Random Variable

• Concept of Random Variable

• Types of Random Variables
• Discrete
• Continuous
• Probability Mass Function (pmf) of a Discrete Random Variable
• Expectation and Variance of a Discrete Random Variable
• Probability Density Function (pdf) of a Continuous Random Variable
• Expectation and Variance of a Continuous Random Variable
• Cummulative Distribution Function

Gaurav Garg (IIM Lucknow)

Concept of Random Variable
• Suppose a person plays a game, where
• he wins Rs. 100 if he gets at least one head in two tosses of a fair coin, and
• he loses Rs. 200 otherwise.
• If the person keeps playing the game for a long-term, is he going to gain some
money?
• What is the long-term gain or loss to the person.

• The probability of getting at least one head in two tosses is P({HH, HT ,TH})=3/4.
• So, he has 75% chance of winning Rs. 100.
• He could lose Rs. 200 with 25% chance.

• Now how to incorporate this idea of his gain (or loss) along with the probability.

• In many situations, Probability fails to provide the quantitative measurement of the

uncertainty.
Gaurav Garg (IIM Lucknow)
• What is lacking in previous example, is some quantitative measurement of
his gain.
• We define as his gain or profit in the same game, then

• Now each of these outcomes is a part of the sample space.

• So, each of them has some specific probability to happen.
• We can write it more systematically as below:
•, or

• This variable is called random variable.

Gaurav Garg (IIM Lucknow)

• Variable changes its values depending on some other factor (unlike
constant).

• Random Variable changes its values depending on what we get as an

outcome of a random experiment.

• Thus, each value of a random variable is associated with a probability.

• In other words, random variable takes a value with some probability.

Gaurav Garg (IIM Lucknow)

• Mathematical Definition
• A random variable, is a function that assigns a numeric value to the outcomes of a
random experiment
• i.e., ,
• where is the sample space, is the real line .
• Example:
• Consider the random experiment of tossing two coins
• Sample space:
• is number of Heads appeared
• It is a random variable
• It can take any value out of 0, 1 and 2.

Gaurav Garg (IIM Lucknow)

• Example:
• Consider the random experiment of tossing a fair die.
• Sample space: S = {1, 2, 3, 4, 5, 6}
• P(getting 1) = 1/6, P(getting 2) = 1/6
• P(getting 3) = 1/6, P(getting 4) = 1/6
• P(getting 5) = 1/6, P(getting 6) = 1/6

• Let X be the value on the upper face of the die

• X could be 1, 2, 3, 4, 5 or 6 (X is a random variable)

• The event of “getting 1” is same as saying “X=1”

• So, P(getting 1) = P(X=1) = 1/6

• Similarly,
• P(X=2) =1/6, P(X=3) = 1/6, P(X=4) = 1/6, P(X=5) = 1/6, P(X=6) = 1/6

Gaurav Garg (IIM Lucknow)

• Example:
• Consider the random experiment of tossing a fair coin thrice.
• Sample space S={HHH, HHT, HTH, THH, HTT, THT, TTH, TTT}
• X: no. of heads appearing in three tosses
• X could be 0, 1, 2, or 3. (X is a random variable)

• P(X = 0) = P(TTT) = 1/8

• P(X=1) = P(HTT, THT, TTH) = 3/8
• P(X=2) = P(HHT, HTH, THH) = 3/8
• P(X=3) = P(HHH) = 1/8

Gaurav Garg (IIM Lucknow)

• Consider the same example again.
• S = {HHH, HHT, HTH, THH, HTT, THT, TTH, TTT}
• X: no. of heads appearing in three tosses
• X belongs to {0, 1, 2, 3}

• Event of “getting at least one head” is same as “X ≥ 1”

• P(X ≥ 1) = P(HHH, HHT, HTH, THH, HTT, THT, TTH) = 7/8

• Similarly, we can write

• P(X ≤ 1) = P(HTT, THT, TTH, TTT) = 4/8
• P(X < 2) = P(HTT, THT, TTH, TTT) = 4/8
• P(X ≤ 2) = P(HHT, HTH, THH, HTT, THT, TTH, TTT)= 7/8

Gaurav Garg (IIM Lucknow)

• Example:
• Consider the random experiment of drawing two cards at random from a well
shuffled pack of 52 cards with replacement.
• The sample space:
• S={ CC, CD, CH, CS,
DC, DD, DH, DS,
HC, HD, HH, HS,
SC, SD, SH, SS }
• X: number of Diamond Cards selected
• X is a random variable.

