Cambridge International As & A Level Thinking Skills Sample

Mark Dawes, Ruth Matthews,
Andrew Roberts and Geoff Thwaites
Cambridge International AS & A Level
Thinking Skills
Coursebook
Third edition
Andrew Roberts, Mark Dawes,
Ruth Matthews, Geoff Thwaites
and John Butterworth
Cambridge International AS & A Level
Thinking Skills
Coursebook
Third edition
University Printing House, Cambridge CB2 8BS, United Kingdom
One Liberty Plaza, 20th Floor, New York, NY 10006, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India
79 Anson Road, #06–04/06, Singapore 079906
Cambridge University Press is part of the University of Cambridge.
It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781108441049 (Paperback)
© Cambridge University Press 2018
This publication is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2018
20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1
Printed in the XXXX by XXXX

A catalogue record for this publication is available from the British Library
ISBN 978-1-108-44104-9 Paperback
ISBN 978-1-108-44110-0 Cambridge Elevate Edition (2 years)
Cambridge University Press has no responsibility for the persistence or accuracy
of URLs for external or third-party internet websites referred to in this publication,
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate. Information regarding prices, travel timetables, and other
factual information given in this work is correct at the time of first printing but
Cambridge University Press does not guarantee the accuracy of such information
thereafter.
notice to teachers in the uk

It is illegal to reproduce any part of this work in material form (including
photocopying and electronic storage) except under the following circumstances:
(i) where you are abiding by a licence granted to your school or institution by the
Copyright Licensing Agency;
(ii) where no such licence exists, or where you wish to exceed the terms of a licence,
and you have gained the written permission of Cambridge University Press;
(iii) where you are allowed to reproduce without permission under the provisions
of Chapter 3 of the Copyright, Designs and Patents Act 1988, which covers, for
example, the reproduction of short passages within certain types of educational
anthology and reproduction for the purposes of setting examination questions.
Contents
Part 1 9 Summarised data 70
1 What do we mean by a ‘problem’? 2 9.1 Summarising data 71
1.1 Routine activities or problems 3 9.2 Working backwards: data 76
1.2 What skills are needed to solve problems? 3 9.3 Working backwards: numbers 79
1.3 Selecting a problem-solving strategy 6 Exam-style questions 82
End of chapter questions 10
Exam-style questions 10 10 Identifying features of a model 85
10.1 Different types of representation 86
2 Selecting and using information 11 10.2 Real graphs are messy 87
2.1 Understanding information in various forms 12 10.3 Features of a model: non-standing representations 88
2.2 Understanding what information is required 10.4 Adjusting a model 90
to solve the problem 16 Exam-style questions 92
Exam-style questions 18
11 Necessary and Sufficient Conditions 94
3 Processing data 21
11.1 Necessary and sufficient 95
3.1 Processing information to solve problems 22
11.2 ‘All’ 96
11.3 Necessity, sufficiency and problem solving 98
4 Working with models 27 11.4 Euler diagrams 99
4.1 What is a model? 28 11.5 Making estimates 100
4.2 Simple models 28 Exam-style questions 102
4.3 Other types of models 29
4.4 Complex models 31 12 Changing the scenario of a problem 104
3
Exam-style questions 33 12.1 Changing scenarios 105
12.2 What types of changes can be made? 105
5 Solving problems by searching 35 12.3 Changing quantities 105
5.1 When is searching an appropriate method? 36 12.4 Changing the timescale 106
5.2 Searching in a systematic way 37 12.5 Changing prices, including introducing
5.3 Selecting an appropriate search method 39 special offers 106
5.4 Identifying the most efficient solution 39 12.6 Changing the law 107
Exam-style questions 42 12.7 Other changes: extrapolation 108
6 Finding methods of solution 44 Exam-style questions 111
6.1 Intermediate stages in problem solving 45
6.2 Drawing conclusions from relationships between Part 2
data 45 13 An introduction to critical thinking 114
6.3 Making inferences about patterns in data 48 13.1 What makes some thinking critical, and
Exam-style questions 50 other thinking uncritical? 115
13.2 Critical thinking (and critical thinking) 115
7 Trends in data 52 13.3 The objects of critical thinking 115
7.1 Identifying trends in data 53 13.4 Activities 116
7.2 Fitting models based on observed patterns 54
13.5 Attitude 116
7.3 Deducing values for one or more parameters 55
13.6 Reasons and evidence 117
14 Claims, statements and assertions 120
8 Transforming data 59
14.1 Claims and statements 121
8.1 Alternative representations of data 60
14.2 Assertion 122
8.2 Relationships between features of an object 62
14.3 Fact, falsity and opinion 122
8.3 Identifying features from different types of
representation64 14.4 Varieties of claim 123
Exam-style questions 67 Exam-style question 129
Cambridge International AS & A Level Thinking Skills
15 Assessing claims 130 22 Interpretation 234

15.1 Separating truth from fiction 131 22.1 Elements of an argument 235
15.2 Justification 131 22.2 Dispersed conclusions 240
15.2 Levels and standards 133 22.3 The need for caution 243
15.3 How do we judge claims? 135 Exam-style questions 246
15.4 Complex claims 136
23 Assumptions 247
15.5 Strong and weak claims 137
23.1 The meaning of ‘assumption’ 248
16 Grounds, reasons and evidence 140
24 Evaluating argument 257
16.1 Reasons in the form of evidence 141
24.1 Sound and unsound arguments 258
16.2 Criteria for assessing evidence 148
24.2 Generalisation 260
Exam style questions 153
24.3 Describing flaws 262
17 Evaluating evidence: a case study 155 24.4 Recognising and explaining flaws 265
17.1 Article 1 156 24.5 Appeals 269
17.2 Article 2: Further investigation 161 Exam-style questions 272
25 Applying logic 274
18 Statistical evidence 168 25.1 Formal and informal logic 275
18.1 What are statistics? 169 25.2 Conditions and conditionals 278
18.2 Trends 173 25.3 Critical thinking and logic 284
18.3 Presentation 179 25.4 Deductive standards – and limitations 287
Exam-style question 185 Exam-style questions 288
4 19 Uses of evidence: inference and explanation 186 26 Longer texts 289

19.1 What is inference? 187 26.1 Introduction 290
19.2 An example from science 189 Exam-style questions 306
19.3 Explanation 191
27 Multiple sources 308
19.4 Inference and explanation, a case study 196
27.1 Introduction 309
27.2 The documents 309
20 Identifying argument 203 27.3 Activities 315
20.1 Recognising an argument 204 27.4 Constructing a reasoned argument 320
20.2 An example from history 205 Exam-style questions 322
20.3 Identifying arguments in dialogue 213
21 Analysing argument 218

21.1 Critical assessment 219
21.2 Mixed arguments 223
21.3 Complex arguments 224
21.4 More about context: targets and opposing views 229
5
Part 1
Problem solving
6
Chapter 1
What do we mean by a ‘problem’?
Learning objectives
This chapter introduces the idea of a problem and considers some simple examples
to help you identify the key information that is required to solve problems and apply
some simple problem-solving strategies.
Chapter 1: What do we mean by a ‘problem’?
1.1 Routine activities or problems

There are many things that we automatically do every day and would not consider to
be ‘problems’. For example, consider the action of making a cup of instant coffee – if you
analyse the processes you need to go through, they are quite complicated. Just the list of
items you need is quite long: a cup, a teaspoon, a jar of coffee, a kettle, water – and milk
and sugar if you take them. Having found all these items, you fill the kettle and boil it; use
the teaspoon to put coffee into the cup; pour the boiling water into the cup, just to the right
level; stir; add milk and sugar; then put all the things you used away again. We could, if
necessary, break this down even further, for example by explaining in detail what needs to be
done to boil the kettle.
Although this is complicated, it is an everyday task that you probably do without thinking
(assuming that you drink coffee that is, although the same principle applies for making tea
or preparing your breakfast). To continue the coffee example, if over the years you have
made many cups of coffee, you may have tried adding different quantities of coffee granules,
milk and sugar to work out the strength of coffee that you prefer. Once you have established
this, you simply make the coffee to your preferred specifications from that point onwards.
When you encounter something new, even a task which is no more complicated than others
that you have previously completed, the processes required to achieve the task may need
considerably more thought and planning. Experience of similar problems is often helpful,
however: when trying a new type of hot drink you may be able to more readily work out what
the best combination of ingredients is for you.
Problem-solving takes place when you are required to perform a task that you have not 7
previously completed. These tasks may be very similar in nature to ones that you have done
in the past or may be completely new. Experience of a wide range of different problems and
their solutions is therefore very useful in developing problem-solving skills.
ACTIVITY 1.1
Imagine you are going to book tickets for a concert. List the pieces of information you
need and the processes that you need to go through in order to book the tickets and get
to the concert. In what order should you do them? First list the main things, then try to
break each down into smaller parts.
1.2 What skills are needed to solve problems?

There are a range of different skills that can be required to solve problems, which can be
divided into the following categories:
1.2.1. Understanding the problem that needs to be solved

Information might be presented in a range of different forms, using tables or diagrams in
addition to text. Different pieces of information will be related to each other in many different
ways, which might be explained through a simple mathematical model.
Being able to identify the appropriate pieces of information, and understand how they are
related to each other, is necessary if the solution to a problem is going to be found.
1.2.2. Processing the information that is available to solve the problem

Reaching the solution to the problem may require some simple mathematics to be
performed, based on the relationships that exist between the different pieces of information.
There may also be certain criteria that need to be met – for example, a maximum weight that
can be transported over a bridge in one journey.
Processing the information allows deductions to be made which lead to the solution to the
problem.
1.2.3. Analysing data and representing it in different ways

Presenting information in a different way often allows different insights to be made about
the data that is available. Recognising patterns that exist in data can be very helpful for
solving problems.
When considering problems that involve objects that can be moved into different positions
and orientations, it can be very useful to be able to identify key features that confirm that the
shape is unchanged.
1.2.4. Developing a problem that has already been solved to consider

more complex situations that might occur
Many problems are solved by first considering much simpler versions of the problem. Once
simplified versions of problems have been solved, it is important to be able to look at how
that solution helps with the solutions to more complicated problems.
8 In addition to these problem-solving skills, there are a number of mathematical skills which
are needed – many tasks will require some simple arithmetic and, as mentioned above,
some familiarity with methods for representing data, such as bar charts and pie charts, is
very useful.
EXAMPLE ACTIVITY 1.1
Imagine that you have to work out the best way to fit a number of rectangular packages
into a box. Which of the four categories of skill identified above would be relevant to
solving the problem?
Commentary
This problem involves mainly categories 1 and 2, but could involve some of category 4
as well:
1 Understand the problem that needs to be solved:
The dimensions of the box and the packages need to be known. Once these have
been worked out you also need to think about the different orientations in which the
packages could be placed into the box. You might make an initial assumption about
the best orientation, which may turn out later to be wrong.
2 Process the information that is available:
Once you know the relevant dimensions, some appropriate calculations should
enable you to work out how many boxes can be fitted in for each orientation.
3 Consider more complex situations:

