1
00:00:00,714 --> 00:00:03,810
Back in 2003,

2
00:00:03,834 --> 00:00:06,343
the UK government carried out a survey.

3
00:00:07,494 --> 00:00:10,643
And it was a survey that measured
levels of numeracy

4
00:00:10,667 --> 00:00:11,904
in the population.

5
00:00:11,928 --> 00:00:13,571
And they were shocked to find out

6
00:00:13,595 --> 00:00:16,959
that for every 100 working age
adults in the country,

7
00:00:16,983 --> 00:00:20,484
47 of them lacked Level 1 numeracy skills.

8
00:00:20,892 --> 00:00:25,004
Now, Level 1 numeracy skills --
that's low-end GCSE score.

9
00:00:25,410 --> 00:00:28,658
It's the ability to deal with fractions,
percentages and decimals.

10
00:00:28,682 --> 00:00:33,310
So this figure prompted
a lot of hand-wringing in Whitehall.

11
00:00:33,334 --> 00:00:34,962
Policies were changed,

12
00:00:34,986 --> 00:00:36,708
investments were made,

13
00:00:36,732 --> 00:00:39,770
and then they ran
the survey again in 2011.

14
00:00:39,794 --> 00:00:41,999
So can you guess
what happened to this number?

15
00:00:44,021 --> 00:00:45,465
It went up to 49.

16
00:00:45,489 --> 00:00:46,938
(Laughter)

17
00:00:46,962 --> 00:00:49,411
And in fact, when I reported
this figure in the FT,

18
00:00:49,435 --> 00:00:51,106
one of our readers joked and said,

19
00:00:51,130 --> 00:00:54,891
"This figure is only shocking
to 51 percent of the population."

20
00:00:54,915 --> 00:00:57,201
(Laughter)

21
00:00:57,225 --> 00:01:00,382
But I preferred, actually,
the reaction of a schoolchild

22
00:01:00,406 --> 00:01:03,501
when I presented
at a school this information,

23
00:01:03,525 --> 00:01:05,056
who raised their hand and said,

24
00:01:05,080 --> 00:01:07,596
"How do we know that the person
who made that number

25
00:01:07,620 --> 00:01:09,435
isn't one of the 49 percent either?"

26
00:01:09,459 --> 00:01:10,713
(Laughter)

27
00:01:10,737 --> 00:01:14,787
So clearly, there's a numeracy issue,

28
00:01:14,811 --> 00:01:16,921
because these are
important skills for life,

29
00:01:16,945 --> 00:01:20,812
and a lot of the changes
that we want to introduce in this century

30
00:01:20,836 --> 00:01:23,277
involve us becoming
more comfortable with numbers.

31
00:01:23,301 --> 00:01:25,149
Now, it's not just an English problem.

32
00:01:25,173 --> 00:01:30,103
OECD this year released some figures
looking at numeracy in young people,

33
00:01:30,127 --> 00:01:32,907
and leading the way, the USA --

34
00:01:32,931 --> 00:01:37,601
nearly 40 percent of young people
in the US have low numeracy.

35
00:01:37,625 --> 00:01:38,922
Now, England is there too,

36
00:01:38,946 --> 00:01:44,479
but there are seven OECD countries
with figures above 20 percent.

37
00:01:44,503 --> 00:01:47,262
That is a problem,
because it doesn't have to be that way.

38
00:01:47,286 --> 00:01:49,294
If you look at the far end of this graph,

39
00:01:49,318 --> 00:01:52,278
you can see the Netherlands and Korea
are in single figures.

40
00:01:52,302 --> 00:01:56,718
So there's definitely a numeracy
problem that we want to address.

41
00:01:57,510 --> 00:02:00,440
Now, as useful as studies like these are,

42
00:02:00,464 --> 00:02:05,864
I think we risk herding people
inadvertently into one of two categories;

43
00:02:05,888 --> 00:02:07,664
that there are two kinds of people:

44
00:02:07,688 --> 00:02:12,037
those people that are comfortable
with numbers, that can do numbers,

45
00:02:12,061 --> 00:02:14,297
and the people who can't.

