微信客服：xiaoxionga100
微信客服：ITCS521
FIT1006-无代写
时间：2024-05-25
Information Technology
Semester 2, 2016Welcome to FIT 1006
Business Information Analysis
Semester 1, 2024
FIT1006 Business Information Analysis
Teaching Team:
Lecturer and Chief Examiner: A./Prof. David Dowe
Tutors (or Teaching Assistants, TAs) : Dr Heshan Kumarage (head
tutor or admin’ tutor), Anthony Wong, Dulan Perera, Emily Yap Sin-Yee,
Flora Jin
Please use your Monash account for all e-mails to lecturer
and tutors. E-mail addresses should be clear from
Moodle. As per the “Welcome to FIT1006” post in Ed
Discussion, it will typically be better to send a post to Ed
Discussion, mindful of
https://learning.monash.edu/course/view.php?id=14294&s
ection=58 Expectations for Forum Communication.
If e-mailing, then the first 8 characters of the e-mail subject
line should be `FIT1006 ‘ (FIT1006 followed by a space).
2FIT1006 Business Information Analysis
Lectures
Lectures
currently recorded unless we are subsequently notified
otherwise
Most lectures will be best if done with pre-reading, which will be
posted on Moodle
Lecture slides will be available after the lecture on Moodle as
PDF.
Lectures recordings available on Moodle.
Some topics will run across multiple lectures.
3FIT1006 Business Information Analysis
Applied Sessions (or Labs)
Enrol for tutorials on Allocate+ :
– https://my-timetable.monash.edu/odd/student
If you need help with your timetabling:
– https://connect.apps.monash.edu/students/timetables/allocat
e/help/
Applied sessions start in Week 2 (2-hour session per week)
Download the applied session sheet (or tutorial sheet) each week
Arrive prepared for each applied session. This means: revising
lectures + reading + practice questions.
Applied sessions will typically comprise: review of lecture, worked
examples, practice questions, manual & computer calculations.
+ 8 hours study/practice/research each week!
4FIT1006 Business Information Analysis
Consultation and Help
Consultations:
– Start from approx Week 3. Consultation times will be posted on
Moodle
Forum – should be your starting point: Ed Discussion Forum
– We can often reply fairly promptly from amongst the
teaching team.
• If someone hasn’t answered already.
Applied Sessions.
Support and help: See the “Welcome to FIT1006” post on Ed
Discussion and various Moodle pages for pointers to help
5FIT1006 Business Information Analysis
Resources
Subject Resources
– Download all lecture notes, applied session sheets, and
other materials from Moodle
Recordings – also on Moodle,
Assignment instructions – released on Moodle.
Support and help: Again, please see the “Welcome to FIT1006”
post on Ed Discussion and also various Moodle pages for
pointers to help
6FIT1006 Business Information Analysis
Textbook and References
Prescribed text:
Business Statistics Abridged, (7th Ed., 2017, or) 8th edition
Selvanathan et al., Cengage, South Melbourne, 2021.
(5th and 6th editions are purportedly quite similar in content)
For any of you wishing to buy a copy, see link within Moodle of
discount offer from publisher
Additional Reading
– Statistics Without Tears, Derek Rowntree, Penguin,
Harmondsworth, 1981.
– Statistics Explained: An introductory guide for life scientists,
Steve McKillup, Cambridge U.P., 2006.
Wikipedia – a handy reference for at least many topics.
Software: MicroSoft Excel and SYSTAT
7FIT1006 Business Information Analysis
Mathematics!
This unit has a mathematics prerequisite of:
A study score of 25 in VCE Mathematics Methods or
Specialist Maths units 3 & 4 or 30 in Further Maths units 3
& 4 or MTH1010 or equivalent.
If you have come to university via a non-VCE pathway, or
are just curious, you can see some sample questions on
Moodle.
Topics covered in this unit will include (e.g.) probability and
hypothesis testing, so please be mindful that at least some
of the mathematics might be – and/or become - challenging
8FIT1006 Business Information Analysis
Assessment
To pass the unit you must obtain:
– An overall unit mark of 50% or more.
9
Unit Assessment
Assessment Due Date % of Mark
Online quiz currently planned for Week 4 2
Data Collection currently planned for Week 5 18
Data Analysis
Report
currently planned for Week 9 40
In-class quiz TBC, currently planned in-
class in week 12
15
Take home test TBC, currently planned for
week 14
25
FIT1006 Business Information Analysis
Quiz 1 – 2%
- Quiz 1, 2%.
- To be done individually and as per given instructions.
- Intend to test on material covered up to the end of the
week before.
10FIT1006 Business Information Analysis
Assignment 1: Data Collection - 18%
- To be done individually and as per given instructions
- Intend to test on material covered up to the end of the
week before it is due
- We’ll go through this in more detail later.
- For students who have work in progress that they’d like to
discuss, we hope to have time for students to have 1-to-1
discussions in Applied sessions or consultation sessions
11FIT1006 Business Information Analysis
Assignment 2: Data Analysis - 40%
- To be done individually and as per given instructions
- Intend to test on material covered up to the end of the
week before it is due
- Write a report following the assignment instructions. We
anticipate that it will involve at least some of:
- Data visualisation
- Hypothesis testing
- Reflective commentary.
- We plan to go through this in more detail before it is due.
- We hope to give everyone some time to talk with a tutor
about at least part of an assignment draft
12FIT1006 Business Information Analysis
In-Class Test - 15%
- To be done individually and as per given instructions
- Intend to test on material covered up to the end of the
week before.
- Make sure to arrive punctually at your allocated applied
session and to follow instructions.
13FIT1006 Business Information Analysis
Take-Home Test - 25%
- To be done individually and as per given instructions
- We’ll explain in more detail closer to the time.
- Intend to test on material covered up to the end of the
week before it is due (i.e., the entire syllabus)
- There will possibly be some selective use of generative AI
permitted, but stay tuned for further instructions
14FIT1006 Business Information Analysis
General Observations
- Make sure you follow the course and are learning as you
go along!
- We are here to help you with this course and want
you to learn and do well.
- Try to avoid the crunch:
- Start assignments as early as possible.
- We intend to give you at least 10 calendar days in each case
(and perhaps more in the case of Assignment 2).
- Towards the end of the Semester, units tend to all pile up!
Trying to do these things at the last minute is usually not a good idea.