If we know how to solve the problem in situations where all of the packages are the
same shape and size then similar methods may help in situations where there is more
than one shape of package.
If you had to solve the problem, you would need to be systematic and have some sort of
strategy. In the case of problems such as this one it may also be difficult to be confident
that you have the best solution – what if the best solution that you have found leaves a
gap and the volume of the gap is larger than the volume of the box, how can you be sure
that a different arrangement would not have managed to pack an extra box?
With some problems the method of finding an answer might be quite clear. With others
there may be no systematic method and you might have to use trial and error from the
start. Some problems will require a combination of both methods or can be solved in
more than one way.
The example below is a simple problem; you can give either a simple answer or a more
complicated one, depending on the degree of detail you consider necessary.
Luke has a meeting, in a town that is 50 miles away, at 3 p.m. tomorrow. He is planning
to travel by train from the town where he lives to the town where the meeting is, walking
to and from the station at both ends. 9
List the pieces of information Luke needs in order to decide what time he must leave
home.
Commentary
Let us start by thinking of everything that Luke does from leaving his house to arriving
at the meeting:
1 He leaves his house.
2 He walks to the station.
3 He buys a train ticket.
4 He goes to the platform.
5 He boards the train when it arrives.
6 He sits on the train until it reaches its destination.
7 He leaves the train.
8 He walks to where his meeting is to be held.
We can construct the pieces of information that he needs from this list. They are:
1 The time taken to walk from his house to the station.
2 The time needed to buy a ticket (including the time waiting in a queue).
3 The time to walk to the platform.
4 The train timetable.
5 The time taken to walk from the station to where the meeting is being held.
Suppose that we are given the following additional information:

It takes Luke 15 minutes to walk to the station and will take him 20 minutes to walk to
the meeting. The trains leave the station every 30 minutes from 9 a.m. and train journey
takes a total of 50 minutes. Luke estimates that it will take him 10 minutes to buy his
ticket and walk to the platform.
What time should Luke leave home in order to get to his meeting on time?
Commentary
We could choose a start time based on when we think Luke should leave home and
then work out what time Luke would arrive, but a better strategy would be to work out
when Luke should leave by working backwards from the meeting start time.
The sections of the journey can be considered one at a time. The start time of one
activity is always the time by which the previous activity must have been completed by.
Section of journey Must be Time required Latest start

completed by time
Walk to meeting 3:00 p.m. 20 minutes 2:40 p.m.
Train 2:40 p.m. 50 minutes 1:30 p.m.
10
The latest train that arrives by 2:40 p.m. must leave no later than 1:50 p.m., and the
1:30 p.m. train is the latest such train.
Walk to station, buy ticket, 1:30 p.m. 25 minutes 1:05 p.m.
walk to platform
Of course, you could do the whole thing by guesswork, but you might get it all wrong
and, more to the point, you cannot be confident that you will have got it right.
1.3 Selecting a problem-solving strategy

In the sense we are using the word in this book, a ‘problem’ means a situation where we
need to find a solution from a set of initial conditions. In some cases, as in the example
above, the method for solving the problem is not difficult to work out once you have
identified the information that is important.
In order to solve problems we must use the information that we are given in a certain way.
The way in which we use it may be quite straightforward; it may for example be simply a
matter of searching a table for a piece of data that matches given conditions. In other cases,
instead of searching for a piece of data, we may have to search for a method of solution. The
important thing in either case will be to have a strategy that will lead to the solution.
Imagine you are going out and can’t find your house keys. Finding them is a problem in the
sense meant by this section of the book.
• One method (and sometimes the quickest) is to run around all the likely places to see if
they are there.
• After the likely places, you start looking at the less likely places, and so on until they turn
up or you have to resort to more systematic methods.
There are two systematic ways of searching.

• The first (using experience) involves thinking carefully about when you last came into the
house and what you did: this can be the quickest method.
• The other (which in mathematical terms is often known as the ‘brute force’ method) involves
searching every room of the house thoroughly until they are found. This is often the most
reliable method but can take a very long time and most people will use it as a last resort.
When people are solving problems similar to the one identified above, they may use any one
or more of these methods. One of the prime skills that you need is to make a good judgement
of which method is the most appropriate one to use in each particular situation.
In any problem, you will be presented with some initial pieces of information – these may
be in the form of words, a table of numbers, a graph or a picture. You will also know what
question it is that you need to answer. The first thing to do is to identify which pieces of
information are most likely to be useful in proceeding to the solution and to try to work out
how these pieces of information may be used. In many situations much more information
than is needed is available and identifying which of it is redundant (not needed for the
solution to the problem) is an important part of the problem-solving process.
The example below is relatively easy. It is not difficult to find a way of approaching the
problem, and the necessary calculations are clear and simple.
Julia has been staying in a hotel on a business trip. When she checks out, the hotel’s
computer isn’t working, so the receptionist makes a bill by hand from the receipts, totalling 11
$471. Julia thinks she has been overcharged, so she checks the itemised bill carefully.
Room: 4 nights at $76.00 per night
Breakfast: 4 at $10.00 each
Dinner: 3 at $18.00 each
Telephone: 10 units at $1.70 per unit
Bar: Juices and soft drinks totalling $23.00
Laundry: 3 blouses at $5.00 each
It appears that the receptionist miscounted one of the items when adding up the total.
Which item has Julia been charged too much for?
List the pieces of information Luke needs in order to decide what time he must leave home.
Commentary
Although this example is simple, it illustrates many of the methods used in solving
problems:
• Look at the data provided. Identify which pieces are relevant and which are
irrelevant. In this case all of the information is relevant to the solution as we will
need to know how much Julia should have been charged.
• Make an intermediate calculation before you can reach the answer. In this case
it was necessary to calculate the value that the bill should have been in order to
identify where Julia had been overcharged. The sum of the charges on the itemised
bill is $453, so Julia has been overcharged by $18.
• Search the data for a piece of information that helps to solve the problem. In this
case, we need to find an item on the bill that could cost $18 (either singly or for
some number of them). The only item on the itemised bill for which this is true is
dinner, so Julia must have been charged for one extra dinner.
This is an example of using a systematic procedure to solve the problem.
The example below, whilst still being relatively simple, involves a slightly different type
of problem where the method of solution is less obvious.
The SuperSave supermarket sells Sudsy washing up liquid for $1.20 a bottle. At this
price they are charging 50% more than the price at which they buy the item from the
manufacturers. Next week SuperSave is having a ‘Buy two, get a third free’ offer on
this item. The supermarket does not want to lose money on this offer, so it expects the
manufacturers to reduce their prices so SuperSave will make the same actual profit on
every three bottles sold.
By how much will the manufacturers have to reduce their prices?
Commentary
12
There are several pieces of information that we will need in order to solve this problem.
1 How much do SuperSave normally buy a bottle of Sudsy washing up liquid for?
To make a 50% profit on the sale of each bottle, SuperSave must pay 80¢ for each
bottle (two-thirds of the price for which they sell it).
2 How much profit do SuperSave normally make on the sale of three bottles of Sudsy
washing up liquid?
The profit is 40¢ for each bottle, so the profit is normally $1.20 on three bottles.
3 How much would SuperSave need to pay for three bottles in order to make this
profit under their new offer?
Under the offer, SuperSave would only be paid $2.40 for three bottles, so need to pay
just $1.20 for three bottles.
4 By how much do the manufacturers need to reduce the price?
The price needs to be reduced to 40¢ per bottle, which is a reduction of 40¢ per bottle
(or half of the normal price).
In fact, it is possible to deduce that the manufacturers would have to reduce the price
to half the normal price without knowing that SuperSave’s normal selling price is $1.20:
1 Since the amount that SuperSave must pay for the bottles is two-thirds of the price
for which they sell it, the profit that SuperSave make on the sale of three bottles
must be the price for which they sell one bottle. Additionally, SuperSave currently
buy three bottles for the same price at which they sell two.
2 Once the offer is in place, customers will only be paying for two bottles when they
buy three. To make the same profit on the sale of three bottles, SuperSave will
therefore need to have bought the three bottles for the price at which they sell one.
3 The price that SuperSave pay for bottles must therefore have reduced to half of its
previous value.
There are other ways in which the solution could be found. For example, it would be
possible to choose a value for the reduction, see whether it would give the required
result and then adjust it accordingly. The systematic approach is more efficient
however (unless you are lucky enough to choose the correct value on your first guess).
ACTIVITY 1.2
Match the key terms (1–7) with the correct definition (A–G).
Note: you will not be asked to give definitions in the problem-solving exam, but it will
help you develop your problem-solving skills if you understand these key terms.
Key terms Definitions
1 Data A Using logic, mathematics or a systematic process in order
2 Problem-solving to work out a way to do a task that you have not done
before.
3 Information
B Plan for how to achieve something.
4 Redundant 13
C Involving a method, plan or step-by-step process to do
5 Systematic
something.
6 Strategy
D Way of solving a problem, or finding the best way to
7 Trial and error achieve a desired result, by trying one thing or another
until something succeeds.
E Knowledge, facts or news about a person, situation or
event.
F Facts, numbers and statistics used to calculate, plan or
analyse something.
G Information that is not useful or necessary.
Summary
Having read this chapter you should be able to:
■ identify the information that is important in solving a ■ identify an appropriate method to solve the problem.
problem
End of chapter questions

‘The questions, example answers, marks awarded and/or comments that appear in this
book/CD were written by the author(s). In examination, the way marks would be awarded to
answers like these may be different.’
1 You want to buy a new mobile phone. You will use it for texting, making a few phone
calls and keeping in touch with friends and family on social media. You have found two
phones:
• a ‘Mobes4U’ phone that costs $36 if you make a two-year contract with the company
• a ‘Pay’N’Go’ phone that costs $50 but you are not tied to a contract.
a What other pieces of information should you consider before you decide which phone
to buy?
b Explain how you would use the information to make a decision.
Exam-style questions
1 Nazim is planning to cook curries and sell them. He will take orders for parties. He will
cook the curry and rice beforehand, then put the cooked food in serving dishes that will
keep it hot. Nazim will pay his friend Raj, who has a van, to make deliveries and to fetch
the serving dishes back after the party.
Nazim’s first order is for a party of 27 people. He has a recipe using vegetables, chick
14
peas, onions and tomatoes. He is now trying to work out the right quantities of spices.
For 15 people he would normally use 100 grams of spices, made up of: ginger (20%),
turmeric , coriander (35%), and cumin seeds (20%).
a What weight of turmeric must Nazim put in a curry for 27 people?
b The food has been ordered to arrive at the party at 7.30 in the evening. Nazim knows
he needs 30 minutes to prepare the vegetables. The curry takes 2 hours to cook and
the rice takes 20 minutes. Putting the food in the dishes takes 10 minutes and putting
the dishes in the van is another 10 minutes. Raj’s journey to the party will take 25
minutes.
What is the latest time Nazim can start preparing the food?
c Nazim is buying rice. The supermarket sells 10 kilograms of basmati rice for $15.45.
Today they have a ‘buy 3, get the 4th free’ offer. The wholesaler also offers 50 kilogram
sacks of rice for $59.00.
Based on price, which rice should Nazim buy?
d Nazim wants to calculate the prices he should charge to customers to make sure he
can make a profit. State three pieces of information that he needs.
168
Chapter 18
Statistical evidence
Learning objectives
The focus of this chapter turns from information in general to statistical information in
particular. Statistics are an important source of evidence in many fields of inquiry, and
it is for this reason that they need to be viewed and assessed critically.
Chapter 18: Statistical evidence
18.1 What are statistics?