46
00:02:14,321 --> 00:02:16,422
And what I'm trying
to talk about here today

47
00:02:16,446 --> 00:02:19,488
is to say that I believe
that is a false dichotomy.

48
00:02:19,512 --> 00:02:21,380
It's not an immutable pairing.

49
00:02:21,404 --> 00:02:25,052
I think you don't have to have
tremendously high levels of numeracy

50
00:02:25,076 --> 00:02:26,804
to be inspired by numbers,

51
00:02:26,828 --> 00:02:29,937
and that should be the starting point
to the journey ahead.

52
00:02:30,387 --> 00:02:34,698
And one of the ways in which
we can begin that journey, for me,

53
00:02:34,722 --> 00:02:36,448
is looking at statistics.

54
00:02:36,472 --> 00:02:39,967
Now, I am the first to acknowledge
that statistics has got somewhat

55
00:02:39,991 --> 00:02:41,309
of an image problem.

56
00:02:41,333 --> 00:02:42,380
(Laughter)

57
00:02:42,404 --> 00:02:43,936
It's the part of mathematics

58
00:02:43,960 --> 00:02:47,019
that even mathematicians
don't particularly like,

59
00:02:47,043 --> 00:02:51,055
because whereas the rest of maths
is all about precision and certainty,

60
00:02:51,079 --> 00:02:53,363
statistics is almost the reverse of that.

61
00:02:53,793 --> 00:02:58,448
But actually, I was a late convert
to the world of statistics myself.

62
00:02:58,472 --> 00:03:00,554
If you'd asked my undergraduate professors

63
00:03:00,578 --> 00:03:05,337
what two subjects would I be least likely
to excel in after university,

64
00:03:05,361 --> 00:03:08,128
they'd have told you statistics
and computer programming,

65
00:03:08,152 --> 00:03:11,091
and yet here I am, about to show you
some statistical graphics

66
00:03:11,115 --> 00:03:12,317
that I programmed.

67
00:03:12,745 --> 00:03:14,500
So what inspired that change in me?

68
00:03:14,524 --> 00:03:18,172
What made me think that statistics
was actually an interesting thing?

69
00:03:18,196 --> 00:03:20,462
It's really because
statistics are about us.

70
00:03:20,869 --> 00:03:23,451
If you look at the etymology
of the word statistics,

71
00:03:23,475 --> 00:03:26,084
it's the science of dealing with data

72
00:03:26,108 --> 00:03:28,538
about the state or the community
that we live in.

73
00:03:28,562 --> 00:03:31,916
So statistics are about us as a group,

74
00:03:31,940 --> 00:03:33,615
not us as individuals.

75
00:03:33,639 --> 00:03:35,109
And I think as social animals,

76
00:03:35,133 --> 00:03:39,077
we share this fascination about how
we as individuals relate to our groups,

77
00:03:39,101 --> 00:03:40,489
to our peers.

78
00:03:40,513 --> 00:03:43,623
And statistics in this way
are at their most powerful

79
00:03:43,647 --> 00:03:44,948
when they surprise us.

80
00:03:45,477 --> 00:03:48,684
And there's been some really wonderful
surveys carried out recently

81
00:03:48,708 --> 00:03:50,422
by Ipsos MORI in the last few years.

82
00:03:50,446 --> 00:03:53,154
They did a survey of over
1,000 adults in the UK,

83
00:03:53,178 --> 00:03:56,958
and said, for every 100 people
in England and Wales,

84
00:03:56,982 --> 00:03:58,852
how many of them are Muslim?

85
00:03:58,876 --> 00:04:01,522
Now the average answer from this survey,

86
00:04:01,546 --> 00:04:04,958
which was supposed to be representative
of the total population, was 24.

87
00:04:04,982 --> 00:04:08,658
That's what people thought.

88
00:04:08,682 --> 00:04:12,321
British people think 24 out of every 100
people in the country are Muslim.

