UNIVERSITY OF REGINA

Department of Mathematics and Statistics

Elementary Statistics For Applications

Stat 100

Final Exam, Fall 2013

Time: 3 hours Full Name:

Pages: 10 Student Number:

Instructor: (check one)

◻ Peter Douglas (100-001)

◻ Robert Petry (100-C01)

◻ Vijayaparvathy Agasthian (100-L01/L02)

◻ Patrick Maidorn (100-991)

INSTRUCTIONS

1. All work and answers are to be placed in this booklet. If you require more space for
an answer, work on the back of the sheet and indicate where the work is to be found.

2. Place final answers in blanks at right when provided.

3. Each question will be graded for both the correct answer and supporting work.

4. The marks are placed beside each question.

5. Only calculator, formula sheet, and the provided probability tables and scrap
paper are permitted in the exam.

For instructor use only:

Page Marks Score

2 15
3 10
4 15
5 10
6 11
7 11
8 10
9 10
10 8
Total: 100
 STAT 100 Final Exam Page 2 Fall 2013

Marks
(15) 1. Joe’s Nut Bar sells snacks to office workers. A random selection of Nut Bar products
had the following prices:

7.25, 7.50, 7.50, 7.50, 8.25, 11.00, 12.50, 13.00, 15.25 ($)

(a) Is the variable discrete or continuous? discrete/continuous

(Circle your answer at right.)

Solution: Discrete as prices do not come in fractions of a cent.

(b) Determine the mode. Answer:

Solution:
mode = $7.50

(c) Determine the mean. Answer:

Solution:
$89.75
x̄ = = $9.9722222 ≈ $9.97
9

(d) Determine the standard deviation. Answer:

Solution:
¿
Á 968.1875 − (89.75)2 √ 73.18055556 √
Á
À
s= 9
= = 9.147569445 = 3.024494907 ≈ $3.02
9−1 8

(e) Determine the first and third quartiles. Answer: Q1 =

Q3 =

Solution:
Q1 = x 1 (9+1) = x2.5 = $7.50
4

since x2 = x3 = $7.50
x7 + x8 $12.50 + $13.00
Q3 = x 3 (9+1) = x7.5 = = = $12.75
4 2 2

(f) Is $15.25 an outlier in the above dataset? Circle your answer at right yes/no
and show the relevant calculation used to determine your result below.

Solution: The interquartile range is IQR = Q3 − Q1 = $12.75 − $7.50 = $5.25

The lower fence and upper fence are

Lower Fence = Q1 − 1.5(IQR) = $7.50 − 1.5($5.25) = −$0.375

Upper Fence = Q3 + 1.5(IQR) = $12.75 + 1.5($5.25) = $20.625

Since $15.25 lies within this range it is not an outlier.
 STAT 100 Final Exam Page 3 Fall 2013

(10) 2. A statistics instructor tracks the calls and texts that he receives on his cell phone over a
month. His results are as follows.

Contact Source
Family Opinion Scam Wrong
Type of Message Member Survey Artist Number Total
Phone Call 7 10 13 6 36
Text 8 4 8 4 24
Total 15 14 21 10 60

Using this data, for a randomly selected message . . .

(a) What is the probability that it came from a Answer:

family member?
Solution:
15 1
P (F ) = = = 0.25
60 4

(b) What is the probability it is a phone call that is a Answer:

wrong number?
Solution:
6 1
P (P and W ) = = = 0.10
60 10

(c) What is the probability that it is a text message Answer:

or a scam artist?
Solution:
24 21 8 37
P (T or S) = P (T ) + P (S) − P (T and S) = + − = = 0.6167
60 60 60 60

(d) What is the probability that it is a family member Answer:

contacting him, given that he received a text
message?
Solution:
8 1
P (F ∣T ) = = = 0.3333
24 3

(e) Are the events “receives a text message” and “is contacted yes / no
by a family member” independent? Why?
(Circle your answer at right and show your work below.)

Solution: N o since (any of the following):

8 2 24 15 1
ˆ P (T and F ) = = = 0.1333 ≠ P (T ) ⋅ P (F ) = ⋅ = = 0.10
60 15 60 60 10
8 24
ˆ P (T ∣F ) = = 0.5333 ≠ P (T ) = = 0.4
15 60
8 15
ˆ P (F ∣T ) = = 0.3333 ≠ P (F ) = = 0.25
24 60
 STAT 100 Final Exam Page 4 Fall 2013

(10) 3. Katniss Everdeen is an expert archer who can hit the bulls-eye on a target in 90% of
her shots.
(a) In a particular training session Katniss takes 8 shots at Answer:
the target. What is the probability that she hits the
bulls-eye at least 6 times?
Solution:

P (x = 6) = 8 C6 (0.9) (0.1)
6 2
= 0.14880348 ≈ 0.1488
P (x = 7) = 8 C7 (0.9) (0.1)
7 1
= 0.38263752 ≈ 0.3826
P (x = 8) = 8 C8 (0.9) (0.1)
8 0
= 0.43046721 ≈ 0.4305

So P (6 ≤ x) = P (x = 6) + P (x = 7) + P (x = 8) = 0.96190821 ≈ 0.9619

(b) Over a two-week period Katniss takes 240 shots at the Answer:
target. Use the normal approximation to the binomial
to estimate the probability that she hits the target at
least 220 times.
Solution: The√ mean for the binomial
√ is µ = (240)(0.9) = 216 with standard
deviation σ = (240)(0.9)(0.1) = 21.6 = 4.647580015. The probability is

219.5 − 216
P (220 ≤ xbinomial ) = P (219.5 < xnormal ) = P ( < z)
4.64 . . .
= P (0.753080095 < z) = P (z < −0.75) = 0.2266

Note if no continuity correction is used the (incorrect) answer is 0.1949 .

(5) 4. A Toronto newspaper wishes to conduct a poll to estimate Answer:

the current approval rating of the mayor. The question
asked will be “Do you approve of the actions of the
mayor?” with the only choice of answers being Yes or No.
How large a sample should be selected if the newspaper
wants its poll results to be accurate within 4% of the true
population proportion, 19 out of 20 times?

Solution: E = 0.04, z = 1.960 for 1 − α = 19/20 = 0.95, and proportion estimate (since
none given) is 0.50. Sample size is:

1.960 2
n = (0.50)(0.50) [ ] = 600.25 = 601 residents
0.04
 STAT 100 Final Exam Page 5 Fall 2013

(10) 5. The weight of male reindeer of the R. Santaclausus subspecies is normally distributed
with mean 102.4 kg and standard deviation 13.9 kg.
(a) What proportion of these reindeer would weigh more Answer:
than 118.0 kg?
Solution:
118.0 kg − 102.4 kg
P (118 kg < x) = P ( < z)
13.9 kg
= P (1.122302158 < z)
= 1 − P (z < 1.12)
= 1 − 0.8686
= 0.1314

(b) Rudolph is a fairly small reindeer. If exactly 10% of Answer:

the reindeer weigh less than Rudolph, how much does
Rudolph weigh?

Solution: 10% of the probability in a normal distribution is less than z = −1.28.

Hence
x = µ + zσ = 102.4 + (−1.28)(13.9) = 84.608 kg ≈ 84.6 kg

(c) If 36 reindeer are randomly selected and their average Answer:

weight calculated, what is the probability that the
mean weight is less than 100.0 kg?
√ √
Solution: Standard error of the mean is σx̄ = σ/ n = 13.9/ 36 = 2.31666666 kg
and so
100.0 kg − 102.4 kg
P (x̄ < 100 kg) = P (z < )
2.316 . . . kg
= P (z < −1.035971223)
= P (z < −1.04)
= 0.1492
 STAT 100 Final Exam Page 6 Fall 2013

(11) 6. Fish oil is a popular supplement taken to support cardiac and cognitive health. A
pharmaceutical company is investigating a new process to extract raw fish oil from
salmon. In a limited test run, the process was used on 18 one-kilogram batches of
salmon. On average, the process was able to extract 50.7 grams of fish oil per batch,
with sample standard deviation 5.9 grams.
Use this information to construct a 90% confidence interval for the true mean amount
of fish oil that this process could extract per one-kilogram batch of salmon. As part of
your answer state any necessary assumptions that have to be made.

Solution:
5.9 g
ˆ Standard error = √
18
= 1.390643336 g

ˆ t = 1.740 for df = 17 and α = 0.90

ˆ Maximum error E = 2.419719405 g ≈ 2.4 g

The 90% confidence interval is 50.7 g − 2.4 g < µ < 50.7 g + 2.4 g or equivalently
48.3 g < µ < 53.1 g
Assumption: the amount of oil extracted per batch is approximately normally dis-
tributed.
 STAT 100 Final Exam Page 7 Fall 2013

(11) 7. Statistics Canada reports that half of all households have at least one pet, with pet
ownership higher in Alberta, Manitoba, and Saskatchewan and lower in the Maritimes,
British Columbia, Ontario, and Quebec.
(a) In a random sample of 300 households across Canada, 126 households said that
they owned at least one pet. Does this data provide sufficient evidence to indicate
that the proportion of households with at least one pet is different from that
reported by the Statistics Canada? Test using a level of significance of α = 0.05 .

Solution: Critical z-values for the two-tailed test (“different”) are ±1.960 .
Sample proportion p = 126/300 = 0.42. Standard error of proportion is
√
(0.5)(0.5)
= 0.028867513 .
300
Calculated z-value for the data is therefore
0.42 − 0.50
z= = −2.771281292 = −2.771
0.0288 . . .
Therefore we reject H0 at α = 0.05 and accept Ha . The evidence supports that
the proportion of households with at least one pet differs from the 50% reported.

