Stat170 - Introductory Statistics Semester 1 Assignment 2

Stat170 - Introductory Statistics

Semester 1, 2015 Assignment 2

Due Wednesday 20th May, 5.00pm via the iLearn link only

Submit via the Assignment 2 link in the ASSESSMENTS area of iLearn.

Instructions:

1) The answers to all questions are to be word processed into this document. You can either type formulae into your solution or you can use the equation editor in Word or you may include hand-written equations and diagrams by photographing them so that you have the image saved as a picture file and then pasting (inserting) the image/s into your solution.

2) Your assignment should be uploaded as a Word (.doc or .docx) or PDF file (created from a Word processed document) ONLY. Other formats, including a PDF created from an image, will not be accepted by the system.

3) Please note that iLearn is fairly reliable, but, like any website, it is occasionally unavailable for short periods. Therefore, do not wait until the due date/time to start trying to submit your work. If you have technical difficulties in submitting your Assignment, it is your responsibility to ensure these difficulties are resolved by the submission deadline. (The University has Help services to support you in the event of a problem; see http://ilearn.mq.edu.au/login/MQ/help.htm.)

4) No submissions will be accepted by email or other means.

5) When you think you have submitted your Assignment, please check to make sure that you have submitted the correct file: In iLearn, click on the Assignment, and you should see your uploaded files.

6) This Assignment is worth 12% of your total grade for STAT170. You may discuss the Assignment with your classmates if you wish, but must then complete all of the calculations and writing by yourself. Note: every semester, several students receive a mark of 0 for this Assignment, due to plagiarism. Ensure that you do not copy parts of your friends' Assignment submissions. Submitting your assignment via iLearn is a declaration that the work is your own.

This assignment is worth 12% and is marked out of 45. Penalties apply as follows:

Submitted on Wednesday after 5pm: Deduct 10 marks

Submitted on Thursday: Deduct 15 marks

Submitted on Friday: Deduct 25 marks

No assignments submitted after Friday will be accepted.

7)

Question 1 (12 marks)

The times taken to assemble a car in a manufacturing plant follow a normal distribution with a mean of 20 hours and a standard deviation of 2 hours. Use this information to answer the following questions. Show your working, including an appropriate diagram.

a. Find the probability that a randomly selected car takes between 21 and 24 hours to assemble.

b. Find the probability that the average assembly time for a randomly selected sample of 5 cars manufactured in the plant will be longer than 22 hours.

c. 69% of new cars purchased in Australia have automatic transmission. Find the probability that less than 40 new cars from a random sample of 50 have automatic transmission.

Question 2 (15 marks)

In 2013, a study of 50 cruise ships was undertaken to investigate the variables which determine the crew size of a cruise ship. For each of the 50 cruise ships, information was recorded on their most recent cruise.

Some of the variables recorded on each cruise ship are described below.

Variable Name Variable Description

Ship Name of cruise ship

Age Age of the cruise ship in years as of 2013

Cabins Number of cabins

Passengers Number of passengers on most recent cruise

Crew Number of crew on most recent cruise

The following output was constructed from information recorded on the age and the crew size of the 50 cruise ships in the study. Consider this output for parts a. to f.

Regression Analysis: Crew versus Age

Model Summary

S R-sq R-sq(adj) R-sq(pred)

235.954 49.67% 48.62% 45.38%

Coefficients

Term Coef SE Coef T-Value P-Value VIF

Constant 1301.3 87.2 14.92 0.000

Age -35.74 5.19 -6.88 0.000 1.00

Regression Equation

Crew = 1301.3 - 35.74 Age

Question 2 continued

a. Comment on the relation between the age of a cruise ship and the number of crew employed on the ship:

b. Did the oldest cruise ship have the largest crew size?

c. What was the approximate age and crew size of the oldest cruise ship?

d. 6 of the cruise ships were 12 years old. What was the approximate crew size of the 12 year old ship with the smallest crew size?

e. The 12 year old cruise ship with the smallest crew size had the largest residual. Calculate the approximate value of this residual, using your results from part d.

f. Write down one reason (referring specifically to relevant part of the output on the previous page) why a linear model may not be a good model for predicting the number of crew required on a cruise ship from the age of the cruise ship:

Question 2 continued

Now consider both the number of cabins and the number of passengers as possible predictors of the crew size. The following output was constructed from the 50 cruise ships in the sample. Use this information for the remainder of Question 4.

Crew vs Cabins:

Regression Analysis: Crew versus Cabins

Model Summary

S R-sq R-sq(adj) R-sq(pred)

81.3511 94.02% 93.89% 93.52%

Coefficients

Term Coef SE Coef T-Value P-Value

Constant 96.6 26.3 3.67 0.001

Cabins 0.7791 0.0284 **** *****

Crew vs Passengers:

Regression Analysis: Crew versus Passengers

Model Summary

S R-sq R-sq(adj) R-sq(pred)

94.2836 91.96% 91.80% 91.15%

Coefficients

Term Coef SE Coef T-Value P-Value

Constant 118.3 29.9 3.95 0.000

Passengers 0.3726 0.0159 23.44 0.000

Question 2 (continued)

g. Carry out an appropriate hypothesis test, using any relevant information from the previous page, to answer the following research question. (You should comment on any test assumption/s but you can assume that assumption/s are reasonably well satisfied.)