89
00:04:12,345 --> 00:04:16,755
Now, official figures reveal
that figure to be about five.

90
00:04:17,732 --> 00:04:21,719
So there's this big variation
between what we think, our perception,

91
00:04:21,743 --> 00:04:23,781
and the reality as given by statistics.

92
00:04:23,805 --> 00:04:25,349
And I think that's interesting.

93
00:04:25,373 --> 00:04:28,663
What could possibly be causing
that misperception?

94
00:04:29,212 --> 00:04:31,066
And I was so thrilled with this study,

95
00:04:31,090 --> 00:04:34,570
I started to take questions out
in presentations. I was referring to it.

96
00:04:34,594 --> 00:04:35,812
Now, I did a presentation

97
00:04:35,836 --> 00:04:38,146
at St. Paul's School for Girls
in Hammersmith,

98
00:04:38,170 --> 00:04:40,310
and I had an audience rather like this,

99
00:04:40,334 --> 00:04:44,202
except it was comprised entirely
of sixth-form girls.

100
00:04:44,226 --> 00:04:46,622
And I said, "Girls,

101
00:04:47,598 --> 00:04:52,141
how many teenage girls do you think
the British public think

102
00:04:52,165 --> 00:04:53,913
get pregnant every year?"

103
00:04:53,937 --> 00:04:56,613
And the girls were apoplectic when I said

104
00:04:57,453 --> 00:05:01,366
the British public think that 15
out of every 100 teenage girls

105
00:05:01,390 --> 00:05:02,683
get pregnant in the year.

106
00:05:03,429 --> 00:05:05,660
And they had every right to be angry,

107
00:05:05,684 --> 00:05:08,442
because in fact, I'd have to have
closer to 200 dots

108
00:05:08,466 --> 00:05:10,036
before I could color one in,

109
00:05:10,060 --> 00:05:12,575
in terms of what
the official figures tell us.

110
00:05:12,599 --> 00:05:16,399
And rather like numeracy,
this is not just an English problem.

111
00:05:16,423 --> 00:05:20,927
Ipsos MORI expanded the survey
in recent years to go across the world.

112
00:05:20,951 --> 00:05:23,901
And so, they asked Saudi Arabians,

113
00:05:23,925 --> 00:05:26,446
for every 100 adults in your country,

114
00:05:26,470 --> 00:05:29,343
how many of them are overweight or obese?

115
00:05:30,526 --> 00:05:35,859
And the average answer from the Saudis
was just over a quarter.

116
00:05:36,402 --> 00:05:37,604
That's what they thought.

117
00:05:37,628 --> 00:05:40,196
Just over a quarter of adults
are overweight or obese.

118
00:05:40,220 --> 00:05:45,001
The official figures show, actually,
it's nearer to three-quarters.

119
00:05:45,025 --> 00:05:46,481
(Laughter)

120
00:05:46,505 --> 00:05:48,797
So again, a big variation.

121
00:05:48,821 --> 00:05:53,267
And I love this one: they asked in Japan,
they asked the Japanese,

122
00:05:53,291 --> 00:05:55,251
for every 100 Japanese people,

123
00:05:55,275 --> 00:05:57,876
how many of them live in rural areas?

124
00:05:58,521 --> 00:06:03,422
The average was about a 50-50 split,
just over halfway.

125
00:06:03,446 --> 00:06:07,593
They thought 56 out of every 100
Japanese people lived in rural areas.

126
00:06:07,617 --> 00:06:09,304
The official figure is seven.

127
00:06:10,259 --> 00:06:14,709
So extraordinary variations,
and surprising to some,

128
00:06:14,733 --> 00:06:17,122
but not surprising to people
who have read the work

129
00:06:17,146 --> 00:06:21,538
of Daniel Kahneman, for example,
the Nobel-winning economist.

130
00:06:21,562 --> 00:06:26,654
He and his colleague, Amos Tversky,
spent years researching this disjoint

131
00:06:26,678 --> 00:06:29,823
between what people perceive
and the reality,

132
00:06:29,847 --> 00:06:33,598
the fact that people are actually
pretty poor intuitive statisticians.