(b) What is the P -value for the hypothesis test?

Solution: The area from the normal curve table for z = −2.77 is 0.0028 . Dou-
bling for the two tails gives

P -value = 2(0.0028) = 0.0056

which is less than α = 0.05 as expected for a test in which the null hypothesis
was rejected.

Answer:
 STAT 100 Final Exam Page 8 Fall 2013

(10) 8. Boston Bobby is a die-hard baseball fan and supports his hometown team the Boston
Red Sox. Bobby has an ongoing argument with New York Nat over the defensive
capabilities of the Boston Red Sox and the New York Yankees team. Bobby claims that
the Red Sox have a superior defense and uses the following statistics to back up his
claim.

Runs Allowed per Game

Games Mean Standard Dev.
Team Played (runs) (runs)
Boston Red Sox 35 3.96 1.07
New York Yankees 32 4.25 1.15

Bobby claims that Boston will allow fewer runs on average than New York. Test this
hypothesis using a 5% level of significance.

Solution: Assuming a left-tailed test (µB − µN√< 0) the critical z-value is −1.645.
(1.07)2 2
The standard error of the difference of means is 35 + (1.15)
32 = 0.272102101 runs.
So the calculated z-value for the data is
3.96 − 4.25
z= = −1.065776407 ≈ −1.066
0.2721 . . .
Therefore we fail to reject H0 at α=0.05 . The evidence does not support Bobby’s
claim. Alternatively the P -value for z = −1.07 is 0.1423 which is greater than α = 0.05
resulting in the same conclusion. Note that a z-test was valid here due to the large
sample sizes (n ≥ 30).
 STAT 100 Final Exam Page 9 Fall 2013

(10) 9. A mathematics drop-in help centre keeps track of the number of students it helps on a
given day with the following results:

Monday Tuesday Wednesday Thursday Friday

4 3 6 1 4
7 6 5 3 2
8 7 8 3 1
4 3 0 5 1
7 4 3 3 1
6 0 2 3

Assuming that the underlying populations are normal with common standard
deviation, can one conclude that the average number of students helped in the centre
on a given day differs from any other at a level of significance of α = 0.01?
Use the following partially completed ANOVA table in your analysis:

Source df Sum of Squares Mean Squares F

Treatments 55.8
Error 99.0
Total 154.8

Solution: The exact ANOVA table is

Source df Sum of Squares Mean Squares F

Treatments 4 55.7586 . . . 13.9396 . . . 3.379. . .
Error 24 99.0000 4.1250
Total 28 154.7586 . . .

For df1 = 4 and df2 = 24 at α = 0.01 we have Fcritical =4.218 . Since the calculated
F =3.379 (3.382 for the rounded values given in the table) is less than this we fail
to reject H0 at α = 0.01; the evidence does not support that the mean number
on any particular day differs from any other. (Note that for the data the actual
P -value=0.025 .)
 STAT 100 Final Exam Page 10 Fall 2013

(8) 10. Data from a study on the number of absences and the final grades of seven randomly
selected students from a statistics class is given in the following table.

Number of absences (x) Final Grade (y)

6 82
2 86
15 43
9 74
12 58
5 90
8 78

The following summations for the data have already been calculated:
Σx = 57, Σy = 511, Σx2 = 579, Σy 2 = 38993, Σxy = 3745

(a) The correlation coefficient for the relationship between number of absences and final
grade is calculated to be r = −0.9442 . Interpret the meaning of this value in plain
English.

Solution: The negative sign indicates a negative correlation between the vari-
ables. (As absences x increase the grade y decreases.) The magnitude of r
indicates that the correlation is extremely high. Therefore, in summary, there
is an extremely high negative correlation.

(b) Find the equation of the best fit (least-squares) regression line.

Solution: Using the following (or other) standard formulas gives:

(∑ x2 ) ⋅ (∑ y) − (∑ x) ⋅ (∑ xy)
a =
n ⋅ (∑ x2 ) − (∑ x)
2

(579) ⋅ (511) − (57) ⋅ (3745) 82404

= = = 102.4925373 ≈ 102.5%
(7) ⋅ (579) − (57)
2
804
n ⋅ (∑ xy) − (∑ x) ⋅ (∑ y)
b =
n ⋅ (∑ x2 ) − (∑ x)
2

(7) ⋅ (3745) − (57) ⋅ (511) −2912

= = = −3.621890547 ≈ −3.62 %/absence
(7) ⋅ (579) − (57)
2
804

The linear regression line ŷ = a + bx is therefore

ŷ = 102.5 − 3.62x [%]

Answer: