ECON1151: Exercises Lecture 4∗

1. Suppose you are interested in a random variable Y which takes on values y1 , . . . , yk with
probabilities p1 , . . . pk . Restate the formula for computing the expected value of Y. Then
assume you were told that each outcome yi is equally likely to occur. Can you restate and
simplify the formula for the expected value?
Solution:
The formula for the expected value of a random variable with outcomes y1 , . . . , yk occurring
with probabilities p1 , . . . pk is
k
X
E[Y ] = yi pi
i=1
Since we know that each outcome is equally likely, we know that all probabilities p1 , . . . pk are
the same and are equal to p. We also know that the sum over all probabilities is equal to 1:
p1 + p2 + · · · + pk = 1 and since they are all the same np = 1 which yields p = 1/k.
We can plug this into our formula
k k k
X X 1X
E[Y ] = yi p = p yi = yi
k
i=1 i=1 i=1

So it becomes clear why the expected value is also called the mean of a random variable be-
cause if all events occur with the same probability the expected value is obtained by calculating
a "standard" mean over all outcomes.
2. Show that the formula for the variance of a random variable V ar[Y ] = E[(Y − µY )2 ], can be
simplified to V ar[Y ] = E[Y 2 ] − (µY )2 .
Solution:
Starting with V ar[Y ] = E[(Y − E[Y ])2 ] we multiply out the square:

E[(Y − E[Y ])2 ] = E[Y 2 − 2Y E[Y ] + E[Y ]2 ] = E[Y 2 ] − 2E[Y ]E[Y ] + E[Y ]2 =
= E[Y 2 ] − 2E[Y ]2 + E[Y ]2 = E[Y 2 ] − E[Y ]2 = E[Y 2 ] − (µY )2

3. Let Y denote the number of heads that occur when two loaded (meaning unfair) coins are
tossed. Assume the probability of getting heads is 0.4 on either coin.
(a) Derive the probability distribution of Y.
(b) Derive the mean and variance of Y.
Solution:

(a) First it is important to realize that we are talking about two unfair coins. We know
that Y is the number of heads after tossing two coins. So the possible outcomes are:
(y1 , y2 , y3 ) being (no heads, one head, two heads). So what are the probabilities of
these events happening? No heads occurs if both coins show tails, which they do with
probability of 0.6. So p1 = 0.6 ∗ 0.6 = 0.36. Then we have the outcome of one head:
p2 = 0.6 ∗ 0.4 + 0.4 ∗ 0.6 = 0.48 (note that an outcome of one head can occur when the
first coin shows head or when the second coin shows head). The third outcome of two
heads occurs with probability p3 = 0.4 ∗ 0.4 = 0.16.
Since the probability distribution for a discrete random variable is a list assigning prob-
abilities to the outcomes we have:
∗
I am continuously updating these exercises. Please let me know of any mistakes or typos at [email protected]

1
 outcome Y=0 Y=1 Y=2
probability 0.36 0.48 0.16
(b) For the mean we use the formula from exercise 1: E[Y ] = ni=1 yi p = 0 ∗ 0.36 + 1 ∗ 0.48 +
P
2 ∗ 0.16 = 0.8
For the variance we use a trick namely the fact that V ar[Y ] = E[Y 2 ] − (E[Y ])2 . So for
E[Y 2 ] = 02 ∗ 0.36 + 12 ∗ 0.48 + 22 ∗ 0.16 = 1.12 and (E[Y ])2 = 0.82 = 0.64 and therefore
V ar[Y ] = E[Y 2 ] − (E[Y ])2 = 1.12 − 0.64 = 0.48

4. In July, Fairtown’s daily maximum temperature has a mean of 65◦ F and a standard deviation
of 5◦ F. What are the mean, standard deviation, and variance in ◦ C? [Hint: The relationship
between Celcius and Fahrenheit is C = (F − 32) 59 ].
Solution:
This is a linear transformation of a random variable:
Let’s use F to denote temperature in ◦ F and C to denote it in ◦ C. The relationship between
Celcius and Fahrenheit is C = (F − 32) 95
Now for the mean we know if Y = a + bX then E[Y ] = a + bE[X] and for the variance
V ar[Y ] = b2 V ar[X]
We know E[F] and Var[F] and want to obtain E[C] and Var[C]. Let’s use these formulae
5 5 5 5 1
E[C] = E[F ] − 32 = ∗ 65 − 32 = 18
9 9 9 9 3
And for the variance:
5
V ar[C] = ( )2 V ar[F ] = 7.716
9
and then
σC = 2.777

5. The following table gives the joint probability distribution between employment status and
college graduation among those either employed or looking for work (unemployed) in the
working-age population of country A.

unemployed (Y=0) employed (Y=1) total

no college (X=0) 0.078 0.673 0.751
college (X=1) 0.042 0.207 0.249
total 0.12 0.88 1

(a) Compute E(Y).

