CPS420 Discrete Structures

4.5 Independence of Events

The copyright to this original work is held by Sophie Quigley and
students registered in course CPS420 can use this material for the
purposes of this course but no other use is permitted, and there can be
no sale or transfer or use of the work for any other purpose without the
explicit permission of Sophie Quigley
 2
4.5 Independence of Events
➢Independent Events
• Independence and Set Operations
• Pairwise and Mutual Independence
 3
Recall Examples P36 and P37
A family has two cisgendered children of different ages, each of whom
is equally likely to be a boy (B) or girl (G).
• P36: You meet the oldest child who is a girl. What is the probability
that the youngest child is a boy?
• P37: You find out that one of the children is a girl. What is the
probability that the other child is a boy?
In both cases we had to calculate P(B|A)
• In P36, identifying A and B was straightforward from the problem
description
• In P37, work was required to come up with events A and B
 4
Conditional Probability – Example P36
You meet the oldest child who is a girl. What is the S A AC
probability that the youngest child is a boy?
P(B|A) = P(A  B)/P(A) = (1/4)/(1/2) = ½ B GB BB
P(B) = ½ BC GG BG

Event Description Set N() P()

S All possible configurations {GG, GB, BG, BB} 4 1
A The oldest child is a girl {GB, GG} 3 1/2
B The youngest child is a boy {GB, BB} 2 1/2
E=AB Oldest is a girl, youngest is a boy {GB} 1 1/4
 5
Conditional Probability – Example P37
You find out that one of the children is a girl. What S A AC
is the probability that the other child is a boy?
P(B|A) = P(A  B)/P(A) = (1/2)/(3/4) = 2/3 B GB, BG BB
P(B) = 3/4 BC GG

Event Description Set N() P()

S All possible configurations {GG, GB, BG, BB} 4 1
A At least one child is a girl {GB, BG, GG} 3 3/4
B At least one child is a boy {GB, BG, BB} 3 3/4
E=AB There is one child of each gender {GB, BG} 2 1/2
 6
Examples P36 and P37
P36: P37:

S A AC S A AC
B GB BB B GB, BG BB
BC GG BG BC GG

• P(B) = 1/2 • P(B) = 3/4

• P(B|A) = 1/2 = P(B) • P(B|A) = 2/3  P(B)
• Event A has no effect on event B • Event A has an effect on event B
 7
Independent Events
Let A and B be events in a sample space S.

A and B are independent iff P(A  B) = P(A).P(B)

This is another way of saying that P(B) = P(B|A):

P(B|A) = P(B)  P(A  B) / P(A) = P(B)  P(A  B) = P(A).P(B)
 8
Independent Events – Example P42
A jar contains 4 red and 6 blue marbles. 2 are picked without
replacement. What is the probability that the second marble is red
given that the first one is blue? i.e, what is P(R2|B1)?
• P(R2|B1) = 4/9 (remove a blue marble)
• P(R2)=2/5 (previously calculated) → R2 and B1 are not independent.
• The fact that R2 and B1 are not independent can be verified with
previously calculated values:
• P(B1 R2) = 4/15
• P(B1)×P(R2) = (6/10)×(2/5) = 6/25
 9
Independent Events – Example P42
A jar contains 4 red and 6 blue marbles. 2 are picked without
replacement. What is the probability that the second marble is red
given that the first one is blue? i.e, what is P(R2|B1)?
• Why are R2 and B1 not independent?
• Removing a blue marble from the jar without replacing it changes the
proportion of marbles in the jar. The odds of picking a specific colour
for the next marble need to be adjusted each time a marble is
removed from the jar.
• If the marbles were replaced, the events would be independent,
much like repeated toin cosses or repeated rolling of a die.
 10
4.5 Independence of Events
✓Independent Events
➢Independence and Set Operations
• Pairwise and Mutual Independence
 11
Independence and Complement
Example P43 – rolling a die twice
Look at the independence of the events in these two questions
• what is the probability that the 2nd roll is a 5 given that the 1st is a 6?
• what is the probability that the 2nd roll is not a 5 given that the 1st is a 6?

Event Description Set N() P()

S sample space {1,2,3,4,5,6}2 36 1
A First roll is a 6 {6}×{1,2,3,4,5,6} 6 1/6
B Second roll is a 5 {1,2,3,4,5,6}×{5} 6 1/6
Bc Second roll is a not 5 {1,2,3,4,5,6}×{1,2,3,4,6} 30 5/6
AB {(6,5)} 1 1/36
ABc {6}×{1,2,3,4,6} 5 5/36
 12
Independence and Complement
Example P43
• P(A).P(B) = (1/6).(1/6) = 1/36 = P(AB) → A and B are independent.
• P(A).P(Bc) = (5/6).(1/6) = 5/36 = P(ABc) → A and Bc are independent.

