How much would you gamble to get a new pill, pesticide or field
experimental result? Consider the typical losses in
the Winter 2009 issue of Biological Theory 4
(Konrad Lorenz Institute for Evolution and Cognition Research)
+
This way to the t-table
+
Guinness:
250 years of clever counting
Click on
Lavoisier to discover the "significance" of Guinness to science
Read
more at:
Economist's
View,
Economist.com,
Salon.com,
Schumpeter's
Century
[Noted as a
"Highlight" of the 2009 Joint Statistical Meetings (JSM 2009)]
"The
Cult of Statistical Significance"
from Amstat.org,
[Statisticians
of Interest]:
"Steve Ziliak's
session at JSM is titled the same as his book [with Deirdre N.
McCloskey]: The Cult of Statistical Significance: How the
Standard Error Costs Us Jobs, Justice, and Lives. The book,
published in 2008, has been reviewed by Science, Nature Medicine, BBC
Radio, Financial Times, the Economist, Cato Journal, and many others.
Ziliak will be available for interviews between August 1, Saturday
morning, and August 3, Monday afternoon."
Reviews of Ziliak's and McCloskey's
The
Cult of Statistical Significance
(University of Michigan Press, 2008)

in
Science, by Theodore M. Porter
in
Nature Medicine, by Jessica Ancker
in
Journal of Economic
Literature, by
Saul Hymans
in
SIAM News,
by James Case
in
Times Higher Ed, by Steve
Fuller
in
Law & Social Inquiry, by Richard
Lempert
in
London Book Review
in
EH-Net (Economic History),
by
Philip Coelho
in
Administrative Science Quarterly,
by Xueguang Zhou
in
European Journal of Epidemiology,
by Olli Miettinen
in
Canadian Journal of Sociology, by
Victor Thiessen
in
Skeptical Inquirer, by Peter Lamal
in
Journal of Economic Issues, by Ron
Smith
in
Journal of Economic Methodology,
by Tom Engsted
in
Cato Journal, by Peter Van Doren
in
Economic Affairs, by Art Carden
in
Erasmus Journal for Economics &
Philosophy, by Aris Spanos
in
Hungarian Economic Review
(Közgazdasági
Szemle), by
Tamás Dusek
"Guinnessometrics"
and
The
Cult discussed on BBC Radio's "More or Less," with Tim
Harford. Listen to a podcast of the Jan. 23rd, 2009 episode.
Image
copyright: BBC Radio 4

The Cult of Statistical Significance
lands on the top shelf at
StatLit
. . .
and at the Eastern Book Company, "Outstanding
Academic Title of 2008, Social and Behavioral Sciences"

What do Guinness and "Student's" t-test have in common? Read
about W. S. Gosset (aka "Student"), R. A. Fisher, and
The Cult of Statistical Significance
in Tim Harford's "The Undercover Economist," Financial Times,
Feb. 7., '09.
"Lovely day for a Regression!"
"Gosset for Strength!"
Image copyright: Financial Times
Listen
to Deirdre McCloskey's
podcast lecture on The Cult
at the National Economists
Club (NEC), Washington, DC, Dec. 4, 2008.

Blog reviews and interviews:
by
Andrew Gelman,
at
Statistical Modeling
by
Mark Thoma,
at
Economist's View
by
Leland Teschler,
at
Machine Design
by
John D. Cook,
at
The Endeavor
by
Arnold Kling,
at
EconLib
by
Cliff Norman
at
Profound Knowledge (Deming)
by
Jacob Grier, at
Liquidity Preference
and:
World Association of Medical Editors
Voice of the Employee
Converge
Jules and James (Climate Change)
Chance
Trinidad & Tobago
Review
("Throwing
a Book at Crime," by Kevin Baldeosingh)
LSE's Cognition & Culture Blog
(Olivier Morin)
John Myles White
University of Michigan Press
The Bayesian Heresy
William M. Briggs
The Bayesian Investor
Christopher Hayes
Homo Phileconomicus
Coert Visser
David Pannell
Common Tragedies (Environment)
History of Economics Playground
ALISE (Library and Information Science)
[The
Haiku Connection]
  
"Economists
embrace haiku,"
by Erica Alini, The Wall Street Journal, Real Time Economics blog, July
2, 2009
[Read
more at Mark Thoma's blog,
Economist's View]
[Read
"Drinking the Haiku Economics Kool
Aid," at
Open Economics]
[Read
Fred Lee's
Heterodox Economics Newsletter]
[Read
Laura Janota's article in
Roosevelt
Review]
[Read
Tina Owen's article in
Iowa Insider]
See Ziliak's
"Haiku Economics,"
in The Wall Street Journal (Dec. 31, '08, page one): "Fannie, Freddie, Bear & Hard Times:
Wall Street's Collapse, Told in Rhymes," by Mary Pilon.
Post a
poem at Mary's blog, "The Wallet"
. .
. and at Steve Levitt's and Stephen
Dubner's
Freakonomics-New York Times blog.
Enjoy a little haiku Q & A in "Verses
of Economy," by Steve Kolowich, The Chronicle of
Higher Education (May 22, 2009, A6),
Step up to NPR's "Recession Haiku Challenge," at
Planet Money
Keep the flow going with West Wing Writers of haiku, at
Podium Pundits
. . .and read
Haiku Economics:
Little Teaching Aids for Big Economic Pluralists, forthcoming in
International Journal of Pluralism
and Economics Education
Wall Street Journal; New York Times and itchys
"Limeraiku" is the marriage - or elopement - of haiku and limerick. See
page 291, The New Oxford Book of English Light Verse (1978),
edited by Kingsley Amis. What is financial or economic limeraiku? As
Shaw and Amis remind us, limerick poems (unlike haiku) are strictly made
by breaking taboo.
There was an
old man
from Lehman,
who yanks Big Hank's . . .
Jewels? No
way, man.
Recessional
plea
from Treasury:
"Fax the tax,
int'rest,
usury!"
There's a
crude old cow
from Camdentown,
who spits chips:
"Curry! Mayo!
Now!"
Recent articles
by Ziliak
about
Gosset, Fisher, &
the history of "statistical significance":
Guinnessometrics: The
Economic Foundation of Student's t, Journal of Economic
Perspectives, Fall 2008
Size Matters: The Standard Error of
Regressions in the American Economic Review, Journal of
Socio-Economics
On
R. A. Fisher and the Copyright History of Student's t (Note: an early version of this
article was titled "The Great Skew")
The Fallacy of the Transposed
Conditional in Medical and Biological Research

McCloskey and Ziliak reply to critics
Hoover and Siegler, in Journal of Economic
Methodology.
Ziliak and McCloskey reply to critics
Elliott, Granger, Horowitz, Leamer, Thorbecke, Wooldridge, Zellner, and
others, in Journal of Socio-economics
Ziliak and McCloskey reply to Schelling,
in Econ Journal Watch
McCloskey and Ziliak reply to Spanos,
in Erasmus Journal for Philosophy and Economics
Recent articles by Ziliak on the
collapse of the welfare state, on the collapse of the fact/value split,
and on teaching pluralism in economics:
Ziliak (with Klamer and McCloskey) launches unusual economics textbook
What is
the history of self-reliance? Has abolishing welfare helped out in the
past?
On the
positive/normative distinction:
what's new about the old collapse of it? "Nothing," as
Ziliak explains in the new
International Encyclopedia of the Social Sciences, edited by William
Darity, Jr.
Ziliak (with Joan Hannon) on
400 years of public welfare in America:
perspective by incongruity!
Ziliak, working in collaboration with the economic historians Joan
Hannon and Price Fishback, published there a "time-line" of significant
events and legislation in U.S. social welfare history, from 1588 to
1997. The unusual table will be of interest to any worker in the areas
of social welfare history and philanthropy.
U.S. Welfare History Time-line
New Course, Roosevelt
University,
Spring
2008:
ECON 426 THEORIES
OF JUSTICE IN ECONOMICS (Ziliak)
CHANGE: IT'S AN M.A.-LEVEL COURSE
Syllabus
The course will ask and examine fundamental
questions about economic justice in a dialogical and inter-disciplinary
context. Students will read selections from classic texts (Aristotle to Sen) and
original journal articles by contemporary theorists. Ideas about
economic "self reliance" - what it is and how to achieve it -- will be
central to our inquiry.
Prerequisites: ECON 323
New Book, March
2008:

The
Cult of Statistical Significance
is the place to
begin your post-Fisherian, post-Kuhnian statistical education.
See
Contents
/
Order
Now
Expert scientists believe
they're testing hypotheses with their conventional tests of statistical
significance. They’re not.
They think the “existence” of
one kind of precision under conditions of random error--namely,
statistical
significance--can answer the scientific, quantitative question of estimation, which is a
question of “size-matters/how much.” It can’t.
They think the null
procedure routine since Fisher can answer the pragmatic question about the
distribution of reasonable degrees of belief over a range of possibly believable if
radically different hypotheses. It can’t, won’t, and never will. Today’s
scientist neither tests nor estimates—he “testimates.”

"Student" (aka William Sealy Gosset, 1876-1937) worried about careless uses
of his test. He had reason to.
After Fisher, testimation is the outcome. Testimation is the
unhappy marriage of the fallacy of the transposed conditional to the sizeless stare of statistical significance. It is ruining the quantitative solidity of
the sciences descended from Galton and Pearson and especially from Ronald A.
Fisher. And its reckless policy recommendations are costing us jobs,
justice, ecology, and even human lives.
In
1908 "Student" changed some sciences with his small sample test of
significance. Now those sciences are stagnating or declining through
their reckless and illogical deployments of "Student's" test, just as
"Student" himself always warned they could.
Click
here
to learn about the neglected "Student" and the strange evolution
of his test after Fisher picked it up and made it central to the
sciences.
Stephen T. Ziliak is an economist whose work spans the fields of history, philosophy, rhetoric, statistics, and social policy.
He experiments with a kind of economic poetry
or criticism, too, in
what he calls
haiku economics
- "as if economics is so efficient".
He learned haiku from a friend, the late
Etheridge Knight, a great American poet & haiku master:
Invisible
hand:
Mother of
inflated hope,
Mistress of
despair.
Cooling
temperatures
Teens, chilly
away from Lake
Ghetto kids
rejoice!
From first
principles
you find an
end-state result:
The state
should stop now.
Corrugated
steel
Fence-links
cap off prison wall/
Blackbird
pecks at chains.
Click here to order Haiku Economics,
No. 1 (Rethinking Marxism, vol. 14(3), 2002)
Etheridge Knight (1931-1991)
"America's greatest poet in the oral tradition."
-
Robert
Bly to Steve Ziliak, Denoument Gallery, Indianapolis, Jan. 1991
Ziliak earned graduate degrees from the University of Iowa in both Economics(Ph.D.,
1996) and the
Rhetoric of the Social Sciences
(Ph.D. Certificate, 1996). A frequent collaborator with
Deirdre McCloskey, Ziliak is the author or editor of three books
(with McCloskey) and over sixty scholarly articles.
CV

An expert at historical archival research, Ziliak is best known for his work on American social welfare and
on the
history, theory, and practice of hypothesis testing in the life and
human sciences.
Ziliak has taught at a number of institutions around the United States,
including Emory University and the
Georgia Institute of Technology where, in 2002, he was awarded "Faculty Member of the Year" and, in the following year, "Most Intellectual Professor." He is currently Professor of Economics at
Roosevelt University. [If you are enrolled in a course taught by Prof. Z, please refer to the
Blackboard site.]
A prize-winning teacher, Ziliak is the author of an unusual micro and macro textbook,
The Economic Conversation(Palgrave/MacMillan), which aims to change the way economics is taught. Co-authored with
Arjo Klamer and Deirdre McCloskey, nearly a third of The Economic Conversation is written in Socratic dialogue.
Ziliak's main project now is on the scientific character of William Sealy
Gosset (1876-1937) aka "Student" of "Student's" t. Most scientists have learned about "Student's" t from Fisher or a Fisher disciple. This has been bad for "Student"--and science.  Fisher misled some sciences in the 20th century. From the very beginning, in 1904, "Student" took an economic approach to the logic of uncertainty. Karl Pearson, and then especially Ronald Fisher, wouldn't listen to "Student." In the 1920s and 1930s Gosset, this "Student," improved upon his own economic approach, inventing the statistical ideas of power and loss, which he gave to Egon S. Pearson and Jerzy Neyman to formalize. Gosset was in these and other regards a great scientist. His economic approach to for example the test of statistical significance and the statistical design of experiments can repair the damage done in science and policy by today's Fisherian methods. |