Forum
Bissantz ponders
Logbook
No. 04.06
Rules of Quotation
When faced with the task of saying something smart about statistics in general, we often find ourselves quoting great people. There's a well known remark by Winston Churchill that is particularly popular. The editor-in-chief of a important business newspaper used it recently, just as many others have in the past. He had to write an introduction to a highly useful compendium packed with statistics related to international numbers that was being published by the same company. To emphasize the quality of the data gathered in the book, he referred to Churchill, trying to reassure him post-humously by saying that the data came from absolutely trustworthy sources and that he would certainly have appreciated the potentially surprising connections the book comes up with.
The problem is that this well-known saying isn't Churchill's at all, but was attributed to him by Joseph Goebbels.
Anyone still wanting to use dead British statesmen to bolster their arguments are better off finding someone else. Benjamin Disraeli, for example, coined the equally oft-quoted "There are three kinds of lies: lies, damned lies and statistics." Or at least, Mark Twain claims Disraeli said it. Plenty of others attribute it to Twain himself.
No. 03.06
Data without a scale for comparison is no use at all
In principle, each analysis is a comparison, because only classification in context provides useful conclusions.
What can a picture of Saturn tell us if we don't have a relative idea of its size? How big is it? Bigger or smaller than...? What kind of scale of comparison would be helpful to us? The Earth? Now, how big might Saturn be in comparison to the Earth? Bigger? A lot bigger?
I have to admit, it blew me away to see our little Earth (see that blue dot to the right of Saturn?) next to this gigantic mega-planet. As a matter of fact, Saturn, the second-largest planet in our solar system, has a diameter 10 times the size of the Earth and is 95 times heavier than our planet.
Comparisons are similarly effective and necessary for each form of economic analysis. The statement that "ACME, Inc. moves 10 million euro" is hardly useful if we know nothing else about the company. However, if we discover that ACME, Inc. increased their turnover by 30 % in comparison with the previous year, and as a result has moved up to second place in their branch while branch volumes fell by a total of 10 %, a very interesting story begins to develop.
No. 02.06
London Tube Map edged out by Concorde
The map of the London Underground ("Tube Map") runner-up in the competition to determine Britain's greatest design icon. Between Concorde in first place and the Spitfire in third, we have an example of groundbreaking information design. You'll find more on the subject in the column Do we have to falsify data?.
No. 01.06
Freakonomics
The unconventional Business Professor Stephen D. Levitt shows how an economic framework coupled with unorthodox thinking can provide new answers to questions of data analysis. The fact that his answers aren't always "politically correct" only increases the enjoyment for critical readers, but also means Levitt receives numerous threatening letters. His thesis that there is a connection between the legalization of abortion in the USA in 1973 and the sinking crime rate 21 years later, has provoked particularly controversial reactions.
All in all, can't-miss reading for all those who mistrust all-too-simple explanations.
Interviews
with Thomas Liehr
TNS Infratest
about the renaissance of the gut feeling
with Johann Härtl
Schott Glas
about how data analysis can be used for the precision control of production processes
with Jens Ritterhoff
SWISS
about sparklines on headbands
with Helmut Wagner
Labor Wagner
about transparency, transparency, transparency
with is report
about Business Intelligence Suites
The interviews are currently available in German only.
Other Interviews
Edward R. Tufte
Beauty and Data Analysis
Join the discussion
We cordially invite you to join the discussion.
Simply write us!
Essays
Can we drive companies like we do cars?
Musings on dashboards, controlling speedometers and traffic light functions.
Do pictures really say more than a thousand words?
The renaissance of table-form design
Do we have to falsify data?
How lies differ from necessary clarifications.
Telling lies with statistics (PDF, in German)
How news magazines make use of figures and how far Microsoft Excel is to blame.
Analyzing means comparing
Data without a scale for comparison is no use at all
The myth of data mining
Why men don't buy beer and diapers at the same time
Think tanks
Visionaries
Authors you should know if you're keen on handling figures right:
