Solution Exercise 10.

1
a. U(−0.5, 0.5); f(x) = 1 if −0.5 < x < 0.5; ); f(x) = 0 otherwise.
b. 0 and 0.2887
c. P (−0.2 < X < 0.4) = area above (−0.2, 0.4) under f , which is 0.6
d. The command “Rand − 0.5” yields a random observation in (−0.5, 0.5).

Solution Exercise 10.2

a. f(x) = 1/15 for −5 < x < 10; f(x) = 0 otherwise.
P (−2 < X < 8) = 10×(1/15) = 2/3
b. E ( X ) = (−5 + 10)/2 = 2.5; V ( X ) = (10 – (−5))2/12 = 18.75
c. For −5 < x < 10: F(x) = area above (−5, x] under f, which is (x + 5)/15
For x ≤ −5: F(x) = 0
For x ≥ 10: F(x) = 1
d. P (−2.5 < X < 12.5) = P ( X > −2.5) = 1 – F(−2.5) = 1 – 2.5/15 = 0.8333

Solution Exercise 10.3

a. f(x) = 0.1e−0.1x for x > 0; f(x) = 0 for x ≤ 0
b. E ( X ) = 10 and V ( X ) = 102 = 100
c. P ( X > 15) = 1 – F(15) = e−15 / 10 = 0.2231
d. P ( X > 15 | X > 5) = P ( X > 5 + 10 | X > 5) = P ( X > 10) (memoryless)
Answer: e−1 = 0.3679

Solution Exercise 10.4

a. P ( Z > 1.4) = 1 – P ( Z ≤ 1.4) = 1 – 0.9192 = 0.0808 (*)
P ( Z < −2.3) = 0.0107 (*)
b. P (−2.3 < Z < 1.4) = P ( Z < 1.4) − P ( Z < −2.3) = 1 – 0.0808 – 0.0107 = 0.9085
P (−1.4 < Z < 1.4) = 1 – 2 P ( Z > 1.4) = 1 – 2×0.0808 = 0.8384
c. P (0 < Z < 2.1) = P ( Z < 2.1) − 0.5 = 0.9821 – 0.5 = 0.4821 (*)
P (−1.8 < Z < 0) = 0.5 – P ( Z < −1.8) = 0.5 – 0.0359 = 0.4641 (*)
d. P (−1.8 < Z < 1.8) = 1 – 2 P ( Z > 1.8) = 1 – 2×0.0359 = 0.9282

Solution Exercise 10.5

a. P ( X < 0) = 0.0478 (*)
P ( X > 2.5) = 0.7977 (*)
P ( X > 5) = 0.5 (symmetry)
P ( X > 11) = 0.0228 (*)
b. P (2 < X < 8) = P ( X < 8) − P ( X < 2) = 0.8413 – 0.1587 = 0.6826 (*)
P (8 < X < 11) = P ( X > 8) − P ( X > 11) = 0.1587 – 0.0228 = 0.1359
c. P ( X = 2.5) = 0 (continuous rv)
P ( X ≥ 2.5) = P ( X > 2.5) = 0.7977
d. P (2.5 < X < 7.5) = 1 – 2 P ( X < 2.5) = 1 − 2×0.2023 = 0.5954
P ( X > 2) = P (2 < X < 8) + P (8 < X < 11) + P ( X > 11) = 0.8413

1
Solution Exercise 10.6
a. Solve: P ( Z > b) = 0.25
Since P ( Z ≤ 0.6745) = 0.75, (*)
it follows that b = 0.6745
b. Solve: P (−b < Z < b) = 0.60
Since P (−b < Z < b) = 1 – 2 P ( Z > b) , we have to solve: P ( Z > b) = 0.20
Since P ( Z ≤ 0.8416) = 0.80, (*)
it follows that b = 0.8416
c. Solve: P (−1.5 < Z < b) = 0.85
0.85 = P ( Z < b) − P ( Z < −1.5) = P ( Z < b) − 0.0668 (*)
Solving P ( Z < b) = 0.9168 yields b = 1.3839 (*)
d. Solve: P (0 < Z < b) = 0.30
Same solution as P (−b < Z < b) = 0.60. Answer: b = 0.8416

Solution Exercise 10.7

a. Solve: P (3.2 < X < b) = 0.65
0.65 = P ( X < b) − P ( X < 3.2) = P ( X < b) − 0.1711 (*)
So, we have to solve: P ( X < b) = 0.8211. Answer: b = 10.6776 (*)
−b b
b. P (7 − b < X < 7 + b) = 0.40 ⇔ P( < Z < ) = 0.40
4 4
b b b
⇔ 1 – 2 P( Z > ) = 0.40 ⇔ P( Z > ) = 0.30 ⇔ P( Z ≤ ) = 0.70
4 4 4
Answer: b / 4 = 0.5244, so b = 2.0976 (*)

Solution Exercise 10.8

a. N(12, 10.80); np = 12 and n(1 − p) = 108, both ≥ 5.
11.5 − 12 12.5 − 12
b. P( X = 12) = P(11.5 < X < 12.5) = P( <Z< )
10.8 10.8
= P( −0.1521 < Z < 0.1521) ≈ 0.5605 – 0.4395 = 0.1209 (*)
c. P( X ≤ 13) = P( X ≤ 13.5) = P( Z ≤ 0.4564) ≈ 0.6759 (*)
P( X > 11) = ( X > 11.5) = P( Z > −0.1521) ≈ 1 – 0.4395 = 0.5605 (*)
d. P(6 < X < 15) = P( −1.6736 < Z < 0.7607) ≈ 0.7766 – 0.0471 = 0.7295 (*)

Solution Exercise 10.9

a. P(−0.91 < Z < 1.51) = P( Z < 1.51) − P( Z < −0.91)
= P( Z < 1.51) − P( Z > 0.91)
= P( Z < 1.51) − 1 + P( Z < 0.91)
= 0.9345 – 1 + 0.8186 = 0.7531
b. Solve P( Z > a ) = 0.66. Notice that a is situated at the left-hand side of the
axis of symmetry, so a < 0. Also note that P( Z ≤ −a ) = 0.66. From the
table it follows that −a is between 0.41 and 0.42; −a = 0.4125 (linear
interpolation) . Hence: a = −0.4125

2
 0 .5 − 6 8 .2 − 6
c. P(0.5 < X < 8.2) = P( <Z< ) = P(−1.10 < Z < 0.44)
5 5
= P( Z < 0.44) − 1 + P( Z < 1.10) = 0.6700 – 1 + 0.8643 = 0.5343
b−6
d. Solve P( X > b) = 0.66. That is, solve P( Z > ) = 0.66 or
5
6−b
(equivalently) P( Z < ) = 0.66.
5
In b. we obtained: P( Z < 0.4125) = 0.66
Hence: (6 – b) / 5 = 0.4125, so b = 3.9375

Solution Exercise 10.10

a. P( R < 0) = P( Z < −12 / 20) = P( Z < −0.60) = 0.2743 (*)
b. P( R > 3) = P( Z > −0.45) = P( Z < 0.45) = 0.6736 (*)
c. P( R < −3) + ( R > 27) = 2 P( R > 27) = 2 P( Z > 0.75)
= 2×(1 – 0.7734) = 0.4532 (*)
b − 12
d. 0.60 = P( R > b) = P( Z > )
20
b − 12
Hence, solve b in P( Z ≤ ) = 0.40. It follows: (b – 12)/20 = −0.2533
20
So, b = 6.9340.

Solution Exercise 10.11

Weekly net sales: S ∼ N(10, 9).
3.5 − 10
a. P( S < 3.5) = P( Z < ) = P( Z < −2.1667) = 0.0151 (*)
3
b. Determine b in the equation P( S < b) = 0.10.
b − 10
That is: 0.10 = P( Z < ) . Hence, (b – 10) /3 = −1.2816 (*)
3
So, b = 6.1552; the level is 6155 euro.

