The Infinite Actuary 1/P Sample Exam 1 Solutions

1. 75% of the customers of ACME Mutual Insurance have auto insurance, and 40% have homeowners
insurance. What is the maximum possible probability that a randomly selected customer with auto
insurance does not have homeowners insurance?

A. 20% B. 40% C. 60% D. 80% E. 100%

Let A be the set of customers that have auto insurance.

Let H be the set of customers that have homeowners insurance.
If a customer is selected at random and we are told that the selected individual has auto insurance,
then the probability that this individual also has homeowners insurance is

P[H C and A]
P[H C | A] = .
P[A]

We know that the probability of A is 0.75. What, then, is the maximum possible value of P[H C ∩ A]?
Well, H C is a larger set than A, since it takes up the larger proportion of all the policy holders. So,
the largest overlap will occur when A is wholly contained in H C . In this case, P[H C ∩ A] = P[A] = 0.4.
Now,
P[H C and A] 0.4
P[H C | A] = ≤ = 0.8 ,
P[A] 0.75
leading to answer choice D.

2. Suppose Z is a normal random variable with E[Z] = 2 and coefficient of variation 3. Find P[Z > 1].

A. 0.05 B. 0.34 C. 0.57 D. 0.66 E. 0.95

From the definition of the coefficient of variation, we have

SD of Z SD of Z
3= =
E[Z] 2

So, SD of Z = 6.
Having the mean and standard deviation of Z in hand, we know how to transform it into a standard
normal, and then use the normal table to finish.
 
Z −2 1−2
P[Z > 1] = P >
6 6
= P[Standard Normal > −1/6]
= 1 − Φ(−1/6)
= 1 − (1 − Φ(1/6)) = Φ(1/6) ≈ 0.57

TIA 1/P Seminar p. 1 Sample Exam 1

3. Let X and Y represent the lifetimes in hours of two components in an electronic device, rounded up
to the nearest hour. The joint distribution of X and Y is uniform on possible values, subject to the
constraints that X and Y are both positive, and their sum is less than 6. What is the variance of X?

A. 1 B. 5/4 C. 4/3 D. 3/2 E. 2

There are 10 possible values for (X, Y ), namely (1, 1), (1, 2), (1, 3), (1, 4), (2, 1), (2, 2), (2, 3), (3, 1), (3, 2)
and (4, 1). In particular, P[X = 1] = 4/10, P[X = 2] = 3/10, P[X = 3] = 2/10 and P[X = 4] = 1/10,
so

E[X] = 0.4 · 1 + 0.3 · 2 + 0.2 · 3 + 0.1 · 4

E[X] = 2
E[X 2 ] = 0.4 · 12 + 0.3 · 22 + 0.2 · 32 + 0.1 · 42
E[X 2 ] = 5
Var[X] = E[X 2 ] − (E[X])2 = 5 − 22
Var[X] = 1

4. The probability that Rafael Nadal wins a tennis match in straight sets is 70%. Assuming that the
outcome of each match is independent, what is the probability that in his next 7 matches that he will
win in straight sets at least 5 times?

A. 0.13 B. 0.33 C. 0.44 D. 0.65 E. 0.96

Let X be the number of times that Nadal wins in straight sets. Then X is the sum of 7 independent
trials, in which each trial will be 1 with probability 0.7, and 0 with probability 0.3. That means X is
a binomial random variable with n = 7 and p = 0.7.
Now we can compute.

P[X ≥ 5] = P[X = 5] + P[X = 6] + P[X = 7]

   
7 5 2 7
= (0.7) (0.3) + (0.7)6 (0.3)1 + (0.7)7
5 6
= 0.318 + 0.247 + 0.082 ≈ 0.65 ,

leading to answer choice D.

TIA 1/P Seminar p. 2 Sample Exam 1

5. Suppose that Z is a normal random variable with mean 5 and variance σ 2 . Let fZ (z, σ) denote the
density of z for a given σ, and let FZ (z, σ) denote the CDF of Z for a given σ. Which of the following
is increasing as σ increases?

A. fZ (5, σ)
B. FZ (3, σ)
C. FZ (8, σ)
D. The 25th percentile of Z
E. None of the above

Let’s break down each choice:

−(5−µ)2
A. fZ (5, σ) = √1 1 e 2σ 2 = c
which is decreasing as σ increases.
2π σ σ

3−5 −2 −2
will go to 0, and FZ (3, σ) will increase to 21 .



B. FZ (3, σ) = φ σ =φ σ . As σ increases, σ

8−5
will decrease down to φ(0) = 21 .


C. For FZ (8, σ) = φ σ

D. The 25th percentile will be roughly 5 − 0.67σ and will go to −∞ as σ increases. So it decreases
rather than increases.

6. X and Y are integer valued random variables such that for x and y integers, the joint cdf of X, Y is
given by
F (x, y) = 1 − e−x − (1 − e−x(y+1) )/(y + 1)
for x ≥ 0, y ≥ 0. Find P[1 < X ≤ 2, 1 < Y ≤ 3].

A. 0.05 B. 0.17 C. 0.19 D. 0.41 E. 0.61

Recall that F (s, t) = P[X ≤ s, Y ≤ t]. The fact that X and Y must be integers means we want
P[X = 2, Y = 2 or 3]. F (2, 3) − F (1, 3) = P[X ≤ 2, Y ≤ 3] − P[X ≤ 1, Y ≤ 3] = P[X = 2, Y ≤ 3].
That’s almost what we want, but includes {X = 2, Y ≤ 1}. We can use a similar expression to take
that out:

P[X = 2, Y ≤ 1] = F (2, 1) − F (1, 1).

Combining, we can say

   
P[X = 2, 1 < Y ≤ 3] = F (2, 3) − F (1, 3) − F (2, 1) − F (1, 1)
= F (2, 3) − F (1, 3) − F (2, 1) + F (1, 1)
≈ 0.05

leading to answer choice A.

TIA 1/P Seminar p. 3 Sample Exam 1

7. A fair die is rolled repeatedly. Let X be the number of rolls needed to obtain 6, and let Y be the
number of rolls needed to obtain an even number.

Find E[X | Y = 5].

A. 5 B. 6 C. 7 D. 8 E. 9

Method 1: Start with the definition, and switch to the survival function method.
X
E[X | Y = 5] = xP[X = x | Y = 5]
x
∞
X
= P[X ≥ n | Y = 5].
n=1

Notice that it is possible that X is the same as Y , since 6 is an even number, but that X cannot be
smaller than Y . Therefore, P [X ≥ n | Y = 5] = 1 for n = 1, 2, 3, 4 and 5.
Since we know Y = 5, then the 6th roll is an even number. If the 6th roll is a 6, then X = 5. Given that
we roll an even number, the probability that this number is a 6 is 1 in 3. Thus, P[X ≥ 6 | Y = 5] = 2/3,
as it is the complementary event.
Next, P[X ≥ 7 | Y = 5] = P[X 6= 5 and X 6= 6] = 2/3 · 5/6.
One more to find the pattern:

P[X ≥ 8 | Y = 5] = P[X 6= 5 and X 6= 6 and X 6= 7] = 2/3 · 5/6 · 5/6.

Now we can compute our sum:

∞
X
P[X ≥ n|Y − 5] = 1 + 1 + 1 + 1 + 1 + 2/3 · 5/6 + 2/3 · (5/6)2 + · · ·
n=1

With what we know of geometric series (see the one that starts at term 6?), we now have
∞
X 2/3
P[X ≥ n | Y − 5] = 5 + = 9,
1 − 5/6
n=1

leading to answer choice E.

Method 2: Split cases according to whether X = 5 or not.

E[X | Y = 5] = E[X | Y = 5, X = 5]P[X = 5 | Y = 5]

+ E[X | Y = 5, X ≥ 6]P[X > 5 | Y = 5]

If we are given that X = 5, then E[X | Y = 5, X = 5] had better be 5.

As we mentioned above, P[X = 5 | Y = 5] = 1/3, and P[X > 5 | Y = 5] = P[X ≥ 6 | Y = 5] = 2/3.
Finally, given X > 5, X is at least 6, so (X − 5 | X > 5) is at least 1. The memoryless property
says that (X − 5 | X > 5) is a geometric starting at 1, with the same p as X, namely p = 1/6, so
E[X − 5 | X > 5] = 1/(1/6) = 6, and E[X | X > 5] = E[X − 5 | X > 5] + 5 = 5 + 6.
Putting it all together, E[X | Y = 5] = 1/3 · 5 + 2/3 · (5 + 6) = 5 + 2/3 · 6 = 9

TIA 1/P Seminar p. 4 Sample Exam 1

8. Suppose that X and Y are random variables with E[X] = E[Y ] = 1 and Var[X] = Var[Y ] = 4. If
Cov(X, Y ) = 3, find Var[3X − 2Y > 0].

A. 0 B. 8 C. 16 D. 34 E. 52

Var(2X − 3Y ) = 9Var(X) − 2 · 3 · 2 · Cov(X, Y ) + (−2)2 Var(Y )

= 9 · 4 − 12 · 3 + 4 · 4 = 16

9. Suppose that I roll two independent dice, one red and one blue. Let A be the event that the blue die
is even, B the event that the red die is even, and C the event that the sum is even. Which of the
following is true?

A. None of them are independent.

B. A and B are pairwise independent, but neither is pairwise independent of C.
C. A and C are pairwise independent, as are B and C, but A and B are not pairwise independent.
D. All three pairs are pairwise independent, but it is not true that all three are mutually independent.
E. All three are mutually independent.

We can easily see that both P[A] and P[B] are 1/2.
To see that P[C] = 1/2 as well, note that this is true regardless of what happens with the first die. If
the first die is even, then there are three possibilities for the second die to be even, so C will occur half
of the time. On the other hand, if the first die is odd, then there are three possibilities for the second
die to be odd and form an even sum, so C will still occur half of the time.
Events A and B are independent because

P[A ∩ B] = P[Red die is even AND Blue die is even] = 1/2 · 1/2 = P[A]P[B]

So answer choice A is false, and answer choice C is also false.

Next, consider that

P[A ∩ C] = P[Red die is even AND Sum is even]

= P[Red die is even AND blue die is even]
= 1/2 · 1/2 = P[A]P[C].

So events A and C are also pairwise independent, eliminating answer choice B. In fact, events B and C
are also pairwise independent by a similar argument, but both of our remaining answer choices already
contain this information, so we know it must be true.
Finally, consider

P[A ∩ B ∩ C] = P[Red die is even AND Blue die is even AND sum is even]
= P[Red die is even AND Blue die is even]
= 1/2 · 1/2 6= P[A]P[B]P[C].

This means that the events are not mutually independent, which confirms answer choice D and elimi-
nates E simultaneously.

TIA 1/P Seminar p. 5 Sample Exam 1

10. Beedle is a salesman and each month has a chance of getting a bonus of 20. During the 3 summer
months, he has a 40% chance of getting a bonus, while during the other months he only has a 30%
chance of getting a bonus. Whether or not he gets a bonus in a month is independent of what happened
during other months. What is the variance of the sum of Beedle’s monthly bonuses, summed over a
whole year?

A. 52 B. 56 C. 60 D. 1044 E. 1125

The number of bonuses he will receive during the summer is binomial with n = 3 and p = 0.4. The
number of non-summer bonuses is binomial with n = 9 and p = 0.3. If N is the total number of
monthly bonuses he gets, then (using the fact that the binomial variance is np(1 − p)):

Var[N ] = 3(0.4)(1 − 0.4) + 9(0.3)(1 − 0.3)

Var[N ] = 2.61
Var[20N ] = 202 · 2.61
= 1044

11. A small shipping company has 5 trucks. The number of accidents that each truck has per year has
a Poisson distribution with mean 1.5. If each accident costs the company $250, and the number of
accidents per truck are independent, what is the standard deviation of the annual cost of accidents to
the shipping company?