133
00:06:33,622 --> 00:06:35,382
And there are many reasons for this.

134
00:06:35,406 --> 00:06:38,521
Individual experiences, certainly,
can influence our perceptions,

135
00:06:38,545 --> 00:06:42,503
but so, too, can things like the media
reporting things by exception,

136
00:06:42,527 --> 00:06:44,223
rather than what's normal.

137
00:06:44,855 --> 00:06:46,981
Kahneman had a nice way
of referring to that.

138
00:06:47,005 --> 00:06:49,090
He said, "We can be blind
to the obvious" --

139
00:06:49,114 --> 00:06:50,752
so we've got the numbers wrong --

140
00:06:50,776 --> 00:06:53,098
"but we can be blind
to our blindness about it."

141
00:06:53,122 --> 00:06:56,388
And that has enormous
repercussions for decision making.

142
00:06:56,412 --> 00:06:59,264
So at the statistics office
while this was all going on,

143
00:06:59,288 --> 00:07:01,200
I thought this was really interesting.

144
00:07:01,224 --> 00:07:03,234
I said, this is clearly a global problem,

145
00:07:03,258 --> 00:07:05,693
but maybe geography is the issue here.

146
00:07:05,717 --> 00:07:09,626
These were questions that were all about,
how well do you know your country?

147
00:07:09,650 --> 00:07:13,643
So in this case, it's how well
do you know 64 million people?

148
00:07:13,667 --> 00:07:16,399
Not very well, it turns out.
I can't do that.

149
00:07:16,423 --> 00:07:17,747
So I had an idea,

150
00:07:17,771 --> 00:07:20,894
which was to think about
this same sort of approach

151
00:07:20,918 --> 00:07:23,023
but to think about it
in a very local sense.

152
00:07:23,047 --> 00:07:24,238
Is this a local?

153
00:07:24,262 --> 00:07:26,203
If we reframe the questions and say,

154
00:07:26,227 --> 00:07:28,349
how well do you know your local area,

155
00:07:28,373 --> 00:07:30,476
would your answers be any more accurate?

156
00:07:31,817 --> 00:07:33,579
So I devised a quiz:

157
00:07:33,603 --> 00:07:35,462
How well do you know your area?

158
00:07:36,454 --> 00:07:38,343
It's a simple Web app.

159
00:07:38,367 --> 00:07:39,550
You put in a post code

160
00:07:39,574 --> 00:07:42,281
and then it will ask you questions
based on census data

161
00:07:42,305 --> 00:07:43,844
for your local area.

162
00:07:44,305 --> 00:07:46,428
And I was very conscious
in designing this.

163
00:07:46,452 --> 00:07:50,561
I wanted to make it open
to the widest possible range of people,

164
00:07:50,585 --> 00:07:53,413
not just the 49 percent
who can get the numbers.

165
00:07:53,437 --> 00:07:55,192
I wanted everyone to engage with it.

166
00:07:55,216 --> 00:07:56,741
So for the design of the quiz,

167
00:07:56,765 --> 00:08:00,380
I was inspired by the isotypes

168
00:08:00,404 --> 00:08:03,006
of Otto Neurath from the 1920s and '30s.

169
00:08:03,030 --> 00:08:07,378
Now, these are methods
for representing numbers

170
00:08:07,402 --> 00:08:09,175
using repeating icons.

171
00:08:09,640 --> 00:08:12,805
And the numbers are there,
but they sit in the background.

172
00:08:12,829 --> 00:08:15,552
So it's a great way
of representing quantity

173
00:08:15,576 --> 00:08:18,560
without resorting to using terms
like "percentage,"

174
00:08:18,584 --> 00:08:19,814
"fractions" and "ratios."

175
00:08:19,838 --> 00:08:21,540
So here's the quiz.

176
00:08:22,310 --> 00:08:23,957
The layout of the quiz is,

177
00:08:23,981 --> 00:08:26,800
you have your repeating icons
on the left-hand side there,

178
00:08:26,824 --> 00:08:29,947
and a map showing you the area
we're asking you questions about

179
00:08:29,971 --> 00:08:31,138
on the right-hand side.

