Quantitative Analysis for
Management - I
Prof. Tamalika Koley
Decision Sciences Area
Indian Institute of Management Lucknow
Module-III
Binomial Distribution Poisson Distribution
Negative Binomial
Distribution
Discrete Distribution
A discrete random variable can take either finite or countably infinite set of
values.
Probability Distribution: Specify the possible values the random variable can
assume and the probability associated with each value.
Example: Two coins are tossed simultaneously, X = number of heads.
Values of X 0 1 2
p(x) 0.25 0.5 0.25
Graphical Representation
Common Distributions
Binomial Distribution
Poisson Distribution
Negative Binomial Distribution
Trial
An attempt to produce a particular event which is neither impossible nor
certain. For example, throwing a dice to get a six is a trial.
Trials are repeated to form a series of trials.
Trials are independent if the probability of the event in any trial is not affected
by the supplementary information about the results of other trials.
Example: Dependent Trial
Drawing cards without replacement.
Introduction
Many experiments result in dichotomous responses - Yes/No, Pass/Fail,
Defective/Non-defective, Male/Female.
Example: Coin tossing - Head/Tail.
Toss coin 10 times, say.
Each toss - Head/Tail.
Interest: Probability distribution of the number of heads (tails).
Many real life experiments are equivalent to tossing a coin fixed number of times,
say, n, and observing the number of times one of the two events (x) occurs.
Applications
Quality Control: To determine the likelihood of defective items in a production batch.
Analyzing the number of defective items in a sample enables manufacturers to make informed
decisions to enhance their production processes.
Risk Assessment: To estimate the likelihood of project success/ failure.
Analyzing the number of successful projects will help in quantifying risks.
Characteristics of a Binomial random variable
Experiment consists of n independent and identical trials.
Two possible outcomes one each trial - Success (S), Failure (F).
The probability of success (p) remains same from trial to trial.
The Binomial random variable X is the number of successes in n trials.
Exercise
Before marketing a new product on a large scale, many companies will conduct a consumer preference
survey to determine whether the product is likely to be successful.
Suppose a company develops a new diet soda and then conducts a taste preference survey in which 100
randomly chosen consumers state their preferences among the new soda and the two leading sellers.
X is the number of customers choosing the
new brand over the two others.
Is X a binomial random variable?
Probability Distribution - Example
According to the latest police reports, 80 % of all petty crimes are unresolved, and in your town,
at least three of such petty crimes are committed.
The three crimes are all independent of each other.
What is the probability that one of the three crimes will be resolved?
Solution
Trials are commitment of the crimes.
Three crimes → Three trials (independent).
Event is resolving the crime. Success (S) if resolved, Failure (F) if unresolved.
Probability of success is probability of resolving the crime = 1-0.8 = 0.2 (independent of trial).
Solution-Cases
Case 1: Trial 1 – S, Trial 2 – F, Trial 3 – F.
Probability: .
Case 2: Trial 1 – F, Trial 2 – S, Trial 3 – F.
Probability: .
Case 3: Trial 1 – F, Trial 2 – F, Trial 3 – S.
Probability: .
Required probability is
General
A series of n independent Bernoulli trials.
p is the probability of success. This implies 1-p is the probability of failure.
Find the probability of getting x successes.
Case 1: P Probability:
How many cases?
Equivalent to choosing x trials from n trials which will result in successes .
Required probability:
Binomial Distribution
Definition: For the general binomial experiment, with n trials and probability of success p on
each trial, the probability distribution of X, the number of successes is
; x = 0,1,2,… , n.
Notation:
Example
A machine that produces stampings for automobile engines is malfunctioning and producing 10%
defectives. The defective and non - defective stampings proceed from the machine in a random
manner. If the next five stampings are tested, find the probability that three of them are defective.
Solution
X is the number of defective in five stampings.
Here p = 0.10 and n = 5.
Required probability is
Exercise
Ashok, Theo and Sadie will each visit the local club to swim one evening next week but have made
no arrangement between themselves to meet or go on any particular day.
The random variable X is the number of three who go for swimming on Wednesday.
Find the probability distribution of X.
Moments
Expectation: Let then the expectation/ mean of X is
Variance: Let then the variance of X is
The standard deviation of X is
Recall is the number of trials, is the probability of success and is the probability of failure.
Example
There is a 90% chance that a particular type of component will perform adequately under high
temperature conditions. The device involved has four such components. Define the events
All of the components perform adequately and therefore the device is operative.
The device is inoperative because exactly one of the four components fails.
The device is inoperative because one or more of the components fail.
Determine the probability of each of the above events.
Calculate the mean and standard deviation of the number of components performing
adequately out of the four components.
Solution
X is the number of components performing adequately out of the four components.
where and
1. and
Exercise
According to a 2010 study conducted by the Toronto – based social media analytics firm Sysomos,
71% of all tweets get no rejection. That is, there are tweets that are not replied to or retweeted.
Suppose we randomly select 100 tweets.
1. Is the selection a binomial experiment? Explain.
2. What is the expected number of these tweets with no reaction?
3. What are the variance and standard deviation for the number of these tweets with no
reaction?