• X ≤ 1 is the event that at most one diamond is selected

• P(X ≤ 1) = 15/16

Gaurav Garg (IIM Lucknow)

• Example:
• Metro trains run every half hour between 7:00 AM to 7:00 PM on a certain line .
• A man enters the station at a random time during this period.

• Let X be the random variable which denotes the waiting time in minutes.
• X can take any value between 0 and 30.
• Sample space: S = [0,30)

• The event that “he has to wait for at most 10 minutes” is same as “X ≤ 10”
• P(X ≤ 10) = (10 – 0)/ (30 – 0) = 1/3

Gaurav Garg (IIM Lucknow)

Types of Random Variables
• Random Variables are of two types:
• Discrete Random Variable
• Continuous Random Variable

• Discrete Random Variable takes only a countable (finite or countably infinite)

number of values.
• no. of accounts opened, no. of cars sold, Friday night attendance at a cinema, no. of
defective items in a box, etc.
• Discrete random variables usually counts the items or individuals.
• Range set (the collection of all possible values of X) of a discrete random variable is countable.

• Continuous Random Variable takes on any values within an interval.

• weight of fertilizer packed in a bag, amount of sugar in an orange, time required to run a
mile, price of a stock, etc.
• Continuous random variable usually measures the characteristics.
• It can take an uncountably infinite number of values.
• Range set of a continuous random variable is an interval.
• It can be of types (-∞, ∞), (-∞, 0),Gaurav
(0, ∞)Garg (IIM Lucknow)
or (a, b).
More Examples
Random Variable Values Type
Flip a coin three times; {0, 1, 2, 3} Range Set is Finite
X = the total number of heads. Discrete Random Variable

Select a mutual fund; {2, 3, 4, ...} Range Set is Countably Infinite

X = the number of companies in the Discrete Random Variable
fund portfolio.
Measure the length of an object; Any positive Range Set is Uncountable
X = its length in centimeters. number Continuous Random Variable

Throw two dice over and over until { 1, 2, 3, 4, ...} Range Set is Countably Infinite
you roll a double six; Discrete Random Variable
X = the number of throws.
Take a true-false test with 100 {0, 1, 2, 3, Range Set is Finite
questions; X = no. of questions 4, ..., 100} Discrete Random Variable
answered correctly.
Invest $10,000 in stocks; {$0, $1, $2, $3, Range Set is Countably Infinite
X = the value, to the nearest $1, of …} Discrete Random Variable
your investment after an year.
Select a group of 50 people at Any positive Range Set is Uncountable
random; number Continuous Random Variable
X = exact average height (in m) of the
group.

Gaurav Garg (IIM Lucknow)

Probability Distribution of a Discrete Random Variable
• Also called as Probability Mass Function (pmf).
• It is a mutually exclusive listing of all possible values of the discrete
random variable along with corresponding probabilities.

All possible values of random variable X

Corresponding probabilities

• p1, p2, p3,…. are non-negative.

• Together they sum up to 1.

Gaurav Garg (IIM Lucknow)

• Example:
• X is the number of heads appearing in two fair coin-tossing experiments.
• X is a discrete random variable having the probability distribution:
probability mass function of X is
1 / 4, if x  0

p ( x) 2 / 4, if x  1
1 / 4, if x  2

0.6

0.5

0.4

0.3

0.2

0.1

0
x=0 x=1 x=2

Gaurav Garg (IIM Lucknow)

• Example:
• In throwing a die, X is the values appeared
• X is a discrete random variable having the probability distribution

probability mass function of X is

p ( x) 1 / 6, if x 1,2,3,4,5,6.