A statistic is a single item of numerical data – for example, the number of students in a
certain college whose second language is English, or the number of different species of
trees in the world. Statistics, in the plural, are sets or collections of such items collected
for a purpose, such as educational research or predicting election outcomes. ‘Statistics’
(as a singular noun) is also the name given to the science of collecting, analysing, and
evaluating data.
Statistical data must be assessed for credibility and reliability for the same reasons as
evidence in general must be assessed, and by many of the same criteria:
• What is the source of the data; who provided it?
• What was the purpose for which it was collected?
• How was it collected; what was the methodology?
• Is it corroborated by (and consistent with) other available evidence?
• Is it relevant, plausible, informative?
18.1.1 Sources of data

The first question is crucial for obvious reasons. It is also inseparable from the second
question, about purpose. People do not normally go to the trouble and expense of gathering
data without a reason – to further an investigation, for example, or to confirm a hypothesis,
or to justify a government policy. Unless the source of a data set is known and trusted, there
is a risk of bias, whether due to vested interest, or to unscientific methods of collection and
169
presentation. If the source is a research team and their purpose is to argue for a particular
outcome, then questions regarding independence and impartiality have to be asked. This
is not to say that when bias occurs it is necessarily deliberate or dishonest: researchers can
unintentionally focus on evidence which supports their viewpoint or which challenges an
opposing one without setting out to do so.
It is also important for critical purposes to have information about the methodology used to
gather and process the data for the project in question. Suppose one such project were to
investigate foreign language aptitude in a given population by conducting a survey in which
a number of people were asked how many languages they could speak. As it stands this
would be a crude and dubious procedure. First, we have no information about the number
of people questioned, or of the makeup of the sample. The sector of the community who
were questioned might have had linguistic skills well above or well below the average for the
population. A reliable survey needs to be accompanied by some explanation as to how such
deficiencies have been guarded against.
Second, the question, ‘How many languages can you speak?’ is hopelessly vague and
uninformative. There is nothing to indicate what level of ability counts as ‘ability to speak’
a language. Indeed, setting and defining such a level is notoriously difficult. Third, some
respondents might not answer truthfully, for various reasons including perhaps not wanting
to appear uneducated. Last of all the evidence, such as it is, stands alone, with no mention
of any corroborating evidence. It is a recognised reliability criterion for any scientific
experiment or survey that it must be repeatable (with a similar outcome). This means that
its findings can be confirmed by other comparable experiments; and at the same time that
comparable experiments or surveys do not contradict the findings.
18.1.2 Collecting data

Statistics are basically numbers, but numbers relating to a particular population.
‘Population’ is the technical term for a whole collection of people or objects that fall under a
particular description. Whales form a population; so do eligible voters, tomatoes, ten-dollar
bills, autonomous vehicles, days since the turn of the century, species of trees in the world,
and so on. The number of any one of these is a statistic.
Some statistics can be obtained and verified by simply counting. The number of days
between two given dates can be calculated easily enough. The number of citizens eligible
to vote in the next election can be obtained from an electoral register. A count of this kind is
known as a census; the term ‘census data’ simply means the entire number of individuals
in a given population. So long as it can be shown that the counting was done carefully and
honestly, census data can be taken as fact. Of course mistakes can be made, especially when
the numbers are large. (Votes are often recounted for this reason.) Also some populations
are too unstable for a complete, up-to-the-minute count to be made. The world’s human
population is a case in point: by the time any count has been made, a large number of births
and deaths will have already occurred. Nonetheless, and whatever the practical difficulties, a
census is in theory a complete population count.
By contrast there are populations which are uncountable, even in theory. Future populations
are an obvious example. Consider the number of votes that Party × will win in a forthcoming
election. Clearly it is impossible to count votes before they are cast, and predicting the
numbers in advance is notoriously difficult. The usual procedure for forecasting an election
result is to ask a lot of people how they will vote when the time comes. But as experience has
170 all too often shown, there are serious problems with this procedure:
1 There is the practical impossibility of contacting and questioning all eligible voters.
2 There is the fact voters can (and frequently do) change their minds.
3 There will always be some respondents who reply untruthfully.
18.1.3 Sampling
When census data is unavailable, the standard solution is to take a sample of the
population, and to use data drawn from it as an indication of some facts about the whole
population. If successful this solves the first problem listed above, though not the second or
third. But it is not a solution even to the first problem unless the sample is representative
of the population concerned. If it is biased or selective the sample cannot be relied upon
to reflect population-wide trends or patterns. For instance, if a sample of voters contains
a greater proportion of over-35s than is found in the general population, then the sample
will not be representative with respect to age. This would have serious implications for
the reliability of the forecast if younger voters tended to favour Party X, and over-35s Party
Y. A similar risk applies if the sample contains a disproportionate number of people from
one part of the country, or from a different class background. Any of these factors could
undermine the prediction.
The standard way to avoid biased or unrepresentative data is to take a random sample.
This is sometimes referred to as a probability sample, because it supposedly allows
every member of the population a fair chance of being selected. For example, if university
graduates make up 20% of all voters, then – according to probability theory – random
selection should result in 20% of the sample having university degrees; and likewise for the
whole range of variables in the population.
18.1.4 Sample size

Random selection, whereby each individual member of the sample is plucked out purely
by chance, will not ensure representativeness for all sectors of the population, but it will
increase the probability of obtaining a representative sample. There is one more (big)
proviso, however: the sample must be large. The probability of the sample mirroring the
population increases in accordance with its size.
Suppose you have a bag containing 20 red balls and 80 blue ones, and you pick out a ball at
random – that is, without looking. There is a probability of 0.8 that the ball you pick out will
be blue. If you repeat this ten times, putting the ball back each time, you will have performed
a random selection. On the basis of probability the selected sample should contain 2 red
balls. But in reality the outcome could be any combination – even 10 red balls in extreme
cases. The possibility of such improbable outcomes decreases the larger the sample is, and
increases when the sample is small. (Ten is a very small sample by any statistical standard.)
If you repeat the experiment with a hundred balls, then a thousand, then ten thousand, the
ratio of blue to red balls will get closer and closer to 80:20. By the same token, the reliability
of the claims made on the basis of the sample increases. As a rough guide, a genuinely
random sample of 1,000 items could be expected to result in a margin of error of around 5%,
giving a 95% probability of a reliable outcome.
Statisticians mathematically calculate how large a sample has to be to give a reasonably
reliable forecast in different kinds of inquiry. They also research the best methodologies
for obtaining a truly random sample. The mathematics is complex, and is beyond the
scope of this book. For most critical thinking purposes it is enough to understand the
principle involved, and to be able to recognise the most common weakness in the statistical 171
reasoning that you encounter in the media and elsewhere. Small and/or unrepresentative
samples are always defects to look out for.
Comment critically on the following statistical claims. How reliable do you consider the
evidence to be in each case?
1 A survey was conducted to assess public opinion on the recent government
proposal to improve air quality by making a charge on diesel vehicles entering major
cities. Around half of all vehicles on the roads currently run on diesel, but out of two
thousand car drivers randomly questioned in car parks in cities across the country,
61% said they were opposed to the charge.
2 72% of retail spending is by card payment, according to a survey at the checkouts
of a massive hypermarket in central France. 2 900 payments were recorded during a
single day, 2 088 of which were made by card.
3 According to a recent study of the world’s plants by Botanical Gardens Conservation
International (BGCI) there are 60 065 species of trees in the world. A comprehensive
tree list was compiled using data gathered from the BGCI network of 500 member
organisations (http://www.bbc.co.uk/programmes/b08kscc6).
Commentary
The first statistic is based on the responses of a sample of car drivers to a question
about charges on diesel vehicles. (Governments of several countries are planning to
charge drivers of diesel vehicles in a bid to improve air quality in inner cities.) We are
told that the sample was 2 000, which is a respectable size, and that it was randomly
selected. Therefore, given that around half of all drivers have diesel cars, there is no
reason to suppose that the sample had a bias towards those who would be financially
penalised by the charge – 61% is a substantial majority, above the generally accepted
margin of error for a sample of its size. But the sample was drawn exclusively from
drivers questioned in city car-parks, and that is not the whole population – i.e. the
public at large – that the statistic is intended to represent. The outcome might be
very different if pedestrians, users of public transport, parents out with their children,
asthma sufferers, health workers, etc., are all proportionally represented in the sample.
There are also unanswered questions about the source of the evidence and the
motives for collecting it. Likely sources are motoring organisations, who have drivers’
interests at heart; or vehicle manufacturers or consortiums who would have an obvious
vested interest. Such factors could have an influence on the selection and analysis of
the data and/or the way the question was framed. (We should not forget the recent
scandal over falsified emissions tests on some makes of diesel cars.) Reputable sources
(like the BGCI in the third example) publish their identity along with their findings. If the
source is not given, the reliability of the evidence is further open to question.
The second statistic is drawn from a sample of payments at retail checkouts to
172
ascertain the proportion that are made by card (as opposed to cash or cheque). The
information here is objective, unlike the previous example which concerned public
opinion. It is a plain fact that 72% of the customers in the survey paid by card. The
sample size was large and selection was random insofar as every payment on the day,
in that store, was counted.
These are positive points. But is it safe to assume that data collected from one store,
on one particular day, is representative of the retail-spending population as a whole.
You will recall that a criterion of reliability for experimental or observational data is its
repeatability. To test the findings from the hypermarket survey, similar samples should
be taken from high-street (town-centre) shops, restaurants, bars, bookshops, music
stores, street-markets, and so on, to discover whether payment habits are the same
across all retail outlets. Also the growing volume of E-commerce needs to be taken into
account, especially as online purchases are predominantly paid for by credit or debit
card. It might be argued that a hypermarket, with its enormous range of goods under
one roof, is a good indicator of retail trade in general. But against this there is the fact
that payments at hypermarket checkouts tend be larger than those in smaller shops,
because people are typically buying in quantity. All in all, it would be unsafe to assume
that the survey is either sufficiently random, or properly representative. Without the
corroboration of findings from other samples, the evidence on its own must be treated
with caution.
The third example is census data. It is not a sample, but a comprehensive survey of the
world’s plants to ascertain all and only the number of species of trees. It alleged that
there were over 60 000. How do we judge such a claim? Since we as readers cannot
repeat the experiment for ourselves, our main concern must be with the reliability
we place on the source – or sources. Several relevant criteria were discussed in the
previous chapter, in particular reputation, expertise, and corroboration. The reliability
of evidence of this kind obviously depends on extensive botanical knowledge, and the
skill to identify and classify specimens. On the strength of its reputation it is reasonable
to assume that the BGCI researchers possess the requisite expertise and that, with 500
member organisations contributing data, the findings are well corroborated.
These observations on the three surveys do not mean that the findings discussed
above are necessarily right or wrong. What they are intended to demonstrate is the
need for critical questioning of statistical data before rushing to general conclusions.
18.2 Trends
A trend is a pattern of more or less gradual change in a set of data. A single statistic or value
cannot represent a trend, nor can a sudden leap from one value to another, though either of
these can be part of a trend. When a single statistic is part of a trend we can refer to it usefully
as a data point, and plot it on a graph. If a line drawn between a sequence of points can be
seen to travel in a particular direction, then it represents a trend. Compare the two graphs
below, giving sales figures for two business over the course of one year:
example data points
Sales 173
30,000
25,000
20,000
Sales (US$)
Sales (US$)
15,000
10,000
5,000
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Fe 2
M 12
Ap 2
M 12
Ju 2
Ju 2
Au 2
Se 12
Oc 2
No 2
De 2
12
1
-1
-1
1
l-1
1
t-1
1
n-
b-
r-
n-
g-
p-
v-
c-
ar
ay
Ja
Figure 18.1 Figure 18.2
The lines connecting the data points in both graphs change direction frequently, but whilst
there is no general direction of travel recognisable in Fig. 18.1, there is in Fig. 18.2, as shown
by the red line. If a similar straight line were drawn on Fig. 18.1 it would give a very false
impression of how sales had performed over the year.
18.2.1 Trends and predictions