Research Question: Is the number of cabins on a cruise ship a useful predictor for the crew size?

Hypothesis Test:

h. Now consider the following research question:

Research Question: Is there a significant linear relation between the number of passengers on a cruise ship and the crew size required?

Calculate a 95% confidence interval to estimate the population slope of the line which relates the number of passengers and the crew size of a cruise ship. Interpret your interval in order to answer the research question. Note that the assumptions are satisfied for this linear model (see scatter plot and residual plots) – you do not need to comment on these here).

i. Confidence Interval:

ii. Conclusion:

.

Medical researchers carried out a study to compare various methods of pain relief for patients suffering different types of pain. 115 patients who took part in this study to test the effectiveness of slow release ‘pain patches’ applied directly to the skin. 52 of these subjects suffered from chronic back pain while the other 63 subjects were chronic migraine sufferers. Each subject had pain scores recorded before and again 2 hours after applying the ‘pain patch’ to the skin. Pain scores were a self-assessed score out of 10. A score of 0 indicated no pain and a score of 10 would indicate the most extreme possible pain. Questions 3 to 5 are based on this study.

Variable Description

ID ID number for patient

Condition Medical Condition: 1 = Back pain, 2 = Migraine

PainScoreBefore Self-assessed pain score before pain patch treatment applied

PainScoreAfter Self-assessed pain score 2 hours after pain patch treatment applied

Question 3 (8 marks)

Using appropriate parts of the following output which was constructed using the results of this study to carry out a hypothesis test in the space provided on the following page to answer this research question:

Research Question: Do self-assessed pain scores two hours after applying pain patches differ, on average, between patients suffering back pain and patients suffering with migraine.

Descriptive Statistics: PainScoreAfter_Back, PainScoreAfter_Migraine

Variable N Mean SE Mean StDev Minimum Median Maximum

PainScoreAfter_Back 52 5.933 0.155 1.116 3.500 6.000 8.500

PainScoreAfter_Migraine 63 4.999 0.118 0.938 3.000 5.000 7.000

Question 3

Hypothesis Test:

Question 4 (10 marks)

Research Question: Is there a significant change in self-assessed pain scores two hours after applying ‘pain patches’ to the skin for migraine sufferers?

a. The file, PainScores.mtw, which is available on iLearn, contains some of the data from the study described on page 6. Use these data to carry out an appropriate hypothesis test to answer this research question above.

b. Copy and paste appropriate Minitab output for this test, along with appropriate graphical summaries onto the following page. All Minitab output should be neatly presented and clearly labelled. Marks will be deducted for summaries which do not have appropriate titles and labels.

c. Write a report addressing the research question above on the page provided. Refer to the document ‘A Guide to Report Writing’ and, in particular, to the section headed ‘A Short Guide to Report Writing for STAT170 students’ which is provided on iLearn. Your report should be based on the results of an appropriate hypothesis tests as directed in this guide. Your report not exceed one A4 page and should be word-processed and presented on the page provided. It should include the following headings:

Introduction

Methods

Results

Conclusion

Assignment 1

8

STAT170 Assignment 2 Session 1 2015

2000-200-400-600

14

12

10

8

6

4

2

0

Residual

F

r

e

q

u

e

n

c

y

Histogram

(response is Crew)

180016001400120010008006004002000

1400

1200

1000

800

600

400

200

0

Cabins

C

r

e

w

Scatterplot of Crew vs Cabins

1400120010008006004002000

200

100

0

-100

-200

Fitted Value

R

e

s

i

d

u

a

l

Versus Fits

(response is Crew)

160800-80-160

16

14

12

10

8

6

4

2

0

Residual

F

r

e

q

u

e

n

c

y

Histogram

(response is Crew)

3500300025002000150010005000

1400

1200

1000

800

600

400

200

0

Passengers

C

r

e

w

Scatterplot of Crew vs Passengers

1400120010008006004002000

200

100

0

-100

-200

-300

Fitted Value

R

e

s

i

d

u

a

l

Versus Fits

(response is Crew)

1000-100-200

16

14

12

10

8

6

4

2

0

Residual

F

r

e

q

u

e

n

c

y

Histogram

(response is Crew)

108642

20

15

10

5

0

108642

16

14

12

10

8

6

4

2

0

PainScoreAfter_Back

F

r

e

q

u

e

n

c

y

PainScoreAfter_Migraine

Histogram of PainScoreAfter_Back, PainScoreAfter_Migraine

PainScoreAfter_MigrainePainScoreAfter_Back

9

8

7

6

5

4

3

D

a

t

a

Boxplot of PainScoreAfter_Back, PainScoreAfter_Migraine

403530252015105

1400

1200

1000

800

600

400

200

0

Age

C

r

e

w

Scatterplot of Crew vs Age

120010008006004002000

400

200

0

-200

-400

-600

Fitted Value

R

e

s

i

d

u

a

l

Versus Fits

(response is Crew)