A. 685 B. 839 C. 1022 D. 1286 E. 1531

Let N denote the total number of accidents. The distribution of N is N ∼ Poisson(5 × 1.5) because
there are 5 trucks. We want to know the standard deviation of the annual cost; the annual cost is
Cost = 250N and the variance is Var[Cost] = 2502 Var[N ] = 2502 (7.5) = 468,750. That means that
√
our standard deviation is SD[Cost] = 468,750 = 684.7

12. A student who is taking a 30 question multiple choice test knows the answers to 24 of the questions.
Whenever the student doesn’t know the answer to a question, he chooses uniformly from one of the
5 choices. Given that the student gets a randomly chosen question right, what is the probability that
the student guessed on the question?

A. 1/25 B. 1/21 C. 1/18 D. 1/11 E. 1/5

We are interested in determining, for a question chosen at random,

P[Guessed | Answered correctly].

TIA 1/P Seminar p. 6 Sample Exam 1

Use the definition of conditional probability to obtain
P[Guessed AND Answered correctly]
P[Guessed | Answered correctly] = .
P[Answered correctly]
In this case the intersection is easier. We can just condition the “other” way.
P[Guessed AND Answered correctly]
= P[Answered correctly | Guessed]P[Guessed]
= 1/5 · 6/30.
To find the total probability that he answered the randomly chosen question correctly, note that
P[Answered correctly] = P[Guessed AND Answered correctly]
+ P[He Did NOT Guess AND Answered correctly].
The first of these we know. The second is easy, since he correctly answers each of the 24 questions on
which he does not guess. Putting it all together, then,
P[Guessed AND Answered correctly]
P[Guessed | Answered correctly] =
P[Answered correctly]
6/30 · 1/5 0.04 1
= = = ,
6/30 · 1/5 + 24/30 0.84 21
leading to answer choice B.

13. An insurance company sells exactly two types of insurance, homeowners and auto insurance. 55% of
their customers have homeowners insurance and 30% have both types of insurance. What fraction of
the customers have auto insurance?

A. 15% B. 25% C. 70% D. 75% E. 85%

Let A denote the number of customers who have auto insurance and let H denote the number of cus-
tomers who have homeowners insurance. We are told that 30% of customers have both types of
insurance, P[A and H] = .3. We are also told that 55% of customers have homeowners. Since we know
30% of the 55% is customers who have both, the number of customers who have only homeowners
insurance is P[H only] = .55 − .3 = .25. Now, we can find the fraction of customers who have auto
insurance:
P[A] + P[A0 ] = 1
P[A] + .25 = 1
P[A] = 0.75

14. If X, Y , and Z are i.i.d. Poisson random variables with mean 3, what is E[(X + Y + Z)2 ]?

A. 27 B. 54 C. 63 D. 90 E. 121

TIA 1/P Seminar p. 7 Sample Exam 1

The sum of independent Poissons is Poisson, so X + Y + Z is Poisson with mean 3 · 3 = 9. The variance
of a Poisson equals its mean, so is also 9, and the second moment is 92 + 9 = 90

Or because X is Poisson, E[X] = 3 = Var[X]. E[X 2 ] = Var[X] + (E[X])2 = 3 + 9 = 12. Now we can
use the independence and compute:

E[(X + Y + Z)2 ] = E[X 2 + Y 2 + Z 2 + 2XY + 2XZ + 2Y Z]

= E[X 2 ] + E[Y 2 ] + E[Z 2 ] + 2(E[XY ] + E[XZ] + E[Y Z])
= 3E[X 2 ] + 2 · 3E[X]E[Y ]
= 3 · 12 + 2 · 3 · 3 · 3 = 90

Answer choice D

15. Let U and V be independent uniform random variables on the set {1, 2, 3, 4, 5}. Find P[min{U, V } ≤
2 | max{U, V } > 2].

A. 2/5 B. 4/7 C. 2/3 D. 8/11 E. 12/25

There are 12 cases that work: U ∈ {1, 2} and V ∈ {3, 4, 5}, which is 6 cases, and another 6 from swap-
ping U and V . There are 4 cases in which max{U, V } ≤ 2 (namely both U and V are 1 or 2, so 21
cases in which max{U, V } > 2. As all cases are equally likely (by the uniform statement), we get 12/21
as our final answer.

