Chance Magazine Articles
Review of Chances, by Jackie Collins
Vol. 1, No. 1, 1988.
Extra Statistical Perception
I recently learned that there's a whole field of applied statistics
devoted to extrasensory perception (ESP) and the paranormal. It has
gone unnoticed by many, because the statistical theory and methods that
are applied are not the usual ones. For example, many applications
use the principle of minimum likelihood. And, of course, statistical
tests are based on the paranormal distribution. It's hard to place
the shape of that distribution, but when you see it, it does ring a bell.
Everyone makes a big fuss about ESP, but I've had it for years.
For example, some time ago, I was thinking about a paper I submitted to
the Annals of Statistics, entitled "A practical application of the Prokhorov-Varadarajan
inequality to sampling inspection," when a feeling came over me that the
paper would be rejected. Two-and-one half years later, the editor
informed me that it was indeed rejected.
Another psychic experience happened when one year I had a vivid dream
that I was at an annual meeting of the American Statistical Association
listening to the presidential address when all of a sudden I woke up and
found that I really was listening to the ASA presidential address.
A few years ago, I started getting into psychic experiential kinds of
things in a big way. I took up transcendental meditation, but soon
had to give it up, because I kept forgetting my mantra. Undaunted,
I began to use my powers to entertain at parties. I billed myself
as "Miron the Magnificent," so I could use the same poster that was designed
to announce my talk on economic forecasting at the Bureau of Economic Analysis.
When I made an appearance at a Washington Statistical Society wine and
cheese gathering, for example, people would be selected at random by my
accomplice and I would divine personal information about them ("You are
a statistician. You work for the government. Nobody understands
what you do.") and the answers to questions they would have on their minds
("The reason they serve such cheap wine here is that's all they can afford
for your lousy two bucks!") After a while, however, it became repetitious.
Many in our military believe in certain parapsychological phenomena.
The Army Research Institute even asked the prestigious National Academy
of Sciences to form a committee to examine the potential value of parapsychological
techniques. The committee describes in its report some of the claimed
phenomena and applications:
The "antimissile time warp," for example, is supposed to somehow deflect
attack by nuclear warheads so that they will transcend time and explode
among the ancient dinosaurs, thereby leaving us unharmed but destroying
many dinosaurs (and, presumably, some of our evolutionary ancestors). .
. . Many of the sources cite the claim that Soviet psychotronic weapons
were responsible for the 1976 outbreak of Legionnaires' disease, as well
as the 1968 sinking of the nuclear submarine Thresher. (Druckman
and Swets, 1988)
One of the psychotronic weapons the report names is the "hyperspatial
nuclear howitzer." Can you imagine a gunnery sergeant coming home
at the end of a bad day on the hyperspatial nuclear howitzer range, turning
to his wife, and saying, "Honey, I shrunk the Marines."
The Committee found that there was "no scientific justification from
research conducted over a period of 130 years for the existence of parapsychological
phenomena." I'm not surprised, because my colleague who directed
the study has been working on these matters for well over 130 years.
With the results of all those experiments being inconclusive, I decided
to try an experiment of my own. One evening, when we were entertaining
a group of high-powered Washington economists, I asked them all to spend
ten minutes thinking very intensely about how to get out of the current
recession. The next morning I called several colleagues at Stanford
and asked them if they could find anything unusual that happened on campus
during that time.
First, my colleagues noted that nearly half the faculty had left for
or was in Washington then to serve on some advisory committee or another--but
that was not unusual. The next day, I got a call from the chief dishwasher
of the campus cafeteria who reported getting an unusually large number
of bent spoons returned. Just as I thought I had a publishable result,
the dishwasher informed me that there was possible confounding.
The Stanford cafeteria generally loses several hundred spoons a day,
some of which are eventually returned. We couldn't be sure the returned
bent spoons were bent on the evening in question. (I couldn't believe
that there were so many spoons in circulation on the campus: if I
collect more than a dozen cafeteria spoons, they begin to clutter up my
drawer.) Consequently, the dishwasher said, the result was not statistically
significant. As a matter of fact, the Washington economists didn't
come up with anything significant either.
