1

Final The three questions (below) refer to the table given below,
which breaks down 100 people based upon their gender and
age.
Set B/D
Male Female Total
Name . . . . . . . . . . . . . . . . . . Young 18 27 45
Middle Aged 18 17 35
Section . . . . . . . . . . . . . . . . . . Senior 4 16 20
Total 40 60 100
Email ID . . . . . . . . . . . . . . . . . .
1. Suppose that the experiment consists of randomly select-
• You are allowed to use a simple calculator. ing a person. What is the probability that the person
selected is Young or a Male?
• No copying, cheating, collaboration, computers, or cell
phones are allowed. Use of mobile phone will lead to im- (a) 0.67
mediate cancellation of your exam.You are not allowed to (b) 0.45
borrow a calculator at any point during the exam. If you
are found cheating, you will be failed. If you are found (c) 0.40
talking with students, you shall be failed. (d) 0.18

• Answers should be exact unless an approximation is asked

for. All parts will be weighted equally within each prob- (a) 0.67
lem.
• Table of normal probabilities: The last page of the exam 2. If a Male is selected, what is the probability that he is
contains a table of standard normal cdf values. Young?

• Simplifying expressions: You don’t need to simplify com- (a) 0.67

1 2 (b) 0.45
plicated expressions. For example, you can leave × +
4 3
1 2 20! (c) 0.40
× exactly as it is. Likewise for expressions like .
3 5 18!2! (d) 0.18

Question Score Question Score (b) 0.45

1 / 2 points 11 / 3 points
3. Which of the following is true about the events Young and
2 / 2 points 12 / 3 points Male?

3 / 2 points 13 / 3 points (a) They are independent.

/ 2 points / 3 points (b) They are mutually exclusive.

4 14
(c) They are both independent and mutually exclusive.
5 / 2 points 15 / 3 points
(d) They are neither independent nor mutually exclusive.
6 / 2 points 16 / 3 points
(a) They are independent.
7 / 2 points 17 / 4 points

8 / 2 points 18 / 4 points 4. In a hypothesis test the decision was made to not reject
the null hypothesis. Which type of mistake could have
9 / 2 points 19 / 4 points
been made?
10 / 2 points Total /50 points
(a) Type 1
(b) Type 2
(c) Type 1 if it’s a one-sided test and Type 2 if it’s a
two-sided test
(d) Type 2 if it’s a one-sided test and Type 1 if it’s a
two-sided test