Or using the definition of conditional probability,

P[min{U, V } ≤ 2 and max{U, V } > 2]
P[min{U, V } ≤ 2 | max{U, V } > 2] =
P[max{U, V } > 2]

For maximums, we prefer less than or equal statements, so

P[max{U, V } > 2] = 1 − P[max{U, V } ≤ 2]

= 1 − P[U ≤ 2, V ≤ 2]
= 1 − (2/5) · (2/5) = 21/25

In order that the minimum be less than 2 while the max is greater than 2, either U is small and V is
large or vice versa.

P[min{U, V } ≤ 2 and max{U, V } > 2] = P[U ≤ 2, V > 2] + P[U > 2, V ≤ 2]

2 3 3 2 12
= · + · =
5 5 5 5 25
Finally,
12/25
P[min{U, V } < 2 | max{U, V } > 2] =
21/25
12 4
= =
21 7

TIA 1/P Seminar p. 8 Sample Exam 1

which is answer choice B.

16. Suppose that X and Y are independent, Poisson random variables with E[X] = 2 and E[Y ] = 2.8.
Find P[X + Y < 3].

A. 0.14 B. 0.15 C. 0.21 D. 0.25 E. 0.29

The key idea for this problem is to remember that independent Poisson random variables sum to a
Poisson whose mean is the sum of the means. In this case that means X + Y ∼ Poisson(2 + 2.8).
Now it’s easy.

P[X + Y < 3] = e−λ [1 + λ + λ2 /2]

= 0.0082 · (1 + 4.8 + 4.82 /2)
= 0.0082 · 17.31
≈ 0.14

which is answer choice A.

17. Let N be a randomly chosen integer with 1 ≤ N ≤ 1, 000. What is the probability that N is not
divisible by 7, 11, or 13?

A. 0.66 B. 0.69 C. 0.72 D. 0.75 E. 0.78

We seek the probability of the complement of a union of 3 sets, the set of integers less than 1000 that
are divisible by 7 (Set A), 11 (Set B), or 13 (Set C). To compute the probability of this union, we use
the inclusion-exclusion principle with three sets:

P[A ∪ B ∪ C] = P[A] + P[B] + P[C] − P[A ∩ B] − P[A ∩ C] − P[B ∩ C] + P[A ∩ B ∩ C]

Since 7, 11, and 13 share no common divisors (other than 1), it follows that in order for a number to
be divisible by all three, it would need to be divisible by the product of all three. But 7 · 11 · 13 = 1001,
which means that no number of size 1000 or below can be easily divisible by all three numbers. So
P[A ∩ B ∩ C] = 0.
Similarly, any number that is to be divisible by any two of 7,11, and 13 must be divisible by the
respective product of the two. From this it follows that

P[A] = (b1000/7c)/1000 = 0.142

P[B] = (b1000/11c)/1000 = 0.090
P[C] = (b1000/13c)/1000 = 0.076
P[A ∩ B] = (b1000/(7 · 11)c)/1000 = 0.012
P[A ∩ C] = (b1000/(7 · 13)c)/1000 = 0.010
P[B ∩ C] = (b1000/(11 · 13)c)/1000 = 0.006

Now P[A ∪ B ∪ C] = 0.280, so our answer is 1 − 0.280 = 0.720 , answer choice C.

TIA 1/P Seminar p. 9 Sample Exam 1

18. Insurance losses L in a given year have a lognormal distribution with L = eX , where X is a normal
random variables with mean 3.9 and standard deviation 0.8. If a $100 deductible and a $50 benefit
limit are imposed, what is the probability that the insurance company will pay the benefit limit given
that a loss exceeds the deductible?

A. 0.10 B. 0.27 C. 0.43 D. 0.66 E. 0.88

In order to pay the benefit limit of $50, the losses L must be at least $50 more than the $100 deductible,
or L ≥ 150. Being given that L > 100,

P[L ≥ 150 and L > 100]

P[L ≥ 150|L > 100] =
P[L > 100]
P[L ≥ 150]
=
P[L > 100]
P[eX ≥ 150]
=
P[eX > 100]
P[X ≥ ln(150)]
=
P[X > ln(100)]
h i
ln(150)−3.9
P X−3.9
0.8 ≥ 0.8
= h i
ln(100)−3.9
P X−3.9
0.8 > 0.8
 X−3.9 
P ≥ 1.39
=  0.8
P X−3.9

0.8 > 0.88
1 − Φ(1.39)
= ≈ 0.43 ,
1 − Φ(0.88)

answer choice C.

