case, being something of an out- Wheelan’s illustration of the the thought of statistics.

If Naked
Naked Statistics: sider is a virtue. Wheelan’s experi-
ence in journalism gives his writing
central limit theorem is ludicrous,
but I think this is essential to its
Statistics can successfully appeal to
these readers, it will have been an
Stripping the Dread a flair and verve that are rarely
found in books on mathematics. As
strength and charm. It shows us
that you do not have to be a whiz
achievement worth undertaking.
Stephanie Kovalchik
from the Data he delves into a range of introduc-
tory statistical concepts – from
with numbers to think statistically.
If you were surprised by the mean
Maryland

the difference between a mean weight of the supposed marathon

and a median to the distinction competitors, it means you had an
between correlation and causation
– Wheelan sticks to plain-language
instinctive sense of the large-sam-
ple properties of sums of random
Practical Business
explanations with a minimum of
technical details. Where a typical
variables without taking a single
limit! Which highlights another
Statistics, Sixth Edition
introductory statistics textbook important point. There is a lot of
might work out a proof, Wheelan talk about how misguided the hu-
offers creative, even whimsical, man mind can be when it comes to
hypotheticals to demonstrate a quantifying risk or judging uncer-
fundamental principle. For Whee- tainty. In his book, The Monty Hall
lan, the technical trappings he Problem, Jason Rosenhouse says
strips away are just the nitty-gritty that for most people “Probabilistic
that too often obscures the subject reasoning is just not something
matter and leaves one without any that comes naturally”. With Naked
intuition about the how or why of Statistics, Wheelan makes a differ-
the principle at hand. ent and, I think, more optimistic
One of my favourite examples argument. What his descriptive
comes in Chapter 8 where Wheelan analogies and examples suggest is
Charles Wheelan; W. W. Norton; 320 introduces what he calls “the Leb- that our struggles with probabil-
pp.; February 2013; £18.89, $26.95; ron James of statistics”: the central istic reasoning are less a failure
ISBN 9780393071955 (hardback) limit theorem. (Lebron James, in of our intuition than a failure of
case you did not know, is a basket- our imagination. Naked Statistics Andrew F. Siegel; Academic Press;
Google’s chief economist Hal Var- ball player who has been described wants to show us how the logic xix+603 pp.; March 2011; $59.97,
ian infamously predicted that the as “the second most influential that seems so mysterious when £60.99; ISBN 9780123852083
2010s would be the decade when athlete behind Lance Armstrong”). in the form of equations on paper
statistics got sexy. Here is statistics Wheelan asks the reader to imag- becomes everyday when we see it This large volume presents the
as striptease. Leonard Mlodinow’s ine that a marathon is about to at work in the world around us. The business user with a range of
The Drunkard’s Walk, Malcolm get underway but runners have to real difficulty is in being able to statistical applications from the
Gladwell’s What the Dog Saw, and be bused from surrounding cities recognise that this logic is one and very simple to the complex. It can
Nate Silver’s The Signal and the to get to the starting line. (I did the same. be seen as a workbook for anyone
Noise are just a few of the recent warn you that these hypotheticals Some may criticise Naked Statis- who is interested in any business
popular works that have success- were sometimes whimsical.) The tics as hokey and overly informal. area of data to apply statistical
fully proven that statistics is not organisers of the competition hit Admittedly, when simplifying any rigour to collection, analysis and
so drab or intimidating as we might a snag when one of the buses goes subject it is difficult to avoid being reporting.
once have thought. Yet no author astray. A search team heads out to accused of inaccuracy. One could ar- The writing is clear and easy
before Charles Wheelan has tried to find the errant bus. Just outside gue that the example of the lost bus to understand. Starting with
convince us that statistics is most the city the team comes upon sells the central limit theorem short descriptive statistics, the reader
approachable and most interesting a broken-down bus. The search because it is really an illustration of is led through basic concepts
when fully undressed. In Naked team leader is convinced that they the law of large numbers. Still, what introducing questions as sections
Statistics Wheelan strips a crash have located the lost bus of racers is likely to cause more of the quants progress. Following issues with
course in statistics of all formulae until he scans the passengers and among us to shake an angry finger data, probability and distributions
and jargon, putting in their place surmises that their average weight at the little man on the cover of are introduced together with some
humorous anecdotes and imagina- must be close to 220 pounds (100 Naked Statistics is the hypocrisy of of the more standard problems of
tive analogies that create a reading kg). With this new bit of informa- trying to persuade us that statistics data handling. Summary statistics
experience not as exciting as the tion, the team leader, like the rest is only fun when she is stripped of follow, with measures of disper-
title promises but nevertheless of us, is no longer so certain that all her numbers. To be fair, quants sion. The book deals extensively
entertaining. he has found the right bus. The – the kind of people who think the with probability and introduces
Former correspondent for The reason is that the average com- best part of a scientific paper is random variables. Complexity then
Economist and one-time candidate petitive marathon runner is quite usually found in the appendix – are increases significantly into sta-
for a seat in the US House of Rep- slim. (In fact the broken-down not Wheelan’s intended audience. tistical inference, regression
resentatives, Wheelan may seem bus, Wheelan informs us, was actu- The readers Wheelan wants to reach and time series. The final part is
ill-suited to take on the task of ally headed to the International are the people who are either bored, titled “Methods and applications”
popularising statistics. But in this Festival of Sausage.) baffled, or cannot be bothered by and will rarely be touched by a

