CONCEPTUAL TOOLS
By: Neil E. Cotter PROBABILITY
CONDITIONAL PROBABILITY
Discrete random variables
EXAMPLE 4
EX: The following formulas define the behavior of conditional probabilities:
P(A,B) P(A and B) P(A ∩ B)
P(A | B) = ≡ ≡ (always true)
P(B) P(B) P(B)
P(A | B) = P(A) (if A and B independent)
P(A,B) = P(A)P(B) (if A and B independent)
€
€ For the following formulas, determine whether the formula is always true, could be
true in certain cases (be sure to consider all possible ways events A and B might be
€
defined), or is never true. Justify each answer mathematically.
a) P(A | B') = 1− P(A | B)
P(A,B,C)
b) P(A,B | C) =
P(C)
€
c) P(A | B,C) = P(A,B | B,C)
€ d) P(A,B,C) = P(A | B,C)P(B | C)P(C)
€
SOL'N: a) We first rearrange the stated formula:
€
P(A | B) + P(A | B') = 1
Using the formula for conditional probability, we substitute for each of the
terms on the left side of the equation:
€
P(A ∩ B) P(A ∩ B')
+ =1
P(B) P(B')
The numerators resemble the law of total probability:
€ P(A ∩ B) + P(A ∩ B') = P(A)
We also have P(B') = 1− P(B) . If P(B) = 12 , then P(B') = 12 and we have
€ P(A ∩ B) P(A ∩ B')
1
+ 1
= 2P(A)
€ 2 2
€ €
which is valid if P(A) = 12 . Thus, it is possible for the equation to
be valid.
€
Another case that works is when A = B. Then we have the following:
€
CONCEPTUAL TOOLS
By: Neil E. Cotter PROBABILITY
CONDITIONAL PROBABILITY
Discrete random variables
EXAMPLE 4 (CONT.)
P(A | B) + P(A | B') = P(A | A) + P(A | A') = 1+ 0 = 1
The equation need not always be true, however. Consider the following
example of rolling a fair, 6-sided die:
€
A ≡ The number rolled on the die is 1
B ≡ The number rolled on the die is odd
B' ≡ The number rolled on the die is even
In this case, P(A | B) = 13 and P(A | B') = 0 . The sum is not equal to 1, and
the equation is invalid. Thus the equation is sometimes true.
Exploring further, we
€ might ask if there are restrictions on possible values
€
of P(A) and P(B). We consider the case that P(B) = 13 . Then P(B') = 23
and we have the following equation:
P(A ∩ B) P(A ∩ B')
1
+ 2
=1 € €
3 3
This will be satisfied if, for example, P(A ∩ B) = 16 and P(A ∩ B') = 13 .
Using the example of rolling a fair, 6-sided die again, we could define A
€
and B as follows:
€ €
A ≡ The number rolled on the die is odd
B ≡ The number rolled on the die is 1 or 2
B' ≡ The number rolled on the die is 3, 4, 5, or 6
We conclude that any value of P(B) should work if event A is chosen
properly. From the examples given above, it also appears likely that P(A)
may have any value if event B is chosen properly. Thus, giving a precise
definition of the relationship between P(A) and P (B ) that is more
informative than the equation in the problem statement is a difficult task.
b) We use the first equation for calculating conditional probability:
P((A,B),C) P((A ∩ B) ∩ C)
P(A,B | C) = =
P(C) P(C)
P(A ∩ B ∩ C) P(A,B,C)
= =
P(C) P(C)
€
CONCEPTUAL TOOLS
By: Neil E. Cotter PROBABILITY
CONDITIONAL PROBABILITY
Discrete random variables
EXAMPLE 4 (CONT.)
P(A,B,C)
We see that P(A,B | C) = is always true. We read (A, B) as
P(C)
A and B or as A ∩ B . We may define this as a new event that is the
intersection of two events. It may be helpful to think of a concrete
€
example:
€
A ≡ It will snow today
B ≡ It will be very cloudy today
C ≡ It is cold today, (as the high temperature today will be < 32° F)
In this case, (A, B) is the event that it will snow and be very cloudy today.
c) We use the first equation for calculating conditional probability applied to
the expression on the right side:
P((A,B),(B,C)) P((A ∩ B) ∩ (B ∩ C))
P(A,B | B,C) = =
P(B,C) P(B ∩ C)
P(A ∩ (B ∩ B) ∩ C) P(A ∩ B ∩ C)
= =
P(B ∩ C) P(B ∩ C)
P(A ∩ (B ∩ C))
= = P(A | B,C)
P(B ∩ C)
Thus, the equation is always true. The key step in the derivation is
realizing that B ∩ B = B .
€
Using the concrete example from (b), we can translate P(A,B | B,C) into
words as the "probability that it will snow today and be very cloudy today
€
given that it is very cloudy and cold today". It makes sense that, since we
know it is cloudy today, the probability that it €
will snow and be cloudy is
the same as the probability that it will snow. An even simpler example of
this would be the equation P(B | B) = 1. If we are given that it is very
cloudy today, the probability that it is very cloudy today is equal to 1.
€
d) This equation is always true, and it follows from repeated application of
the first equation for conditional probability rearranged slightly:
P(A,B) = P(A | B)P(B)
€
CONCEPTUAL TOOLS
By: Neil E. Cotter PROBABILITY
CONDITIONAL PROBABILITY
Discrete random variables
EXAMPLE 4 (CONT.)
Working from right to left in the formula as stated in the problem we have
the following:
P(B,C) = P(B | C)P(C)
Proceeding to the next step, we have the final result we seek:
€ P(A | B,C)P(B | C)P(C) = P(A | B,C)P(B,C) = P(A,B,C)
Thus, we see that the equation stated in the problem always holds.