Statistical trends are often used as evidence to justify predictions, and/or decisions. If a
company’s sales have been steadily rising over a sufficient period of time a company may feel
entitled to assume the trend will continue, and make optimistic business plans accordingly.
The crucial proviso, however, is in the word ‘sufficient’. Making the wrong decision could
result in a mountain of unsold stock if the trend changes in the near future. A trend, in this
respect, functions in the same way as a sample – a sample of data points – and is open to
the same critical questions about size, representativeness, and repeatability.
Like a very small sample in a survey, a very short sequence of data points is weak evidence
of a lasting trend. If the company featured in Fig. 18.1 had based its forecasts on the upward
trend from July to October, and planned for a bumper Christmas, it would have been in
for a shock, because the data for those months were not representative of the whole year.
This also emphasises the need for repeatability: a trend can only be good grounds for an
important prediction if it is consistent with other relevant trends – for example, the trends in
other similar companies, and/or trends over longer periods of time. Suppose Fig. 18.1 was
matched with data from previous years, or repeated over several more years, and that the
graph looked very similar each time. Then the company’s analysts could conclude with some
confidence that their sales were likely to peak in the spring and autumn, and dip in summer
and winter.
In the chart below, the top of each column marks a data point, corresponding to the
number of new houses sold in the United States in that year. Joining the points does
not result in a single continuous trend, but it does produce a pattern of sorts that can be
interpreted and described.
New houses sold 1000 units
1500
174
1200
900
600
300
0
‘90 ‘91 ‘92 ‘93 ‘94 ‘95 ‘96 ‘97 ‘98 ‘99 ‘00 ‘01 ‘02 ‘03 ‘04 ‘05 ‘06 ‘07 ‘08 ‘09 ‘10 ‘11 ‘12 ‘13 ‘14 ‘15 ‘16
Source: U.S. Census Bureau
Figure 18.3
Write a short paragraph describing the fluctuation in sales of new houses in the United
States in the 1990s and first 16 years of the twenty-first century. What does the data
provide by way of evidence for predicting house sales in the next years of the decade?
Commentary
The key word here is ‘fluctuation’. The sales of house prices year on year were not
stable nor was there a single direction of change. The broken line indicates that sales
were practically the same in 2016 as they had been in 1990, but they rose and fell above
and below that level by many hundreds of thousands. For the first fifteen years of the
period there were small ups and downs and short periods of flat-lining. There was also
a marked acceleration in the rate of increase – commonly referred to as a spike – in the
five years leading up to the peak in 2005. All the same, the 15 year trend from 1990 to
2005 was relentlessly upward.
Following the spike there was an even more dramatic plunge in sales. Anyone familiar
with recent economic history will understand the significance of this. In the years
leading up to the 2007–2008 financial crisis, property prices in the US, the UK, and
many other countries rose to unsustainable levels, fuelling (amongst other things) the
demand for new homes. In retrospect the bubble was bound to burst, but even the
experts at the time could not (or chose not) to see it coming. Whatever the reason, the
market collapsed in value, with consequences clearly reflected in Fig. 18.3.
Around 2011 the downward trend bottomed out and by 2016 had returned to roughly
the same rate of increase as was seen in the 1990s. Is it safe to predict that house sales
in the US will continue to rise again post 2016, or that the cycle of boom and bust
will be repeated? By the time you are reading this page you might be able to give a
particle answer. But that is not really the point of the question because by then you
will have the benefit of hindsight. The question is about the value of past trends as
evidence for predictions. The lesson to be learnt from Fig. 18.3 should be clear enough:
such evidence on its own is inherently unreliable. It has a role to play, but only if
corroborated by evidence of other kinds.
It should be added that some trends are naturally more predictable that others. The
rise and fall of tides can be predicted with near certainty, not just because this has
always happened in the past but because we know why it happens as well as that it
happens. Trends that are governed by human behaviour – like economic fluctuations –
are not so fully understood. If they were, prediction could be improved.
175
18.2.2 Correlation
One of the reasons for studying trends and patterns in data is to identify correlations. A
correlation is a measure of the extent to which two or more variables change in relation
to each other. A correlation is called positive if the variables increase or decrease in line
with each other. In a negative (or inverse) correlation one variable increases as the other
decreases. Correlation can also be described as strong or weak, which is best understood
by examples. There is a strong, positive correlation between age and height during the early
years of a person’s life. The trend, in other words, is to grow taller with age, not necessarily at
the same rate, but more or less continually. Fig. 18.4 depicts one child’s development over
time, correlating their age (on the horizontal axis) with their height (on the vertical axis).
130
120
110
Height (cm)
100
90
80
70
60
1 2 3 4 5 6 7 8
Age (years)
Figure 18.4
There is also a strong correlation between height and weight, not just over time, but between
people of different sizes. The data points in Fig. 18.5 – known as a scatter graph – represent
20 young people ranging between 1.34 m and 1.6 m in height and weighing between 28 kg
and 70 kg. If there was a perfect correlation the data points would form a single straight line.
But humans do not conform in that way: there are variations in the ratio of height to weight
from one person to the next. We can say there is a positive correlation because the points
cluster around a straight line (known as a line of best fit) which angles upwards from the
bottom left-hand corner towards the top right. Because the sample is small this cannot be
taken as the average weight/height ratio for the whole population. However, the cluster is
close enough to a straight line to confirm that there is a strong correlation between height
and weight generally – something we knew, or could have guessed, anyway.
66
64
62
60
58
56
54
52
Weight (kg)
50
48
46
44
42
40
176
38
36
34
32
30
0 134 136 138 140 142 144 146 148 150 152 154 156 158 160
Height (cm)
Figure 18.5
The closer the points are to a single line the stronger the correlation, whether positive or
negative. If there is no recognisable line of fit, we must assume there is no correlation.
ACTIVITY 18.3
Suggest a pair of variables that would illustrate each of the examples of different
patterns of data shown in Fig. 18.6. For example, which pattern might reflect the
sales of umbrellas in relation to hours of rainfall on a given day?
Strong positive Weak positive Strong negative

correlation correlation correlation
Weak negative Moderate negative No

correlation correlation correlation
177
Figure 18.6
18.2.3 Relations between trends

When two or more trends can be identified at the same time they may or may not be related.
For instance, if seasonal air temperature is rising it can be expected that water temperature
will rise accordingly. Hence we say that there is a positive correlation between the two events.
Neither one causes the other: both have the same cause, namely the heat from the sun.
Not all correlations are positive. An increase over time in interest rates is likely to be
accompanied by a reduction in borrowing and spending, again for fairly obvious reasons.
Between these trends there is an inverse correlation, one going up as the other goes down.
In this particular case it is a reasonable hypothesis that one of the factors is responsible for
the other: the higher cost of borrowing usually makes people more reluctant to borrow, and
with less borrowing there is less money to spend. However, the connection is not a cast-iron
one or applicable in all cases: some people will continue to borrow even when interest rates
are high. They ‘buck the trend’, as the saying goes. Nonetheless, in the population as a whole
– or over a large sample – the correlation between interest rates and reduced spending is a
strong one.
It is often tempting when we discover a correlation to look for a causal connection between
the correlates, or to see one as the explanation for the other. (The noun ‘correlate’
means any one of the elements in a correlation.) Likewise it is tempting to use a statistical
correlation as evidence of a causal or explanatory connection. There is an extremely
important critical point to take on board here: just because two things happen at the same
time, or one follows soon after the other, it does not mean that one has caused the other.
This is even true when one thing is deliberately done with the intention of bringing about the
other: it may have the desired effect, but on occasions the desired effect may just be a happy
coincidence.
Consider an imaginary report published by a city police authority on a five-year initiative

to reduce local crime. Three trends are identified in the report: 1) an increase of 15% over
the five years in the number of police officers regularly and visibly patrolling the streets;
2) a 7% reduction in the number of crimes committed (including a 12% reduction in
violent crime); 3) an increase in public fear of crime – as evidenced by a survey of 1,200
residents. Of those questioned 64% said that they were growing more afraid of becoming
a victim of crime (as opposed to less afraid or just the same), compared with only 46% of
those questioned five years earlier.
What can be concluded about the trends and correlations identified in this ‘report’?
Discuss or think about this question before reading the following commentary.
Commentary
There are a number of statistical facts stated in the report. The question we are
concerned with is how they are inter-related. Take the first two trends: the five-year
increase in visible police patrols and a concurrent reduction in crime. (‘Concurrent’
178 means happening at the same time, or alongside, as opposed to ‘consecutive’
which means one after the other.) Numerically speaking there is an inverse correlation
between the two trends. However, in qualitative terms (as opposed to purely
numerical) a fall in crime would be seen as a positive outcome; a success (though
evidently not enough of a success to prevent a growth in fear of crime). A rise in fear of
crime, qualitatively speaking, is a negative outcome. But as already stated we
have to be careful how we use these terms. As soon as we start talking about ‘success’
or positive (or negative) outcomes we are already running the risk of jumping to
unsupported conclusions. There is no question about the existence of a correlation,
but looking beyond that can we attribute the fall in crime to the police initiative? Is
the report evidence that higher numbers of police on the streets has the effect of
reducing crime?
The short answer is no. All we can conclude is that on this occasion there was a fall
in crime that coincided with increased policing. The danger here is that because the
increased police presence is so plausible as an explanation for the correlated fall in
crime, we assume that it is the explanation. It is perfectly possible, however, that crime
rates might have been dropping everywhere in the country during the same period and
for altogether different reasons, even in cities where police numbers were being cut; so
that although the two trends are matters of fact one may have had no impact at all on
the other.
What about the third trend, the increase in fear of crime? Initially this might seem a
negative outcome in terms of what was hoped for from the initiative. Moreover, it is a
surprising fact. The natural expectation would be that better policing and a fall in crime
would reduce anxiety, not raise it. Because the correlation is a surprising one, we find
ourselves looking for an explanation for it, and assuming that if it fits the facts it must
be the case. This too is a temptation, and one to be wary of. It is not too far-fetched
to imagine that the sight of more police might have the effect of making people more
aware of the threat posed by criminals, and this has increased their fears. But just
because this would explain the correlation, if true, it does not mean that it is true.
So what can be concluded from a single initiative like the imaginary example above?
The answer is: very little. Without more research and corroborating evidence the report
provides little more than anecdotal evidence: a single experiment, conducted in one
city and over a relatively short period of time.
The issues introduced here are discussed in more detail in the coming chapters on
inference, argument, and explanation. In this section they relate to statistical evidence
in particular, but they have much wider application to critical thinking in general.
18.3 Presentation
We end this chapter by looking at modes of presentation: ways in which statistical
information is commonly put before the audience. One might think that data is data, and
so long as it is accurate, representative, etc. it is merely a matter of preference as to how it is
displayed. But how a statistic is displayed can make a big difference to the effect it has on its
audience. On the positive side a well-chosen mode of presentation can make a statistic more
comprehensible or digestible; but on the negative side there are ways in which essentially
179
accurate data can create a false or misleading impression – intentionally or otherwise.
The plainest and generally the most neutral way to present a body of numerical information
is in a table. Here, for example, is a table showing the weekly fluctuation in the exchange rate
of the euro against the US dollar over a five-week period in 2017.
Table 18.1
Date EUR USD

