Continuous probability

distributions
Normal, Gamma and Chi-squared
probability distributions.
 Normal probability distribution:
• The most widely used probability distribution
of a continuous random variable is normal
probability distribution. This distribution was
discovered by De-Moivre in 1733 as a limiting
case of binomial distribution. Gauss used this
theory to describe the theory of accidental
error of measurement involve in the
calculation of orbits of heavenly bodies so this
is also known as Gaussian distribution.
 Normal distribution is limiting case of binomial
distribution under the following conditions;

• (i) when the number of trials are infinitely

large i.e.n→ ∞.
(ii) When the probability of success/ failure is
neither too large nor to small.
 Definition:
• A continuous random variable X with
parameters mean(𝜇) and variance (𝜎 2 ) is said
to follows a normal probability distribution if
it’s probability density function is denoted by
f(x) or f(x;𝜇, 𝜎 2 ) and given by a relation as,
𝟏 𝒙−𝝁 𝟐
𝟐 𝟏 − ( )
f(x)=f(x; 𝝁, 𝝈 )= 𝒆 𝟐 𝝈 for −∞ < 𝑋 < ∞
𝝈 𝟐𝝅
Where 𝜇 = Mean such that −∞< 𝜇 <∞
𝜎 2=𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝑠𝑢𝑐ℎ 𝑡ℎ𝑎𝑡 𝜎>0.
Normal probability curve:
 Properties:

1. Normal probability curve is bell shaped and

symmetrical about the line X=𝝁.
2. Being symmetrical in nature it’s mean, median and
mode are coincident i.e. mean=median=mode.
3. It’s probability density function is
𝟏 𝑿−𝝁 𝟐
𝟏
f(x)= 𝒆 𝟐 𝝈 )
− (
𝝈 𝟐𝝅
4.Having only one highest pick the distribution is
unimodal.
5.Due to symmetricity, median is equidistance from it’s
𝑸₁+𝑸₃
quartiles i.e. median-Q₁=Q₃-median ∴ 𝒎𝒆𝒅𝒊𝒂𝒏 =
𝟐
 continue
𝟒
6. It’s mean deviation from mean= 𝝈 and
𝟓
𝟐
Quartile deviation= 𝝈.
𝟑
7. X-axis is the asymptote of the curve.
8.Normal distribution is a limiting case of
binomial distribution when the number of trials
are infinitely large and probability of
success/failure is neither too large nor too
small.
9. It is the limiting case of Poisson distribution when the mean
𝜆 of the distribution is infinitely large i.e.𝜆→ ∞.

10
.If X₁,X₂,……,Xn are n normal variables with means µ₁,µ₂,µ₃,……,µn and
variances𝝈₁𝟐 , 𝝈₂𝟐 , 𝝈₃𝟐 , … … … … . , 𝝈n𝟐 𝒕𝒉𝒆𝒏 𝒊𝒕′ 𝒔 𝒍𝒊𝒏𝒆𝒂𝒓 𝒄𝒐𝒎𝒃𝒊𝒏𝒂𝒕𝒊𝒐𝒏
X= a₁X₁+a₂X₂+a₃X₃+………..+anXn is also a normal variable with mean
µ=a₁µ₁+a₂µ₂+a₃µ₃+…………..+anµn and variance 𝝈𝟐 =a₁𝝈₁𝟐 + 𝒂₂𝝈₂𝟐 +a₃𝝈₃𝟐 +
⋯ . +𝒂n𝝈n𝟐
11.
Area property
(i) It covers 68.26% area within the range µ±𝜎.
Thus p(µ-𝜎 < X < µ + 𝜎)=68.28%=0.6828.
(ii) It covers 95.44% area within the range µ±2𝜎.
Thus p(µ-2𝜎<X<µ+2𝜎)=95.44%=0.9544.
(iii)It covers 99.73% area within the range µ±3𝜎.
Thus p(µ-3𝜎<X<µ+3𝜎)=99.73%=0.9973
 Standard normal distribution:

𝑿−𝝁
• In normal distribution if Z= then Z follows a normal
𝝈
distribution with mean 0 and variance 1 is known as
standard normal distribution and Z is called standard normal
variate. Hence the probability density function of standard
normal distribution is
𝟏
𝟏 −𝟐𝒛𝟐
f(z) = 𝒆 for −∞ < 𝒁 < ∞
𝟐𝝅
standard normal curve
 Properties of standard normal distribution:

1. it’s curve is symmetrical about the line Z=0.

