one were to
construct a table, a very large number of tables would be required,
separately for each value of m and σ. This problem has been solved by
making the substitution
z = (x – m)/ σ. If we do so, the value of the integral is
 Normal Distribution
•   under he standard normal curve, we refer to the Table T1, giving area
Now, since the value of an integral between two limits is equal to the area

under the standard normal curve, and find the area therein from – to –
1.67.

It may be noted that the table gives the area from z = 0 to any positive value f
z. For example, the area from 0 to 1.67 is 0.4525.
 Normal Distribution
•  
Properties of Normal Distribution
(i) The range of the variable is form – to + .
(ii) It is bell shaped curve symmetrical with respect to its men.
(iii) Being symmetrical with respect to its mean, its measure of
skewness is zero.
(iv) The measure of peakedness of the curve i.e kurtosis is 3. This
value is taken as reference point for indicating whether a
curve is more flatter than this (platykurtic) or more peaked
than this (leptokurtic). The normal curve, itself, is referred to
as meso kurtic.
(v) The mean, median and mode are all equal.
(vi) The mean being equal to median divides the curve in two
equal parts.
 Normal Distribution
(vii) The mean being equal to mode, the curve has maximum value
at the mean.
(viii) The Mean Deviation of Normal distribution is
σ √ 2 π = (4/5) σ = 0.8 σ (approximately)
(ix) The area under the curve from m – σ to m + σ is 68.26%.
(x) The area under the curve from m – 2σ to m + 2σ is 95.44%.
(xi) 50% of the area under the curve lies between m – 0.645 σ (first
quartile) and m + 0.645 σ (third quartile).
(xii) 95% of the area under the curve lies between m – 1.96σ to m +
1.96σ .
(xiii) 99.73% of the area under the curve lies between m – 3σ to m
+ 3σ
The properties (ix), (xii) and (xiii) are depicted in the following
figure.
Thank You