Sept 08 Am Stat News

SPECIAL CAREER ISSUE • SPECIAL CAREER ISSUE • SPECIAL CAREER ISSUE
September 2008 • Issue #375
The Membership Magazine of the American Statistical Association
Making a Name:
Up-and-Coming Statisticians
on the Verge of Great Things
Statisticians in History
Publications Agreement No. 41544521
SEPTEMBER AMSTAT FINAL.indd A 8/20/08 2:26:50 PM

SEPTEMBER AMSTAT FINAL.indd B 8/20/08 2:26:52 PM
VISION STATEMENT September 2008 • Issue #375
To be a world leader in promoting statistical
practice, applications, and research; publishing
statistical journals; improving statistical education; STATISTICIANS IN HISTORY
and advancing the statistics profession.
4 Kimiko Bowman
Executive Director Never Say Never
Ron Wasserstein: ron@amstat.org
Associate Executive Director and Director of Operations 7 Irene Hess
Stephen Porzio: steve@amstat.org
Sampling Is in the Details
Director of Programs
Martha Aliaga: martha@amstat.org 11 Jack Chao-sheng Lee
Director of Science Policy
Steve Pierson: pierson@amstat.org HISTORY 13 Paul Dixon Minton
Managing Editor Read about the careers and Building a Department
Megan Murphy: megan@amstat.org lives of statisticians whose
Production Coordinators/Graphic Designers names are recognizable 15 Mollie Orshansky
Melissa Muko: melissa@amstat.org as both published Author of the Poverty Thresholds
Lidia Vigyázó: lidia@amstat.org authors and scientists
Publications Coordinator
Val Snider: val@amstat.org
Advertising Manager
Claudine Donovan: claudine@amstat.org DAY IN THE LIFE
Special Contributors
21 Life as a Stochastic Modeler
Martha Aliaga • Keith Crank • Rosanne Desmone
Rebecca Nichols • Rick Peterson Alan Gelfand, Duke University
Gladys Reynolds • Fritz Scheuren
Amstat News welcomes news items and let- 25 On Becoming a Teacher
ters from readers on matters of interest to the asso- Daren Starnes, The Lawrenceville School
ciation and the profession. Address correspondence to
Managing Editor, Amstat News, American Statistical
Association, 732 North Washington Street, Alexandria 30 What Does a Teaching Associate Professor Do?
VA 22314-1943 USA, or email amstat@amstat.org. Pam Arroway, North Carolina State University
Items must be received by the first day of the preced-
ing month to ensure appearance in the next issue (for
example, June 1 for the July issue). Material can be sent
as a Microsoft Word document, PDF, or within an
email. Articles will be edited for space. Accompanying CAREER GUIDE
artwork will be accepted in graphics file formats
only (.jpg, etc.), minimum 300 dpi. No material in 34 Snakes and Ladders: Building a
WordPerfect will be accepted.
Amstat News (ISSN 0163-9617) is published Career in Statistics
monthly by the American Statistical Association, 732 David L. Banks, Duke University
North Washington Street, Alexandria VA 22314-1943
USA. Periodicals postage paid at Alexandria, Virginia,
and additional mailing offices. POSTMASTER: Send 43 Are You Media Shy…or Media Savvy
address changes to Amstat News, 732 North Washington Rosanne Desmone,
Street, Alexandria VA 22314-1943 USA. Send Canadian
address changes to Station A, P.O. Box 54, Windsor ON CAREER ASA Public Relations Specialist
N9A 6J5; returnsIL@imex.pb.com. Annual subscriptions Statisticians and other experts
are $50 per year for nonmembers. Amstat News is the
offer career advice for 46 Making a Name:
member publication of the ASA. For annual membership
rates, see www.amstat.org/join or contact ASA Member those looking to sharpen Up-and-Coming Statisticians on the
Services at (888) 231-3473.
their quantitative careers Verge of Great Things
American Statistical Association
732 North Washington Street
Alexandria, VA 22314–1943 USA
(703) 684–1221 • FAX: (703) 684-2036
ASA GENERAL: asainfo@amstat.org SPECIAL MATERIAL
ADDRESS CHANGES: addresschange@amstat.org
28 Where Are Your Colleagues?
AMSTAT EDITORIAL: amstat@amstat.org
ADVERTISING: advertise@amstat.org 38 Top Five Favorites
WEB SITE: www.amstat.org
Printed in USA © 2008
American Statistical Association
54 Math Is Music; Statistics Is Literature
(Or, Why Are There No Six-Year-Old Novelists?)
MISSION STATEMENT
Support excellence in statistical practice, research,
journals, and meetings. Work for the improvement DEPARTMENTS
of statistical education at all levels. Promote the
1 President's Invited Column
proper application of statistics. Anticipate and
meet the needs of our members. Use our discipline
to enhance human welfare. Seek opportunities to Cover design by Megan Murphy 61 Professional Opportunities
advance the statistics profession.
SEPTEMBER AMSTAT FINAL.indd 1 8/20/08 2:26:53 PM