11.3.17 1.00 1.066
18.3.17 1.00 1.076
25.3.17 1.00 1.078
1.4.17 1.00 1.066
8.4.17 1.00 1.064
15.4.17 1.00 1.061
The data from the table can alternatively be presented in visual form.
Compare the next two figures. Do they represent the same, or different, information?
2.500
2.000
1 EUR = (US$)
1.500
1.000
0.500
0.000
17
17
7
.1
.1
.1
.1
4.
4.
.3
.3
.3
.4
1.
8.
11
18
25
15
Figure 18.7 Euro–dollar exchange rates
1.080
1.075
1.070
1 EUR = (US$)
180
1.065
1.060
1.055
1.050
7
17
17
7
.1
.1
.1
.1
4.
4.
.3
.3
.3
.4
1.
8.
11
18
25
15
Figure 18.8 Euro–dollar exchange rates
Commentary
Both bar-charts convey the same raw data as the table in Fig. 18.7, and as each other.
The difference between the two figures is purely presentational. However, that is not
to say that there is no difference in the impression given by each one. Fig. 18.8 suggests
that the exchange rate practically flat-lined for the five weeks in question, whereas
Fig. 18.9 depicts sharp fluctuations over the same period. It is easy to see why the
difference occurs: it is due entirely to the range of values on the vertical axis – in
other words, scale. In Fig. 18.8 the scale begins at zero, and ranges up to 2.5 dollars,
rendering the 1 or 2 cent variation barely visible. The range on the axis in Fig. 18.9 is
itself only 3 cents ($1.05–$1.08), creating a ‘zoom-in’ effect on the narrow range within
which the rate varies.
Is one of the graphs the ‘right’ visualisation, and the other ‘wrong’? Not exactly: both
give a somewhat distorted picture. One exaggerates the fluctuations in the rate, the
other registers no significant movement at all. Unless you are exchanging very large
amounts of currency, a cent either way is not going to have much impact on your life,
but still it is useful to know the direction in which the rate is changing, and Fig. 18.8 fails
to do that. Fig. 18.9 is certainly more sensitive to variation, but perhaps unnecessarily
alarmist, if used to prompt headlines such as:
EURO PLUNGES TO 5-WEEK LOW AGAINST DOLLAR

Ideally we want a happy medium: visualisation which is sensitive to variation, but not
so much so as to distort or exaggerate.
Some problems with presentation

Sometimes the nature of the data itself makes presentation difficult. In the previous
example the problem was the low level of variation, but large variations, and/or
major anomalies can also make trouble. The following sales were reported by a small
publishing house after its first year of trading. A total of five titles were marketed, one
of which proved much more popular than the other four. They are listed in the table
below as books A–E.
Table 18.3
A 357
B 488 181
C 361
D 580
E 16,676
Examine the three graphs below (Figs. 18.9–11), each of which is meant to depict the
sales figures in Table 18.3 visually. Explain the problem with each of the visualisations.
18000
16000
14000
12000
10000
8000
6000
4000
2000
0
A B C D E
Figure 18.9
1800
16676
1600
1400
1200
1000
800
600
400
200
0
A B C D E
Figure 18.10
20000
10000
1000
800
600
182 400
200
0
A B C D E
Figure 18.11
Commentary
In Fig. 18.9 an ordinary scale is used on the vertical axis, with a range from 0 to 20 000.
This range is needed because the volume of sales of one of the books is more than
27 times greater than next-best seller in the list. The effect is to squash the columns
representing the other titles, so that the differences between them are almost
indistinguishable.
The solution adopted by the next two figures is to truncate some of the data in one
of two ways. This just means shortening, or condensing, one or more of the columns.
For example in Fig 18.10 the column representing Book E is broken, to show that there
is a missing section. The actual number of sales is displayed above the column to
compensate. This permits visual discrimination between the sales of the other titles,
but at the cost of losing any sense of the gulf between Book E and the rest. Fig. 18.11
has a similar purpose, but achieves it by compressing the vertical axis. Above 1 000 the
intervals change from 200 to 5 000. But unless this is noticed and adjustment made in
how the graph is interpreted, it could be misleading.
The message to take from this is the need for vigilance: to be aware of the purpose and
usefulness of such techniques but also of their potential for abuse. Imagine a politically
sensitive report on, say, funding for schools by different educational authorities.
Depending on who is publishing the report, and why, there may be a motive for
presenting the data in a manner that either maximises or minimises the impression
created by a particular mode of presentation without actually falsifying the data.
18.3.1 Multiple trends

There are further difficulties when it comes to showing correlations between different trends
in a graphic mode. Recall the earlier example of police numbers and criminal offences
before and after a five-year programme to reduce crime. The first figure rose by 15%, the
second fell by 7%. Without actual numbers we learn little from these statistics, beyond the
fact that there was an inverse relation between the two trends. We are left with the question:
‘Percentage of what?’ The table below provides an answer:
Table 18.4
Mar. 2012 Mar. 2017 % change

Police numbers 420 483 +15%
Crimes 19,002 17,672 –7%
As with the previous example, there is a large discrepancy between the numbers, making 183
a meaningful graph difficult to construct. But in this instance it is not only the size of
numbers that are different, it is what kind of set each one represents: one a number of
people, the other a number of acts or events. It is, as the saying goes, like ‘comparing apples
and oranges’.
A common solution to such discrepancies is to use a different scale for each variable, as
below. The scale corresponding to police numbers (blue) is on the left; crime numbers (red)
on the right. Time, as usual, is on the horizontal axis.
Police Crimes
numbers
20 000
400 19 000
18 000
200 17 000
16 000
0 1 2 3 4 5
Years
Figure 18.12
Comment critically on the presentation of the data in Fig. 18.12. Does it accurately
represent the statistical information in Table 18.4? Are there any respects in which you
think the graph could give a misleading impression? If so, does it matter?
Commentary
The graph employs two different scales in order to depict the rising trend in police
numbers and the simultaneous fall in crime. The former increases from 420 to 483, a
rise of 15% as recorded in the table. The latter decreases from 19 002 to 17 672, a fall of
7%. Individually both trends are correctly plotted, but due to the choice of scales, the
steepness of the decline in crime is exaggerated. In percentage terms the increase in
policing is over twice that of the fall in crime, yet in visual terms it appears that the fall
in crime is relatively greater. Does it matter? Yes, if the impression given by the graph
has the effect of persuading the reader to draw an incorrect or overstated conclusion;
and if the purpose of the report were to support the view that the initiative had been
highly effective, when in fact it had a minimal effect.
Graphs and diagrams have a useful role to play in depicting trends and patterns that
are hard to recognise in large and complex tables. But they can also distort data,
sometimes harmlessly, sometimes to the point of falsifying evidence.
Be on guard.
184
Country Average income 200 metres national

($ per annum) record (seconds)
USA 30 000 18
Italy 25 000 20
Vietnam 15 000 22
Bangladesh 9000 23
Cameroon 8000 25
Honduras 6000 27
Suggest two explanations for the correlation revealed between national records and
average income in the table.
CI AS Thinking Skills Paper 2 Critical Thinking 9694/23 May/June 2017
Summary
Following on from the discussion of evidence in ■ sources of data
general, this chapter examined some important aspects ■ collecting and sampling
of statistical evidence. In particular it considered the ■ trends, and their relation to prediction
importance of: ■ correlation
■ modes of presentation.
End of chapter question

1 Refer to Source C Trend on page 167. Figure 17.1 shows data relating to unprovoked shark
attacks worldwide.
2 a How does the graph in Source C support the claim that beach safety and other factors
have advanced over the decades?
b Does the graph give a misleading impression of the relation between shark attacks
and fatality rates? (Give a reason for your answer.)
Exam style questions

1 Study the information in Figure 18.13 below and answer the question that follows.
Our recent drive to cut prices is paying off. We compared last month’s prices
on all food items at four supermarkets. Guess who was the cheapest?
Cotes 662
Dasamart 1983
Shrewsbury 553
185
Hendriks 1035
Number of items on sale that were less

expensive than at the other three supermarkets
To save money on your weekly shopping, come to Dasamart!
Figure 18.13
Make five criticisms of the statistics in the advertisement and/or the support they give to
the claim,
“To save money on your weekly shopping, come to Dasamart!”
CI A Level Thinking Skills 9694/42 Paper 4 Applied Reasoning May/June 2016
2 Figure 18.14 shows reported reactions to common allergens. A newspaper article has
published this chart under the headline, ‘Food additives make children more hyperactive.’
Identify and explain two weaknesses in the way the newspaper has used the data.
Penicillin
allergy 150
Hayfever 100
Asthma 88
Dermatitis 80
General food
sensitivity 20
Sensitivity to 2
food additives
0 50 100 150
Figure 18.14
Adapted from CIE AS Level Thinking Skills 9694/02 Critical Thinking Specimen Paper 2 (published 2010)
186
203
Chapter 20
Identifying argument
Learning objectives
This chapter introduces the major topic of argument. It examines standard and
non-standard forms of argument, and argument in the sense of dispute. Activities in
the chapter include identifying arguments and their components, and first steps in
analysis and evaluation.
Reading through this chapter and attempting the Activities will help you identify an
argument, in the critical thinking sense, and the key components of an argument: one or
more reasons and a conclusion. In this textbook a more specialist word is generally used
instead of ‘reason’. ‘Premise’ has a very similar meaning to ‘reason, as this chapter explains.
If you are preparing for the Cambridge International A Level Thinking Skills, you will find that
the questions in those examinations use the term ‘reason’ rather than ‘premise’.
This chapter also introduces the concept of the ‘soundness’ of an argument. Soundness is
important if you are interested in the wider application of critical thinking, but you will not
be tested on it in the Cambridge International A Level Thinking Skills examinations.
In the CI Critical Thinking exams you will need to be able to:
20.1 Recognising an argument