There is a physical theory that explains why ESP works, at least if
my cleaning lady is to be believed. According to this theory, the
brain, being an electrical network, emits radio waves that other brains
can pick up. A problem my colleagues who subscribe to this theory
worry about is that the EMF resonance from active brains might cause cancer.
They warn me that I'm placing my health in danger by having lunch at Berkeley
with Peter Bickel, David Blackwell, and David Freedman. And I thought
all along it was just indigestion.
My colleagues also say that, because there is no threshold, getting
close to any active, intelligent mind on a regular basis is dangerous.
They tell me that living in Washington, however, should pose no problem.
--------------------
REFERENCE
Druckman, Daniel and Swets, John A., eds. (1988). Enhancing Human
Performance: Issues, Theories, and Techniques. Washington,
D.C.: National Academy Press
TOP
Shlepping Statisticians to
Siberia
The Chairman of the Council of Economic Advisers, the Chief Statistician
in OMB, and the Commissioner of Labor Statistics recently returned form
an official visit to the Soviet Union, and all I heard from them was what
an exciting time they had. So when I was asked to join a delegation
of social scientists to the Soviet Academy of Sciences in Moscow, I quickly
accepted. Little did I know what would be in store. The Soviets
have developed to a finely honed skill our custom of dumping on American
social scientists.
Between glasnost and perestroika, the Soviet Academy has a choice of
three hotels to offer its guests in Moscow. One, Boondockski, is
a clean, modern facility located somewhere near the Transvaal. It's
about two hours and 45 minutes from Moscow--by phone. The second
hotel, Nogoodski, is affectionately called "Roach Hotel" (La cukarachski
zhestina), after the architect who designed it. It's the hotel of
choice for American guests of the Soviet Academy because of the location,
on the confluence of the Moscow sewer and the Metro green line. The
third hotel, Worsethanadumpski, was actually torn down several years ago,
but continues in operation principally to house students from third world
countries who are homesick for a room without running water. The
students call it the Hyatt Exorcy. Here is where we were deposited.
Until then, I had never thought I could envy living in our son's college
dormitory room. I had not seen toilets like we had since 1954, when
I was in a Trailways depot in Dry Gulch, Wyoming. Filth was everywhere.
It was the only place I've seen where guests were wiping their feet before
going outside.
I should have had a clue about some of the cheap Soviet hotels from
my Russian phrase book. Most foreign language phrase books begin
with useful phrases, such as "Hello." "How are you?" "Please." "Thank
you." and "Where is the post office?" The Russian phrase book starts
off with "The toilet isn't working." "There's no hot water!"
"Where is the bed?" "What do you mean that's a bed!" "What
do you mean that's a toilet!"
I felt right at home, however, the moment I hopped into a Moscow taxicab.
Just like in Washington, the cab drivers there didn't speak English either.
After several hours of listening to a Berlitz tape, I mastered "da" and
"nyet," which I had thought meant "yes" and "no." In spoken Russian,
however, these words do not simply mean yes or no. "Da" usually means
"I understand." Tell a waiter that you are thirsty and would like
something to drink, he may reply "Da," but do not take that to mean that
he will bring you something to drink.
Compared with "da," the word "nyet" is used with great frequency in
Russian conversation. In french, for example, the frequency of yes
to that of no in conversation must be at least five to one (Mais, Oui!).
The French even use no to mean yes (N'est ce pas?). In Russian, however,
the ratio must be less than one in ten. And "nyet" is rarely used
to mean no. For example, if you ask if you can make an international
phone call, the response may be "Nyet. Is absolutely impossible,"
which only means that you must try harder. The most frustrating use
of "nyet" is the common phrase "Nyet problem." You need a taxi?
"Nyet problem," you're told. And you wait until you give up and take
the metro. There's no hot water? "Nyet problem!" And
you take a cold shower. You want some wine in a restaurant?
"Nyet problem!" And the waiter disappears for hours. It wasn't
until I returned to the States and discussed my experiences with a Russian
scholar that I learned, from him, the translation of "Nyet problem."
It means "That's your problem."