- Whether or not you might possibly get extensions, please do not
rely on this happening
Support and help: See the “Welcome to FIT1006” post on Ed
Discussion and various Moodle pages for pointers to help
15FIT1006 Business Information Analysis
Unit Learning Outcomes
16FIT1006 Business Information Analysis
01
Perform and justify the use of
appropriate descriptive and inferential
statistical analyses such as standard
deviation, t-tests, and Pearson's
correlation coefficients;
02 Communicate the results of statistical analysis effectively and accurately;
03
Create graphics and tables for effective
communication of the results of data
analysis;
04 Analyse the factors that make statistics reliable, credible and trustworthy.
Brief outline – First 3 or 6 Weeks or so
• Week 1: Intro + Surveys and Data Collection
• Week 2: Data Representation and Presentation
• Summary stats and some charts and plots.
• Week 3: Intro to statistical software and report writing.
• This is useful for Assignments 1 and 2.
• We then come to the following topics:
• Probability
• Correlation and Regression
• Hypothesis testing
• ANOVA (Analysis Of Variance), Classifier Performance, Fair AI
• We also look at aspects including estimation and time-series data.
17FIT1006 Business Information Analysis
Week 1 Learning Outcomes
• Week 1: Intro + Surveys and Data Collection
After completing the work in this week and attending the applied
session in Week 2, you should be able to:
Understand the purpose of statistics and why it is beneficial to study
them.
Understand the concept of sampling and relevant terminology
associated with this (e.g., sampling, population, observation, variable,
etc.)
Be able to explain how to design a survey sampling strategy that is
appropriate with respect to a given scenario.
Understand how to conduct a survey and the types of errors that may
be made when designing a survey.
18FIT1006 Business Information Analysis
About FIT1006…
Quantitative analysis
Some real-world issues where quantitative analysis may lead to
improved outcomes:
• Health: patterns of disease, effectiveness of treatments, the
inequality of nations;
• Civilization: the science of food production, the risk of
catastrophe – the global financial crisis (GFC), climate
change, global warming, social disadvantage;
• Commerce: the volatility of investments, the risk of
borrowers defaulting on loans, …;
• Sport: player statistics and rankings, is something a good
bet?, ...;
• etc.
19FIT1006 Business Information Analysis
About FIT1006… Quantitative analysis
In IT as well as Computing, we need this too!
– Fair AI – does my system unfairly (or unlawfully)
discriminate based on protected attributes?
– A/B testing – I’ve made a change to my system, does it
have the effect I want? (used by at least many of, e.g.,
Twitter, Google, Microsoft, etc. on a day-to-day basis).
– Probability – underpins many algorithms (e.g., in Networks).
– Writing reports – making a persuasive case to sell software,
or to make decisions on what to buy within an organization.
As citizens, being statistically literate is really important too. For
example, it can help you make life decisions (should I rent or get
a mortgage?), understand politics (especially polling) and
interpret media articles (many can mislead with numbers, …).
20FIT1006 Business Information Analysis
Quantitative analysis
Quantitative analysis shows us the big picture.
Using data (usually numerical) to:
• Generalise: what has happened now and in the past (and
may happen in the future), what is typical behaviour?
• cf. N. Goodman’s ``grue’’ paradox
• Normalise: what is normal? What is exceptional?
• Contextualise: how do we compare with others around us?
How are things changing over time?
21FIT1006 Business Information Analysis
Quantitative analysis
Quantitative analysis for business:
• Customer behaviour, randomness, risk;
• Demand patterns, trends – change over time;
• Determining public sentiment;
• Cause and effect relationships;
• Identifying the systematic from the ad hoc, …
• A good example of a company using quantitative analysis to
substantial effect is Google. Their page rank algorithm has
substantially changed the world. Search Engine
Optimisation (SEO) was an industry relatively unknown
approximately 15 or so years ago
22FIT1006 Business Information Analysis
Overview
Quantitative techniques include:
• Descriptive techniques applicable to a wide range of data –
sample, financial, Internet, etc.;
• Conducting surveys and analysing the results;
• Modelling relationships and trends;
• Forecasting based on historical data;
• Working with large(ish) data sets;
• Presenting results as summaries and reports;
• Machine learning;
• Deep learning;
• Testing hypotheses;
• etc.
23FIT1006 Business Information Analysis
FIT1006: The big picture
24
Describing
Data
Probability
Modelling
relationships
Reporting
results
Application of theory
to practical problems
Quantifying
(Un)certainty
FIT1006 Business Information Analysis
FIT1006: A bigger picture
25
FIT3158
Business
decision
modelsFIT1006
Business
information
analysis
FIT3152
Data analytics
The wide world!
Big Data, Analytics, ...
FIT1006 Business Information Analysis
Descriptive statistics
• A statistic is a summary of data. For example, the number of
students enrolled in this subject is a statistic. Statistics are a
way of summarizing the essential features of a large quantity of
data.
• Statistics gets its name from the collection of data about the
State.
• Data is a plural noun. Datum is the singular form.
• Descriptive statistics describe the features of a data set; for
example, the mean and the standard deviation.
26FIT1006 Business Information Analysis
Describing data (a)
• How was the data sample obtained; could this introduce doubt into –
or increase the amount of doubt about - our conclusions?
• Representing the data graphically; calculating summary statistics:
e.g., average height.
27FIT1006 Business Information Analysis
Describing data (b)
• What are the major differences between the two groups as suggested
by the summary statistics?
• Can we represent this graphically?
28FIT1006 Business Information Analysis
Statistical reports
• Obtaining data sometimes requires obtaining ethics clearance (and
filling in forms, etc.)
• Having responsibly obtained our data and accurately calculated
statistics, perhaps the next most important skill is the interpretation
and communication of the results.
• Reports should discuss the significance and consequences of findings
(and some sort of measure of uncertainty) as well as any assumptions
made in analysis.
• Statistical summaries should be presented in an easy to read form -
such as a table.
• Your report should ideally be understandable to a person without a
detailed statistical knowledge.
29FIT1006 Business Information Analysis
Probability
• What is the chance that a certain person selected at random will
exhibit a certain property?
• Does that chance change if the person belongs to a subgroup of the
population?
30FIT1006 Business Information Analysis
Inferential statistics
• Having observed some differences between groups of data we may
wish to know whether these differences are due to some systematic
cause or just due to random fluctuations.
• This is an important part of product testing, medical treatments (e.g.,
vaccinations), etc.
• We generally use descriptive statistics to observe differences between
groups and then use inferential statistics to test whether the difference
is significant.
31FIT1006 Business Information Analysis
Being certain - or, at least, quantifying this
• Are the two groups different?
• What measure are we comparing?
• What level of confidence and/or certainty?
• Are you 99% sure – or only 50.1% sure, or …?
32FIT1006 Business Information Analysis
Why is this unit important?
We are constantly being told certain things are true. This subject gives
you the tools to track through the evidence, apply the theory, verify for
yourself, and endeavour to quantify (un)certainty.