0.18
0.16
Probabilities

0.14
0.12
0.1
0.08
0.06
0.04
0.02
0
x=1 x=2 x=3 x=4 x=5 x=6
Gaurav Garg (IIM Lucknow)
• Recall the first example where X is the random variable denoting the gain
and is given by

• In tabular format, the probability distribution of X is given by

3 / 4 if x  100
• Probability mass function of X is written as p ( x) 
1 / 4 if x   200

Gaurav Garg (IIM Lucknow)

• Sometimes, the pmf can be expressed as a mathematical function.
• Example
• denotes the number of tosses needed to get the first tail.

• is a discrete random variable with probability distribution:

probability mass function of X is

 1/ 2 if x 1
1 / 2 2 if x 2

• pmf can also be written as p ( x ) 
 
1 / 2 3
if x 3
• Clearly,   

Gaurav Garg (IIM Lucknow)

• Example:
• Consider the previous example again.
• The discrete random variable has the pmf:

• The probability of getting a tail in less than 3 tosses is given by

Gaurav Garg (IIM Lucknow)

• Example:
• Suppose an individual purchases two electronic components
• Each of which may either be defective or acceptable.
• Sample space: S = {DD, DA, AD, AA}
• Given that P(DD) = 0.09, P(DA) = 0.21, P(AD) = 0.21, P(AA) = 0.49.
• X: number of acceptable components.
• Obtain the probability distribution of X.
• What is the probability of getting at least one acceptable components?

• Ans: 0.91

Gaurav Garg (IIM Lucknow)

Expectation of a Discrete Random Variable
• is a discrete random variable having the pmf
• Expectation or Expected value or mean of is given by

• Example:
• Recall the first example where is the random variable denoting the gain
• pmf of is
• Expected gain of the person is
• Rupees

Gaurav Garg (IIM Lucknow)

• Expectation of a random variable can be interpreted as “some sort of
average” of all the values that the random variable takes.
• So, expected value of a random variable is the center of its distribution.
• The concept of expectation is analogous to the physical concept of the
center of gravity of a distribution of mass.
• If and are constants, and are random variables, then we have

• For any function ,

Gaurav Garg (IIM Lucknow)

• Example
• Suppose X is a discrete random variable having following probability distribution
• x: -2 -1 0 1
• p(x): 0.4 k 0.2 0.3
• Find E(X).
• Ans: -0.6

• Example
• A shipment of 6 television sets contains 2 defective sets.
• A hotel makes a random purchase of 3 of the sets.
• If X is the no. of defective sets purchased by the hotel.
• How many defective sets are expected?
• Ans: 1

Gaurav Garg (IIM Lucknow)

• Example
• Consider the random experiment of tossing a fair coin repeatedly till we get the
first tail.
• What is the expected number of tosses needed to get the first tail.
• Random variable X is the number of tosses needed to get the first tail.
• pmf of X is given as

• We have to find E(X)

• Ans: 2

Gaurav Garg (IIM Lucknow)

Variance of a Discrete Random Variable
• Standard deviation is a universally accepted measure of dispersion.
• It is obtained as a positive square root of variance.
• Variance of a random variable is defined as

• where for a discrete random variable with pmf ,

• Note that Variance is always non-negative.

• If and are constants, and are random variables, then we have

• is the covariance between and It would be(IIMdiscussed

Gaurav Garg Lucknow) later.
• Example
• Suppose X is a discrete random variable having following probability distribution
• x: -2 -1 0 1
• p(x): 0.4 k 0.2 0.3
• Find Var(X).
• Ans: 1.64

• Example
• A shipment of 6 television sets contains 2 defective sets.
• A hotel makes a random purchase of 3 of the sets.
• If X is the no. of defective sets purchased by the hotel.
• Find Var(X).
• Ans: 2/5

Gaurav Garg (IIM Lucknow)

Joint Probability Distribution of two Discrete Random Variables
• Joint probability distribution or joint probability mass function of two
discrete random variables X and Y is
• for all possible values
• It can be written in a tabular form as below:
Possible values of Y
X Y b1 b2 ……… bm Total
Possible values of X
a1 p11 p12 ……… p1m p1.
a2 p21 p22 ……… p2m p2.
. . . . . .
. . . . . . Marginal pmf of X

an pn1 pn2 ……… pnm pn. Marginal pmf of Y

Total p.1 p.2 ……… p.m 1

pij  P ( X ai and Y b j ), i 1,2,  , n; j 1,2,  , m.