2. It’s mean ,median and mode are coincident and equal to zero.
3. It’s probability density function is
𝟏
𝟏 −𝟐𝒛𝟐
f(z)= 𝒆 where −∞ < 𝒁 < ∞
𝟐𝝅
4. The curve changes it’s direction at Z=±1. So this is the point of inflection.
5.it’s mean deviation from mean is 0.8 and quartile deviation is 0.667.
6.It covers 68.26% area within the range Z=∓1
i.e. p(-1< Z< 1)=68.26%=0.6826
7. It covers 95.44% area within the range Z=∓2
i.e. p(-2<Z<2)=95.44%=0.9544
8.It covers 99.73% area within the range Z=±3.
i.e.p(-3<Z<3)=99.73%=0.9973
Importance of normal distribution:
1.Most of the discrete probability distributions like binomial, Poisson
etc. can be approximated by normal distribution.
2.For large n almost all sampling distribution like Z. F.
etc. follows normal distribution.
3. The whole theory of exact sample tests like t, F, χ𝟐 etc for small
samples is based on the assumption that it’s parent population is
taken from normal.
4. For n→ ∞ central limit theorem follows normal distribution.
5. Most of the distributions which are not normal can be made
normal by simple transformation.
6.It is used in statistical quality control to set quality limits in
industry.
Standard Normal probability distribution table
Continue.
continue
contnue
 Worked out problems

1.If a random variable has the standard normal distribution, find the
probability that it will take on a value (a) less than 1.50 (b) less than -1.20
(c) Greater than 2.16 (d) greater than -1.75 (e) between -1.35 and -0.35
Solution;
(a) Less than 1.50
Z<1.50
Now, 1.50
p(Z<1.50)=0.93319
(from table)
(b)Less than -1.20
Z<-1.20
Now, p(Z<-1.20)=0.11507 (from table)

-1.20
 continue
(c) Greater than 2.16
Here, Z>2.16
P(Z>2.16)=1-p(Z≤2.16)=1- F(2.16)
=1-0.98461
=0.01539 2.16
(d) greater than -1.75
Here, Z>-1.75
p(Z>-1.75)=1-p(Z≤-1.75)=1-F(-1.75)
=1-0.040006
=0.959994 -1.75
(e)Between -1.35 and -0.35
Here , -1.35<Z<-0.35
p(-1.35<Z<-0.35)=F(-0.35)-F(-1.35)
=0.36317-0.08851
=0.27466
+
-1.35 -0.35
 2.Find Z if the probability that a random variable having the standard normal
distribution will take on a value (a) less than Z is 0.9911 (b)greater than Z is 0.1093
(c) greater than Z is0.6443 (d)less than Z is 0.0217 (d) between –Z and Z is 0.9298

Solution;
(a) Less than Z is 0.9911
Here, p(Z<z)=0.9911 0.9911

or F(z)=0.9911 ∴ 𝒛 = 𝟐. 𝟑𝟕
(b)Greater than Z is 0.1093
Here , p(Z>z)=0.1093
or, 1- p(Z≤ 𝒛) = 𝟎. 𝟏𝟎𝟗𝟑
or, p(Z≤ 𝒛) =0.8907
or, F(z) =o.8907
0.1093
∴ z =1.23
(c) and (d) do yourself
 continue

(e) Between –Z and Z is 0.9298

P(-z<Z<z)=0.9298
Let p(Z<-z)=p(Z>z)=x
Then we have .
P(Z<-z)+p(-z<Z<z)+p(Z>z)=1
Or, x +0.9298+ x =1
Or, 2 x = 1-0.9298
0r, x =0.0351 x x
Or, p(Z<-z)=0.0351 -z z
0r -z=-1.81 ∴ 𝑧 = 1.81
 3. A corporation installs 10,000 electric lamps in the street of a city. If these lamps
have average life 1850 hours and standard deviation 200 hours,what number of
lamps may be expected to burn for (a) more than 2000 hours (b) less than 1600
hours (c) between 1540 and 1800 hours?
Solution; Given that mean(𝝁)=1850 hours, S.D.(𝝈)=200 hours. Number of
lamps to be installed (N)=10,000
(a) more than 2000 hours:
Let x denotes a normal variate and z be it’s corresponding standard normal
𝒙−𝝁 𝟐𝟎𝟎𝟎−𝟏𝟖𝟓𝟎 𝟓𝟎
variate then, X>x=2000, and z= = = = 0.25
𝝈 𝟐𝟎𝟎 𝟐𝟎𝟎
Now,
p(X>2000)= p(Z>0.25)
=1-p(Z≤ 𝟎. 𝟐𝟓)
=1-0.5987
=0.4013
Required number of lamps=N p(X>2000)
=10,000x0.4013
= 4013
 continue