(b) Type 2
1 Contact: amit.kr.goyal@gmail.com

1
 5. The board of examiners that administers the real estate (a) −1.96
broker’s examination in a certain state found that the (b) +1.96
mean score on the test was 493 and the standard devi-
(c) −1.645
ation was 72. If the board wants to set the passing score
so that only the best 10% of all applicants pass, what is (d) +1.645
the passing score? Assume that the scores are normally
distributed. (c) −1.645
(a) 400.73
(b) 585.27 9. Next step is to compute the z-test statistic. What is the
(c) 550.75 z-test statistic?
(d) 425.12 (a) −2.218
(b) +2.218
(b) 585.27
(c) −140.51
(d) +140.51
6. The Central Limit Theorem implies that
(a) All variables have approximately bell-shaped sample (a) −2.218
distributions if a random sample contains at least 30
observations.
10. Is the proportion of female students in the sample from the
(b) Population distributions are normal whenever the
rural areas significantly lower than the population propor-
population size is large.
tion at a 0.05 significance level?
(c) For large random samples (larger than 30), the sam-
pling distribution of x is approximately normal, re- (a) Yes, because the z-test statistic is more than the crit-
gardless of the shape of the population distribution. ical value for a .05 significance level.
(d) The sampling distribution of x looks more like the (b) Yes, because the z-test statistic is less than the crit-
population distribution as the sample size increases ical value for a .05 significance level.
above 30. (c) No, because the z-test statistic is more than the crit-
ical value for a 0.05 significance level.
(c) For large random samples (larger . . . . . . . . . (d) No, because the z-test statistic is less than the critical
value for a 0.05 significance level.
Next four questions refer to the following information: An As-
sociation of American Medical Colleges report stated that, in (b) Yes, because the z-test statistic is less . . .
2011-2012, 47.0% of all matriculated medical school students in
the United States were female. It is thought that, perhaps, the
number of matriculated female students may be lower in rural 11. A casino wants to estimate on average how much a cus-
areas. Researchers collect a random sample of 1000 2011-2012 tomer loses after a night’s play by recording the amount
matriculated medical students from rural areas and find that loss Xi (in dollars) of 100 customers. They assume that
435 of the students were female. Were there proportionately X1 , X2 , . . . , X100 are i.i.d. normal, and found the sample
fewer females matriculated medical students in rural areas as mean X and sample standard deviation s to be respec-
compared to the national proportion of female matriculated tively
medical students? 100
1 X
7. What are the null and alternative hypotheses? X = Xi = 80
100 i=1
(a) H0 : p = 0.435, Ha : p 6= 0.435 v
u 100
u1 X
(b) H0 : p = 0.435, Ha : p > 0.435 s= t (Xi − X)2 = 100
99 i=1
(c) H0 : p = 0.47, Ha : p 6= 0.47
(d) H0 : p = 0.47, Ha : p < 0.47 Find a 95% confidence interval [a, b] (in dollars) of the
average nightly loss amount. (Give numerical answers ac-
(d) H0 : p = 0.47, Ha : p < 0.47 curate to the integer.)
Confidence interval is
8. Using a 0.05 significance level, what critical value(s)
should be used to determine whether there are propor- [a, b] = [60, 100]
tionately fewer matriculated female medical students in
rural areas as compared to the national proportion of ma- Assuming the sample standard deviation s does not
triculated female students. change much as more customers are sampled, how many

2
 customers would the casino need to sample in order to (b) Use the Central Limit theorem to approximate
shrink the confidence interval to half the width of the con- X25
!
fidence interval above? Pr 15 < Xi < 35 ≈ 0.9545
i=1
(a) 50
(b) 200 16. Suppose X and Y are i.i.d with MGFs MX (t) = MY (t) =
1
(c) 400 1−t where t ∈ (−1, 1), evaluate the following:

(d) None of the above  2

1
MX+Y (t) =
1−t
(c) 400
Pr(X ≤ 2) = 1 − e−2
12. Suppose Z1 ∼ N (0, 1) and Z2 ∼ N (0, 1) are i.i.d. stan-
dard normal. Use standard normal CDF table to find the
following probabilities.
Pr(min(X, Y ) > 2) = e−4

√
Pr(Z1 + Z2 ≤ 2) = 0.8413 17. Suppose the joint density of X and Y is
(
c if 0 ≤ y ≤ 1, 0 ≤ x ≤ 2
Pr(Z12 ≤ 2.56) = 0.8904 fX,Y (x, y) =
0 elsewhere

(a) Find c.
Pr(max(Z1 , Z2 ) ≤ 2) = 0.9550 (b) Find the marginal pdf of Y , fY (y).
(c) Find the conditional pdf of Y given X = x, where
13. Let U1 , U2 , . . . , U192 be i.i.d. U(0, 4) and X = U1 + U2 + 0 ≤ x ≤ 2.
· · ·+U192 . According to Central Limit Theorem, the distri- (d) Are X and Y independent?
bution of X is approx. normal with mean µ and variance
σ 2 . Find µ and σ 2 . Also use the normal table to find an (e) Find the E(Y ).
approximation for Pr(X > 400). Let Z ∼ N (0, 1).
Solve this for c :
(µ, σ 2 ) = (384, 256)

