PPOL 501 - 04 Answer Key

Problem Set #2

1. Answer problem 1.39 “Associations” on page 65 of the Diez et al. textbook.

Positive Association, Linear
No Association
Positive Association, Non-Linear
Negative Association, Linear

2. Answer problem 1.42 “Sleeping in college” on page 65 of the Diez et al. textbook.

6.25 represents the sample mean and 5.5 represents the claimed population mean.

3. Answer problem 1.43 “Parameters and statistics” on page 66 of the Diez et al.
textbook.

a)$58 represents the sample mean and $52 represents the claimed population mean
b) 3.59 represents the sample mean and 3.37 represents the population mean.

4. Answer problem 1.44 “Make-up exam” on page 66 of the Diez et al. textbook.

a):
Step 1:                     Step 2:                             

x/24 = 74                (1776+64)/25 = 73.6

x = 74 * 24

x = 1776

Answer: The new student’s score decreases the average.

b): The new average is 73.6.

c): The standard deviation increases since the new score is more than 1 standard deviation
away from the old mean.

1 standard deviation from the mean = 74 ± 8.9 = [ 65.1 ,82.9]

5. Answer problem 1.56 “Distributions and appropriate statistics, Part II” on page 69 of
the Diez et al. textbook.

1
a) Since they are a meaningful number of houses that cost more than $6,000,000, the
mean would be affected. Since the mean would be greater than the median, the
distribution is expected to be right skewed. The median would best represent the data and
the IQR would best represent the variability of the observations.

b): The distribution is symmetric since there appears to be equal intervals in the housing
prices. Thus, the mean would best represent the data and the standard deviation would
best represent the variability of the observations.

c) : While only a few students drink excessively, these outliers would still affect the
mean. Since the mean would be greater than the median, the distribution is expected to be
right skewed. The median would best represent the data and the IQR would best represent
the variability of the observations.
d): While only a few houses cost more than $1,200,000, these outliers would still affect
the mean. Since the mean would be greater than the median, the distribution is expected
to be right skewed. The median would best represent the data and the IQR would best
represent the variability of the observations.

6 Answer problem 1.61 “Income at the coffee shop” on page 70 of the Diez et al.
textbook.

a) The median would best represent the typical income for the 42 patrons at the
coffee shop. The median is more robust than the mean since outliers have a
smaller effect on the median.
b) In this scenario, the IQR would best represent the amount of variability. The IQR
is more robust than the standard deviation since outliers barely have an effect on
the IQR.

7. High school seniors with strong academic records apply to the nation’s most selective
colleges in greater numbers each year. Because the number of slots remains relatively
stable, some colleges reject more early applicants. According to an article in USA
Today, the University of Pennsylvania received 2,851 applications for early admission.
Of this group, it admitted 1,033 students, rejected 854 outright, and deferred 964 to
the regular admissions pool. Penn admitted about 18 % of the applicants in the
regular admissions pool for a total class size (number of early admissions plus number
of regular admissions) of 2,375 students.

Let E, R, and D represent the events that a student who applies for early admission is
admitted, rejected outright, or deferred to the regular admissions pool; and let A
represent the event that a student in the regular admissions pool is admitted.

Number of students who applied for early admissions = 2,851

Number of these students who received early admissions = 1,033
Number of these students who were rejected = 854
Number of these students who were deferred = 964

2
Percent of students admitted in the regular admissions pool = 18%
Therefore P(A) = 0.18
Number of students admitted total = 2,375

E = student who applied for early admission and was admitted

R = student who applied for early admission and was rejected
D = student who applied for early admission and was deferred
A = student in the regular admissions pool is admitted

(a) Use the data to estimate P(E), P(R), and P(D).

1033
P(E) = 2851 = 0.3623

Therefore, there is a 36.23% chance, or probability of 0.3623, that a student who

applied for early admission is admitted early.

854
P(R) = 2851 = 0.2995

Therefore, there is a 29.95% chance, or a probability of 0.2995, that a student

who applied for early admission is rejected outright.

964
P(D) = 2851 = 0.3381

Therefore there is a 33.81% chance, or a probability of 0.3381, that a student who

applied for early admission is deferred to the regular admissions pool.