(b) Less than 1600 hours

𝒙−𝝁 𝟏𝟔𝟎𝟎−𝟏𝟖𝟓𝟎 −𝟐𝟓𝟎
Here X<x=1600, z= = = = -1.25 Then p(1540<X<1800)
𝝈 𝟐𝟎𝟎 𝟐𝟎𝟎
Now, p(X<1600)= p(Z<-1.25) =p(-1.55<Z<-0.25)
=0.1056 = F(-0.25) – F(-1.55)
Required number of lamps=N p(X<1600) =0.40129 -0.06057
=10000x0.1056 =0.3407
= 1056 Required number of lamps=3407
(c) Between 1540 and 1800 hours
𝒙₁−𝝁 𝟏𝟓𝟒𝟎−𝟏𝟖𝟓𝟎
Let x₁=1540 and x₂=1800 then z₁= = =-1.55
𝝈 𝟐𝟎𝟎
𝒙₂−𝝁 𝟏𝟖𝟎𝟎−𝟏𝟖𝟓𝟎 −𝟓𝟎
and z₂= = = =-0.25
𝝈 𝟐𝟎𝟎 𝟐𝟎𝟎
 5.The distribution of monthly income of 5000 employees of a certain industrial unit
was found to be normally distributed with mean of Rs2000 and standard deviation of
Rs 200. estimate (a) the range of income of middle 60% employees.(b) the lowest
income of richest 10% employees. (c)the highest income of poorest10% employees.
Solution;
Average (µ)=Rs2000, S.D.(𝝈)=Rs200
(a) The range of income of middle 60% employees:
Let x₁ and x₂ be the lowest and highest income of
middle 60% employees, then 20% employees lies
below than x₁ and 20% employees lies above than 20% 60% 20%
x₂. Let z₁ and z₂ are the corresponding standard normal
z₁ z₂
variates of x₁ and x₂ respectively.
𝒙₁−𝟐𝟎𝟎𝟎 x₁ x₂
Now, p(Z<z₁)=20% or, =-0.84 or, F(z₂)=0.8 The required range
𝟐𝟎𝟎
or, F(z₁) =0.2 ∴x₁=1832 From table of income=x₂-x₁
From table, z₁= -0.84 Also, p(Z>z)=20% 0r, z₂= 0.84 =336
𝒙₁−𝝁 𝒙−𝟐𝟎𝟎𝟎
or, 𝝈
= -0.84 0r, 1-p(Z≤z₂)=0.2 or, 𝟐𝟎𝟎
= 0.84
or, p(Z≤z₂)=0.8 ∴ x₂=2168
 continue

(b) Lowest income of richest 10% employees:

Let x be the lowest income of richest 10%
employees and z be it’s SNV then,
P(Z>z)=10%
Or,1-p(Z≤z)=0.1
10%
or, F(z) = 0.9 x
From table z
(c) highest income of poorest 10% employees
z= 1.28 P(Z<z)=10%=0.1
𝒙−𝝁
Or, =1.28 From table z= -1.28
𝝈
𝒙−𝟐𝟎𝟎𝟎 𝒙−𝟐𝟎𝟎𝟎
Or, =1.28 or, =-1.28
𝟐𝟎𝟎 𝟐𝟎𝟎
∴ x=Rs2256 .Hence the lowest ∴x= Rs1744
10% x
income of richest 10% employees =Rs2256 hence highest z
income of poorest 10% employee=Rs1744
 5. In an examination, 10% 0f the students got less than 20 marks and 5%
students got over 75 marks, assuming that the distribution is normal. Find
the mean and variance of the distribution.

Solution:
Let x₁=20 and x₂= 75 and let z₁ and z₂ are corresponding SNV of x₁ and x₂.
Then Also, Solving both equations we get
P(Z<z₁) =10% P(Z>z₂)=5% 𝝁 = 𝟒𝟒. 𝟎𝟔𝟖 and 𝝈 =18.803
Or, F(z₁)=0.1 or, 1- p(Z≤z₂) =0.05
Or, z₁ = -1.28 or, p(Z≤ 𝒛₂) = 𝟎. 𝟗𝟓
𝒙₁−𝝁
Or, 𝝈
= -1.28 or, F(z₂) = 0.95
𝟐𝟎−𝝁
Or, 𝝈
=-1.28 or, z₂ =1.645
𝒙₂−𝝁
∴ µ = 20+1.28𝝈………(i) or, =1.645 10% 5%
𝝈
𝟕𝟓−𝝁 z₁ z₂
0r 𝝈 =1.645
x₁=20 x₂=75
∴ 𝝁 = 𝟕𝟓 − 𝟏. 𝟔𝟒𝟓𝝈……..(ii)