Graphical Representation
Bernoulli Random Variable
Bernoulli distribution is a special case of the binomial distribution
where a single trial is conducted, that is, n=1.
The probability mass function of X is
and
Example: When developing a new drug, pharmaceutical scientists can use
Bernoulli distribution to calculate the probability that a person will be cured or
not cured with the help of the new drug.
Graphical Representation
Exercise
A basketball player can shoot a ball into the basket with a probability of 0.6. What is the
probability that he misses the shot?
If a Bernoulli distribution has a parameter 0.45 then find its mean and standard deviation.
Poisson Distribution
Counts the number of events that will occur in a fixed time interval.
Events occur independently, at a constant average rate (mean number of events per interval)
The Poisson distribution is a useful general model for count data.
Examples:
The number of people entering a supermarket per hour.
The number of cars passing a point on a road every minute.
Possible values of X are 0,1,2,…
Probability Distribution
Definition: Let X be a Poisson random variable with mean rate The distribution of X is given by
Notation:
Example
Suppose the number of a company's employees who are absent on Mondays has (approximately) a
Poisson probability distribution with an average number of 2.6.
Find the probability that on next Monday
Fewer than two employees will be absent.
Exactly two employees will be absent.
Solution
X denote the number of a company's employees absent on Mondays.
Moments
Expectation: Let then the expected value or mean of X is
Variance: Let then the variance of X is
For Poisson distribution, mean and variance are equal.
The standard deviation of X is
Example
US. airlines fly approximately 48 billion passenger-miles per month and average about 2.63 fatalities
per month (Statistical Abstract of the United States: 1998).
What is the probability that no fatalities will occur during any given month?
Find the expected number of fatalities and the standard deviation in a month.
Solution
X denote the number of fatalities per month.
Probability that no fatalities will occur during any given month
and
Exercise
Suppose we are interested in the number of arrivals at the drive-up teller window of a bank
during a 15-minute period on weekday mornings.
An analysis of historical data shows that the average number of cars arriving in a 15-minute
period of time is 10.
Find the probability of at least two arrivals.
Compute the mean and standard deviation of the number of arrivals.
Example
A factory produces 1,000 bulbs a day.
Each bulb has a tiny, p equals 0.002, of being defective.
What is the probability that exactly 3 bulbs are defective?
Solution:
X is the number of defective bulbs in a day.
average number of defective bulbs in a day.
Required probability:
Computationally challenging!!!
Solution
Recall the Poisson assumptions:
• Average number of events fixed
• There are many opportunities is large
Treat X as the Poisson random variable.
Probability: (almost same).
Mathematics:
As (Check!!)
Poisson Approximation to Binomial Distribution
X denote the number of successes resulting from n independent trials with p as the probability
of success in each trial.
Result: If the number of trials, n is very large, and probability of success, p is very small, but is of
moderate size, then the distribution of X can be approximated by the Poisson distribution with
Example
The manufacturer of a certain electronic component knows that 3% of his product is defective. He
sells the components in boxes of 100 and guarantees that not more than 2% in any box will be
defective. What is the probability that a box will fail to meet the guarantee?
Solution
X denote the number of defectives in a box of 100 components.
Required probability is
Solution-Poisson Approximation
n is quite large, p is very small, but (moderate).
Required probability is
Advantage of Poisson distribution over Binomial
Distribution
Calculations tend to be rather more straightforward than those of the binomial.
Indeterminate sample size makes it applicable to a wider range of situations.
Graphical Representation
Negative Binomial Distribution
Repeating trials where each trial results in two possible outcomes - Success, Failure.
Probability of success p is same in every trial.
Trials are independent, that is, outcome on one trial does not affect the outcome on other trials.
The experiment continues until k successes are observed, where k is specified in advance.
Probability Distribution
X denote the number of trials required to get k successes.
Notation:
Geometric Distribution
If k=1, we are interested in the number of trials required to get a single (first) success.
The distribution of X is called Geometric distribution.
The probability distribution is
Notation:
Example
An oil company has 20 % chance of striking oil when drilling a well. What is the probability the
company drills seven wells to strike oil three times?
Solution:
X denote the number of times the company drills wells to strike oil three times.
Required probability is
Exercise
A door-to-door encyclopedia salesperson is required to document five in-home visits each day.
Suppose that she has a 30% chance of being invited into any given home, with each address
representing an independent trial.
What is the probability that she requires fewer than eight houses to achieve her fifth success?
Moments
Expectation: Let then the expected value of X is
Variance: Let then the variance of X is
Example
A sales team has a target to make 100 sales.
The team leader is interested in understanding the probability distribution of the number of calls
that need to be made before achieving 100 sales.
From the historical data, the probability of making a sale on each call is 0.3.
Find the probability distribution of the number of calls to be made.
Calculate the expected number of calls and also the corresponding standard deviation.
Solution
X denote the number of calls that need to be made by the team before achieving 100 sales.
Required probability distribution is
and
Graphical Representation