Gaurav Garg (IIM Lucknow)
Example:
• A fair coin is tossed once.
• Let X denote the number of heads.
• Y denote the number of tails.
• Find the joint probability mass function of X and Y.

• Ans Y
X 0 1 Total
0 0 1/2 1/2
1 1/2 0 1/2
Total 1/2 1/2 1

Gaurav Garg (IIM Lucknow)

• Example:
• A fair coin is tossed three times independently.
• X: no. of heads on the first toss;
• Y: total no. of heads.
• Find the joint pmf of X and Y.

• Ans Y
X 0 1 2 3 Total
0 1/8 2/8 1/8 0 4/8
1 0 1/8 2/8 1/8 4/8
Total 1/8 3/8 3/8 1/8 1

Gaurav Garg (IIM Lucknow)

Covariance between two Discrete Random Variables
• Covariance is a measure of relationship between two variables.
• For two random variables and , it is given by

• For discrete random variables and with joint pmf

•

• Example:
• A fair coin is tossed three times independently.
• Let X denote the no. of heads on the first toss;
• Y denote the total no. of heads.
• Obtain covariance between X and Y.
• Example:
• A fair coin is tossed once.
• Let X denote the number of head.
• Y denote the number of tail.
• Obtain covariance between X and Y.
Gaurav Garg (IIM Lucknow)
• Example:
• Using the given probability distribution of random variables and ,
compute

P(XiYi) X Y
0.2 -100 50
0.4 50 30
0.3 200 20
0.1 300 20

Gaurav Garg (IIM Lucknow)

Independence of two Discrete Random Variables
• Two discrete random variables and are said to be independent of each
other when

• for all possible values

• or,

• When X and Y are independent, then covariance between them is zero.

Gaurav Garg (IIM Lucknow)

Probability Distribution of a Continuous Random Variable
• A continuous random variable
• Has its range set is an interval on real line, e.g. etc.
• Takes uncountably many (infinite) values.
• Can not have countable (finite or infinite) possible values
• Can not have a pmf to give its probabilities.

• The probability distribution of a continuous random variable is described

by probability density function (pdf), which is given as

•,

• provided

Gaurav Garg (IIM Lucknow)

• Consider a continuous random variable having pdf
• Probability that takes the values on an interval is given by

• Example:
• Let is a random variable with pdf
• Show that is a valid pdf.
• Find and
• Example:
• Let is a random variable with pdf
• Show that is a valid pdf.
• Find and

Gaurav Garg (IIM Lucknow)

Expectation of a Continuous Random Variable
• If is a continuous random variable with pdf ,
• Then the Expectation of is given by

• For any function ,

• Example:
• If is a continuous random variable with pdf
• .

• Find .

• 3/4
• 1
Gaurav Garg (IIM Lucknow)
Variance of a Continuous Random Variable
• Variance of a random variable is defined as

• where for a continuous random variable with pdf ,

• Example:
• If is a continuous random variable with pdf
•.

• Find .

• 3/80
• 1/3
Gaurav Garg (IIM Lucknow)
Cumulative Distribution Function (cdf)
• Also called a Distribution Function.
• For a random variable , the cdf is defined as

• Clearly, it is a non-decreasing function.

• When X is a discrete random variable with pmf p(x), then

• When X is a continuous random variable with pdf f(x), then

• When X is continuous random variable, then

• cdf is also useful in obtaining various probabilities

Gaurav Garg (IIM Lucknow)