© 2013 The Royal Statistical Society april2013 43

practitioner. Nevertheless, each have been found to be horsemeat and sunshine may have little or no
of the 18 chapters concludes with
a summary, key words, questions,
The Real Cost of Cheap rather than beef; that at least con-
centrated public and media minds
influence over trade deals. All these
aspects of the problem are difficult
problems exercises and suggested
projects, which would be good
Food on the global system that produces
our cheapest foods, sometimes
to quantify, but Carolan pulls in
facts and figures from a wide range
not only for students in a college through networks that are hard to of sources to build up a compelling
situation but also for the lone user trace to their source. argument: simply turning more land
in a business. A companion web- I enjoyed the many facts that over to cattle and high-yield cereals
site with all of the quantitative went against my preconceptions, will not feed the world, keep it safe
examples and problem data allows and I will offer a few here. Flying or make farming robust (a smart
hands-on learning with the same fresh vegetables from Kenya is terrorist releasing foot-and-mouth
data. The data comes in a zip for- better in terms of carbon footprint disease into intensive livestock
mat – of 11 MB – but does contain than growing them intensively in industries would cost taxpayers
easily accessible data organised by Europe through the winter. (Kenya £60 billion in the USA alone). The
chapter. is singled out as a good news story only part I found hard to accept
Only two requirements are nec- because 60% of exported vegetables at face value was the association
essary: a reasonable understanding came from owner-occupied small- he claims between living near in-
of and ability in mathematics and holdings in 2005, delivering profits tensive livestock farms and various
a desire to learn. The basis of the to local communities.) Transport- health problems such as asthma;
book was a university course but it ing food by large trucks over long as a medical statistician I know it
could be used by many others who distances can actually be better in is not so easy to compare people
have a need to apply statistics in carbon terms than a multitude of in different neighbourhoods and
their place of work. Each chapter local journeys in small vehicles. socio-economic brackets and reach
is a clearly defined topic. For exam- The vicious circle between conclusions about cause and effect.
ple, a complete chapter on random Michael Carolan; Earthscan food security, stable incomes from Carolan’s solution for feeding a
sampling – which includes a little (Routledge); 288 pp.; September agriculture, and conflict emerges growing population in the future
on getting the data – does not go 2011; £19.99, $34.95; ISBN in case studies from around the hinges on eating less meat, reduc-
into product satisfaction or user 9781849713214 world, including prescient details ing subsidies and creating market
surveys. Concepts are discussed about Mali. International interven- conditions that no longer artificially
clearly, though probably too deeply Last Christmas my wife received a tion through investment by large absorb the externalised costs of
for the average business person. calendar with “kitchen tips” from corporations, or food aid, appears production, such as the taxpayer
A short chapter on report writ- a relative. Before sitting down to repeatedly as part of the problem covering the cost of cleaning fer-
ing refers only to communicating write this review I noticed that tiliser-laden groundwater (most of
the results of a multiple regres- today’s tip is to buy dried lentils the UK figures come from http://
sion. Most business people will and beans because they are much bit.ly/ZeRJnI). He suggests
be required to write many reports cheaper than canned ones, and the annual externalised cost in the
containing data and I am sure few only require overnight soaking USA alone can be counted in bil-
will include multiple regression. The and boiling for 2 hours. Perhaps Simply turning more lions of dollars. I was looking for-
book analyses data and displays re-
sults from the producer perspective.
you feel the same nagging doubt
about this as I do; after all, isn’t
land over to cattle and ward to reading his new economics
that would value nutrition, cultural
Some tables do not follow even it more wasteful for us each to high-yield cereals will and ecological resources, and incor-
rudimentary rules of right-justifying boil something for two hours using not feed the world or porate them in the marketplace, but
data, and rounding in results is not natural gas mostly shipped from keep it safe he holds back from proposing how
applied or dealt with. Many of the Qatar and Norway rather than to implement such bold, sweeping
tables and charts could be simpli- have one big industrial process do reforms. Carolan does not allow him-
fied to enhance their clarity. tons at a time and then can them? self to stray into moral or political
One key item missing is statis- Appearances can be deceptive, and recommendations. He suggests an
tics on the measurement of quality. producing food in a quaint, old- increasing role for hydroponics and
The statistical concepts are in the fashioned way is not necessarily the and not the solution, failing to pro- urban market gardening on rooftops
book but would need a brief chapter most ethical choice. vide long-term security or stability. and balconies; as before, the solu-
bringing them together. This is the thrust of sociologist Tax breaks as incentives for foreign tion may not look anything like a
A clear glossary and comprehen- Michael Carolan’s engaging book investment have been similarly pre-industrial cottage industry.
sive index make this a good refer- which challenges us as consumers counter-productive, with examples The book is pitched at an inter-
ence text for those who complete of cheap food to think more broadly from Mexico, Madagascar and Paki- ested general reader who is not put
the course for which this is the base and internationally about who stan. Although the United Kingdom off by a lot of numbers, and it is
text. It would also assist those who bears the cost and, crucially, how managed to pull off this trick to de- assiduously referenced, although it
did not do the course but want to we and our governments should try velop the North Sea oilfields, crops seems that in this highly contested
get to grips with the some of the to count it. It was written before – unlike oil wells – can be relocated field there are few entirely disinter-
statistical concepts in business. the horsemeat scandal broke in the when the tax breaks are no longer ested sources of data.
Ed Swires-Hennessy UK – many cheap burgers, includ- so favourable, and paradoxically the Robert Grant
Newport ing some served as school lunches, country supplying the soil, water London

44 april2013