180
00:08:31,162 --> 00:08:32,443
There are seven questions.

181
00:08:32,467 --> 00:08:36,360
Each question, there's a possible answer
between zero and a hundred,

182
00:08:36,384 --> 00:08:37,733
and at the end of the quiz,

183
00:08:37,757 --> 00:08:40,975
you get an overall score
between zero and a hundred.

184
00:08:40,999 --> 00:08:43,083
And so because this is TEDxExeter,

185
00:08:43,107 --> 00:08:45,432
I thought we would have
a quick look at the quiz

186
00:08:45,456 --> 00:08:47,765
for the first few questions of Exeter.

187
00:08:47,789 --> 00:08:49,194
And so the first question is:

188
00:08:49,218 --> 00:08:52,210
For every 100 people,
how many are aged under 16?

189
00:08:52,784 --> 00:08:56,384
Now, I don't know Exeter very well
at all, so I had a guess at this,

190
00:08:56,408 --> 00:08:58,969
but it gives you an idea
of how this quiz works.

191
00:08:58,993 --> 00:09:02,699
You drag the slider
to highlight your icons,

192
00:09:02,723 --> 00:09:04,958
and then just click "Submit" to answer,

193
00:09:04,982 --> 00:09:08,645
and we animate away the difference
between your answer and reality.

194
00:09:08,669 --> 00:09:12,744
And it turns out, I was a pretty
terrible guess: five.

195
00:09:13,149 --> 00:09:14,573
How about the next question?

196
00:09:14,597 --> 00:09:16,753
This is asking about
what the average age is,

197
00:09:16,777 --> 00:09:19,222
so the age at which half
the population are younger

198
00:09:19,246 --> 00:09:20,920
and half the population are older.

199
00:09:20,944 --> 00:09:24,294
And I thought 35 -- that sounds
middle-aged to me.

200
00:09:24,318 --> 00:09:25,761
(Laughter)

201
00:09:28,206 --> 00:09:30,312
Actually, in Exeter,
it's incredibly young,

202
00:09:30,336 --> 00:09:34,874
and I had underestimated the impact
of the university in this area.

203
00:09:34,898 --> 00:09:36,929
The questions get harder
as you go through.

204
00:09:36,953 --> 00:09:39,336
So this one's now asking
about homeownership:

205
00:09:39,955 --> 00:09:43,654
For every 100 households, how many
are owned with a mortgage or loan?

206
00:09:43,678 --> 00:09:44,958
And I hedged my bets here,

207
00:09:44,982 --> 00:09:48,080
because I didn't want to be
more than 50 out on the answer.

208
00:09:48,104 --> 00:09:50,124
(Laughter)

209
00:09:50,148 --> 00:09:52,614
And actually, these get harder,
these questions,

210
00:09:52,638 --> 00:09:55,497
because when you're in an area,
when you're in a community,

211
00:09:55,521 --> 00:10:00,771
things like age -- there are clues
to whether a population is old or young.

212
00:10:00,795 --> 00:10:03,140
Just by looking around
the area, you can see it.

213
00:10:03,164 --> 00:10:06,555
Something like homeownership
is much more difficult to see,

214
00:10:06,579 --> 00:10:09,187
so we revert to our own heuristics,

215
00:10:09,211 --> 00:10:13,662
our own biases about how many people
we think own their own homes.

216
00:10:13,686 --> 00:10:17,336
Now the truth is,
when we published this quiz,

217
00:10:17,360 --> 00:10:20,896
the census data that it's based on
was already a few years old.

218
00:10:20,920 --> 00:10:24,489
We've had online applications
that allow you to put in a post code

219
00:10:24,513 --> 00:10:26,607
and get statistics back for years.

220
00:10:26,631 --> 00:10:27,820
So in some senses,

221
00:10:27,844 --> 00:10:31,393
this was all a little bit old
and not necessarily new.