Solution Exercise 10.12

a. If X > t, then there was no heavy storm during a period of t years, so Y = 0.
And if Y = 0, then there hasn’t been a heavy storm t years so X > t. Hence:
{ X > t} = {Y = 0} .
But then P( X > t ) = P(Y = 0) , and since Y ∼ Po(2.5t) this last probability is
equal to e−2.5t
b. From a. it follows that: P( X ≤ t ) = 1 − e −2.5t = 1 − e −t / 0.4 , so X ∼ Expo(0.4).
c. P( X ≤ 0.5) = 1 − e −0.5 / 0.4 = 0.7135
d. P( X > 1 | X > 0.5) = P( X > 0.5 + 0.5 | X > 0.5) = P( X > 0.5) = 0.7135
(memoryless)

3
Solution Exercise 10.13
a. The command 90*Rand simulates the random drawing of a real number in
the interval (0, 90). Hence, the command 10+90*Rand can be used to
generate a real number in (10, 100).
b. X is uniformly distributed; parameters α = 10 and β = 100. Notation: X ∼
U(10, 100). Continuous!
c. f (x) = 1/90 for all x in (10, 100) and f (x) = 0 for x otherwise.
d. µ X = (10 + 100)/2 = 55, σ X2 = (100 – 10)2/12 =675 and σ X = 25.9808.
e. P( X = 25) = 0 since X is continuous. For the other probabilities: recall that
probabilities are areas under the pdf. Always make a picture!!
P( X ≤ 25) = (25 – 10)×(1/90) = 15/90 = 0.1667, according to the rectangle
method;
P( X < 25) = P( X ≤ 25) = 0.1667;
P( X > 80.5) = (100 – 80.5)×(1/90) = 19.5/90 = 0.2167;
P(20.6 < X < 90.6) = (90.6 – 20.6)×(1/90) = 70/90 = 0.7778;
P({ X ≤ 20} ∪ { X > 35.1}) = 1 − P(20 < X < 35.1) = 1 – (35.1 – 20)×(1/90) =
1 – 0.1678 = 0.8322;
P({ X ≤ 80} ∩ { X > 20}) = P(20 < X < 80) = 60/90 = 0.6667.

Solution Exercise 10.14

a. This constant b falls in (10, 100) and has to be such that the area on its left-
hand side equals 0.75. That is: (b – 10)×(1/90) = 0.75. It follows that b = 77.5.
b. Recall that F (x) = P( X ≤ x) .
For x < 10 we have F (x) = 0; for x > 100 we have F (x) = 1.
Now suppose that x lies between 10 and 100. Then F (x) is just the area of the
rectangle with sides (x – 10) and 1/90. Hence:
F (x) = (x – 10)/90 for all x in [10, 100]
P(20.6 < X < 90.6) = F (90.6) − F (20.6) = 80.6/90 – 10.6/90 = 0.7778.
c. The graph of F is 0 at the left-hand side of 10 and a straight line increasing
from 0 to 1 between 10 and 100. At the right-hand side of 100, the graph of F
stays at the level 1.

Solution Exercise 10.15

a. This mean is just E ( X ) , which is (10000 + 35000)/2 = 22500 litre. For the
pdf f of X it holds that f (x) = 1/25000 for 10000 < x < 35000.
b. One litre costs 1.40 euro, so X litre costs 1.40×X euro. Hence, Y = 1.4X.
c. E ( X ) = 22500 (see part a.) and E (Y ) = 1.4 E ( X ) = 31500;
V ( X ) = (35000 – 10000)2/12 = 52083333.3333 and SD( X ) = 7216.8784;
V (Y ) = (1.4)2× V ( X ) = 102083333.333 and SD(Y ) = 10103.6297.
d. P( X ≤ 20000) = (20000 – 10000)×(1/25000) = 10/25 = 0.40;
P(Y ≤ 40000) = P(1.4 X ≤ 40000) = P( X ≤ 28571.4286)
= 18571.4286 / 25000 = 0.7429;
P(30000 < Y < 40000) = P(30000 < 1.4 X < 40000)
= P(21428.5714 < X < 28571.4286) = 0.2857