Z 1 Z 2
Pr(X > 400) ≈ Pr(Z > 1) = 0.1587 c dxdy = 1
0 0

14. Let X1 , X2 , . . . be independent Geometric random vari- and we get c = 0.5.

ables with parameter p = 23 . Marginal PDF of X is
X1 + X2 + · · · + Xn Z 1
(a) Let Mn = . What value does fX (x) = 0.5 dy = 0.5, for x ∈ [0, 2]
n 0
the sequence Mn converge to, in probability?
1 Marginal PDF of Y is
2 Z 2
fY (y) = 0.5 dx = 1, for y ∈ [0, 1]
(b) Let Ij = 1 if Xj ≥ 1, and Ij = 0, otherwise and 0
I1 + I2 + · · · + In
consider Yn = . What value does Therefore, conditional PDF of Y given X = x is
n
the sequence Yn converge to, in probability?
fX,Y (x, y) 0.5
1 fY |X (y|x) = = = 1, where y ∈ [0, 1], x ∈ [0, 2]
fX (x) 0.5
3
Since fX,Y (x, y) = fX (x)fY (y) holds for all x and y, X
15. Let X1 , X2 , . . . , X25 be independent Exponential random and Y are independent.
variables with mean 1.
Observe that Y ∼ Unif(0, 1), therefore E[Y ] = 0.5. Alter-
(a) Use the Chebyshev Inequality to obtain a lower natively, you can find it in the following way :
bound on Z 1
25
!
X 3 E[Y ] = y dy = 0.5
Pr 15 < Xi < 35 ≥ 0
i=1
4

3
18. Wait time for the “Airport Rocket” bus (X) is Uniformly
distributed from 0 to 10 minutes. An airport monitor
p̂ − p 0.154 − 0.25
records wait times for 30 randomly selected travelers. z=q =q ≈ −1.79
p(1−p) 0.25(1−0.25)
(The chosen sampling procedure ensures that the inde- n 65
pendence assumption is reasonable.) What is the chance
that the sample mean wait time is less than 4.5 minutes? Since −1.79 < −1.645, we reject the null with a 5% signif-
icance level and conclude that the population proportion
If X ∼ Unif(0, 10), then its expected value and variance is significantly lower than 0.25.
are as follows : Approach 2

E[X] = 5
100 25 p̂ − p 0.154 − 0.25
V[X] = = z=q =q ≈ −2.144
12 3 p̂(1−p̂) 0.154(1−0.154)
n 65
We have sample of size 30, so we can have the follow-
ing approximation of X or sample mean?s distribution by Since −2.144 < −1.645, we reject the null with a 5% signif-
central limit theorem: icance level and conclude that the population proportion

5
 is significantly lower than 0.25.
X ∼N
˙ 5,
18

So we have
 
4.5 − 5 
Pr(X < 4.5) = Pr Z < q = Pr (Z < −0.9487) ≈ 0.17
5
18

19. Some boxes of a certain brand of breakfast cereal include

a voucher for two free movie tickets inside the box. The
company that makes the cereal claims that a voucher can
be found in 25 percent of the boxes. However, based on
their experiences eating this cereal at home, a group of stu-
dents believes that the proportion of boxes with vouchers
is less than 0.25. This group of students purchased 65
boxes of the cereal to investigate the company’s claim.
The students found a total of 10 vouchers for two free
movie tickets in the 65 boxes. Suppose it is reasonable
to assume that the 65 boxes purchased by the students
are a random sample of all boxes of this cereal. Based on
this sample, is there support for the students’ belief that
the proportion of boxes with vouchers is less than 0.25?
Answer by performing all steps of a hypothesis test with
level of significance α = 0.05.

We have the following null and alternative hypothesis:

H0 : p = 0.25
Ha : p < 0.25

We have the sample proportion to be

10
p̂ = ≈ 0.154
65
Then we calculate the z-statistic
(The following two approaches are both considered
correct.)
Approach 1

4
Contact: amit.kr.goyal@gmail.com