222
00:10:31,417 --> 00:10:35,056
But I was interested to see
what reaction we might get

223
00:10:35,080 --> 00:10:37,797
by gamifying the data
in the way that we have,

224
00:10:37,821 --> 00:10:39,228
by using animation

225
00:10:39,252 --> 00:10:43,000
and playing on the fact
that people have their own preconceptions.

226
00:10:43,508 --> 00:10:47,091
It turns out, the reaction was, um ...

227
00:10:48,328 --> 00:10:50,256
was more than I could have hoped for.

228
00:10:50,280 --> 00:10:53,661
It was a long-held ambition of mine
to bring down a statistics website

229
00:10:53,685 --> 00:10:55,093
due to public demand.

230
00:10:55,117 --> 00:10:56,917
(Laughter)

231
00:10:56,941 --> 00:11:00,405
This URL contains the words
"statistics," "gov" and "UK,"

232
00:11:00,429 --> 00:11:03,671
which are three of people's least
favorite words in a URL.

233
00:11:03,695 --> 00:11:07,680
And the amazing thing about this
was that the website came down

234
00:11:07,704 --> 00:11:09,797
at quarter to 10 at night,

235
00:11:09,821 --> 00:11:13,032
because people were actually
engaging with this data

236
00:11:13,056 --> 00:11:14,595
of their own free will,

237
00:11:14,619 --> 00:11:16,654
using their own personal time.

238
00:11:16,678 --> 00:11:19,165
I was very interested to see

239
00:11:19,189 --> 00:11:22,902
that we got something like
a quarter of a million people

240
00:11:22,926 --> 00:11:26,198
playing the quiz within the space
of 48 hours of launching it.

241
00:11:26,222 --> 00:11:30,149
And it sparked an enormous discussion
online, on social media,

242
00:11:30,173 --> 00:11:32,210
which was largely dominated

243
00:11:32,234 --> 00:11:36,227
by people having fun
with their misconceptions,

244
00:11:36,251 --> 00:11:39,310
which is something that
I couldn't have hoped for any better,

245
00:11:39,334 --> 00:11:40,494
in some respects.

246
00:11:40,518 --> 00:11:43,744
I also liked the fact that people started
sending it to politicians.

247
00:11:43,768 --> 00:11:46,357
How well do you know the area
you claim to represent?

248
00:11:46,381 --> 00:11:47,543
(Laughter)

249
00:11:47,567 --> 00:11:49,127
And then just to finish,

250
00:11:49,992 --> 00:11:52,322
going back to the two kinds of people,

251
00:11:52,346 --> 00:11:54,603
I thought it would be
really interesting to see

252
00:11:54,627 --> 00:11:57,442
how people who are good with numbers
would do on this quiz.

253
00:11:57,466 --> 00:12:00,482
The national statistician
of England and Wales, John Pullinger,

254
00:12:00,506 --> 00:12:02,579
you would expect he would be pretty good.

255
00:12:03,524 --> 00:12:05,973
He got 44 for his own area.

256
00:12:05,997 --> 00:12:08,465
(Laughter)

257
00:12:08,489 --> 00:12:13,438
Jeremy Paxman -- admittedly,
after a glass of wine -- 36.

258
00:12:14,051 --> 00:12:15,512
Even worse.

259
00:12:15,536 --> 00:12:18,737
It just shows you that the numbers
can inspire us all.

260
00:12:18,761 --> 00:12:20,021
They can surprise us all.

261
00:12:20,045 --> 00:12:22,084
So very often, we talk about statistics

262
00:12:22,108 --> 00:12:24,070
as being the science of uncertainty.

263
00:12:24,094 --> 00:12:25,876
My parting thought for today is:

264
00:12:25,900 --> 00:12:28,935
actually, statistics is the science of us.

265
00:12:28,959 --> 00:12:31,747
And that's why we should
be fascinated by numbers.

266
00:12:31,771 --> 00:12:32,961
Thank you very much.

267
00:12:32,985 --> 00:12:36,762
(Applause)