We were warned not to drink the water, but to use bottled mineral water,
a sulphurous, highly gaseous liquid that tastes like a highly concentrated
mixture of Alka-Seltzer. Leaving the bottles of water open overnight
didn't do a bit of good. The Soviets have discovered a way to make
carbonated liquids hold their gas for 15 years after being opened.
And most everything is carbonated. After a few days, our formal diplomatic
introductions sounded like: "Hello [burp], how are [hic] you?"
"It's a [burp] pleasure to be here [belch]."
Despite the difficulties in the introductions, our meetings to exchange
information were informative. In one meeting, we were discussing
the difficulty in determining the amount of discretionary income that people
have to spend. "In the United States," I mentioned, "an average family
earns $25,000 per year. To live comfortably, they require $21,000
per year. How they spend or save that extra $4,000, we have no idea."
One of my colleagues in our delegation replied that the same situation
was true in European countries. In Germany, for example, an average
family earns 35,000 Deutchmarks and must spend 30,000 to live comfortably.
But what they do with that extra 5,000 Deutchmarks is anyone's guess.
"Is the same in Soviet Union," our Soviet counterpart replied. "In
Soviet Union, a worker earns 200 rubles per month. To live, his family
must spend 250 rubles per month. Where they get that extra 50 rubles
we have no idea."
We had time for some sightseeing. We drove by one of the most
popular spots in Moscow--McDonald's. The line to get in extended
for miles. If it were me, I would have used the drive-thru window.
The most exciting spot to see was Lenin's tomb in Red Square. I highly
recommend the souvenir shop there--it's called "Lenins and Things."
Despite the difficulties, when I returned to Washington, I felt as if
I had brought a bit of Moscow back with me. But my doctor says, with
a course of antibiotics, that too should pass.
TOP
On Time
Like I handle most things in life, I get around to writing this column
on the day that it's due, more often on the day after it's due, with my
editor screaming at me for missing another deadline. To placate my
editor, I promised him that this column would be on time. And that's
exactly what this column is about: time. A little late, perhaps,
but timely.
My colleague, John Robinson, has written a book about the surprising
ways Americans use their time. He's conducted surveys over the years
to show that, contrary to what everyone knows, Americans increasingly have
more free time. Where does it all go? Watching television,
Robinson says.
I realized the truth to that. Instead of spending a relaxing evening
arguing with your spouse and kids, you can simply turn on the television
and argue over which station to watch. (I thought the new dual channel,
picture-in-picture technology would resolve our problem, but, especially
with cable, our arguments have increased combinatoricly.)
Robinson says that studies show that brain-wave activity during television-watching
closely resembles that during sleep. I seriously doubt if the measuring
devices are that sensitive, however, because once a psychologist concluded
from an experiment that the same finding held for students in my statistics
lectures.
The fact is that most people watch television while really doing something
else that's not so demanding, like reading the latest issue of AmStat News.
For his time use surveys, Robinson has categorized seemingly every type
of activity you could be doing during the day. Even "doing nothing,"
a category, I must admit, that I waste more time on than television.
But I found that there are some activities that I spend a lot of time
on that Robinson has not categorized. For example, writing notes
to myself on little scraps of paper to remind me of something important
to do that day; waiting at my computer to get onto the world wide web;
and, especially, looking for those little scraps of paper with important
messages on them.
I realized that almost everything I do could be done more efficiently.
I got that inspiration, not from Robinson's book, but from that travelling
troupe that performs Shakespeare in a minute, condensing an entire Shakespeare
play in under 60 seconds. Why stop there I thought. Why not
try to do that for the complete works. I tried my hand:
King Lear, his ship caught in a tempest, lands in Verona, changes clothes
to appear in drag as Rosalin, only to find his brother Mercutio fall in
love with him. Together, they conspire to kill their father, Hamlet.
In the end, all is revealed and they go down to the seashore and live happily
every after.
Taking lecture notes can be done more efficiently. For example,
don't worry about spelling and using all those extra letters. Instead,
write your words fonetikly. Ya c'n even rite da werds as peepal speek
th'm. For example, at a lecture in Washington, write "normal."