For example, from the past and recent press:
• New Vaccine Prevents Severe COVID As Well As Existing Vaccines
Do, Experts Say. (a reputable popular science magazine, 2021)
• Belly fat may be resistant to weight loss when intermittent fasting
(a reputable newspaper, 2021)
• Climate change sceptics are more likely to be from a certain
demographic (gender and/or political views). (a reputable popular
science magazine, 2016)
• One gender is better than another at detecting infidelity from facial
expressions. (a reputable newspaper, 2012)
34FIT1006 Business Information Analysis
… some claims in media from not too long ago
Drinking soft drink linked to cancer risk
By Sarah Wiedersehn
Updated February 22, 2018 — 10.54am first published at 10.42am
https://www.SMH.com.au/lifestyle/health-and-wellness/drinking-soft-drink-
linked-to-cancer-risk-20180222-h0wgvb.html
Accessed in February/March 2024
FIT1006 Business Information Analysis 35
Soft Drink
Quoting from https://www.SMH.com.au/lifestyle/health-and-wellness/drinking-soft-
drink-linked-to-cancer-risk-20180222-h0wgvb.html (Accessed in February/March
2024),
``… the study found a positive association between soft drink consumption and
cancer risk independent of obesity after statistically adjusting for waist
circumference …’’
``… "Initially our hypothesis was that drinking soft drinks would cause obesity
which would then cause an association with obesity-related cancers but we
found that there was more beyond the affect of obesity," said lead researcher,
Associate Professor Allison Hodge of Cancer Council Victoria's Cancer
Epidemiology and Intelligence Division. …’’
`` … According to the research, the more sugary soft drinks participants drank
the higher their risk of cancer.
However, this was not the case with those who drank diet soft drinks,
suggesting sugar could be the key, says Professor Hodge. …’’
FIT1006 Business Information Analysis 37
Quantitative claims using data
For any quantitative argument based on data, you should get in the
habit of thinking about the following three considerations:
1. Data: how was it collected, how much have you got?
2. Model: are you using the correct model for the argument?
(average, linear equation, proportion, time series or forecast,
or some combination thereof, …?)
Do we want an explainable model?
What about a deep learning model (or ``model’’)?
3. Randomness: how does variability in the data affect the
reliability of the conclusion?
38FIT1006 Business Information Analysis
What to do this week (today)
What to do this week (today) if you have not yet done so
• Log in to Moodle and check that you have access to the
FIT1006 page.
• Check your Applied session (and lab) time using Allocate+.
• Obtain a textbook if required
• And see elsewhere for the publisher’s offer of a
discount
• Have a look at FIT1006 Ed Discussion, the
https://learning.monash.edu/course/view.php?id=14294&se
ction=58 Expectations for Forum Communication, and the
``Welcome to FIT1006’’ post – and the links from there and
within Moodle
39FIT1006 Business Information Analysis
Seek assistance as a preventive measure
Take the following relevant preventive measures
as soon as possible, if you are falling behind in
your studies:
Study difficulties: Discuss any difficulties you are experiencing with your
course leader, unit coordinator, lecturer or tutor.
– These staff members can assist you in identifying your problem areas and explore
the options available to you in your course.
Student Academic Success can help you with study methods, work
presentation and a wide range of matters
https://www.monash.edu/student-academic-success
Student (life and) support services can be found at:
http://monash.edu/students/support/
and include: Health services, support and services, clubs and sports, etc.
See also links from Moodle and ``Welcome to FIT1006’’ Ed Discussion post
40FIT1006 Business Information Analysis
Disability Support Services
Do you have a disability, medical or mental health condition
that may impact on your study?
Disability Support Services provides a range of services for registered
students including:
Note takers and Auslan interpreters
Readings in alternative formats
Adaptive equipment and software
Alternative arrangements for exam and class tests
Disability Support Services also support students who are
carers of a person with a disability, medical or mental health
condition, or who is aged and frail.
For further information and details about how to register:
T: 03 9905 5704
E: disabilitysupportservices@monash.edu
https://www.monash.edu/students/support/disability
FIT1006 Business Information Analysis 41
A short break
There might possibly be a short break around about now
FIT1006 Business Information Analysis 46
Information Technology
FIT1006
Business Information Analysis
Sampling and Surveying
Topics covered:
Populations and samples
Collecting data
Sources of data
Designing surveys
Survey errors
Choosing a sample
48
Motivating Problem
A major fictitious hypothetical bank currently offers
Private Banking services to its clients in the Central
Business District (CBD). The bank is investigating the
possibility of extending its Private Banking business to
include several major regional towns including
Geelong, Ballarat, Hamilton and Bendigo.
The bank wants to survey potential clients in these
districts to determine whether likely demand would
make this business viable.
Suggest a sampling/survey design for the bank to use,
outlining the issues you need to consider and any
problems you anticipate.
49
10 (Possible) Survey questions
These questions could all be addressed using survey data:
1. Who is/are settling into university better, country or city students?
2. Are students happy with the quality of food in the Campus Centre?
3. What is/are the average savings of first year students?
4. Do savings differ between students of different faculties?
5. What is the average age of glider pilots?
6. Are OZ Lotto Division 1 winners still happy one year after their
win?
7. Where do the majority of visitors to my website click through from
– DuckDuckGo? Google? Yahoo? Or …?
8. Is Coles cheaper than Woolworths?
9. What proportion of Australians support immigration?
10. Would the Liberal Party win a (Federal/State) election held today?
50
Choosing a Sample
Probability Designs
- Random Sampling - Systematic Sampling
- Stratified Sampling - Cluster Sampling
Non-Probability Designs
- Convenience Sampling - Judgemental Sampling
- Quota Sampling - Snowball Sampling
Non-probability designs are more prone to bias.
51
How does this work? Probability designs
Screening questions
– Then people are thrown out …
FIT1006 Business Information Analysis 53
Populations and Samples
Population: whole collection of what we are
observing.
Sample: subset we actually observe.
We look at the sample to make an inference about
the population.
Population Parameter: the thing of interest.
Sample Statistic: what we actually measure.
We estimate the value of the population parameter
from the sample statistic.
54
The General Picture
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Population (frame)
One Observation
Variable has value X or O
Sample
O
O
O
O O
O
O
55
Reasons for Using Samples
High cost of sampling a whole population.
Length of time required to sample whole
population.
May have too much data.
May not have access to whole population.
Sampling can be destructive
– Testing seat belts for breaking strain.
A census is a survey of a whole population.
– The Australian population census is conducted once
every five years. Cost was perhaps circa $500 million!
– https://en.wikipedia.org/wiki/Census_in_Australia
(accessed Feb/Mar 2024) says 2006 census cost ~
$300ish million and 2011 census cost ~ $440ish million
56
Question 1
I want to know whether Monash University students support
a Rail link from city. Which sample is least biased?