PRESIDENT’S INVITED COLUMN
This month, I have invited Nathaniel Schenker of the National Center for Health Statistics to discuss how that
agency provides information about the nation’s health. Since before 1960, NCHS has monitored health statistics,
initially under the stewardship of Forrest Linder. It is a wonderful resource for data and information.
Statistics in Monitoring the Nation’s Health

Nathaniel Schenker, Senior Scientist for
Research and Methodology, NCHS/CDC
I
’ve spent the past nine years at the health topics, including vital events An NCHS statistician often works
National Center for Health and demography, health conditions as part of a team, together with other
Statistics (NCHS), having previ- and behaviors, access to and qual- statisticians as well as specialists
ously taught biostatistics at UCLA for ity of health care, and injuries and in areas such as computer science,
11 years and worked at the U.S. disabilities. Sources of NCHS data demography, economics, epidemi-
Census Bureau for three years before include birth and death certificates, ology, medicine, public health, and
that. NCHS, which is part of the as well as various sample surveys that sociology. There are ample opportu-
Centers for Disease Control and collect information via review of nities for individual research projects,
Prevention, is the nation’s principal medical records, personal interviews, as well. Moreover, collaborations fre-
health statistics agency. Its motto is physical and dental examinations, quently occur with researchers at other
“Monitoring the Nation’s Health,” and laboratory tests. federal and local government agencies,
and its mission is to provide statistical in industry, and in academia.
information that will guide actions ‘S’ Stands for ‘Statistics’
and policies to improve the health of Given that the “s” in NCHS stands for Mr. Academic Goes (Back) to
the American people. “statistics,” statisticians obviously have Washington
a major role in the work of the agency. When I decided to leave my tenured
Nathaniel What Is Done at NCHS? As do many agencies in the federal academic job in UCLA's Department
Schenker
When I used to take my infant son statistical system, NCHS has separate of Biostatistics in sunny California to
for a medical check-up, the pedia- divisions that focus on specific data col- return to the East Coast and work for
trician would plot his height and lection systems and topics. Statisticians the government, I was often asked,
weight on growth charts. Little did within these divisions work on the “Are you crazy? Why would you do
I know the charts were produced by design and implementation of the that?” Well, there were a number of
NCHS and were based primarily on data collections, analyses of the data, reasons, some of them personal, such
data from NCHS’ National Health and preparation of the data for release as my growing up in Washington and
and Nutrition Examination Survey to the public. Other statisticians (such the East Coast having a lot to offer.
(NHANES). Similarly, when cars as I) work in offices within the agency As a statistician, however, I was excit-
began to use unleaded gasoline, I that focus more on cross-cutting analy- ed by the prospect of working at the
didn’t realize the formulation of poli- ses and research and on the develop- source of health data that are impor-
cy to move away from leaded fuel was ment of statistical methodology, while tant and of interest to many outside
based on analyses of data on blood- also providing consultation to the data the agency. Indeed, at UCLA, I often
lead levels from the NHANES. (For divisions. Examples of methodologi- ran across NCHS data and researchers
more on this policy, see the President’s cal research areas at NCHS include analyzing them. Also, NCHS seemed
Invited Column in the May issue of optimal strategies for sample design, like a relatively small, friendly agency,
Amstat News). questionnaire design, modeling and so I didn’t feel as though I’d get lost
Indeed, NCHS collects and ana- estimation, data mining, developing working there. And, it seemed to have
lyzes data and produces information indexes of health, confidentiality, and a culture of conducting research, so
for the public on a wide range of handling missing data. I thought I’d be able to continue my
2 AMSTAT NEWS SEPTEMBER 2008

