Upload
paul-barsch
View
217
Download
0
Embed Size (px)
Citation preview
8/9/2019 Eight Things You Should Know About Statistics
http://slidepdf.com/reader/full/eight-things-you-should-know-about-statistics 1/3
Eight Things You Should Know About Statistics
By Paul Barsch, originally published in MarketingProfs, April 29, 2010
Statistics have been called ³an engine of knowledge´ by one risk
management expert. And while it¶s true that some business managers
don¶t have a fundamental grasp of statistical concepts, we also know there
is opportunity for misuse of mathematics. Is statistics the ³new grammar´
or are efforts to attach certainty to life¶s events doing more harm than
good?
In May 2010¶s issue of Wired Magazine, author Clive Thompson laments
the poor mathematical literacy of his fellow citizens. For example, he cites
people laughing at the concept of global warming as they face some of the
harsher winters on record, or the extra-vocal debate on vaccines and
possible links to autism. Mr. Thompson would tell us that it¶s the trend lines
that matter, and we too often look at the trees and miss the forest.
The problem, he says, is that ³statistics is hard´ and an overall
understanding of this important discipline is severely lacking. He says, ³If
you don¶t understand statistics, you don¶t know what¶s going on, and you
can¶t tell when you¶re being lied to.´
Thompson is correct that statistics are difficult for most of us, and that
thinking by the numbers takes training and much effort. It¶s also true that
one must understand statistical concepts, especially when percentages,
populations, and probabilities are bandied about in business and technical
press. However, broader acceptance of the power of statistics should be
tempered with limitations of this mathematical science.
Before accepting any statistic, study or experiment as gospel, the following
should be considered (there may be more«):
8/9/2019 Eight Things You Should Know About Statistics
http://slidepdf.com/reader/full/eight-things-you-should-know-about-statistics 2/3
1. Assumptions: What are the assumptions underpinning the research?
As seen from recent debate on CBO numbers for the U.S. health reform
package, assumptions matter tremendously.
2. History: How much historical data was used in the study? What was the
time scale? As seen from the 2008 financial crisis, the models used by
Wall Street mavens often only took into account 10 years of data in judging
the volatility and probability of failure of complex financial instruments.
3. Samples: Are the samples selected randomly? From what populations?
Is there enough data for statistical significance?
4. Data Quality: The output is only going to be as good as the quality of data feeding the analysis. Garbage in, garbage out.
5. Survivorship Bias: Author Nassim Taleb points out ³losers are often
not in the sample.´ Does the analysis include a population of survivors and
those who also failed?
6. Falsification and Omission: Yes, in an era of IPCC¶s Climate Gate,
one needs to ascertain if data are hidden, missing or outliers ignored.
7. Association equals causation fallacy: Correlation does not equate to
causation (a common mistake made by marketing and finance executives
alike).
8. Proper Application of Statistics: The effective use of statistics by
insurance actuaries, scientists, and even casino managers is well-
documented. However, real danger results when mathematical concepts
are used to denote certainty indecision-making and divining behavior of
markets.
Now, please don¶t get me wrong. Statistical analysis is very important for
many industries (e.g., health care, transportation, and manufacturing).
Statistics, however, can give us an illusion of control in a world that¶s much
8/9/2019 Eight Things You Should Know About Statistics
http://slidepdf.com/reader/full/eight-things-you-should-know-about-statistics 3/3
more complex than our models suggest. Nassim Taleb, author of the Black
Swan likes to remind us that ³(real) life isn¶t a casino.´
Statistical analysis is definitely a powerful gadget in the business
manager¶s decision-making toolkit. But one needs to understand the
limitations of this science.
After all, Taleb points out that many of today¶s statistical models work as
though we have ³full knowledge of the probability of future outcomes.´ And
this just isn¶t so, especially when it comes to fat tails, or the ³ten sigma´
event. Indeed, sometimes those rare events have extremely large impacts.
Were he alive today, the former captain of the Titanic, E.J. Smith wouldwholeheartedly agree.
Fortune 500 marketer Paul Barsch writes about the intersection of
markets, marketing, mathematics, technology and globalization.
Twitter: @paul_a_barsch.