A. E-mail all students and record the first 100 responses.
B. Randomly choose 100 students in the Campus Centre
(formerly called Union Building) at lunch time.
C. Randomly choose a subject (say MED1000) and
randomly choose 100 students from that class.
D. Choose 100 students at random from the University’s
register and e-mail them.
57
Examples of surveys
Television ratings (a purpose – help advertisers)
Market research (a purpose – sell more products)
Pre-election polls (a purpose – help political parties work out how to
win or at least how best to focus their efforts)
Product registration information (a purpose, track your personal data)
Product Testing (a purpose – how to design them better or make them
more sellable)
58
Question 2
I want to know whether people with early-onset
dementia have had trouble managing their
personal finances. The best survey method is:
A. Face-to-face interview.
B. Telephone interview.
C. A postal survey.
D. An Internet survey.
Would ethics play a part in choosing the method?
60
Types of Surveys
There are roughly three main types of survey methods:
– Personal (face to face) Interview
– Telephone Survey
– Self-administered survey (postal/Internet)
Two main issues are at stake:
– How to encourage people to respond truthfully and accurately?
– How to conduct the survey with the minimum cost?
61
Stages of a Survey
• Exploratory interview with focus group
– Identify issues, form hypothesis and questions
• Questionnaire designed and tested
• Experimental design and population frame is determined
• Samples are selected
• Questionnaire is administered
• Analysis and reporting
62
Question 3
Sampling 8 (one at a time, they don’t have to be
neighbours) at random. Which sample could not
have come from the population (on the RHS)?
A.
B.
C.
D.
E. All could have.
64
Question 4
Which sample is more likely to have come from the
population (on the right-hand side), with no
requirement about neighbours?
A.
B.
C.
D.
E. All equally likely.
65
Question 5
Which sample is more likely to have come from the
population (on the right-hand side), with no
requirement about neighbours?
A.
B.
C.
D.
E. All equally likely.
67
Survey Errors
Sampling Errors
– The characteristics of the sample do not match those of the
population.
– This is usually addressed by taking a larger sample.
Non-sampling Errors
– Errors in response or in recording data
– Misclassification or inaccurate response
– Bias in the selection of the sample
– Non-response bias, Self-selection of respondents.
68
Survey Errors – Survey Design
The survey questions are also important.
Are they ambiguous or easy to read and understand?
Is the survey too long or onerous to complete (if so, people might
drop out)?
Are the categories designed appropriately, or do they miss out
key cases?
Are the questions leading – e.g., `please tell me how much you
prefer Volvo to Ford’?
Human factors are a big issue in survey design, if you want high
quality data.
69
Motivating Problem (from before)
A major fictitious hypothetical bank currently offers
Private Banking services to its clients in the Central
Business District (CBD). The bank is investigating the
possibility of extending its Private Banking business to
include several major regional towns including
Geelong, Ballarat, Hamilton and Bendigo.
The bank wants to survey potential clients in these
districts to determine whether likely demand would
make this business viable.
Suggest a sampling/survey design for the bank to use,
outlining the issues you need to consider and any
problems you anticipate.
70
Key Ideas
You should be familiar with the following:
• Parameter, Sample Statistic, Variable, Observation;
• The reasons for sampling;
• The 3 main methods of surveying (advantages / disadvantages of
each);
• Causes of error in surveys;
• Methods of choosing samples (4 random / 4 non-random
methods).
71
Key Ideas (cont’d)
Main points:
• Think of sampling as the first stage in the process of
statistical analysis - so errors here undermine any
conclusions we can draw from subsequent analysis.
• Relevant issues are: sampling method, survey design,
potential bias in sample selection.
• In later lectures we look at how sample size affects the
validity of conclusions we draw.
72
Key Ideas (cont’d)
Think about:
• How we sample,
• Why we sample,
• How we conduct a survey and collect the results,
• Problems or considerations that emerge depending
on our choice of sample/survey method.
73
Purportedly from http://www.theaustralian.com.au/national-affairs/newspoll/vic
Purportedly: Australian – News poll (Feb 2021)
75
Question 6 (and we don’t need the poll graphic
from the earlier slide for this)
Based on a poll result of 49% ALP and 51% Coalition
with a sample size of 1,205 we can conclude that:
A. 95% sure ALP would have won an election held over
the sampling period.
B. 95% sure Coalition would have won the election.
C. Results are too close to call.
D. Sample size is too small.
E. Don’t know.
77
We intend to re-visit this
question later in our subject.
It will involve however much
mathematics.
Expert attacks ‘healthy’ food tick
Excerpt:
THE Heart Foundation's tick program gives a ''stamp of
credibility'' to unhealthy foods such as frozen pies and pizzas
and distracts from the key message to avoid processed
foods, a prominent nutritionist says
“ … Dr Rosemary Stanton, in an article published online in
The Medical Journal of Australia, said food manufacturers
paid to use the tick for products that might be less healthy
than alternatives. …’’
…
Design a survey to test whether ticked items are more
expensive than their non-ticked counterparts.
http://www.TheAge.com.au/national/expert-attacks-healthy-food-tick-20110228-1bbqy.html
Kate Hagan March 1, 2011 — 3.00am (Accessed February/March 2024)
79
Reading
General
• Selvanathan, Chapter 1.
Sampling
• Selvanathan, Chapter 2.
• McClennan, W., An Introduction to Sample Surveys, Australian
Bureau of Statistics, 1999.Cat 1299.0
http://abs.gov.au/ausstats/abs@.nsf/mf/1299.0
• Mission Australia Youth Survey 2015
https://www.MissionAustralia.com.au/what-we-do/research-evaluation/youth-survey
80
Week 1 Learning Outcomes
• Week 1: Intro + Surveys and Data Collection
After completing the work in this week and attending the applied
session in Week 2, you should be able to:
Understand the purpose of statistics and why it is beneficial to study
them.
Understand the concept of sampling and relevant terminology
associated with this (e.g., sampling, population, observation, variable,
etc.)
Be able to explain how to design a survey sampling strategy that is
appropriate with respect to a given scenario.
Understand how to conduct a survey and the types of errors that may
be made when designing a survey.
81FIT1006 Business Information Analysis
Questions – for (Week 2) Applied Session 1
General
Selvanathan: Questions 1.1 – 1.7
(For thinking about ...)
Sampling
Selvanathan:
• 2.18, 2.21, 2.22, 2.24, 2.25, 2.26.
Tutorial 1 Questions.
82
Supplementary exercise(s)
Look for examples of surveys and sampling in the media
– Feel free to share nice examples on Ed Discussion
Week 2 plan: Data Representation and Presentation
What is Florence Nightingale (1820-1910) (most) famous for?