At Stony Brook, write "noimal." At Harvard, write "nahhh-mal."
And, at Berkeley, write { }.
As another example, instead of wading through the latest issue of the
Annals of Statistics, you can download pages from its web site (www.brdcglnr.edu)
or purchase the issue on a CD-ROM (look for it on the flip side of a John
Tesh CD). Either way, you can scroll through the pages in a breeze.
I was, for example, able to finish an entire issue in five minutes without
any loss of comprehension.
Time management professionals tell you that the first step in working
more efficiently is to keep track of how you actually spend your time.
So, I kept a log, just like one of Robinson's survey diaries. Here
is how I spent today:
10:00 a.m. Arrive at office. Unpack briefcase filled with material
I didn't get to work on the night before.
10:05 a.m. Workout in the gym
10:30 a.m. Coffee break. Catch up on latest gossip about the Census
Bureau.
10:45 a.m. Read The Washington Post. Catch up on latest stories
about the Census Bureau.
11:00 a.m. Open mail. Arrange important correspondence in a pile.
Read ASA flyers and other junk mail before throwing them away.
11:30 a.m. Cappuccino break.
11:45 a.m. Turn on computer. Scan e-mail headings.
12:00 noon Play computer solitaire until colleagues are ready for lunch.
12:20 p.m. Lunch
2:00 p.m. Return from lunch. Wait to get on the web.
2:10 p.m. On hold with computer services "Help Desk."
2:25 p.m. Finally can surf the net
2:45 p.m. Cafe Latte break
3:00 p.m. Prepare memo that boss will throw away.
3:30 p.m. Prepare tenth revision of project cost estimates.
4:00 pm. Take call from boss. While listening, reorganize desk
drawer (unless fingernails need clipping).
4:10 p.m. Strategic planning exercise
4:15 p.m. Revise performance plan
4:25 p.m. Read latest reorganization memo.
4:30 p.m. Xerox interesting articles from The New York Times for which
there is no time to read.
4:40 p.m. Unwrap latest issue of the Annals of Statistics. Decide
whether to place on bookshelf immediately or leave on desk to shelve tomorrow.
4:41 p.m. Quiet time for creative brainstorming: Think of clever
ideas for next CHANCE column.
5:00 p.m. Personal calls.
5:15 p.m. Call for weather report.
5:18 p.m. Call for traffic report.
5:20 p.m. Call travel agent for vacation reservations.
5:40 p.m. Return phone calls from the day. Leave messages on voice
mail systems that the best time to reach me is "just before 8 o'clock."
5:55 p.m. Delete unread e-mails from computer.
6:00 p.m. Sort trash for recycling bin
6:05 p.m. Fill out time sheet for the day.
6:10 p.m. Repack briefcase with material I didn't get to work on during
the day.
6:15 p.m. Leave office.
At home this evening, my major activity is writing this column.
It's late, but I feel a sense of accomplishment, as my computer "beeps"
me with the message that I have written the required number of column inches.
That means I can finally turn off the television.
---------------
Reference
Robinson, John P. and Godbey, Geoffrey (1997). Time for
Life: The Surprising Ways Americans Use their Time. University
Park, Pennsylvania: Penn State University Press.
TOP
Statistical Partnering
Being at a loss for an idea for a column, I decided to attend an American
Statistical Association conference. I got out my folder of unopened
mail from the ASA and started tearing open envelopes at random. After
several dunning notices for dues and for overdue referees' reports and
a confusing ballot (for the Council of Sections as representative-elect
for the liaison to the Council of Chapters), I finally found an invitation
to a conference. It was addressed to leaders of the statistical profession
calling upon us to gather in Raleigh, North Carolina, for a conference
on Statistical Partnerships in Academe, Industry, and Government (SPAIG).
"What a curious acronym for a conference," I remarked to a couple of
colleagues. "SPAIG sounds like a revolting canned meat product."
"Probably, baloney," one colleague replied.
"They settled for SPAIG," said the other, "because it was originally
called Partnerships In Statistical Sciences."
I could see right away that the conference was founded on three mistaken
premises.
Mistaken premise #1: People will eagerly flock to a conference
that's in a small, out-of-the-way city. It was clear that conference
planners took their cue from the ASA School of Conference Site Selection.
But at least they didn't choose either of the first two sites ASA recommended:
(1) Disneyland and (2) the San Diego Zoo, although the latter would have
been in keeping with the conference.
Mistaken premise #2: It's necessary to begin your meetings promptly
at 8 o'clock every morning. I don't even get through my fourth snooze
alarm until 8:30. If they were really interested in forming partnerships
with academics, they wouldn't start a meeting until after 10 o'clock.
Mistaken premise #3: Academic statisticians are really interested
in what statisticians in industry are doing. In fact, most academic
statisticians aren't even interested in what other academic statisticians
are doing.
Most people, I felt, were going just to be seen and recognized as one
of the leaders of the profession, to whom, after all, the invitation was
addressed. (Many of these same people are still waiting for their
$10 million from Ed Mc Mahon.) I figured that, if I couldn't strut
amongst the hoity-toity, I'd hang out in the bar with the hoi polloi.
At the conference, you could tell whether a statistician was from industry,
government, or academia by the way he or she was dressed. Industry
people wear dark business suits. Government people wear drab business
suits. And academics forget to wear a jacket and at least one other
piece of clothing.
Dress at conferences, in fact, has changed as a result of technology.
The plastic pocket protector has been replaced by a pager. Some old
fashioned engineering statisticians, however, wear their pager on their
belt, next to their slide rule.
The conference began with a series of lectures. Here, too, you
could see a difference between some industry and academic statisticians
in their visual displays. The well prepared speaker from industry
came equipped with a computer projector display driven by a Pentium laptop
(with a backup software disk, backup battery, battery charger, cables,
and extension cords), a laser pointer (with backup batteries), a remote
mouse (with backup batteries and a backup cable mouse with a mouse pad),
a transmitter for the remote mouse (with backup batteries), a set of transparencies
in case of a computer failure (with backup projector bulb), and a set of
handouts with flashlights (and backup batteries). I figured that,
if you laid all his batteries end to end, you might get enough of a charge
to service all of Raleigh, North Carolina, for fifteen minutes--well, certainly
after 10 p.m.
The academic speaker, on the other hand, showed up with transparencies
displaying unintelligible scribbling. Through the overlay of smeared
letters and thumb prints, I could make out, in the lower left hand corner,
a mayonnaise stain from lunch. One transparency, however, showed
signs of mustard. Since the corned beef had been served the previous
day, it was clear that the speaker had given more advance thought to preparing
this transparency.
The irony is that you really learn more from these scribbled view graphs.
Since they are so unreadable, you find yourself listening more carefully
to find out what in the world the speaker is talking about.
Many of the talks were filled with success stories. For example,
one was about an undergraduate statistics student who developed, for a
food company, a procedure to measure the texture of mashed potatoes.
Hearing all these stories, I was skeptical about selection bias.
Shouldn't we also be reporting the failures? For example, Joe Blow,
a mathematically oriented graduate student, went to work for Amalgamated
Chemical, and, on his very first day, managed to wipe out the company's
entire hard disk storage.
Later, we divided ourselves into small groups and were asked to do some
serious brainstorming through an elaborate process involving affinity mapping,
trend diagrams, prioritization matrices, interrelationship digraphs, GANT
charts, and chi-squared tables. I'm afraid I'm not good at these
kinds of things. In one session, I started to write down my idea
on the poster board, when the facilitator yelled at me that he didn't say
"Simon Says."
At the final dinner, one fellow academic sat next to a young woman in
government and suggested that they follow the recommendation of the keynote
speaker and do some serious partnering. She slapped him.
Even though no one understood whether or not they accomplished anything,
it became clear at the end why everyone came. Industry and government
statisticians came to seek greater attention and recognition from academics.
And the academics came to seek more money from industry and government.
It was a stand-off.
At the end of the conference, we all engaged in trading business cards.
I got quite a deal. I traded one DuPont for two Stanfords and a Princeton.
TOP
HOME