In logic an argument is defined as a set of sentences, one of which – the conclusion – is
claimed to follow from the others which are its premises. Premises are reasons given to
support or justify a conclusion. To say that a conclusion ‘follows from’ one or more premises
(or reasons), is just to say that if the premises are true, the conclusion must also be true. If
it is judged that both these conditions are met – the premises are true and they justify the
conclusion – then logicians (people who study logic) say that the argument is sound.
In natural, real-life contexts, arguments are typically used to persuade the audience of the
truth or the rightness of a claim or point of view, by giving supportive reasons. Take the
following very simple example:
204 (1) The tide is out, so we can get to the island.
We can imagine the circumstances in which someone might put forward this argument. The
speaker wants to persuade one or more companions that it is possible and/or safe to cross to
an island that is usually cut off by the sea. The reason the speaker gives is that the tide is out.
The argument consists of two claims – a premise and a conclusion – joined by the word
‘so’. ‘So’ is one of several words and phrases which are used to mark the conclusion of an
argument. Others include: ‘therefore’, ‘thus’, ‘consequently’ and ‘for this reason’. Because
drawing a conclusion is very similar to inferring (see previous chapter), these words are often
called inference indicators (or argument indicators). What the author (or proponent) of
this argument is doing is inferring from the observed state of the tide that the island can be
reached. That is what makes (1) an argument.
Note: a proponent is someone who propounds, i.e. puts forward, an argument.
Figure 20.1
Chapter 20: Identifying argument
To judge whether or not (1) is a good argument two criteria must be met. First, it must be true
that if the tide is out the island truly can be reached. This, as stated above, is what it means
for the conclusion to follow from the reasons. If it does, the argument can be approved as
valid. If this condition is not met, then it could be the case that the tide is out, but the island
is still beyond reach, which would make the inference unsafe and the argument invalid.
Second, the tide must actually be out. Only then can the argument be judged sound.
These are quite technical points, which may be unfamiliar to some readers. Here they are
introduced briefly, but they will be revisited at various points in the coming chapters 22–24.
20.2 An example from history

Until a few hundred years ago it was generally believed that the world was flat. This was a
natural belief to have because the Earth’s surface looks flat. But people had also observed
(and been puzzled by the fact) that ships sailing away from land appeared to get lower and
lower in the water, as if they were sinking, and appeared to rise up again as they approached
land. Some argued – from this and other observations – that the Earth’s surface could not be
flat, but was curved. They drew this conclusion because if the Earth were flat, a ship would
just appear to get smaller and smaller until it was too small to see. Expressed as an argument
the reasoning went like this:
(2) If the Earth were flat, ships sailing away from land would appear to get smaller and
smaller until they disappear. Instead they appear to drop out of sight after a few
miles. Therefore the Earth cannot be flat.
205
This argument consists of two premises, corresponding to the first two sentences, and a
conclusion at the end indicated by ‘therefore’. However, when we refer to something as an
argument, that does not necessarily mean the text. The text expresses the argument, but the
argument itself – like any other claim or assertion – could be expressed in other ways without
being a different argument. For example:
(2a) The Earth cannot be flat because, if it were flat, ships sailing away from land would
appear to get smaller and smaller until they disappeared. Instead they appear to drop
out of sight after a few miles.
In this version of the argument the sentences are in a different order, with the conclusion at
the front and the reasons (premises) following. But they have the same logical function as
they do in (2). Notice, too, that as the sentences are reversed in (2a), a different connective
is needed, i.e. one that identifies the reasons. In this case the proponent has used ‘because’,
but could just have well used ‘since’. These words can also function as argument indicators,
but in a different way.
There is a third option which is to use no connective or indicator at all.
(2b) The Earth cannot be flat. If it were flat, ships sailing away from land would appear
to get smaller and smaller until they disappeared. Instead they appear to drop out of
sight after a few miles.
You will often find natural-language arguments without argument indicators, simply because
the context and sentence-meaning make it clear which sentence is the conclusion and which
are the premises.
20.2.1 The form of an argument

In each of the examples above, the argument is expressed differently. But it is still the
same argument, with the same reason and same conclusion. Because there are many
ways in which an argument can be expressed, it is convenient to have one standard or
regular form for setting arguments out. The standard form of argument can be represented
by the schema:
R1, R2, . . . Rn ; C
in which ‘C’ stands for the conclusion and ‘R’ reason. (Use ‘P’ for ‘premise’ if preferred.) As
you can see (2) is already expressed in a very standard form, with its sentences in the same
order as the schema. (2a) and (2b) are variant forms, but still very recognisable as argument
texts. Some arguments, however, are less obvious, and need to be extracted from the text
and reconstructed to show their argument form.
The customary way do this, both in logic and critical thinking, is to place the reasons
(premises) in a list, and to separate them from the conclusion by a horizontal line. So, with
some abbreviation, you could set out argument (2) as follows:
(2.1) R1 If the Earth were flat, ships sailing away would appear to get smaller . . .
R2 But they appear to drop out of sight . . . .
C: The Earth cannot be flat.
Reconstructing an argument to give it this standard form helps to make the reasoning clear,
206
and assist with its subsequent evaluation. It also gives the critic a convenient way of referring
to the components of the argument. Instead of having to write out a premise or conclusion
each time it is mentioned, the labels can be used instead. For example:
R1 and R2 are given as reasons for C.
It may seem a lot of trouble to go to when an argument is as brief, and as recognisable, as
(2). But with more complex reasoning, which you will encounter as you progress through the
chapters, this kind of formal reconstruction is a valuable tool.
20.2.2 Argument and dispute

Of course, not everyone has to accept an argument when it is put to them. Sometimes, even
when you have given your reasons, people may still disagree with your conclusion; or they
might challenge the claims you have given as reasons. This may have been quite common
hundreds of years ago when the first ‘round-earthers’ set out to persuade sceptics that the
world was spherical.
(3) Kris: Did you know that Earth is a huge ball hanging in space?
Bart: Don’t be ridiculous. The Earth is flat. Can’t you see that?
Kris: It can’t be flat. If you just let me explain—
Bart: There’s nothing to explain. Just use your eyes.
Kris: I am using my eyes, and they tell me the Earth is round. Look at that ship.
Soon you will only see the top of the mast. That wouldn’t happen if the
Earth was flat.
Bart: (in a lowered voice): Well let me tell you something. If you go around talking
this kind of nonsense, someone is going to lock you up and throw away the
key. Or worse.
Kris: But just listen—

Bart: No, you listen. The Earth is flat, and so is the sea.
Kris: It’s round.
Bart: Flat. F-L-A-T, flat!
Kris: ROUND . . .
Texts (2) and (3) are both examples of argument, though in different senses of the word. This
is due to the fact that in English the word ‘argument’ has two distinct meanings, as seen in
the following dictionary definition:
argument (noun)
1 a reason or reasons supporting a conclusion; a case made for or against a point of view.
2 a debate or dispute, especially a heated one; disagreement, quarrel, row.
As you can see, example (2) is an argument of the first sort. It can be called a reasoned
argument to distinguish it from mere dispute. Example (3) is quite obviously an argument in
the second sense – a dispute, bordering on a quarrel. There is very little reasoned argument
going on in (3). Kris tries to explain his position, but his opponent shouts him down. The two
speakers are mostly just exchanging opinions, without giving any serious reasons to back
them up. There is no single conclusion to which the text as a whole is leading.
Interestingly, the English language is unusual in giving these two senses to the same word.
In Spanish, for example, argumento has the reason-giving sense; the disputational sense is
given by discusión, controversia or disputa. 207
However, it would be wrong to think that the two meanings of ‘argument’ are completely
unrelated. As stated at the beginning of the chapter, arguments typically exist to persuade,
and it is clear that in a dispute like (3) each of the participants is trying to change the mind
of the other. For (2) there is no such context available, but the most likely context would be
some real or imagined debate. Why else would its proponent feel any need to give reasons
or evidence to support the claim? Arguments in the reason-giving sense occur typically in the
face of opposition, i.e. in the context of dispute or disagreement.
Moreover, most arguments of the disputational kind have some elements of reason-giving in
them. Even in (3), which is predominantly a quarrel, both participants draw upon the evidence
of their senses (i.e. sight) to support their differing conclusions. Bart says that anyone can
see the Earth is flat; that all you need to do is use your eyes. Kris responds that his eyes tell a
different story, and tries to explain by bringing in the evidence of disappearing ships.
If you wanted to represent Bart’s contribution to the dispute as a standard argument, you
could summarise it as follows:
(B) R The Earth looks flat. (You can see that.)
C The Earth is flat.
No one today would rate this a good argument. Though the premise is true – the Earth does
look flat – the conclusion is false. However, in the fictional dispute between Bart and Kris
it was not known that the Earth was spherical. Nobody had sailed round it or seen it from
space. But even without that knowledge it can still be seen that Bart’s argument is invalid
because the evidence – though true – does not justify the inference from R to C. In general,
the premise that something looks to have some property –

flatness, roundness, etc. – is not a sufficient reason to conclude
that it has that property. Appearances can be deceptive. A stick
looks bent when part of it is under water but that does not mean it
is bent. The same goes for the apparent flatness of the Earth. The
surface of any sphere with a circumference of 40 000 kilometres will
look flat to someone on the surface. That does not make it flat.
To accept (B) as a valid argument it would have to be assumed that
things are just the way they look. Since we can’t assume that, we
have to reject Bart’s reasoning. Figure 20.2
20.2.3 Analysing and evaluating argument

We have seen then that an argument consists of reasons (premises) that are offered in
support of a conclusion. In a good argument the reasons do justify the conclusion. In a poor
one they do not. Evaluating argument means distinguishing good examples from bad ones.
Much of the remaining content of this book is about the critical evaluation of reasoned
argument. Here is a first taste.
We have established that Bart’s argument (B) is a bad one. Once the solitary premise
and conclusion are identified it is very easy to see that the grounds are inadequate. But
not all arguments that you hear will be as simple or as obviously defective as (B).
208
Compare (B) with the argument Kris attempts in the course of the dispute:
(K) I am using my eyes, and they tell me the Earth is round. Look at that ship.
Soon you will only see the top of the mast. That wouldn’t happen if the Earth
was flat.
Is (K) a good argument, or not? Is it better than (B)? Would it persuade you that the
Earth’s surface was curved if you had previously believed it was flat?
Commentary
(K) might seem like a reasonable argument, because we now accept that the Earth is
roughly spherical. But, as we also know from history, arguments like (K) – or (2), which
makes the same point – were not enough to convince the general public straight away.
People needed more reasons if they were going to give up a belief that had persisted
for centuries. Judged critically it becomes clear that (K) is has much the same weakness
as (B) because, like (B), it also argues from appearances. If the flat appearance of the
Earth does not mean that it is flat, then can we be sure that the appearance of ships
dropping out of sight does not prove, on its own, the surface of the Earth is curved. It
could be an optical illusion; some kind of mirage perhaps. It isn’t a mirage: it is perfectly
true both that ships appear to sink and that the Earth’s curvature is the reason. But
today we know that fact independently of the argument. The single reason given in (K)
does not, on its own, establish its conclusion.
ACTIVITY 20.2
Find a short argument published in a newspaper or magazine or on the internet. Copy it

down and underline its conclusion.
20.2.4 Strengthening reasoned argument

For an effective argument we usually need more than one reason, or a single piece of
circumstantial evidence. Imagine you were sent back in time several hundred years and had
to convince people that the Earth was not flat. What would you take with you: pictures from
space; stories of people who have sailed round the world? These would seem like a good
start. Armed with such evidence, you could supplement (2) and thereby make it stronger, for
example:
(4) Ships appear to sink lower and lower in the water the further they are from land.
But they cannot actually be sinking, or they would not come back. Also, sailors have
proved that if you set off in one general direction, for example east or west, and keep
going, you eventually arrive back where you started from. These facts show that the
Earth cannot be flat. Besides, photographs have been taken from space that show the
Earth’s curvature.
Here four reasons are given in support of the conclusion. The conclusion is introduced by
the phrase: ‘These facts show that’ – another way of saying, ‘so’ or ‘therefore’. Three of the 209
reasons are given first; then the conclusion; then a further, seemingly indisputable premise.
So the structure of the argument could be depicted as follows:
(4.1) P1 Ships appear to sink as they sail away.

P2 They can’t actually be sinking or they wouldn’t come back.
P3 Ships can sail in one direction but return to their starting point.
P4 Pictures from space show the curvature of the Earth.
C The Earth cannot be flat.
Obviously (4) is a much stronger argument than (2). Whether it actually convinces its
audience will still depend on their willingness to accept the evidence. But if they understand
and believe the claims you are making, then it would be irrational of them not to accept the
conclusion also.
Of course, the ‘if’ is a big one. In all probability the audience from that time would not accept
your claims because they would not understand them. What could ‘pictures from space’
mean to a fourteenth-century fisherman? They would lock you up or declare you mad (or
both), and carry on believing what they had always believed, and could see with their own
eyes: a flat Earth surrounded by flat sea.
This is why ‘claim’ is the right word for the statements that appear in arguments. Some of the
claims made in an argument may be known facts, but others may be forecasts, suggestions,
beliefs or opinions. Claims may also be false. It is perfectly possible to construct an argument
from false claims, either out of ignorance, or to misinform the audience. We hear all too
much of this practice these days under the heading of ‘fake news’. (That is probably what
people hundreds of years ago would have suspected you of doing, as they slammed the
dungeon door.)
ACTIVITY 20.3
1 Think of a suitable conclusion that you could add to the following to make it into an
argument:
Police forces the world over face a dilemma. On top of dealing with murders and
other major incidents, they have to divide their limited time and finite resources
between tackling minor crimes such as shoplifting and street robbery, and traffic
offences such as speeding or careless driving. Of course, the consequences of
speeding can be as bad or worse than the theft of a wallet or a mobile phone. They
can be fatal. But there is a big difference of another sort. The thief intends to do harm
and to deprive people of their rightful property, whereas any harm that is done by a
car-driver, however serious, is usually accidental.
20.2.5 Grammatical note

It was noted in Chapter 14 that claims can sometimes have the form of rhetorical questions,
or other sentence types such as instructions, or exclamations. When interpreting and
analysing an argument in which one or more of the sentences is not a statement, but is
making a claim nonetheless, it is good practice to transform it into a statement. For example,
when Bart says; ‘Can’t you see that?’ he is not really expecting an answer as one would
210 to a normal question. He is effectively saying: ‘You can see that (the Earth is flat)’. And he
is making the same point when he uses the grammatical command: ‘Just use your eyes’.
The reason why we should interpret premises as claims, statements, or assertions is that
ultimately we are assessing their truth, and the only expressions that can literally be true or
false are statements.
20.2.6 Identifying arguments

Before an argument can be analysed and evaluated it must first be established that it is
an argument. This can be harder than it sounds if the argument is expressed in natural
language, and in a non-standard form; and if the context in which it is embedded is a longer
text of which argument is only a part. It is also made harder if the argument is a poor one. In
a good argument the conclusion can be seen to follow from the reasons. In a bad argument
it does not follow – or at least it does not follow convincingly. Establishing that some piece
of text is an argument then comes down to deciding whether or not the author meant or
intended one of the claims to be a conclusion, and the others to be reasons for it. Judging an
author’s intention, from a text alone, is not a very exact science!
Matters are made easier if the conclusion or reasons are marked by indicators – ‘therefore’,
‘so’, ‘since’, ‘because’, etc. However, these connectives have other functions in the language
besides signalling argument. They occur frequently, for example in explanations, but in other
expressions too. Just finding two sentences joined by ‘so’ or ‘since’ does not automatically
identify a reasoned argument. Think of the words of the song:
(5) Since I was twelve

I’ve done my homework
There is no argument here. ‘Since’ in the phrase means ‘ever since’, which is different from
the meaning it has in front of a premise. For another example, compare,
(6) Julia is ambitious; so is her daughter.
With
(7) Julia is ambitious so her daughter will be, too.
There is no argument in (6). The word ‘so’ in that context means ‘also’, not ‘therefore’. But
with very little change (7) is a simple argument, an inference from Julia’s alleged ambitious
nature to prediction about her daughter. The point to take from this is that although the
argument indicators can help to identify possible arguments in a text, it cannot be assumed
that wherever the words occur they have that function.
Besides, as observed earlier in the chapter, there are plenty of examples of natural-
language arguments which contain no connectives. An argument may be expressed by a
plain sequence of statements (or claims), yet obviously not every sequence of claims is an
argument. There it is no argument here, for example:
(8) Photographs from space show the Earth’s surface as curved. The curvature does not
show when a photograph is taken from ground level.
Being able to identify arguments means being able to distinguish confidently between
211
argument and non-argument. It is not always as straightforward as it might seem.
20.2.7 The ‘therefore/so’ test

Despite what was said just now about such connectives having multiple meanings, they do
at least provide a useful test for ruling out some sets of sentences as arguments. The test
is performed by inserting ‘therefore’ or ‘so’ between the sentences and asking: does the
resulting expression make sense? If it does not make sense, then there is no argument. Here
is the test applied to (8):
• Photographs from space show the Earth’s surface as curved. Therefore the curvature
does not show when a photograph is taken from ground level.
• The curvature does not show when a photograph is taken from ground level, so
photographs from space show the Earth’s surface as curved.
Neither of these makes sense. So (8) is not an argument. Unfortunately it is less informative
when the result is positive: if inserting ‘therefore’ or ‘so’ or ‘since’ does make sense it does
not automatically mean that the resulting expression is an argument. The usefulness of the
test is in eliminating non-arguments, leaving the others as possible arguments.
The same test can be applied to the next example, only as there are more claims there will
be more rearrangements to try out.
(9) Completed tax forms and payments must be received by 31 July. Late payment may
result in a fine not exceeding $100. Your payment did not reach the tax office until
12 August.
There are three possible configurations of the sentences, using ‘so’ or ‘therefore’:
• Completed tax forms and payments must be received by 31 July. Late payment may
result in a fine not exceeding $100. Therefore your payment did not reach the tax office
until 12 August.
• Late payment may result in a fine not exceeding $100. Your payment did not reach the tax
office until 12 August. So completed tax forms and payments must be received by 31 July.
• Completed tax forms and payments must be received by 31 July. Your payment did
not reach the tax office until 12 August. Therefore late payment may result in a fine not
exceeding $100.
In each rearrangement the attempt to use an argument indicator sounds unnatural, which
indicates that none of the sentences is the kind of claim that could follow from the others in
the way that a conclusion follows from reasons.
Using the ‘therefore/so’ test, and the definition of an argument as reasons given for a
conclusion, decide which of the following could be interpreted as arguments.
For those that are arguments, identify the conclusion and note what kind of claim it is.
Lastly, discuss how well supported you think the conclusion is, given the reasons.
1 The Tokyo train leaves at 4:24. It can take up to 40 minutes to get to the station if the
traffic is bad. We should leave for the station by 3:40.
212
2 Raisa is the only person with a key to the safe. The police are bound to treat her as a
suspect. The money went missing when she was in the building on her own.
3 You are likely to get a fine. Completed tax forms and payments must be received by
31 July and people who miss the deadline are usually fined $100. Your payment did
not reach the tax office until 12 August.
4 From the fifteenth century European sailors reached the lands of the east by sailing
west. Those who sailed on and survived eventually arrived back in Europe. When
they claimed they had sailed around the world, few people believed them.
5 There are only three possible causes of the leak in your system: the pump could be
worn, a hose could be split or one of the connections could be loose. I’ve checked
the hoses and tightened all the connections, but the machine still leaks.
Commentary
(1) can be understood as an argument. The conclusion, which is at the end, is a
recommendation. This also is a useful clue: recommendations are often accompanied
by reasons. Here there are two: the time of the train’s departure and the possibility of
a 40-minute journey to the station. If they are both true, then clearly they justify the
conclusion.
(2) The same applies: the conclusion is a prediction that the police will (definitely)
suspect Raisa, firstly because she is the only key-holder, and secondly because she
was alone in the building. The argument is perhaps not quite as solid as (a). Do police
always treat people as suspects in these circumstances? The words ‘bound to . . .’ make
the conclusion a very strong claim. Even if both premises are true, there may be other
factors – CCTV footage for instance – that show Raisa was nowhere near the safe, and
therefore make it less than definite that she will be treated as a suspect. Nonetheless, it
is a safe interpretation of the text to read it as an argument – good or bad.
(3) is also an argument. The conclusion is another prediction (of sorts). You could
also have described it as a statement of probability: ‘You are likely to get a fine.’ The
reasoning for the conclusion is that payment did not reach the tax office until 12
August, together with the second sentence which establishes that the payment was
late and that late payment usually results in a fine. The argument is quite sound,
mainly because the conclusion is a fairly weak claim. If fines are usual for lateness, then
a fine is likely.
(4) is not an argument. None of the three sentences makes sense with ‘therefore’ in
front of it, e.g. ‘From the fifteenth century European sailors reached the lands of the
east by sailing west. Those who sailed on and survived eventually arrived back in
Europe. Therefore when they claimed they had sailed around the world, few people
believed them.’ The connective that makes most sense is ‘but’, not ‘therefore’. None of
the claims is a conclusion drawn from either or both of the other two; and it is the same
whichever order the claims are placed in.
(5) is not an argument either – at least not an explicit one – because, as in (4) – none 213
of its actual sentences is a natural conclusion. However, (5) does point towards a
conclusion, even though it is not stated. In fact there is really only one conclusion
that you could draw from (5) – that the pump must be worn – because both the
other possibilities are ruled out. What we can say about (5) is that it is not a complete
argument. It is left to you (the reader or listener) to draw a conclusion – though in this
case it leaves you in little doubt as to what the conclusion would be if the author had
meant (5) to be an argument. We could say therefore that (5) is an implicit argument, or
that it has an implicit conclusion.
ACTIVITY 20.5
Explain why the ‘so/therefore test’ is useful but not the last word when identifying
arguments.
20.3 Identifying arguments in dialogue

Earlier in the chapter there was some discussion of arguments in the disputational sense.
Read the following passage – preferably aloud with a partner, taking a part each – and then
answer the question that follows.
SCENE: a table for two in a restaurant

Anita: What are you going to have?
(Sound of a mobile phone)
Bara: Just a minute. I’ve got a message.
Anita: Not another!
Bara: I need to answer it.
Anita: Why don’t you just switch it off? Restaurants are places for conversation. They’re
so antisocial, those things.
Bara (texting at the same time): You wouldn’t say that if you had one. You’d be on it all
the time.
Anita: I wouldn’t have one as a gift.
Bara: Yes, you would. I’ll give you my old one.
Anita: Keep it. I’m better off without it. In fact the whole world would be better off if the
wretched things had never been invented.
Bara: How do you work that out?
Anita: Well for a start you can’t sit anywhere quietly anymore without having to listen
to one end of someone else’s shouted conversation. Secondly, they’re a health risk
because they pour out microwaves that cook your brain. Thirdly, they distract
drivers and cause road accidents. So, like I said: they do more harm than good.
Bara: You just can’t say that. No one thinks they are a health risk anymore. They don’t
214 distract drivers unless the drivers are stupid enough to have them switched on
in the car. Not everybody shouts into their phones, and not everyone finds them
irritating. They help people to keep in touch. They save lives in emergencies. They
access information when you need it. What more do you want?
Anita (shouting): I’m sorry, but people do shout into them. They don’t even know they’re
doing it. And they do use them when they’re driving, whatever the law does to
stop them. If someone smashed into you because she was reading a text message,
you would soon change your tune.
Bara: Hang on, you’re blaming an inanimate object for what people do with it. Of course
there are always some idiots who misuse stuff. It’s like guns, isn’t it? Guns don’t
kill, it’s the people who fire the guns. You’re making the same mistake.
Anita: I’m not making a mistake. The machines are to blame. I agree, a gun can’t kill
you until someone fires it, but you can’t get shot either if there are no guns to do
it with. And people couldn’t be distracted by their phones when they’re driving
if there were no mobile phones. And you wouldn’t still be sending that text and
spoiling our lunch.
Bara: That’s just silly. You’ve lost that one.
Anita: No I haven’t.
Bara: You have. You’re just old-fashioned, so you can’t see the value of the new
technology.
Anita: I’m not old-fash—
Bara: Be quiet, and let me finish this message. I’ll be quicker if you just stop talking.
Is the conversation above just a quarrel, or is there reasoned argument going on here
as well? If there is, identify some examples of arguments propounded by one or other of
the participants.
Commentary
Overall, this conversation is a quarrel, and parts of it are no more than exchanges of
opinion, laced with mild insults. But in the course of the exchange there are examples
of developed, reasoned arguments as well, coming from both sides.
The clearest example is Anita’s first long paragraph. This is practically a standard
argument, with three numbered reasons and a conclusion signalled by ‘so’. Bara
responds with a counter-argument. This gives three reasons which challenge or
contradict Anita’s claims, then two further reasons (the value of keeping people in
touch, and of saving lives in emergencies) to support a position which is the complete
opposite of Anita’s. Bara’s conclusion is expressed by the first sentence of the
paragraph: ‘You just can’t say that.’ In other words: ‘It is not true that mobile phones
do more harm than good’ (as Anita has just asserted). In natural-language arguments,
conclusions may not always be spelled out in full, as they are in a standard argument.
Expressions such as ‘Yes’, ‘No’ and ‘You’re wrong!’ can be understood as conclusions, if
it is clear what they refer to, and they are supported by reasons. 215
In the three paragraphs that follow we see Anita and Bara each trying to reinforce their
arguments with further reasons and objections. Then, as their tempers begin to fray,
they go back to mere quarrelling and personal remarks.
In some textbooks the impression is given that critical thinking is concerned only with
arguments in the standalone, reason–conclusion form, and not with argument in the
sense of dispute. But as suggested earlier, we miss something important about the
practical meaning and purpose of argument if we ignore the most obvious context in
which it occurs. Much of our reasoning – perhaps all of it – arises in or from differences
of belief or opinion. An argument that the Earth is not flat makes practical sense only if
someone – past or present – thinks that it is flat, or needs proof that it is. A defence of
the use of mobiles phones is unlikely to be called for unless there is understood to be
some opposition or objection to it.
20.4 Writing your own reasoned arguments

So far we have looked at the elements of reasoning and how they work together to form
arguments. We have also considered criteria that can be applied in deciding whether
reasoning is good or not. If you want to improve your skills as a critical thinker, it is important
that, when you produce your own reasoning, you aim to apply the same standards as you
would when assessing other people’s reasoning.
The next three Activities are designed to help you start writing your own reasoned
arguments. However, you should also remember that working through the Activities and
questions in this textbook is not the only way to develop as a critical thinker: you can apply
critical thinking skills and techniques whenever you are attempting to present a point of
view, either verbally or in writing.
ACTIVITY 20.7
Think back over what you have learnt up till now about good and poor reasoning. As a
group discuss what you think makes reasoning persuasive. Can you identify any writers
or speakers who you think are persuasive? Is it their use of reasoning that is persuasive
or is it something else?
ACTIVITY 20.8
Think of one or two reasons that could be used to support each of the following claims,
making them into arguments.
a It is wrong to charge foreign students higher fees than other students.
b Private cars with fewer than four occupants should be banned from city centres.
c The stars of football, baseball and other popular sports deserve every cent of the
millions that they are paid.
ACTIVITY 20.9
Use newspapers or the internet to find a current issue that interests you. Write your own
216
short argument for or against a conclusion in the form, ‘The Government should ...’
Your argument should contain at least three reasons as well as the conclusion.
Follow the steps below to help you develop your argument.
1 Clearly state your conclusion. (You can use argument indicators and you could put
your conclusion either at the beginning or at the end of the argument.)
2 Make sure your reasons link to your conclusions.
3 Decide whether you should make your conclusion stronger or less strong. (If it is less
strong, it may link more closely to your reasons.)
Summary
The topic of the chapter was argument. ■ Argument indicators were discussed as means of
identifying and testing texts for the presence of
■ An argument was defined as a complex construction
argument.
in which a conclusion is claimed to follow from stated
■ The distinction between reasoned argument and a mere
reasons (alternatively referred to as premises).
dispute was discussed.
■ A good argument was defined as one in which the
■ The critical activities of simple analysis and evaluation
premises are true and follow from the reasons, meaning
that if the reasons are true then the conclusion cannot were introduced.
be false. ■ The activity of writing reasoned argument was
■ Standard and non-standard forms of argument were introduced

examined, with examples.
End of chapter questions

1 For each of the following short texts, decide whether or not it is an argument. If it is an
argument, say which claim is the conclusion.
a Since the last earthquake in California, engineers have been investigating what
happens to man-made structures during a large seismic event. They were surprised
that a section of the Bay Bridge, which connects Oakland to San Francisco, fell like a
trapdoor. They also discovered that in some of the older double-decker freeways, the
joints that connect the lower column to the upper column may be suspect.
b The public should not expect the safety of drugs to be guaranteed by animal testing.
Aspirin, which is a safe and effective painkiller for most humans, is fatal to the
domestic cat. Penicillin poisons guinea pigs. These examples show that different
species react to drugs differently.
c If more cash machines start making a fixed charge for each withdrawal, people who
draw small amounts will pay more in the long run than those who make larger but
fewer withdrawals. People with low incomes tend to make smaller withdrawals, but
are more willing to look for machines that don’t charge.
d Short range air-travel may be cheap but it makes no sense. Flying is responsible for ten
times more carbon emissions than rail travel. And it is immeasurably more stressful for
passengers. Also trains take you to the heart of a city, not to some far-flung airport.
e You will be lucky if you get to see even one solar eclipse in your lifetime. The
phenomenon occurs only when the Moon crosses between the Earth and the
Sun. This happens roughly once a month, but because the Moon’s orbit is angled
differently from the Earth’s, it is usually too high or too low in the sky to pass right in 217
front of the Sun. Even when this does happen, every eighteen months or so, it is only
visible in certain parts of the world, different every time.
2 Which one of the following is an argument? Why are the others not?
a Although some people think that microwave radiation from mobile phones may
damage the user’s brain, no hard evidence has been found of harmful effects on
health. ‘Hands-free’ kits are supposed to make the phones safer because the user
does not need to hold the phone as close to their head.
b Land animals such as tortoises often spread to small islands aboard ships. However,
the tortoises on the Galapagos Islands must have arrived by some other means. The
first people to visit the Galapagos found tortoises already there.
c In most countries cyclists are not legally required to wear a helmet, in the way car-
drivers are required to wear a seat belt. This inconsistency in the law persists despite
the growing number of road accidents in which a cyclist is injured or even killed.
Exam-style questions
People are always fascinated by the question of whether there are extra-terrestrial life forms.
By this they usually mean ‘aliens’ as portrayed in popular science fiction series like Star Trek.
However, if there were other life forms of this type in outer space they would have visited us or
they would have managed to send some sort of radio signal which we would have intercepted.
As this has not happened, it looks as if we are the only life form in the universe. So the dream of
meeting people from other galaxies needs to be dismissed as a childish fantasy.
Adapted from CI Thinking Skills 8435/1 Paper 1 October/November 2002
State, using the author’s words, the conclusion of the above argument.

Cambridge International As & A Level Thinking Skills Sample

Uploaded by

Copyright:

Available Formats

Cambridge International As & A Level Thinking Skills Sample

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cambridge International As & A Level Thinking Skills Sample

Uploaded by

Copyright:

Available Formats

Mark Dawes, Ruth Matthews,

Andrew Roberts and Geoff Thwaites

Cambridge International AS & A Level

Cambridge International AS & A Level

© Cambridge University Press 2018

This publication is in copyright. Subject to statutory exception

Printed in the XXXX by XXXX

notice to teachers in the uk

15 Assessing claims 130 22 Interpretation 234

4 19 Uses of evidence: inference and explanation 186 26 Longer texts 289

21 Analysing argument  218

1.1 Routine activities or problems

1.2 What skills are needed to solve problems?

1.2.1. Understanding the problem that needs to be solved

1.2.2. Processing the information that is available to solve the problem

1.2.3. Analysing data and representing it in different ways

1.2.4. Developing a problem that has already been solved to consider

EXAMPLE ACTIVITY 1.1

3 Consider more complex situations:

EXAMPLE ACTIVITY 1.2

EXAMPLE ACTIVITY 1.3

Suppose that we are given the following additional information:

Section of journey Must be Time required Latest start

1.3 Selecting a problem-solving strategy

There are two systematic ways of searching.

EXAMPLE ACTIVITY 1.4

EXAMPLE ACTIVITY 1.5

End of chapter questions

18.1 What are statistics?

18.1.1 Sources of data

18.1.2 Collecting data

18.1.4 Sample size

EXAMPLE ACTIVITY 18.1

example data points

Figure 18.1 Figure 18.2

18.2.1 Trends and predictions

EXAMPLE ACTIVITY 18.2

New houses sold 1000 units

Strong positive Weak positive Strong negative

Weak negative Moderate negative No

18.2.3 Relations between trends

EXAMPLE ACTIVITY 18.1

Consider an imaginary report published by a city police authority on a five-year initiative

Date EUR USD

EXAMPLE ACTIVITY 18.5

Figure 18.8 Euro–dollar exchange rates

EURO PLUNGES TO 5-WEEK LOW AGAINST DOLLAR

Some problems with presentation

EXAMPLE ACTIVITY 18.6

18.3.1 Multiple trends

Mar. 2012 Mar. 2017 % change

EXAMPLE ACTIVITY 18.7

EXAMPLE ACTIVITY 18.8

Country Average income 200 metres national

End of chapter question

Exam style questions

Number of items on sale that were less

To save money on your weekly shopping, come to Dasamart!

20.1 Recognising an argument

15 Assessing claims 130 22 Interpretation 234

4 19 Uses of evidence: inference and explanation 186 26 Longer texts 289

21 Analysing argument 218