research career. (After coming to NCHS, race—estimates of the counts that would types (continuous, categorical, count) were
I also became an adjunct professor in the have been obtained had the prior standards used as predictors, but they sometimes had
Joint Program in Survey Methodology at been in effect (www.cdc.gov/nchs/about/ small amounts of missingness as well.
the University of Maryland, so I didn’t have major/dvs/popbridge/popbridge.htm). The I was involved in many aspects of this
to abandon academia completely.) Finally, bridging process used models predicting project (as with the bridging project), but
the Washington area has an exceptionally responses under the prior standards from especially in developing and evaluating the
active statistical community. For example, responses under the current standards and methods for imputation, writing technical
it is home to the largest chapter of the ASA: covariates. These bridging models were documentation, and consulting with ana-
the Washington Statistical Society. developed using data from NCHS’ National lysts using the multiply imputed data.
Do I miss being a full-time professor at a Health Interview Survey (NHIS), which has
fine university? Yes. Am I glad that I changed allowed multiple-race responses for many An Exciting, yet Sometimes
jobs? Absolutely. Besides the fact that years, but which also asks multiple-race ‘Scary,’ Job
changing jobs every once in a while helps to reporters for a single race group that best The two projects just described have sever-
'keep one’s blood flowing,' I’ve enjoyed the describes the person in question. al general characteristics in common. First,
atmosphere and work at NCHS. I’m less of I worked on many aspects of this project, they required teamwork, both among staff
a free agent in some ways now than I was but perhaps my biggest contributions were at NCHS and between NCHS and other
in academia (although I also don’t have to to the formulation of the overall approach organizations. Second, they required sta-
hustle for grant support the way I did), and used to solve the problem, the development tistical research. In fact, both projects led
there are some bureaucratic hassles involved of the bridging models, and the derivation to publications in JASA [98(464) and
in working for the government; but every of methods for assessing uncertainty in anal- 101(475)]. Third, the problems were
job has its own administrative headaches, yses that use the bridged data. complicated, and they required making
and any annoyances due to governmental A second project involved multiple assumptions and implementing approxi-
red tape have been far outweighed by the imputation of missing income data. The mations for their solution. Finally, there
opportunities I’ve had for interesting work. NHIS provides a rich source of data for was great interest in the data, and people
studying relationships between income both inside and outside the agency were
A Small Sample of My Projects and health and for monitoring health clamoring for release of the data while the
To illustrate some of the statistical work work was going on.
and health care for persons at different
that goes on at NCHS, I’ll briefly outline Projects with the second through fourth
income levels. However, the nonresponse
characteristics create the need to “push the
a couple of major projects in which I’ve rates are high (roughly 30%) for two key
envelope” with regard to methodology
been involved. items: total family income in the previous
under time constraints while maintaining
In 1997, the Office of Management calendar year and personal earnings from
and Budget revised the standards for clas- the integrity of the data. This results in a
employment in the previous calendar year.
sifying federal data on race and ethnicity. job that is exciting, yet sometimes ‘scary.’
To handle this missing-data problem and
A key provision was that respondents to allow analysts of the data to assess the Challenges and Opportunities
federal data collections be given the option uncertainty due to missing data, mul- Statistical work at NCHS, and in the
of choosing more than one race group to tiple imputation of these items has been government in general, is challenging
describe the person in question. Because the performed for several years (www.cdc.gov/ for administrative and technical reasons.
previous (1977) standards called for only a nchs/about/major/nhis/2006imputedincome.
For example, budgets are tight; costs are
single race group to be reported, data col- htm) in collaboration with researchers at
increasing for data collection, processing,
lected under the 1977 standards are not the University of Michigan.
and dissemination; there is a desire on the
comparable with data collected under the Several features of the data made the
revised standards. This can cause problems part of policymakers, researchers, and oth-
project particularly challenging and inter-
ers for more information; and the protec-
in trend analysis, the calculation of a vital esting. First, the data are hierarchical
tion of confidentiality is becoming more
rate—for which the number of events in in nature, with one of the key variables difficult. Oh, yes, and let’s not forget the
the numerator and the population size in reported at the family level and the other famous “graying of the federal work force,”
the denominator can come from different reported at the person level. Second, there with many employees nearing retirement
sources using different standards—and in were cases in which the value of one vari- age.
other types of studies that combine data able (e.g., personal earnings) could be Of course, with challenges come
classified under the two standards. restricted by the value of another variable opportunities. Difficulties due to sparse
The decennial census, a widely used (e.g., whether the person was employed), resources and conflicting desires and con-
source of denominators for vital rates, began but the values of both variables were miss- straints with regard to data will need to be
allowing multiple-race reporting in 2000. To ing simultaneously. Third, in some cases, addressed through the creative develop-
make the 2000 census data (and intercensal family income and personal earnings need- ment of efficient methods for collecting
and postcensal estimates) comparable ed to be imputed within bounds because new data and methods for getting the most
with data classified according to the 1977 partial information was available about out of analyses of existing data. And the
standards, NCHS, with assistance from the them (e.g., when a range was provided for ‘graying’ work force will result in openings
U.S. Census Bureau, produced “bridged” family income, rather than an exact dollar for new generations of statisticians to carry
census counts by county, age, sex, and value). Finally, several variables of various out such work. ■
SEPTEMBER 2008 AMSTAT NEWS 3