(b) The unemployment rate is a fraction of the labor force that is unemployed. Show that
the unemployment rate is given by 1 − E(Y).
(c) Calculate E(Y|X = 1) and E(Y|X = 0).
(d) Calculate the unemployment rate for (i) college graduates and (ii) non–graduates.
(e) A randomly selected member of this population reports being unemployed. What is the
probability that this worker is a (i) college graduate, (ii) non–graduate?
(f) Are educational achievement and employment status independent? Explain.

Solution:

(a) In order to compute E[Y] we have to use the marginal distribution of Y and apply our
formula:
E[Y ] = y1 p1 + y2 p2 = 0 ∗ 0.12 + 1 ∗ 0.88 = 0.88
number unemployed
(b) The unemployment rate is given by u = number labor force = P r(Y = 0) = 0.12
(c) Solution see at the end of document. We begin by calculating the conditional probabili-
ties.

2
 (d) Using the solutions to part c the unemployment rates for college graduates is
uc = 1 − E[Y |X = 1] = 1 − 0.831 = 0.169

Similarly for non-graduates

unc = 1 − E[Y |X = 0] = 1 − 0.896 = 0.104

The interesting situation we have is that while the absolute probability of observing an
unemployed college graduate (P r(Y = 0, X = 1) = 4.2%) is smaller than observing an
unemployed non-graduate (P r(Y = 0, X = 0) = 7.8%) that does not directly inform us
about the unemployment rates of these two sub-populations. We also need to take into
account that we have about 3 times as many non-graduates in the sample than graduates.
(So they are three times as likely in our sample but their probability of being unemployed
is only roughly half that of non-graduates).
(e) The probability that a randomly selected worker, who is reported to be unemployed is a
college graduate is:
P r(X = 1, Y = 0) 0.042
P r(X = 1|Y = 0) = = = 0.35
P r(Y = 0) 0.12
(f) Education and employment would be independent if
P r(X = x|Y = y) = P r(X = x)
From part e we calculated P r(X = 1|Y = 0) = 0.35 but P r(X = 1) = 0.249

6. In a given population of two-earning male-female couples, male earnings have a mean of

$50,000 per year and a standard deviation of $15,000. Female earnings have a mean of $48,000
per year and a standard deviation of $13,000. The correlation between male and female
earnings for a couple is 0.90. Let C denote the combined earnings for a randomly selected
couple.
(a) What is the mean of C?
(b) What is the covariance between male and female earnings?
(c) What is the standard deviation of C?
(d) Convert the answers from (a) to (c) from USD to EUR
Solution:
Using the notation C = M + F (couple, male, female) this implies µC = µM + µF and
2 = σ 2 + σ 2 + 2Cov[M, F ]
σC M F

(a) We can use these relationships to obtain µC = 50 + 48 = 98

(b) We now Corr[M, F ] = Cov[M,F ]
σM σF . So we can obtain Cov[M, F ] = 13 ∗ 15 ∗ 0.9 = 175.5 in
units of thousands of USD squared.
(c) From √the above formula σC 2 = σ 2 + σ 2 + 2Cov[M, F ] = 152 + 132 + 2 ∗ 175.5 = 745 Then
M F
σC = 745 = 27.295
(d) The first step is to look up an exchange rate. The European Central Bank reports an
exchange rate of 1EU R = 1.1396U SD on the 17th of January 2019. Thus 0.8775EU R =
1U SD and we can convert the results by multiplying both σC and µC with 0.8775 and
the correlation coefficient is unit free.

7. The random variable Y has mean 4 and variance 1/9. Let Z = 3(Y − 4). Find the mean and
variance of Z.
Solution:

E[Z] = 3E[Y ] − 12 = 3 · 4 − 12 = 0
1
Var[Z] = 32 Var[Y ] = 9 · = 1
9
What we are doing here is to standardize the random variable Y and show that it has indeed
mean 0 and variance 1.

3
 8. A jar of Adams Old Fashioned Peanut Butter contains 453 grams, with a standard deviation
of 10.1 grams. Find the probability that one of these jars contains less than 450 grams if we
can assume a normal distribution.
Solution:
If we assume X is the weight of a randomly selected jar of peanut butter we assume X ∼
N (µ = 453, σ 2 = 10.12 ). In order to find P r(X ≤ 450) we standardize first:
X −µ 450 − µ
P r(X ≤ 450) = P r( ≤ ) = P r(Z ≤ −0.297) = Φ(−0.297)
σ σ
We look up the value Φ(−0.297) in the standard normal table and find Φ(−0.297) = 0.3859.
So the probability that a randomly selected jar weights less than 450g is 38.59%.

