Department of Statistics

STATS 101/101G/108 Introduction to Statistics

Assignment 2, Second Semester 2016
Due: 3pm Monday 12th September

Question 1. [7 marks] [Chapter 3]

In a recent test of the effectiveness of a new sleeping pill, 75 volunteers were randomly assigned
to three groups of 25. The first group was given the new drug, the second group was given a
placebo and the third group was given no treatment at all. Whether or not the volunteer managed
to fall asleep in under 30 minute was recorded.

Read these instructions carefully

Marks
Assignment 2 is worth 5% of your final mark.
It will be marked out of 45 marks, 40 marks for the questions and 5 marks for communication
and presentation. See below for how communication and presentation marks are allocated. Your
final mark will be converted to a mark out of 10 which will be recorded towards your course
work.
Statistics is about summarising, analysing and communicating information. Communication is
an important part of statistics. For this reason you will be expected to write answers which
clearly communicate your thoughts.
Communication and Presentation marks
Demonstrate clear sentence structure: this includes correct use of full stops and capital
letters; not writing overly long or complicated sentences; attention to spelling and grammar.
Demonstrate ability to communicate information clearly in sentences: this includes
sentences clearly conveying the correct idea; sentences making sense; comments not being
excessively long or short; conclusions following logically from previous statements.
Assignment tidily set out and easy to follow: this includes the answers being clearly set
out in the correct order; the assignment not being messy; graphs and plots tidy with correct
labelling of axes; the assignment (including the correct cover sheet) being clipped together
or stapled.
Follow the Step-by-Step Guide to Performing a Confidence Interval by Hand as
required.
Student ID number shown somewhere on assignment: this can be on the inside of the
cover sheet or on the top of the first page of the assignment.
Handing in
Hand in to the appropriate assignment drop-off box to the left of the counter in the
Student Resource Centre, ground floor of building 301, by the plaza that connects buildings
301 and 303. Do not hand your assignment in to the unsecured assignment return boxes!
Assignments handed in to the wrong place or received after the due time will not be marked.
Question guide
Attempt question 1 when chapter 3 has been covered.
Attempt question 2 when chapter 4 has been covered.
Attempt question 3 when chapter 5 has been covered.
Attempt questions 4 and 5 when chapter 6 has been covered.
Question 1 and 3 will require use of VIT. Hand in the required computer output.
Notes
The format and handing in of Assignment 2 is the same as that for Assignment 1. Refer to the
instructions on page 1 of Assignment 1.
Refer to the Worked Examples under Assignment Resources on Canvas for examples of how to
set out your answers.
Refer to the Lecture Workbook, Section A (Course Information), page 3, Assignment Rules:
Working together versus cheating.

The data for this study is stored in the file Sleep.csv which can be downloaded from Canvas.
The data contains 2 variables:
Treatment
Sleep

The treatment group the volunteer was assigned to (either Drug, Placebo or
Neither).
Whether the volunteer fell asleep within 30 minutes (either Yes or No).

Download the VIT Guide for Randomisation Tests from Canvas.

Use the Guide while working on this question.
(a)

Briefly explain why this study is an experiment.

(b)

(i)

Run the iNZightVIT software and load the file Sleep.csv into it.

Run a randomisation test to compare the proportion of volunteers who fell asleep within
30 minutes between the three treatment groups Include the output from this in your
assignment answers.
Notes: Variable 1 needs to be Sleep and Variable 2 needs to be Treatment. Before you select
Record my Choice in the Analyse window, change the level of interest to Yes.
(ii)

When chance is acting alone, would it be unusual to get an average deviation from the
overall proportion who fell asleep at least as big as the observed difference?
(Use your randomisation test output to answer this.)

(iii) Is it plausible that the observed average deviation from the overall proportion who fell
asleep can be explained by chance acting alone? Briefly justify your answer.
(iv) Can we conclude that different treatments caused differences in the sleep rate? If so,
justify why with two reasons. If not, what can we conclude?

 Department of Statistics

Question 2. [7 marks] [Chapter 4]

(a)

Many research organisations give their interviewers exact scripts to follow when conducting
interviews to measure opinions on controversial issues.
(i) Give one type of bias which the research organisations are trying to minimise by using
exact scripts for the interview.
(ii) Will the use of scripts completely eliminate this form of bias? Briefly justify your answer.

(b)

In March 2001, Princeton Survey Research Associates conducted a poll asking about the 2000
presidential election in America. The poll asked a random sample of 1200 adults Did you vote
in the 2000 presidential election? 72% said yes and 28% said no (with a margin of error of
3 percent). The actual voter turnout was 51%.
What non-sampling error is most likely to explain the difference in the polled and actual figures.
Justify your answer.

(c)

In a U.S. court case, Bristol Myers was ordered by the Federal Trade Commission to stop
advertising that twice as many dentists use Ipana than any other toothpaste. Bristol Meyers
had based their claim on a survey of 10,000 randomly selected subscribers to two dental
magazines. They received 1,983 responses, with 621 saying they used Ipana and only 258
saying they used the second most popular brand.
(i)

Explain how selection bias may be a potential problem with the survey.

(ii)

Explain why self-selection bias is not a potential problem with the survey.

Question 3. [10 marks] [Chapter 5]

How long do you have to wait for your coffee after placing an order? And does the waiting time
depend on the gender of the person ordering the coffee? The following data was imputed from a
study Ladies first? A field study of discrimination in coffee shops. by Caitlin Myers. In this
study, the times (in seconds) between ordering and receiving coffee for a random sample of
customers in Boston-area coffee shops was recorded.
The data are stored in the file "Coffee.csv" which can be downloaded from Cecil.
contains 2 variables:
Wait

The wait time between placing the order and receiving the coffee (in seconds)

Gender

The gender of the customer (Male or Female)

Download the VIT Guide for Bootstrapping from Canvas.
Use the Guide while working on this question.

Run the iNZightVIT software and load the file Coffee.csv into it.
(a)

(i)
(ii)

(b)

(i)

(iii) Apart from selection bias, what is the main non-sampling error affecting the survey?
Briefly justify your answer.
(iv) Do the results of the survey provide convincing evidence of Bristol Myers claim? Briefly
justify your answer.

The data

Why would the median be a better estimate of the centre of this data than the mean?
Generate a bootstrap confidence interval for the median wait times of customers. (DO
NOT use the variable Gender at this point.) Include the output in your assignment
answers.
(iii) What is the parameter we are estimating using this bootstrap confidence interval?
(iv) Do we know the true value of this parameter?
(v) Interpret the bootstrap confidence interval.
Generate a bootstrap confidence interval for the difference in the median wait times
between males and females. Include the output in your assignment answers.
(ii) What is the parameter we are estimating using this bootstrap confidence interval?
(iii) Interpret the bootstrap confidence interval.
(iv) Based on the bootstrap confidence interval, is it believable that the median wait time for
males is the same as the median wait time for females? Briefly justify your answer.

Question 4. [7 marks] [Chapter 6]

A statistics student was interested in investigating how long it takes to get a pizza delivered
from the local pizzeria. Over a few weeks, a random sample of 10 delivery times (in minutes)
was recorded. The data are displayed below:
18.5, 27.4, 19.8, 23.8, 23.4, 19.2, 28.4, 21.3, 17.9, 22.2
Summary Statistics: x = 22.19 minutes. And s = 3.61
(a)

Calculate and interpret a 95% confidence interval for the mean delivery time.
Note: You must clearly show that you have followed the step-by-step guide to producing a
confidence interval by hand given in the Lecture Workbook, Chapter 6. Use the t-procedures
tool to find values for t-multipliers and standard errors.

(b)

The student believes that, on average, pizzas from this pizzeria take less than 25 minutes to
arrive. Does the data support this? Briefly justify your answer.

 Department of Statistics

Question 5. [9 marks] [Chapter 6]

Southern Cross Travel Insurance commissioned a survey about New Zealanders travel in late
2014. Some of the results are shown below:
Respondents who had travelled overseas
before were asked what was the most
annoying thing a nearby traveller can do
on a flight.
Results, classified by age, are:
Annoyance
Under 40
Let children misbehave
149
Talk loudly/incessantly
89
Recline seat
75
Take over your armrest
41
Smell
234
Other reason
42
Not sure
58

40+
208
183
119
55
251
54
75

Total

945

688

621 males and 938 females said they

would take out travel insurance if they
were to travel for a holiday. They were
asked which of the following were
reasons for taking out travel insurance.
To cover cost of:
Males
lost luggage
492
emergency medical
treatment
571
stolen items
480
car accidents
387
missed connections 397
cancelled activities
286

Females
795
897
750
607
634
465

(a)

State the sampling situation (a, b or c) for calculating the standard error of the difference in the
following scenarios:
(i) for people who would take out travel insurance, estimating the difference between the
proportion of females who did so to cover the cost of lost luggage and the proportion of
males who did so to cover the lost luggage.
(ii) for people who would take out travel insurance, estimating the difference between the
proportion of females who did so to cover the cost of car accidents and the proportion of
females who did so to cover the cost of stolen items.
(iii) for people who had travelled overseas before, estimating the difference between the
proportion of people under 40 who thought nearby travellers letting children misbehave
was most annoying and the proportion of people under 40 who thought nearby travellers
reclining their seat was most annoying.

(b)

Consider people who had travelled overseas before. Calculate and interpret a 95% confidence
interval for the difference between the proportion of people under 40 years of age who thought
the most annoying thing a nearby traveller could do was smell and the proportion of people at
least 40 years of age who thought the most annoying thing a nearby traveller could do was
smell.
Note: You must clearly show that you have followed the step-by-step guide to producing a
confidence interval by hand given in the Lecture Workbook, Chapter 6. Use the t-procedures
tool to find values for t-multipliers and standard errors.