for Economics, has been so effective that one student your life. Our happiness or misery often literally
wrote in his/her evaluation, “It is like taking a course depends on our understanding of statistics. Statistics,
in Christianity and Jesus himself is teaching.” (If or quantitative evidence, is being used in the media,
you can come up with more impressive praise than scientific publications, and elsewhere to persuade us
that, email me at
chair@stat.harvard.edu.) Another to buy products, arguments, theories, etc. Some of
colleague, Joe Blitzstein, doubled the enrollment of the claims are statistically and scientifically sound,
Introduction to Probability in just three years. He is but many are not. A good percentage of them are
now an international sensation, so to speak. A stu- deliberate lies, intended to deceive the public in
dent was telling her friend in Germany that she was order to make a profit.
taking this cool stat course with Joe and her friend
responded, “Oh, you mean that YouTube stat pro-
Honey, I Know You Are in
fessor?” (You can satisfy your curiosity by googling Excruciating Pain, but Which
“Stat 110 at Harvard.”)
Treatment Do You Want?
Last year, we also launched Real Life Statistics:
Here is another real-life scenario that, depending
Your Chance for Happiness (or Misery). This course
on your understanding of statistics, could literally
was designed by my “happy team”—which con-
make your happiness or misery. Two treatments for
sisted of eight master’s and PhD students from
kidney stones were evaluated in a medical study.
the statistics department—over two years and
Treatment A has a success rate of 78% and treat-
many happy dinners (not happy meals) at the
ment B a success rate of 83%. Which one should
best restaurants Boston offers. The course aims to
you choose? Treatment B, right? Well, what if I
introduce students to the wonderland of statistics
told you that when treatment A and treatment B
by showcasing how it is used (and misused) in
are applied to those who suffer small stones, the
real-life situations every student should be able to
success rates becomes 93% and 87%, respectively?
relate to, either happily or miserably.
And when they are applied to those who carry large
Unlike many traditional statistics courses that
stones, the success rate is 73% for treatment A and
arrange the material by statistical topic in order of
69% for treatment B? That is, regardless of the sizes
complexity, Real Life Statistics arranges the material
of the stones, treatment A has a higher success rate.
by “real-life modules.” Last year, the five modules
Then you should choose treatment A, right?
were finance (e.g., stock market), romance (e.g.,
Confused? You should be if you don’t understand
online dating models), medical sciences (e.g., Viagra
Simpson’s Paradox (no relationship to O. J.). There is
trial), law (e.g., O. J. Simpson trial), and wine and
actually no paradox at all in the mathematical sense.
chocolate tasting (depending on your age). This
The numbers I reported above are from an actual
semester, we are replacing the law module with an
study from the March 29, 1986, issue of the British
election module, given the historic election we all
Medical Journal (Clinical Research Ed). For treatment
just witnessed (and now that O. J. is behind bars).
A, there were 350 patients, 87 carrying small stones,
All these efforts are aimed at making statistics not
among which treatment A was successful for 81. For
just palatable, but delicious to all of you, who, I am
the remaining 263 patients (with large stones), treat-
98% sure (that is the highest assurance any profes-
ment A was successful for 192 of them. For treat-
sional statistician would give), will need statistics
ment B, there were also 350 patients, with 270 suf-
in your own research, regardless of the subject, and
fering small stones, among which treatment B was
successful for 234. For the remaining 80 (with large
stones), treatment B was successful for 55.
Statistics in Defense and National Security Conference Now, you do the math and then think statisti-
The Army Conference on Applied Statistics (ACAS) announces Arizona
cally how this could happen. That is, how could
State University in Tempe, Arizona, as host of its 15th annual meeting,
treatment B have a better success rate overall than
to take place October 21–23, 2009.
treatment A and yet a worse rate in each subgroup
This year’s conference will feature three special sessions. Ed Wegman and defined by the stone size? What could cause such
Yasmin Said of George Mason University are organizing two sessions:
a paradox? What are its general implications? Let’s
“Agent-Based Modeling and Simulation” and “Data Mining Applications.”
label this Puzzle Three and read on.
In addition, Andrew Glen and Rodney Sturdivant from the U.S. Military
Academy are organizing a session titled “Current Army Challenges in
The Best Thing About Being a
Statistics.”
Statistician Is That You Get to Play
The call for papers is extended. Speakers will be notified regarding
in Everyone’s Backyard
paper acceptance no later than September 11.
This line is attributed to John Tukey, a statistical
Information regarding the conference can be found at www.
armyconference.org. Updates are being made as program details are
giant who also coined the terms “software” and
finalized.
“bit.” This is literately true, as many statisticians can
44 AMSTAT NEwS SEPTEMbER 2009
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76 |
Page 77 |
Page 78 |
Page 79 |
Page 80