4
Solution Exercise 10.16
a. The first two probabilities are equal, since P( Z = 1.23) = 0. The answers are:
0.8907, 0.8907, 0.4522, 0.3075, 0.5654. (*)
b. Recall that ( X − 10) / 4 has the same probability distribution as Z. As a
consequence, P( X ≤ 14.92) = P( Z ≤ (14.92 − 10) / 4) = P( Z ≤ 1.23) , which
equals 0.8907 because of a. The other probabilities follow analogously; the
answers are: 0.8907, 0.4522, 0.3075, 0.5654.
c. P( X > 18) = 1 − P( X ≤ 18) = 1 – 0.9772 = 0.0228 (*)
Analogously, P( X > 22) = 0.0013 and P( X > 26) = 0.000032. (*)
Because of the symmetry around 10, the answers of the following three
probabilities are the same: 0.0228, 0.0013 and 0.000032.
d.
0.12

0.1

0.08

0.06

0.04

0.02

0
-10 -5 0 5 10 15 20 25 30

Solution Exercise 10.17

a. Under the pdf of N(10, 4), the area at the left-hand side of a has to be equal to
0.20. It follows that a = 8.3168. (*)
b. b = 7.8394. (*)
c. From the symmetry it follows that P( X < 10 − c) = 0.30.
Hence, c = 1.0488. (*)

Solution Exercise 10.18

µ −σ − µ µ +σ − µ
a. P( µ − σ < X < µ + σ ) = P( <Z< )
σ σ
= P (−1 < Z < 1) = 2 P ( Z < 1) − 1 = 0.6827 (*)
b. P( µ − 2σ < X < µ + 2σ ) = P(−2 < Z < 2) = 2 P( Z < 2) − 1 = 0.9545 (*)
c. P( µ − 3σ < X < µ + 3σ ) = P(−3 < Z < 3) = 2 P( Z < 3) − 1 = 0.9973 (*)
d. There is respectively 68.27%, 95.45% and 99.73% chance that X will fall less than 1
2, or 3 standard deviations from µ.
e. These probabilities are in accordance with the empirical rules for k = 1, 2, 3.

Solution Exercise 10.19

a. The uniform distribution with α = 50 and β = 100; that is: U(50, 100).
For x in the interval (50, 100) it holds that f (x) = 1/50 = 0.02 (and f (x)
= 0 otherwise).

5
 The graph of f has the constant level 0.02 above the interval (50, 100). By
symmetry it follows that the mean is 75, while the variance is (100 – 50)2 / 12
= 208.3333 and the standard deviation is 14.4338.
b. For x between 50 and 100, f (x) = 0.02. That is: with this model, all possible
outcomes have the same likelihood of becoming the realisation of X.
c. P( X ≤ 75) = 0.5, in accordance with the empirical fact.
P(70 < X < 80) is just the area above the interval (70, 80), which is: 0.20.
This is much less than the empirical fact 0.42.

Solution Exercise 10.20

a. Write c for g(75) . This pdf g has to grow linearly from 0 (at x = 50) to c (at x
= 75), and to decrease linearly from c to 0 on the interval (75, 100). Since the
total area under the graph of g has to be equal to 1 and since the two
rectangular triangles are congruent, the total area of the rectangle with sides 25
and c has to be 1. As a consequence, c = 1/25 = 0.04.
Hence, on the interval (50, 75], the linear function g (x) has the slope
0.04/25 = 0.0016 and g (x) = 0.0016x + b1; since g (50) = 0, it follows that b1
= −0.08. On the interval (75, 100), the linear function g (x) has the negative
slope –0.0016 and g (x) = −0.0016x + b2; since g(100) = 0 it follows that b2 =
0.16. That is:

g (x) = 0.0016x – 0.08 for all x in (50, 75]