(b) Are events E and D mutually exclusive?

Yes, the two events are mutually exclusive. At most one of these events can
occur; a student is either admitted early (E) or is deferred to the regular pool (D).
A student cannot be both admitted early and deferred at the same time.

(c) For the 2,375 students admitted to Penn, what is the probability that a
randomly selected student was accepted for early admission?
1033
P(student from incoming class was early admission) = 2375 = 0.4349
Therefore there is a 43.49% chance, or probability of 0.4349, that a student
randomly selected from the Penn incoming class had been accepted for early
admission.

(d) Suppose a student applies to Penn for early admission. What is the
probability the student will be admitted for early admission or be accepted
for admission in the regular admissions pool?

3
 Because this situation is a bit too complex for any of our formulas to be applied
directly, I fall back on the simple f / n approach.
The population of interest here is all students applying for early admission, so
n=2,851.
f will include all those admitted early (1,033) plus (if we assume that being
differed does not affect your chances in the regular process) 18% of the 964
students who were differed. This gives us,
1033 + 18% * 964 = 1207
f/n = 1207/ 2851 = 0.4234 = 42.34%

8. A newspaper circulation department knows that 84% of the households in a

particular neighborhood subscribe to the daily edition of the paper. Let D denote the
event that a household subscribes to the daily edition (so, P(D) = 0.84). In addition,
they know that the probability that a household that already holds a daily
subscription also subscribes to the Sunday edition (event S) is 0.75; that is, P(S | D) =
0.75. What’s the probability that a household subscribes to both the Sunday and daily
editions of the newspaper?

P(D) = 0.84
P(S | D) = 0.75
P(S and D) = P(D)*P(S | D) = 0.84*0.75 = 0.63
The probability that a randomly selected household subscribes to both editions of the
newspaper [that is, P(S & D)] is 0.63.

9. As researchers, we decide to classify a household as prosperous if its income exceeds

$100,000. We decide to classify a household as educated if the householder completed
college. Select an American household at random, and let A be the event that the
selected household is prosperous and B the event that it is educated. According to the
U.S. Census Bureau’s Current Population Survey, 13.8% of all households are
prosperous, 26.1% of households are educated, and 8.2% of households are both
prosperous and educated.

Income Education
Prosperous 13.8% Educated 26.1%
Non-prosperous 86.2% Non-educated 73.9%

(a) What is the probability that a household selected is either prosperous or

educated P(A or B)?

P(A or B) =
P(A) + P(B) – (P(A) and P(B)) = 0.138+0.261-0.082 = 0.317

(b) What is the conditional probability that a household is prosperous, given that it
is educated p(A|B)? Explain in a few sentences why your result shows that
events A and B are not independent

P(B) = We know that 26.1 (0.261) percent of households are educated.

4
 P(A and B) = We know that 8.2% (0.082) of households are educated and prosperous

In our particular case,

P(A/B) = P(A and B)

P(B)

0.082/0.261 = 0.3142

If A and B were independent, then P(A/B) would be equal to P(A), and also P(A and B)
would equal the product P(A)P(B).

10. Answer problem 2.5 “Coin flips” on page 116 of the Diez et al. textbook.
1) 0.510
2) 0.510
3) 1 – (0.5)10
11. Answer problem 2.9 “Disjoint vs. independent” on page 117 of the Diez et al.
textbook.

1) Assuming the professor doesn’t curve, not using the method of normal distribution to
grade, they are independent but not disjoint. The possibilities of you earning A is
highly related to factors including the time you spend on this course, how many classes
you attend during the semester and probably you gift and interest on this subject,
instead of the grades received by other students in the class. This is the similar case for
that random selected student.
2) The possibility of you earning A is neither independent nor disjoint. The efficiency of
your study group is associated with the final grade of you two. If can be interpreted to
The more productive your group study is, the higher possibility of getting A’s for you
two. However, the possibility of you getting A, which is P1, doesn’t equal to 1-P2, ( P2 is
the possibility of you study partner getting A in this course). Therefore, they are not
disjoint.
3) No, only when the outcome of a certain event is not affected by the outcome of the
other event, can we state that these two events are independent. A simple example is
flipping two coins and they are independent.