9. Compute the following probabilities:

(a) If Y is distributed N (4, 9) find P r(Y ≤ 5)

(b) If Y is distributed N (5, 16) find P r(Y > 2)
(c) If Y is distributed N (1, 4) find P r(2 ≤ Y ≤ 5)
(d) If Y is distributed N (2, 1) find P r(1 ≤ Y ≤ 4)

Solution:
A useful result when dealing with the standard normal distribution is:

Φ(−c) = 1 − Φ(c)

It holds because the standard normal distribution is symmetric about its mean.
5−4
(a) P r(Y ≤ 5) = P r(Z ≤ 3 ) = Φ(1/3) = 0.6293
(b) P r(Y > 2) = P r(Z > 2−5 2−5
4 ) = 1 − P r(Z ≤ 4 ) = 1 − Φ(−0.75) = Φ(0.75) = 0.7734
(c) P r(2 ≤ Y ≤ 5) = P r( 2−1 5−1
2 ≤ Z ≤ 2 ) = Φ(2) − Φ(0.5) = 0.9772 − 0.6915 = 0.2857
(d) P r(1 ≤ Y ≤ 4) = P r( 1−2 4−2
1 ≤ Z ≤ 1 ) = Φ(2) − Φ(−1) = 0.9772 − 0.1587 = 0.8185

10. (From: S. Ross, A First Course in Probability Theory, 2nd ed. (New York: MacMillan, 1984)).
An expert witness in a paternity suit testifies that the length (in days) of a pregnancy, from
conception to delivery, is approximately normally distributed, with parameters µ = 270 and
σ = 10. The defendant in the suit can prove that he was out of the country during the period
from 290 to 240 days before the birth of the child. What is the probability that the defendant
was in the country when the child was conceived?
Solution:
Let X represent the duration of pregnancy in days with µX = 270 an dσX = 10. We are
interested in the probability that the pregnancy was either longer than 290 days or shorter
than 240:
P r(X > 290) + P r(X ≤ 240) =?
This corresponds to 1 minus the probability that it lasted between 240 and 290 days:

1 − P r(240 ≤ X ≤ 290)

We can compute that:

240 − 270 290 − 270
1 − P r(240 ≤ X ≤ 290) = 1 − P r( ≤X ) = 1 − (Φ(2) − Φ(−3))
10 10
= 1 − Φ(2) + Φ(−3) = 1 − Φ(2) + 1 − Φ(3) = 0.0228 + 0.0014 = 0.0242
So the probability that the defendant is the father is 2.42%.

11. In a population µY = 50 and σY2 = 21. Use the central limit theorem to answer the following
questions:

(a) In a random sample of size n=50, find P r(y ≤ 51).

(b) In a random sample of size n=150, find P r(y > 49).

4
 (c) In a random sample of size n=45, find P r(50.5 ≤ y ≤ 51).

Solution:
The central limit theorem (CLT) states that as n grows to infinity we can assume that:

Y ∼ N (µY , σY2 )

where
σY2
σY2 =
n
Given the mean and variance for Y we can solve the exercises:
√
(a) P r(Y ≤ 51) = P r( Y −µ
σ
Y
≤ 51−µ
σ
Y
) = Φ(1/ 0.42) = Φ(1.543) = 0.9382
Y Y

(b) P r(Y > 49) = 1 − P r(Y ≤ 49)). Thus = 1 − Φ( 49−µ

σY ) = 1 − Φ(−2.673) = Φ(2.673) =
Y

0.992
51−50
(c) P r(50.5 ≤ Y ≤ 51) = Φ( √ 0.467
) − Φ( 50.5−50
√
0.467
) = 0.1623

12. In any year the weather may cause damage to a home. On a year-to-year basis, the damage
is random. Let Y denote the USD value of damages in any given year. Suppose that during
95% of the year Y = 0 but that during 5% of the year Y = 30, 000.

(a) What are the mean and standard deviation of damages caused in a year?
(b) Consider an insurance pool of 120 people who are sufficiently dispersed so that in any
year, the damage to different homes can be viewed as independently distributed random
variables. Let Y denote the average damage caused to these 120 homes in one year.
What is the expected value of the average damage Y ? What is the probability that Y
exceeds 3000 USD?

Solution:

(a) The mean of Y is computed as:

E[Y ] = 0.95 ∗ 0 + 0.05 ∗ 30, 000 = 1500

The variance is:

V ar[Y ] = E[Y 2 ] − E[Y ]2
So we have E[Y 2 ] = 0.95 ∗ 02 + 0.05 ∗ 30, 0002 = 45, 000, 000 and it follows that V ar[Y ] =
45, 000, 000 − 15002 = 42, 750, 000
(b) We have E[Y ] = µY = 1500. Now what is P r(Y > 3000)? We rely on the CLT and
argue that standardized Y will follow a standard normal distribution. Then:
3000 − 1500
1 − P r(Y ≤ 3000) = 1 − Φ( )
σY

So we need to find σY :
σY √
σY = √ = 6538.35/ 120 = 596.87
n

Then we obtain: Φ(2.51) = 0.9939. And finally:

1 − Φ(2.51) = 1 − 0.9939 = 0.61%

So the probability that the average damage exceeds $3000 is 0.61%.

5
Figure 1: Steps for Exercise 5c