= −0.0016x + 0.16 for all x in (75, 100)

Since the graph of g is symmetric, it follows that the mean of the distribution
is 75 and again it holds that P( X ≤ 75) = 0.5.
b. For x between 50 and 100, g (x) represents the likelihood that x will become
the realisation of X under this model. This likelihood is largest near the centre
75000.
c. Since the graph of g is symmetric, it follows that the mean of the distribution
is 75 and again it holds that P( X ≤ 75) = 0.5.
The probability P(70 < X < 80) is just the sum of the two areas above the
intervals (70, 75] and (75, 80); make a picture yourself. Notice that these two
areas are both equal to 0.18. Hence, the probability is 0.36, slightly less than
the empirical fact 0.42.

Solution Exercise 10.21

a. Of course, take µ = 75. Furthermore, take σ such that 3σ = 25; hence, σ 2 =
69.4444. Here is the model: X ∼ N(75, 69.4444).
b. Again it holds that P( X ≤ 75) = 0.5.
P(70 < X < 80) = 2 P( X ≤ 80) − 1 = 2 P( Z ≤ 0.6000) − 1 = 0.4515 (*)
c. This last model fits best to the empirical facts.
d. With the normal model: P( X > 85) = P( Z > 1.20) = 1 – 0.8849 = 0.1151 (*)

Solution Exercise 10.22

a. f (5) = 1.5958; see also Appendix A1.10. (*)

6
 This number is larger than 1, so it cannot be a probability.
b. (*)
h P ( 6 ≤ X ≤ 6 + h ) f ( 6) h
1 0.116949 0.125794
0.5 0.060904 0.062897
0.25 0.030980 0.031449
0.125 0.015611 0.015724
0.0625 0.007834 0.007862
0.03125 0.003924 0.003931
0.0156 0.001961 0.001962
0.0078 0.000981 0.000981
As h gets closer to 0, the results in the columns 2 and 3 come closer.

Solution Exercise 10.23

a. Since X ∼ N(10, 25), the probability P( X < 0) can be computed; the answer is
0.0228. (*)
b. Y = 50000×(X/100) – 1000 = 500X – 1000
E (Y ) = 500× E ( X ) − 1000 = 500×10 – 1000 = 4000;
SD(Y ) = 500× SD( X ) = 500×5 = 2500
c. P(Y < 0) = P(500 X − 1000 < 0) = P( X < 2) = P( Z < −1.6) = 0.0548 (*)
d. Since Y is a linear transformation of X, it is normal too.
Y ∼ N(4000, (2500)2)

Solution Exercise 10.24

a. P( X < 45) = 0.6497 (*)
b. P(10 < X < 60) = P( X < 60) − P( X < 10) = 0.9380 – 0.0105 = 0.9275 (*)
c. Solve b in P( X > b) = 0.10, computation yields b = 56.6602 euro. (*)
Hence, at least 56.66 euro.
P( X ≥ 70) 0.0105
d. P( X ≥ 70 | X ≥ 56.6602) = = = 0.1050. (*)
P( X ≥ 56.6602) 0.10
Answer: 10.5%.
e. Recall that ( X − 40) / 13 is also standard normally distributed. Hence:
0.975 = P(( X − 40) / 13 ≤ 1.96) = P( X − 40 ≤ 25.4800) = P( X ≤ 65.4800)
That is: 97.5% of the customers pay 65.48 euro or less.

Solution Exercise 10.25

a. It is the probability P(| Y − µ |< 2.5 × σ ) that is asked for. It follows that:
Y −µ
P(| Y − µ |< 2.5 × σ ) = P(| |< 2.5) = P(−2.5 < Z < 2.5) = 0.9876 (*)
σ
(make a picture!)
b. The constant a has to satisfy: P(Y − a < µ < Y + a ) = 0.95.
Recall from (10.19) that: P(Y − 1.96σ < µ < Y + 1.96σ ) = 0.95
Hence, take a = 1.96×σ = 4.12.
c. (36.50 – 4.12, 36.50 + 4.12) = (32.38, 40.62). In view of b. it is likely that µ
falls in this interval.

