STAT320: MATHEMATICAL

STATISTICS

Anwer Khurshid
University of Nizwa
28th April 2021
The shape of the binomial distribution
depends on 𝑛 and 𝑝
The shape of the Poisson distribution depends on 𝜆
1. Assume that 25% of fuses are defective, and the fuses in packages of
six fuses are independently selected. What is the probability that
(exactly) two fuses in a package of six are defective? What is the
probability that fewer than two are defective?
2. A life insurance salesman sells on the average 3 life insurance policies
per week. Calculate the probability that in a given week he will sell
(i) Some policies
(ii) 2 or more policies but less than 5 policies.
(iii) Assuming that there are 5 working days per week, what is the
probability that in a given day he will sell one policy?
Some Discrete Distributions other than Binomial and
Poisson: Geometric Distribution

We start by comparing the binomial and geometric distributions.

Binomial has a FIXED number of trials before the experiment begins and
counts the number of successes obtained in that fixed number, whereas
Geometric has a fixed number of successes (ONE...the FIRST) and counts the
number of trials needed to obtain that first success. It is theoretically
possible to proceed indefinitely without ever obtaining a success.
Examples would be
i) flip a coin UNTIL you get a tail
ii) roll a die UNTIL you get a 5
iii) repeated attempts to start an engine
A random variable 𝑋 is geometric provided that the conditions for
binomial distribution are met but there is a one NEW: the variable of
interest is the number of trials required to obtain the FIRST success.
The shorthand 𝑋~𝑔𝑒𝑜𝑚𝑒𝑡𝑟𝑖𝑐 (𝑝) is used to indicate that the random
variable 𝑋 has the geometric distribution with real parameter 𝑝 .

The pmf is
𝑝 𝑥 = 𝑝(1 − 𝑝)𝑥−1 for 𝑥 = 1,2,∙∙∙ 0 < 𝑝 < 1 and 𝑝 + 𝑞 = 1
Properties
1
(i) The mean is 𝐸 𝑋 = 𝜇 =
𝑝

𝑞
(ii) The variance is 𝑉𝑎𝑟 𝑋 = 𝜎2 =
𝑝2

𝑝𝑒 𝑡
(iii) The mgf of a geometric distribution is 𝑀𝑋 𝑡 =
1−𝑞𝑒 𝑡

(iv) The cumulative distribution function of a geometric random

variable 𝑋 is 𝐹𝑋 𝑥 = 𝑃 𝑋 ≤ 𝑥 = 1 − (1 − 𝑝)𝑥
The geometric distribution can be used to model the number of failures
before the first success in repeated mutually independent Bernoulli
trials, each with probability of success p .

Note: Another form for a geometric distribution is

𝑝 𝑥 = 𝑝(1 − 𝑝)𝑥 ; 𝑥 = 0,1,2,∙∙∙

Here 𝑥 starts from 0 and not 1.

Example #1

It is known that the probability of a defective computer component is 0.02.

Components are randomly and independently selected. Find the probability
that the first defect is caused by the seventh component tested. How many
components do you expect to test until one is found to be defective?

Solution: Let random variable 𝑋 is the number of computer components

tested until the first defect is found. Here 𝑋 takes on the values 1,2,∙∙∙ where
𝑝 = 0.02. So 𝑋~𝑔𝑒𝑜𝑚𝑒𝑡𝑟𝑖𝑐 (𝑝 = 0.02). We need to find 𝑃 𝑋 = 7 .
We have
𝑃 𝑋 = 7 = 0.02(1 − 𝑝)7−1 = 0.0177

The probability that the seventh component is the first defect is 0.0177.
Example #2

A test of weld strength involves loading welded joints until a fracture

occurs. For a certain type of weld, 80% of the fractures occur in the
weld itself, while the other 20% occur in the beam. A number of welds
are tested and the tests are independent. Let 𝑋 be the number of test
at which the first beam fracture is observed.
Either a weld fracture or a beam fracture will occur on each Bernoulli
trial. We’ll call a success a beam fracture. So 𝑋~𝑔𝑒𝑜𝑚𝑒𝑡𝑟𝑖𝑐 (𝑝 = 0.2).
(i) Find the probability that the first beam fracture happens on the third
trial or later.
We are asked to find 𝑃 𝑋 ≥ 3 . We start with complement of what is
required. Why?
We are asked to find 𝑃 𝑋 ≥ 3 . We start with complement of what is
required. Why?

𝑃 𝑋 ≥3 =1−𝑃 𝑋 <3 =1− 𝑃 𝑋 =1 +𝑃 𝑋 =2

which will give 𝑃 𝑋 ≥ 3 =
(ii) Find 𝐸(𝑋) and 𝑉(𝑋) and the standard deviation of 𝑋.
We know
1 1
(i) The mean is 𝐸(𝑋) = 𝜇 = = =5
𝑝 0.2
2 𝑞 0.8
(ii) The variance is 𝑉𝑎𝑟(𝑋) = 𝜎 = = = 20
𝑝2 (0.2)2

(iii) Standard deviation is 𝑉𝑎𝑟(𝑋) = 𝜎 = 20 = 4.47

Negative binomial probability distribution

Suppose that repeated Bernoulli trials are performed until 𝑟 successes occur. The number
of trials required 𝑋, follows the negative binomial probability distribution with following
probability mass function:

𝑥−1 𝑟
𝑃(𝑋 = 𝑥) = 𝑝 (1 − 𝑝)𝑥−𝑟 for 𝑥 = 𝑟, 𝑟 + 1, 𝑟 + 2, . . . 0 < 𝑝 < 1 and 𝑝 + 𝑞 = 1,
𝑟−1

Properties:
𝑟
(i) The mean is 𝐸(𝑋) = 𝜇 =
𝑝
𝑟𝑞
(ii) The variance is 𝑉𝑎𝑟(𝑋) = 𝜎 2 =
𝑝2
𝑟
𝑝𝑒 𝑡
(iii) The mgf of a negative binomial distribution is 𝑀𝑋 (𝑡) =
1−𝑞𝑒 𝑡
(iv) If 𝑟 = 1 then Negative binomial will become geometric.
Example #1 (Earlier we have seen this problem with
geometric distribution)
A test of weld strength involves loading welded joints until a fracture
occurs. For a certain type of weld, 80% of the fractures occur in the
weld itself, while the other 20% occur in the beam. A number of welds
are tested and the tests are independent. Let 𝑋 be the number of test
at which the first beam fracture is observed.
(i) Find the probability that the 3rd beam fracture (success) occurs on
the 6th trial.
(ii) Find 𝐸(𝑋) and 𝑉(𝑋) and the standard deviation of 𝑋.
(i) Recall, 𝑃(𝑆𝑢𝑐𝑐𝑒𝑠𝑠) = 𝑃(beam fracture) = 0.2. We want the probability
that there were 2 successes somewhere within the first 5 trials, and the 6th
trial was a success.

𝑥−1 𝑟 𝑥−𝑟 6−1

𝑃𝑟 𝑜 𝑏𝑎𝑏𝑖𝑙𝑖𝑡𝑦 = 𝑝 (1 − 𝑝) = 0. 23 (0.8)6−3 = 0.04096
𝑟−1 3−1
𝑟 3
(ii) The mean is 𝐸(𝑋) = 𝜇 = = = 15
𝑝 0.2
𝑟𝑞 3×0.8
The variance is 𝑉𝑎𝑟(𝑋) = 𝜎 2 = 2 = = 60
𝑝 (0.2)2

Standard deviation is 𝑉𝑎𝑟(𝑋) = 𝜎 = 60 = 7.746

Hypergeometric probability distribution

Select without replacement 𝑛 from 𝑁 available items of which 𝑟 are

labeled as “successes”, and 𝑁 − 𝑟 are labeled as “failures”. Let 𝑋 be the
number of successes among the 𝑛. Probability mass function of 𝑋:
𝑟 𝑁−𝑟
𝑥 𝑛−𝑥
𝑃(𝑋 = 𝑥) = 𝑁 for 𝑥 = 0,1,2, . . . , 𝑟 0 < 𝑝 < 1 and 𝑁 = 1,2, …,
𝑛
Properties:
𝑛𝑟
(i) The mean is 𝐸(𝑋) = 𝜇 =
𝑁
𝑛𝑟(𝑁−𝑟)(𝑁−𝑛)
(ii) The variance is 𝑉𝑎𝑟(𝑋) = 𝜎 2 =
𝑁2 (𝑁−1)
(iii) The difference between binomial and hypergeometric is sampling
with?????
Example#1

Suppose 5 cards are selected at random and without replacement from

an ordinary deck of playing cars.
(i) Construct the probability distribution of 𝑋, the number of club cards
among the five cards.

(ii) Use (i) to find 𝑃 𝑋 ≤ 1 .

It’s a hypergeometric probability distribution problem. We can make a
table like.
Random variable (𝑋) 𝑃(𝑋 = 𝑥)

13 39
𝑃(𝑋 = 0) = 0 5 = 0.2215
52
0 5

5
Example #2

A purchaser of electrical components buys them in lots of size 10. It is

his policy to inspect 3 components randomly from a lot and to accept
the lot only if all 3 components are nondefective. If 30 percent of the
lots have 4 defective components and 70 percent have only 1, what
proportion of lots does the purchaser reject?

It’s a hypergeometric problem. The probability of accepting the lot is

the probability that all 3 are found nondefective. Therefore the
probability that the lot is rejected is 𝑃(reject the lot) = 1 − 𝑃(accept
the lot). So
 4 6 1 9
0 3 0 3
𝑃(𝑎𝑐𝑐𝑒𝑝𝑡) = (0.3) 10 + (0.7) 10 = 0.54,
3 3
so 𝑃(reject the lot) = 1 − 𝑃(accept the lot) = 1−0.54 = 0.46.
Problems
1. (Geometric) A medical team randomly selects people in an area, until he finds a
person who has a corona virus. Let 𝑝 is the probability that he succeeds in finding
such a person, is 0.2 and 𝑋 denote the number of people asked until the first
success.
(i) What is the probability that the team must select 4 people until he finds one
who has a corona virus?
(ii) What is the probability that the team must select more than 6 people before
finding one who who has a corona virus?

2. (Negative binomial) Suppose 40% of employees at a firm have traces of asbestos

in their lungs. The firm is asked to send 3 of such employees to a medical center for
further testing. Find the probability that exactly 10 employees must be checked to
find 3 with asbestos traces. What is the expected number and variance of
employees that must be checked?
3. (Negative binomial) If the probability is 0.40 that a child exposed to a certain
contagious disease will catch it, what is the probability that the tenth child exposed
to the disease will be third to catch it?
4. (Hypergeometric distribution) In the manufacture of car tires, a particular
production process is known to yield 10tyres with defective walls in every batch of
100 tires produced. From a production batch of 100 tires, a sample of 4 is selected
for testing to destruction. Find: (i) the probability that the sample contains 1
defective tire; (ii) the expectation of the number of defectives in samples of size 4;
(iii) the variance of the number of defectives in samples of size 4.

5. Suppose 2% of items produced from an assembly line are defective. If we sample

10 items, what is the probability that 2 or more are defective? Here the count
follows the binomial distribution. Suppose now that we sample 10 items from a
small collection, like 20 items, and count the number of defectives. The resulting
random variable is not binomial. Why not?