Event Description Set N() P()

S sample space {1,2,3,4,5,6}2 36 1
A First roll is a 6 {6}×{1,2,3,4,5,6} 6 1/6
B Second roll is a 5 {1,2,3,4,5,6}×{5} 6 1/6
Bc Second roll is a not 5 {1,2,3,4,5,6}×{1,2,3,4,6} 30 5/6
AB {(6,5)} 1 1/36
ABc {6}×{1,2,3,4,6} 5 5/36
 13
Independence and Complement Theorem
Let A and B be events in a sample space S.
If A and B are independent events, then so are A and Bc

Proof:
Let A and B be independent events in a sample space S
i.e. such that P(AB) = P(A).P(B)
We want to show than A and Bc are independent,
i.e. that P(A  Bc) = P(A).P(Bc)
 14
Independence and Complement Theorem
Look at relationship between (A  B) and (A  Bc): S A AC
1) (AB)  (ABc) = A  (BBc) = A
P(A) = P(AB) + P(ABc) - P((AB)  (ABc) ) B AB
2) (AB)  (ABc) = (AA)  (BBc) =  BC A  BC
P((AA)  (BBc))=0
Substituting 2) in 1) we get:
P(A) = P(A  B) + P(A  Bc)
P(A  Bc) = P(A)- P(A  B)
=P(A)-P(A).P(B) since A and B are independent
=P(A).(1-P(B))
=P(A).P(Bc) by probability properties 
 15
Independence and Subset Operation
Question:
Let A and B be events in a sample space S with non-zero probability
If B  AC, are A and B independent?
Answer: A AC
Since B  Ac then AB=
 P(A  B) = 0  P(A).P(B) B
 A and B are not independent
 16
Independence and Disjointedness
Theorem:
Let A and B be events in a sample space S with non-zero probability
If AB= then A and B are not independent.
Proof: A AC
AB=
 P(A  B) = 0  P(A).P(B) B
 A and B are not independent
 17
4.5 Independence of Events
✓Independent Events
✓Independence and Set Operations
➢Pairwise and Mutual Independence
 18
Mutual Independence - Example P44
Look at the independence of the events in this question:
When rolling a die 3 times, what is the probability of rolling three 6s?
Event Description Set N() P()
S sample space {1,2,3,4,5,6}3 216 1
A1 1st roll is a 6 {6}×{1,2,3,4,5,6}2 36 1/6
A2 2nd roll is a 6 {1,2,3,4,5,6}×{6}×{1,2,3,4,5,6} 36 1/6
A3 3rd roll is a 6 {1,2,3,4,5,6}2×{6} 36 1/6
A1A2A3 6 in all 3 rolls {(6,6,6)} 1 1/216
P(A1).P(A2).P(A3) = (1/6).(1/6).(1/6) = 1/216 = P(A1A2A3)
 19
Pairwise Independence
Let A, B, C be events in a sample space S.

A, B, and C are pairwise independent iff they satisfy all the following
conditions:
• P(A  B) = P(A).P(B)
• P(B  C) = P(B).P(C)
• P(A  C) = P(A).P(C)
 20
Mutual Independence
Let A, B, C be events in a sample space S.

A, B, and C are mutually independent iff they are pairwise independent

and P(A  B  C) = P(A).P(B).P(C)

Generally: events A1 to An are mutually independent iff the probability

of the intersection of any subset of the events is the product of the
probabilities of the events in the subset.
 21
Pairwise and Mutual Independence
• It is possible for events to be pairwise independent but not mutually
independent:
• N(S)=9, S
• P(A)=P(B)=P(C)=1/3 A
• P(A  B) = P(A).P(B)=1/9 B
• P(B  C) = P(B).P(C) = 1/9
• P(C  A) = P(C).P(A) = 1/9
• P(ABC) = P() = 0  P(A).P(B).P(C) C
 22
Mutual Independence - Example P55
A coin is loaded so that P(head)=P(H)=0.6 and P(tail)=P(T)=0.4. The coin is
tossed 10 times. What is the probability of obtaining 7 heads?
• Count the tosses as i=1 to 10, and define:
• Hi=ith toss is a head P(Hi)=P(H)=0.6
• Ti=ith toss is a tail P(Ti)=P(T)=0.4
 23
Mutual Independence - Example P55
• Look at one possibility: HHHTHHTTHH. because the tosses are mutually
independent P(HHHTHHTTHH)=P(H)7P(T)3
10
• However, there are different ways of tossing 7 heads and 3 tails
7
• They are all mutually exclusive, so we can add their probabilities.
However, they all have the same probability, so we can multiply that
10
probability by instead.
7
10
• I.e. P(7H+3T) = . 0.67 . 0.43
7
 24
Bernoulli Trials
Definitions:
• A Bernoulli trial (or binomial trial) is a random experiment with
exactly 2 possible outcomes, usually labelled “success” and “failure”,
in which the probability of success is the same every time the
experiment is conducted.
• Independent repeated trials of a Bernoulli trial are called Bernoulli
trials.
Theorem:
• In Bernoulli trials with n trials where the probability of success is p,
the probability of exactly k successes is 𝑛𝑘 𝑝𝑘 (1 − 𝑝)𝑛−𝑘