7
Solution Exercise 10.26
Let X be the weight (in gram) of this randomly drawn parcel of butter.

X − 252 255 − 252

a. P( X > 255) = P( > ) = P( Z > 1.5) = 0.0668
2 2
X − 252 250 − 252
b. P( X < 250) = P( < ) = P( Z < −1) = P( Z > 1) = 0.1587.
2 2
The answer is: 15.87%.
X − µ 250 − µ 250 − µ
c. 0.05 = P( X < 250) = P( < ) = P( Z < )
2 2 2
It is given that z0.05 = 1.6449; that is: 0.05 = P( Z > 1.6449) = P( Z < −1.6449)
Hence, µ has to satisfy: (250 − µ ) / 2 = −1.6449. This yields µ = 253.2898
gram.
X − 252 250 − 252 −2
d. 0.05 = P( X < 250) = P( < ) = P( Z < )
σ σ σ
On the other hand: 0.05 = P( Z < −1.6449)
It follows that: −1.6449 = −2/σ, so σ = 1.2159.

Solution Exercise 10.27

a. Since Y ∼ Bin(500, 0.002), it follows that:
 500 
P(Y = 1) =  (0.002)1 (0.998) 499 = 0.3682
 1 
E (Y ) = 1 and V (Y ) = 500×0.002×0.998 = 0.998
b. It is not wise to use a normal approximation since the 5-rule is not satisfied.
 2500 
c. P(Y = 5) =  (0.002)5 (0.998) 2495 = 0.1756
5 
E (Y ) = 2500×0.002 = 5, V (Y ) = 2500×0.002×0.998 = 4.9900 and SD(Y ) =
2.2338.
d. The probability distribution of Y can be approximated by N(5, 4.9900). Below,
Z is a standard-normally distributed rv.
4 .5 − 5 Y −5 5.5 − 5
P(Y = 5) = P(4.5 ≤ Y ≤ 5.5) = P( ≤ ≤ )
2.2338 2.2338 2.2338
≈ P(−0.2238 ≤ Z ≤ 0.2238) = 1 − 2 P( Z > 0.2238)
= 1 − 2×0.4115 = 0.1771 (*)
e. P(Y = 6) = P(5.5 ≤ Y ≤ 6.5) ≈ P(0.2238 ≤ Z ≤ 0.6715)
= P( Z ≤ 0.6715) − P( Z ≤ 0.2238)
= 0.7490 – 0.5885 = 0.1605 (*)
P(Y ≤ 6) = P(Y ≤ 6.5) ≈ PZ ≤ 0.6715) = 0.7490
P(Y > 7) = P(Y > 7.5) ≈ P( Z > 1.1192) = 0.1315 (*)
P(4 ≤ Y < 8) = P(3.5 ≤ Y < 7.5) ≈ P(−0.6715 < Z < 1.1192)
= 0.8685 – 0.2510 = 0.6175 (*)

8
Solution Exercise 10.28
Let Y be the number of headache sufferers in the sample that do get relief with two
aspirins.
a. Since E (Y ) = 165 and V (Y ) = 74.25, the approximating distribution is N(165,
74.25). Hence:
Y − 165 149.5 − 165
P(Y < 150) = P(Y ≤ 149.5) = P( ≤ )
8.6168 8.6168
≈ P( Z ≤ −1.7988) = 0.0360 (*)
b. It suggests that it is likely that the claim of the manufacturer is not correct.
c. P(Y ≤ 145) = P(Y ≤ 145.5) ≈ P( Z ≤ −2.2630) = 0.0118
If the claim is correct, it is very unlikely that 145 (or less) of the 300 persons
get relief.