What work, if any, did Florence Nightingale do on data representation
and presentation?
Semester 2, 2016Welcome to FIT 1006
Business Information Analysis
Semester 1, 2024
FIT1006 Business Information Analysis
Teaching Team:
Lecturer and Chief Examiner: A./Prof. David Dowe
Tutors (or Teaching Assistants, TAs) : Dr Heshan Kumarage (head
tutor or admin’ tutor), Anthony Wong, Dulan Perera, Emily Yap Sin-Yee,
Flora Jin
Please use your Monash account for all e-mails to lecturer
and tutors. E-mail addresses should be clear from
Moodle. As per the “Welcome to FIT1006” post in Ed
Discussion, it will typically be better to send a post to Ed
Discussion, mindful of
https://learning.monash.edu/course/view.php?id=14294&s
ection=58 Expectations for Forum Communication.
If e-mailing, then the first 8 characters of the e-mail subject
line should be `FIT1006 ‘ (FIT1006 followed by a space).
2FIT1006 Business Information Analysis
Lectures
Lectures
currently recorded unless we are subsequently notified
otherwise
Most lectures will be best if done with pre-reading, which will be
posted on Moodle
Lecture slides will be available after the lecture on Moodle as
PDF.
Lectures recordings available on Moodle.
Some topics will run across multiple lectures.
3FIT1006 Business Information Analysis
Applied Sessions (or Labs)
Enrol for tutorials on Allocate+ :
– https://my-timetable.monash.edu/odd/student
If you need help with your timetabling:
– https://connect.apps.monash.edu/students/timetables/allocat
e/help/
Applied sessions start in Week 2 (2-hour session per week)
Download the applied session sheet (or tutorial sheet) each week
Arrive prepared for each applied session. This means: revising
lectures + reading + practice questions.
Applied sessions will typically comprise: review of lecture, worked
examples, practice questions, manual & computer calculations.
+ 8 hours study/practice/research each week!
4FIT1006 Business Information Analysis
Consultation and Help
Consultations:
– Start from approx Week 3. Consultation times will be posted on
Moodle
Forum – should be your starting point: Ed Discussion Forum
– We can often reply fairly promptly from amongst the
teaching team.
• If someone hasn’t answered already.
Applied Sessions.
Support and help: See the “Welcome to FIT1006” post on Ed
Discussion and various Moodle pages for pointers to help
5FIT1006 Business Information Analysis
Resources
Subject Resources
– Download all lecture notes, applied session sheets, and
other materials from Moodle
Recordings – also on Moodle,
Assignment instructions – released on Moodle.
Support and help: Again, please see the “Welcome to FIT1006”
post on Ed Discussion and also various Moodle pages for
pointers to help
6FIT1006 Business Information Analysis
Textbook and References
Prescribed text:
Business Statistics Abridged, (7th Ed., 2017, or) 8th edition
Selvanathan et al., Cengage, South Melbourne, 2021.
(5th and 6th editions are purportedly quite similar in content)
For any of you wishing to buy a copy, see link within Moodle of
discount offer from publisher
Additional Reading
– Statistics Without Tears, Derek Rowntree, Penguin,
Harmondsworth, 1981.
– Statistics Explained: An introductory guide for life scientists,
Steve McKillup, Cambridge U.P., 2006.
Wikipedia – a handy reference for at least many topics.
Software: MicroSoft Excel and SYSTAT
7FIT1006 Business Information Analysis
Mathematics!
This unit has a mathematics prerequisite of:
A study score of 25 in VCE Mathematics Methods or
Specialist Maths units 3 & 4 or 30 in Further Maths units 3
& 4 or MTH1010 or equivalent.
If you have come to university via a non-VCE pathway, or
are just curious, you can see some sample questions on
Moodle.
Topics covered in this unit will include (e.g.) probability and
hypothesis testing, so please be mindful that at least some
of the mathematics might be – and/or become - challenging
8FIT1006 Business Information Analysis
Assessment
To pass the unit you must obtain:
– An overall unit mark of 50% or more.
9
Unit Assessment
Assessment Due Date % of Mark
Online quiz currently planned for Week 4 2
Data Collection currently planned for Week 5 18
Data Analysis
Report
currently planned for Week 9 40
In-class quiz TBC, currently planned in-
class in week 12
15
Take home test TBC, currently planned for
week 14
25
FIT1006 Business Information Analysis
Quiz 1 – 2%
- Quiz 1, 2%.
- To be done individually and as per given instructions.
- Intend to test on material covered up to the end of the
week before.
10FIT1006 Business Information Analysis
Assignment 1: Data Collection - 18%
- To be done individually and as per given instructions
- Intend to test on material covered up to the end of the
week before it is due
- We’ll go through this in more detail later.
- For students who have work in progress that they’d like to
discuss, we hope to have time for students to have 1-to-1
discussions in Applied sessions or consultation sessions
11FIT1006 Business Information Analysis
Assignment 2: Data Analysis - 40%
- To be done individually and as per given instructions
- Intend to test on material covered up to the end of the
week before it is due
- Write a report following the assignment instructions. We
anticipate that it will involve at least some of:
- Data visualisation
- Hypothesis testing
- Reflective commentary.
- We plan to go through this in more detail before it is due.
- We hope to give everyone some time to talk with a tutor
about at least part of an assignment draft
12FIT1006 Business Information Analysis
In-Class Test - 15%
- To be done individually and as per given instructions
- Intend to test on material covered up to the end of the
week before.
- Make sure to arrive punctually at your allocated applied
session and to follow instructions.
13FIT1006 Business Information Analysis
Take-Home Test - 25%
- To be done individually and as per given instructions
- We’ll explain in more detail closer to the time.
- Intend to test on material covered up to the end of the
week before it is due (i.e., the entire syllabus)
- There will possibly be some selective use of generative AI
permitted, but stay tuned for further instructions
14FIT1006 Business Information Analysis
General Observations
- Make sure you follow the course and are learning as you
go along!
- We are here to help you with this course and want
you to learn and do well.
- Try to avoid the crunch:
- Start assignments as early as possible.
- We intend to give you at least 10 calendar days in each case
(and perhaps more in the case of Assignment 2).
- Towards the end of the Semester, units tend to all pile up!
Trying to do these things at the last minute is usually not a good idea.
- Whether or not you might possibly get extensions, please do not
rely on this happening
Support and help: See the “Welcome to FIT1006” post on Ed
Discussion and various Moodle pages for pointers to help
15FIT1006 Business Information Analysis
Unit Learning Outcomes
16FIT1006 Business Information Analysis
01
Perform and justify the use of
appropriate descriptive and inferential
statistical analyses such as standard
deviation, t-tests, and Pearson's
correlation coefficients;
02 Communicate the results of statistical analysis effectively and accurately;
03
Create graphics and tables for effective
communication of the results of data
analysis;
04 Analyse the factors that make statistics reliable, credible and trustworthy.
Brief outline – First 3 or 6 Weeks or so
• Week 1: Intro + Surveys and Data Collection
• Week 2: Data Representation and Presentation
• Summary stats and some charts and plots.