19. A fair 6-sided die is rolled 1,000 times. Using a normal approximation with a continuity correction,
what is the probability that the number of 3’s that are rolled is greater than 150 and less than 180?

A. 0.78 B. 0.81 C. 0.84 D. 0.88 E. 0.95

The number of 3’s rolled out of 1,000 trials is a binomial random variable X with n = 1000, and p = 1/6.
The mean, E[X] is 1000/6, and the variance, Var(X) = 100 · 1/6 · 5/6.
This X may be approximated by a normal random variable N with the same mean µN = 1000/6 and
standard deviation σN = 1000 · 1/6 · 5/6. But the range of this normal random variable is all real
numbers, while the range of values for X is just the integers from 0 to 1000. To adjust for this, we
approximate
P[150 < X < 180] ≈ P[150.5 ≤ N ≤ 179.5].

TIA 1/P Seminar p. 10 Sample Exam 1

That is, we don’t want to include any values of N that would round down to 150 or up to 180, neither
of which is included in our event {150 < X < 180}.
Now we proceed to use our normal table.
 
150.5 − µN N − µN 179.5 − µN
P[150.5 ≤ N ≤ 179.5] = P ≤ ≤
σN σN σN
= Φ(1.09) − Φ(−1.37) ≈ 0.78 ,

answer choice A.

20. Four red dice and six blue dice are rolled. Assuming that all ten dice are fair six sided dice, and rolls
are independent, what is the probability that exactly three of the red dice are even, and exactly two
of the blue dice come up ones?

A. 0.05 B. 0.10 C. 0.16 D. 0.21 E. 0.27

There are 43 = 4 ways for exactly three red dice to be even, and 62 = 15 ways for exactly 2 of the

6 blue dice to come up ones. The probability of each way for exactly three red dice to be even is
(1/2)3 (1/2)1 , while the probability of each way for exactly 2 of the blue dice to come up ones is
(1/6)2 (5/6)4 . Our answer, then, is

(4 · (1/2)3 (1/2)1 )(15 · (1/6)2 (5/6)4 ) ≈ 0.05 ,

answer choice A.

21. A life insurance company classifies its customers as being either high risk or low risk. If 20% of the
customers are high risk, and high risk customers are three times as likely as low risk customers to file
a claim, what percentage of claims that are filed come from high risk customers?

A. 30% B. 37% C. 43% D. 54% E. 60%

Let p be the probability of a claim by a low risk customer. Then the probability of a claim by a high
risk customer is 3p.
The probability of a claim from a randomly chosen customer is, then,

P[claim] = P[High]P[claim | High] + P[Low]P[claim | Low]

1 4
= · 3p + · p.
5 5
Another way to state “the percentage of claims that are filed that come from high risk customers” is
“the probability that given a claim is made, the customer making the claim is a high risk customer.”
We may therefore compute

P[High and claim]

P[High | claim] =
P[claim]

TIA 1/P Seminar p. 11 Sample Exam 1

 1
5 · 3p
= 1 = 3/7 ≈ 0.43 ,
5 · 3p + 54 · p

leading to answer choice C.

22. Suppose that X1 , . . . , X100 are random variables with E[Xi ] = 100 and E Xi2 = 10,100. If Cov(Xi , Xj ) =

−1 for i 6= j, what is Var[S], where S = 100

P
i=1 Xi ?

A. 0 B. 100 C. 1,000 D. 5,050 E. 10,000

The variance of a sum S of 100 random variables Xi may be written

100
X X
Var(S) = Var(Xi ) + Cov(Xi , Xj )
1 i6=j
100
X X
= Var(Xi ) + 2 Cov(Xi , Xj )
1 i>j

This is because in the index collection for the first covariance sum both i = 1, j = 2 and i = 2, j = 1,
for example, are needed, while in the second covariance sum, only i = 2, j = 1 would appear. This
means each pair of numbers below 100 only appears
one time in the index of the second covariance
sum. Thus the number of terms in that sum is 100 2 .
Fortunately, the variance of each Xi and the covariance of each pair Xi , Xj is the same, so

Var(Xi ) = E[Xi2 ] − (E[X])2 = 10,100 − 10,000 = 100,

and
 
100
Var(S) = 100 · 100 + (−1) ·2
2
= 100 · 100 − 100 · 99 = 100(100 − 99) = 100 .

This is answer choice B.

23. Let X be a Poisson random variable with second moment 6. Find P[X = 3].

A. 0.06 B. 0.12 C. 0.18 D. 0.20 E. 0.22

For a Poisson random variable, its distribution is determined by the parameter λ, which is the mean
of the variable. We are given the second moment, so we need to derive a value for λ from the second
moment.

X ∼ Poisson(λ)
E(X) = Var(X) = λ
E[X 2 ] = λ + λ2

TIA 1/P Seminar p. 12 Sample Exam 1

 6 = λ + λ2
0 = λ2 + λ − 6
0 = (λ + 3)(λ − 2)
λ = 2 or −3

We know λ must be greater than or equal to 0, so we get λ = 2. Now, the probability P[X = 3]:

λ3
P[X = 3] = e−λ
3!
−2 8
=e
6
= 0.18

24. The cdf of a random variable X satisfies

2002
F (x) = 1 −
(x + 200)2

for x > 0. Find P[50 < X < 300 | X > 100].

A. 0.22 B. 0.36 C. 0.51 D. 0.64 E. 0.78

Start from the definition, and work toward an expression with the CDF.

P[50 < X < 300 & X > 100]

P[50 < X < 300 | X > 100] =
P[X > 100]
P[100 < X < 300]
=
P[X > 100]
F (300) − F (100)
=
1 − F (100)

2  200 2

1 − 200


500 − 1− 300
=  2 ≈ 0.64 ,
200
100+200

answer choice D.

25. Loss amounts have a continuous distribution that ranges from 0 to 3. The density of that distribution,
when positive, is proportional to the square of the loss amount. Find the 80th percentile of losses.

A. 0.9 B. 1.3 C. 1.8 D. 2.3 E. 2.8

Being proportional to y 2 means that fY (y) = cy 2 for some constant c. Since the total probability must

TIA 1/P Seminar p. 13 Sample Exam 1

be 1,
3
3
cy 3 
Z
2
1= cy dy = = 9c,
0 3 0
so c = 1/9.
To find the 80th percentile,

P[Y < t] = 0.8

Z t
y 2 /9 dy = 0.8
0
t3 /27 = 0.8
t3 = 27 · 0.8
t ≈ 2.8 ,

which is answer choice E.

26. Suppose that (X, Y ) are uniformly chosen from the set of integers given by 0 ≤ X ≤ 3 and X ≤ Y ≤ X 2 .
Find P[Y ≤ 3].

1 1 5 1 2
A. B. C. D. E.
4 3 12 2 3

Since we are choosing uniformly, we want to know how many cases there are. They are:

x=0 y=0
x=1 y=1
x=2 y = 2, 3, 4
x=3 y = 3, 4, . . . , 9

That is 1 + 1 + 3 + 7 = 12 cases. Of those, Y ≤ 3 in 1 + 1 + 2 + 1 = 5 of them, for an answer of 5/12 ,

answer C.

27. If X is a Poisson random variable with P[X = 1] = 2.5P[X = 0], then what is the probability that X
will be within 1 standard deviation of E[X]?

A. 0.08 B. 0.29 C. 0.47 D. 0.63 E. 0.81

We know the forms for these probabilities for Poisson random variables.

P[X = 1] = 2.5P[X = 0]
λe−λ = 2.5e−λ
λ = 2.5

So the mean and the variance of X are both 2.5, and the standard deviation is approximately 1.58.
Now we need to find P[2.5 − 1.58 < X < 2.5 + 1.58] = P[X = 1, 2, 3, 4]. Again, using the Poisson

TIA 1/P Seminar p. 14 Sample Exam 1

distribution,

P[X = 1, 2, 3, 4] = e−λ λ + λ2 /2 + λ3 /3! + λ4 /4!

= e−2.5 2.5 + 2.52 /2 + 2.53 /3! + 2.54 /4! ≈ 0.809 ,

which corresponds to answer choice E.

28. For 0 ≤ x ≤ 2 and 0 ≤ y ≤ 2, the joint probability mass function of X and Y is

P[(X, Y ) = (x, y)] = c(6 − x − 2y),

and the joint probability mass function is 0 otherwise. Find P[X + Y ≥ 3].

A. 1/9 B. 1/8 C. 1/6 D. 1/5 E. 1/3

There are 9 possible cases (3 values of X and Y ) and the total probability is 1, which gives

1 = c(6 − 0 − 2 · 0) + c(6 − 0 − 2 · 1) + c(6 − 0 − 2 · 2)

+ c(6 − 1 − 2 · 0) + c(6 − 1 − 2 · 1) + c(6 − 1 − 2 · 2)
+ c(6 − 2 − 2 · 0) + c(6 − 2 − 2 · 1) + c(6 − 2 − 2 · 2)
= 27c
1
c=
27
P[X + Y ≥ 3] = c(6 − 2 − 2 · 1) + c(6 − 1 − 2 · 2) + c(6 − 2 − 2 · 2)
1
= (2 + 1 + 0)
27
1
=
9

29. The joint probability mass function of X and Y is proportional to (x2 + 2xy) for integers such that
0 ≤ x ≤ 2 and x ≤ y ≤ x + 1 and is 0 otherwise. Find E[Y | X = 1].

A. 10/8 B. 11/8 C. 12/8 D. 13/8 E. 14/8

P[X = x, Y = y] = c(x2 + 2xy) when non-zero

P[X = 1, Y = 1]
P[Y = 1 | X = 1] =
P[X = 1]
P[X = 1, Y = 1]
=
P[X = 1, Y = 1] + P[X = 1, Y = 2]
c(12 + 2 · 1 · 1)
=
c(12 + 2 · 1 · 1) + c(12 + 2 · 1 · 2)

TIA 1/P Seminar p. 15 Sample Exam 1

 3 3
= =
3+5 8
c(12 + 2 · 1 · 2)
P[Y = 2 | X = 1] =
c(12 + 2 · 1 · 1) + c(12 + 2 · 1 · 2)
5 5
= =
3+5 8
E[Y | X = 1] = 1 · P[Y = 1 | X = 1] + 2 · P[Y = 2 | X = 1]
3 5
=1· +2·
8 8
13
=
8

for answer choice D

Note that the proportionality constant cancelled, so we didn’t have to find it. To do so, we would have
summed over all possible cases, set that probability equal to 1, and gotten c = 1/96.

30. Suppose that X1 , . . . , X5 are i.i.d., uniform random variables on the set {3, 4, 5, 6}. Let X denote the
average of X1 through X5 , and let σX and µX denote the standard deviation and mean of X. Find
the probability that the minimum and maximum of X1 , . . . , X5 both differ from µX by at most σX .

A. 0.00000 B. 0.00098 C. 0.00186 D. 0.00274 E. 0.03125

To begin, let’s compute µX and σX .

First, recall that

3+6
E[X1 ] = = 4.5, and
2
Var(X1 ) = [(#Possible values)2 − 1]/12 = (42 − 1)/12 = 15/12.

The Xi are i.i.d., so

 
X1 + · · · + X5 1
µX = E = (5E[X1 ]) = E[X1 ] = 4.5, and
5 5
 
2 X1 + · · · + X5 1 1 15
σX = Var = 2 (5 · Var(X1 )) = · = 0.25.
5 5 5 12
√
Now we√ can proceed with the question at hand. µ X − σX = 4.5 − 0.25 = 4 and µX + σX =
4.5 + 0.25 = 5, so we want the min and max to both be 4 or 5. That happens if and only if all 5 of
our values are 4 or 5, which has probability (2/4)5 = 1/32 = 0.03125 which corresponds to answer
choice E.

TIA 1/P Seminar p. 16 Sample Exam 1