Solution Exercise 10.29

a. Let F be the cdf of X. Since X is discrete, F is a step-function

x 0 1 2 3
f(x) 0.375 0.375 0.188 0.062
F(x) 0.375 0.750 0.938 1

The cdf F (x) passes the levels 0.529, 0.744 and 0.869 respectively at x = 1, x
= 1 and x = 2. Hence, the observations are 1, 1 and 2.
b. Firstly, the continuous cdf of Y is calculated. Notice that, for y between −2
and 2:
G( y) = P(Y ≤ y ) = area under g and above (−2, y)
= ( y − (−2)) × g ( y ) / 2 = ( y + 2)( y + 2) / 16
= ( y + 2)2 / 16
To obtain the first simulation, the equation 0.041 = ( y + 2)2 / 16 has to be
solved. That is:
( y + 2) 2 = 0.656 ⇔ y + 2 = ± 0.656 ⇔ y = −2 ± 0.656
Since y = −2 − 0.656 is not a valid solution (it falls beyond (-2, 2)), the
simulated value is –1.1901. The other two simulations can be obtained
analogously: 0.1541 and 1.5485.

Solution Exercise 10.30

a. P ( X > 5) = 0.09.
b. That the 250 sub-experiments (trials) that check daily on workdays whether
the demand is larger than the supply are independent and that each workday
has the same success probability.
c. Under the assumption of part b., Y has the distribution Bin(250, 0.09).
E (Y ) = 250×0.09 = 22.5 and SD(Y ) = 250 × 0.09 × 0.91 = 4.5249;
P (Y ≥ 25) = 1 − P (Y < 25) = 1 − P (Y ≤ 24) = 1 – 0.6794 = 0.3206. (*)
d. Below, the random variable Z is N(0, 1).
Y − 22.5 24.5 − 22.5
P (Y ≥ 25) = P (Y ≥ 24.5) = P ( ≥ )≈
4.5249 4.5249

9
 ≈ P ( Z ≥ 0.4420) = 1 − P ( Z < 0.4420) = 0.3292. (*)

Solution Exercise 10.31

Let X denote the IQ of a randomly drawn EU25 inhabitant. Then, X ∼ N(100, 162).

a. The probabilities P ( X < 80) and P ( X > 140) are needed. They are
respectively equal to 0.1056 and 0.0062. (*)
The answers to the two questions are
455× 0.1056 = 48.05 million and 455× 0.0062 = 2.82 million.
b. The equation P ( X > a ) = 0.02 has to be solved. Or, equivalently:
P ( X ≤ a ) = 0.98.
It follows that: a = 132.86. (*)
The minimum IQ that gives the right to become a member is 132.86.
c. Notice that this question concerns a conditional probability: given that the
randomly chosen person is a potential Mensa member, what is the probability
that this person’s IQ is above 140? That is, the following probability is
wanted:

P ({ X > 140} ∩ { X > 132.86}) P ( X > 140)

P ( X > 140 | X > 132.86) = =
P ( X > 132.86) P ( X > 132.86)

(The last equality follows since the occurrence of { X > 140} automatically
implies the occurrence of { X > 132.86}.) Since the numerator and the
denominator are already calculated, the proportion that is asked for equals
0.0062 / 0.02 = 0.31. More than 30% of the potential Mensa members have IQ
above 140.

Solution Exercise 10.32

a. Rt = 100 × (Yt − Yt −1 ) / Yt −1
b.
rel frequency

0.20

0.15

0.10

0.05

0.00
-3 -2 -1 0 1 2 3 4 5 6

return

It is hard to see a normal distribution in this relative frequency distribution.

10
c. The normal distribution with mean and variance equal to the mean and
(population) variance of the 51 returns of part a.
Hence: R ∼ N(−0.0015, 5.5114)
d.
Empirical model Normal model
P (−0.5 < R < 0.5) 12/51 = 0.2353 0.1687
P ( R < −3) 6/51 = 0.1176 0.1008
P ( R > 5) 3/51 = 0.0588 0.0166

Indeed, the empirical probabilities are larger than the accompanying normal
probabilities, which supports the arguments that the normal model
underestimates the actual frequencies in the tails and near 0.

11