• Week 3: Intro to statistical software and report writing.
• This is useful for Assignments 1 and 2.
• We then come to the following topics:
• Probability
• Correlation and Regression
• Hypothesis testing
• ANOVA (Analysis Of Variance), Classifier Performance, Fair AI
• We also look at aspects including estimation and time-series data.
17FIT1006 Business Information Analysis
Week 1 Learning Outcomes
• Week 1: Intro + Surveys and Data Collection
After completing the work in this week and attending the applied
session in Week 2, you should be able to:
Understand the purpose of statistics and why it is beneficial to study
them.
Understand the concept of sampling and relevant terminology
associated with this (e.g., sampling, population, observation, variable,
etc.)
Be able to explain how to design a survey sampling strategy that is
appropriate with respect to a given scenario.
Understand how to conduct a survey and the types of errors that may
be made when designing a survey.
18FIT1006 Business Information Analysis
About FIT1006…
Quantitative analysis
Some real-world issues where quantitative analysis may lead to
improved outcomes:
• Health: patterns of disease, effectiveness of treatments, the
inequality of nations;
• Civilization: the science of food production, the risk of
catastrophe – the global financial crisis (GFC), climate
change, global warming, social disadvantage;
• Commerce: the volatility of investments, the risk of
borrowers defaulting on loans, …;
• Sport: player statistics and rankings, is something a good
bet?, ...;
• etc.
19FIT1006 Business Information Analysis
About FIT1006… Quantitative analysis
In IT as well as Computing, we need this too!
– Fair AI – does my system unfairly (or unlawfully)
discriminate based on protected attributes?
– A/B testing – I’ve made a change to my system, does it
have the effect I want? (used by at least many of, e.g.,
Twitter, Google, Microsoft, etc. on a day-to-day basis).
– Probability – underpins many algorithms (e.g., in Networks).
– Writing reports – making a persuasive case to sell software,
or to make decisions on what to buy within an organization.
As citizens, being statistically literate is really important too. For
example, it can help you make life decisions (should I rent or get
a mortgage?), understand politics (especially polling) and
interpret media articles (many can mislead with numbers, …).
20FIT1006 Business Information Analysis
Quantitative analysis
Quantitative analysis shows us the big picture.
Using data (usually numerical) to:
• Generalise: what has happened now and in the past (and
may happen in the future), what is typical behaviour?
• cf. N. Goodman’s ``grue’’ paradox
• Normalise: what is normal? What is exceptional?
• Contextualise: how do we compare with others around us?
How are things changing over time?
21FIT1006 Business Information Analysis
Quantitative analysis
Quantitative analysis for business:
• Customer behaviour, randomness, risk;
• Demand patterns, trends – change over time;
• Determining public sentiment;
• Cause and effect relationships;
• Identifying the systematic from the ad hoc, …
• A good example of a company using quantitative analysis to
substantial effect is Google. Their page rank algorithm has
substantially changed the world. Search Engine
Optimisation (SEO) was an industry relatively unknown
approximately 15 or so years ago
22FIT1006 Business Information Analysis
Overview
Quantitative techniques include:
• Descriptive techniques applicable to a wide range of data –
sample, financial, Internet, etc.;
• Conducting surveys and analysing the results;
• Modelling relationships and trends;
• Forecasting based on historical data;
• Working with large(ish) data sets;
• Presenting results as summaries and reports;
• Machine learning;
• Deep learning;
• Testing hypotheses;
• etc.
23FIT1006 Business Information Analysis
FIT1006: The big picture
24
Describing
Data
Probability
Modelling
relationships
Reporting
results
Application of theory
to practical problems
Quantifying
(Un)certainty
FIT1006 Business Information Analysis
FIT1006: A bigger picture
25
FIT3158
Business
decision
modelsFIT1006
Business
information
analysis
FIT3152
Data analytics
The wide world!
Big Data, Analytics, ...
FIT1006 Business Information Analysis
Descriptive statistics
• A statistic is a summary of data. For example, the number of
students enrolled in this subject is a statistic. Statistics are a
way of summarizing the essential features of a large quantity of
data.
• Statistics gets its name from the collection of data about the
State.
• Data is a plural noun. Datum is the singular form.
• Descriptive statistics describe the features of a data set; for
example, the mean and the standard deviation.
26FIT1006 Business Information Analysis
Describing data (a)
• How was the data sample obtained; could this introduce doubt into –
or increase the amount of doubt about - our conclusions?
• Representing the data graphically; calculating summary statistics:
e.g., average height.
27FIT1006 Business Information Analysis
Describing data (b)
• What are the major differences between the two groups as suggested
by the summary statistics?
• Can we represent this graphically?
28FIT1006 Business Information Analysis
Statistical reports
• Obtaining data sometimes requires obtaining ethics clearance (and
filling in forms, etc.)
• Having responsibly obtained our data and accurately calculated
statistics, perhaps the next most important skill is the interpretation
and communication of the results.
• Reports should discuss the significance and consequences of findings
(and some sort of measure of uncertainty) as well as any assumptions
made in analysis.
• Statistical summaries should be presented in an easy to read form -
such as a table.
• Your report should ideally be understandable to a person without a
detailed statistical knowledge.
29FIT1006 Business Information Analysis
Probability
• What is the chance that a certain person selected at random will
exhibit a certain property?
• Does that chance change if the person belongs to a subgroup of the
population?
30FIT1006 Business Information Analysis
Inferential statistics
• Having observed some differences between groups of data we may
wish to know whether these differences are due to some systematic
cause or just due to random fluctuations.
• This is an important part of product testing, medical treatments (e.g.,
vaccinations), etc.
• We generally use descriptive statistics to observe differences between
groups and then use inferential statistics to test whether the difference
is significant.
31FIT1006 Business Information Analysis
Being certain - or, at least, quantifying this
• Are the two groups different?
• What measure are we comparing?
• What level of confidence and/or certainty?
• Are you 99% sure – or only 50.1% sure, or …?
32FIT1006 Business Information Analysis
Why is this unit important?
We are constantly being told certain things are true. This subject gives
you the tools to track through the evidence, apply the theory, verify for
yourself, and endeavour to quantify (un)certainty.
For example, from the past and recent press:
• New Vaccine Prevents Severe COVID As Well As Existing Vaccines
Do, Experts Say. (a reputable popular science magazine, 2021)
• Belly fat may be resistant to weight loss when intermittent fasting
(a reputable newspaper, 2021)
• Climate change sceptics are more likely to be from a certain
demographic (gender and/or political views). (a reputable popular
science magazine, 2016)
• One gender is better than another at detecting infidelity from facial
expressions. (a reputable newspaper, 2012)
34FIT1006 Business Information Analysis
… some claims in media from not too long ago
Drinking soft drink linked to cancer risk
By Sarah Wiedersehn
Updated February 22, 2018 — 10.54am first published at 10.42am
https://www.SMH.com.au/lifestyle/health-and-wellness/drinking-soft-drink-
linked-to-cancer-risk-20180222-h0wgvb.html
Accessed in February/March 2024
FIT1006 Business Information Analysis 35
Soft Drink
Quoting from https://www.SMH.com.au/lifestyle/health-and-wellness/drinking-soft-
drink-linked-to-cancer-risk-20180222-h0wgvb.html (Accessed in February/March
2024),
``… the study found a positive association between soft drink consumption and
cancer risk independent of obesity after statistically adjusting for waist
circumference …’’
``… "Initially our hypothesis was that drinking soft drinks would cause obesity
which would then cause an association with obesity-related cancers but we
found that there was more beyond the affect of obesity," said lead researcher,
Associate Professor Allison Hodge of Cancer Council Victoria's Cancer
Epidemiology and Intelligence Division. …’’
`` … According to the research, the more sugary soft drinks participants drank
the higher their risk of cancer.
However, this was not the case with those who drank diet soft drinks,
suggesting sugar could be the key, says Professor Hodge. …’’
FIT1006 Business Information Analysis 37
Quantitative claims using data
For any quantitative argument based on data, you should get in the
habit of thinking about the following three considerations:
1. Data: how was it collected, how much have you got?
2. Model: are you using the correct model for the argument?
(average, linear equation, proportion, time series or forecast,
or some combination thereof, …?)
Do we want an explainable model?
What about a deep learning model (or ``model’’)?
3. Randomness: how does variability in the data affect the
reliability of the conclusion?
38FIT1006 Business Information Analysis
What to do this week (today)
What to do this week (today) if you have not yet done so
• Log in to Moodle and check that you have access to the
FIT1006 page.
• Check your Applied session (and lab) time using Allocate+.
• Obtain a textbook if required
• And see elsewhere for the publisher’s offer of a
discount
• Have a look at FIT1006 Ed Discussion, the
https://learning.monash.edu/course/view.php?id=14294&se
ction=58 Expectations for Forum Communication, and the
``Welcome to FIT1006’’ post – and the links from there and
within Moodle
39FIT1006 Business Information Analysis
Seek assistance as a preventive measure
Take the following relevant preventive measures
as soon as possible, if you are falling behind in
your studies:
Study difficulties: Discuss any difficulties you are experiencing with your
course leader, unit coordinator, lecturer or tutor.
– These staff members can assist you in identifying your problem areas and explore
the options available to you in your course.
Student Academic Success can help you with study methods, work
presentation and a wide range of matters
https://www.monash.edu/student-academic-success
Student (life and) support services can be found at:
http://monash.edu/students/support/
and include: Health services, support and services, clubs and sports, etc.
See also links from Moodle and ``Welcome to FIT1006’’ Ed Discussion post
40FIT1006 Business Information Analysis
Disability Support Services
Do you have a disability, medical or mental health condition
that may impact on your study?
Disability Support Services provides a range of services for registered
students including:
Note takers and Auslan interpreters
Readings in alternative formats
Adaptive equipment and software
Alternative arrangements for exam and class tests
Disability Support Services also support students who are
carers of a person with a disability, medical or mental health
condition, or who is aged and frail.
For further information and details about how to register:
T: 03 9905 5704
E: disabilitysupportservices@monash.edu
https://www.monash.edu/students/support/disability
FIT1006 Business Information Analysis 41
A short break
There might possibly be a short break around about now
FIT1006 Business Information Analysis 46
Information Technology
FIT1006
Business Information Analysis
Sampling and Surveying
Topics covered:
Populations and samples
Collecting data
Sources of data
Designing surveys
Survey errors
Choosing a sample
48
Motivating Problem
A major fictitious hypothetical bank currently offers
Private Banking services to its clients in the Central
Business District (CBD). The bank is investigating the
possibility of extending its Private Banking business to
include several major regional towns including
Geelong, Ballarat, Hamilton and Bendigo.
The bank wants to survey potential clients in these
districts to determine whether likely demand would
make this business viable.
Suggest a sampling/survey design for the bank to use,
outlining the issues you need to consider and any
problems you anticipate.
49
10 (Possible) Survey questions
These questions could all be addressed using survey data:
1. Who is/are settling into university better, country or city students?
2. Are students happy with the quality of food in the Campus Centre?
3. What is/are the average savings of first year students?
4. Do savings differ between students of different faculties?
5. What is the average age of glider pilots?
6. Are OZ Lotto Division 1 winners still happy one year after their
win?
7. Where do the majority of visitors to my website click through from
– DuckDuckGo? Google? Yahoo? Or …?
8. Is Coles cheaper than Woolworths?
9. What proportion of Australians support immigration?
10. Would the Liberal Party win a (Federal/State) election held today?
50
Choosing a Sample
Probability Designs
- Random Sampling - Systematic Sampling
- Stratified Sampling - Cluster Sampling
Non-Probability Designs
- Convenience Sampling - Judgemental Sampling
- Quota Sampling - Snowball Sampling
Non-probability designs are more prone to bias.
51
How does this work? Probability designs
Screening questions
– Then people are thrown out …
FIT1006 Business Information Analysis 53
Populations and Samples
Population: whole collection of what we are
observing.
Sample: subset we actually observe.
We look at the sample to make an inference about
the population.
Population Parameter: the thing of interest.
Sample Statistic: what we actually measure.
We estimate the value of the population parameter
from the sample statistic.
54
The General Picture
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Population (frame)
One Observation
Variable has value X or O
Sample
O
O
O
O O
O
O
55
Reasons for Using Samples
High cost of sampling a whole population.
Length of time required to sample whole
population.
May have too much data.
May not have access to whole population.
Sampling can be destructive
– Testing seat belts for breaking strain.
A census is a survey of a whole population.
– The Australian population census is conducted once
every five years. Cost was perhaps circa $500 million!
– https://en.wikipedia.org/wiki/Census_in_Australia
(accessed Feb/Mar 2024) says 2006 census cost ~
$300ish million and 2011 census cost ~ $440ish million
56
Question 1
I want to know whether Monash University students support
a Rail link from city. Which sample is least biased?
A. E-mail all students and record the first 100 responses.
B. Randomly choose 100 students in the Campus Centre
(formerly called Union Building) at lunch time.
C. Randomly choose a subject (say MED1000) and
randomly choose 100 students from that class.
D. Choose 100 students at random from the University’s
register and e-mail them.
57
Examples of surveys
Television ratings (a purpose – help advertisers)
Market research (a purpose – sell more products)
Pre-election polls (a purpose – help political parties work out how to
win or at least how best to focus their efforts)
Product registration information (a purpose, track your personal data)
Product Testing (a purpose – how to design them better or make them
more sellable)
58
Question 2
I want to know whether people with early-onset
dementia have had trouble managing their
personal finances. The best survey method is:
A. Face-to-face interview.
B. Telephone interview.
C. A postal survey.
D. An Internet survey.
Would ethics play a part in choosing the method?
60
Types of Surveys
There are roughly three main types of survey methods:
– Personal (face to face) Interview
– Telephone Survey
– Self-administered survey (postal/Internet)
Two main issues are at stake:
– How to encourage people to respond truthfully and accurately?
– How to conduct the survey with the minimum cost?
61
Stages of a Survey
• Exploratory interview with focus group
– Identify issues, form hypothesis and questions
• Questionnaire designed and tested
• Experimental design and population frame is determined
• Samples are selected
• Questionnaire is administered
• Analysis and reporting
62
Question 3
Sampling 8 (one at a time, they don’t have to be
neighbours) at random. Which sample could not
have come from the population (on the RHS)?
A.
B.
C.
D.
E. All could have.
64
Question 4
Which sample is more likely to have come from the
population (on the right-hand side), with no
requirement about neighbours?
A.
B.
C.
D.
E. All equally likely.
65
Question 5
Which sample is more likely to have come from the
population (on the right-hand side), with no
requirement about neighbours?
A.
B.
C.
D.
E. All equally likely.
67
Survey Errors
Sampling Errors
– The characteristics of the sample do not match those of the
population.
– This is usually addressed by taking a larger sample.
Non-sampling Errors
– Errors in response or in recording data
– Misclassification or inaccurate response
– Bias in the selection of the sample
– Non-response bias, Self-selection of respondents.
68
Survey Errors – Survey Design
The survey questions are also important.
Are they ambiguous or easy to read and understand?
Is the survey too long or onerous to complete (if so, people might
drop out)?
Are the categories designed appropriately, or do they miss out
key cases?
Are the questions leading – e.g., `please tell me how much you
prefer Volvo to Ford’?
Human factors are a big issue in survey design, if you want high
quality data.
69
Motivating Problem (from before)
A major fictitious hypothetical bank currently offers
Private Banking services to its clients in the Central
Business District (CBD). The bank is investigating the
possibility of extending its Private Banking business to
include several major regional towns including
Geelong, Ballarat, Hamilton and Bendigo.
The bank wants to survey potential clients in these
districts to determine whether likely demand would
make this business viable.
Suggest a sampling/survey design for the bank to use,
outlining the issues you need to consider and any
problems you anticipate.
70
Key Ideas
You should be familiar with the following:
• Parameter, Sample Statistic, Variable, Observation;
• The reasons for sampling;
• The 3 main methods of surveying (advantages / disadvantages of
each);
• Causes of error in surveys;
• Methods of choosing samples (4 random / 4 non-random
methods).
71
Key Ideas (cont’d)
Main points:
• Think of sampling as the first stage in the process of
statistical analysis - so errors here undermine any
conclusions we can draw from subsequent analysis.
• Relevant issues are: sampling method, survey design,
potential bias in sample selection.
• In later lectures we look at how sample size affects the
validity of conclusions we draw.
72
Key Ideas (cont’d)
Think about:
• How we sample,
• Why we sample,
• How we conduct a survey and collect the results,
• Problems or considerations that emerge depending
on our choice of sample/survey method.
73
Purportedly from http://www.theaustralian.com.au/national-affairs/newspoll/vic
Purportedly: Australian – News poll (Feb 2021)
75
Question 6 (and we don’t need the poll graphic
from the earlier slide for this)
Based on a poll result of 49% ALP and 51% Coalition
with a sample size of 1,205 we can conclude that:
A. 95% sure ALP would have won an election held over
the sampling period.
B. 95% sure Coalition would have won the election.
C. Results are too close to call.
D. Sample size is too small.
E. Don’t know.
77
We intend to re-visit this
question later in our subject.
It will involve however much
mathematics.
Expert attacks ‘healthy’ food tick
Excerpt:
THE Heart Foundation's tick program gives a ''stamp of
credibility'' to unhealthy foods such as frozen pies and pizzas
and distracts from the key message to avoid processed
foods, a prominent nutritionist says
“ … Dr Rosemary Stanton, in an article published online in
The Medical Journal of Australia, said food manufacturers
paid to use the tick for products that might be less healthy
than alternatives. …’’
…
Design a survey to test whether ticked items are more
expensive than their non-ticked counterparts.
http://www.TheAge.com.au/national/expert-attacks-healthy-food-tick-20110228-1bbqy.html
Kate Hagan March 1, 2011 — 3.00am (Accessed February/March 2024)
79
Reading
General
• Selvanathan, Chapter 1.
Sampling
• Selvanathan, Chapter 2.
• McClennan, W., An Introduction to Sample Surveys, Australian
Bureau of Statistics, 1999.Cat 1299.0
http://abs.gov.au/ausstats/abs@.nsf/mf/1299.0
• Mission Australia Youth Survey 2015
https://www.MissionAustralia.com.au/what-we-do/research-evaluation/youth-survey
80
Week 1 Learning Outcomes
• Week 1: Intro + Surveys and Data Collection
After completing the work in this week and attending the applied
session in Week 2, you should be able to:
Understand the purpose of statistics and why it is beneficial to study
them.
Understand the concept of sampling and relevant terminology
associated with this (e.g., sampling, population, observation, variable,
etc.)
Be able to explain how to design a survey sampling strategy that is
appropriate with respect to a given scenario.
Understand how to conduct a survey and the types of errors that may
be made when designing a survey.
81FIT1006 Business Information Analysis
Questions – for (Week 2) Applied Session 1
General
Selvanathan: Questions 1.1 – 1.7
(For thinking about ...)
Sampling
Selvanathan:
• 2.18, 2.21, 2.22, 2.24, 2.25, 2.26.
Tutorial 1 Questions.
82
Supplementary exercise(s)
Look for examples of surveys and sampling in the media
– Feel free to share nice examples on Ed Discussion
Week 2 plan: Data Representation and Presentation
What is Florence Nightingale (1820-1910) (most) famous for?
What work, if any, did Florence Nightingale do on data representation
and presentation?