STATISTICIANS IN HISTORY
Kimiko Bowman
Never Say Never

Bowman Helps Provide Formula for Approximating the
Distribution of Maximum Likelihood Estimators
S
he could have been a horticulturist. She
had, after all, thought it would be exciting
to create new plants. But Kimiko Bowman,
better known by her byline as K. O. Bowman,
Val Snider chose to be a statistician and help develop a for-
ASA Publications Coordinator mula for approximating the distribution of max-
imum likelihood estimators.
It was by chance, really, that Bowman chose
statistics. Some might even say statistics chose
her. She began her college education majoring in
home economics at Radford College. “I wanted
to make sure I would graduate, and I thought
home economics would be relatively easy,” said
Bowman. “In my second year, the president of
the college called me into his office and advised
me to change my major to the science field. He
thought my future would be much better in sci-
ence. I [had] liked mathematics ever since I was a
small child, so I immediately changed my major
to math and chemistry.”

Kimiko Bowman, who became a U.S. citizen in 1958,
was the first ‘foreigner’ to receive a doctorate from the
University of Tokyo.
At Virginia Tech, Kimiko Bowman worked with L. R. Shenton, her thesis advisor, to provide a
formula for approximating the distribution of maximum likelihood estimators.
About a year later, Bowman graduated with a
bachelor’s degree and won a National Institutes
of Health fellowship in mathematical statistics space,” said Shenton. Bowman not only imple-
from Virginia Tech that sealed her interest in the mented the algorithm, but “it works, thanks
statistics profession. Three years after that, she to Dr. Bowman,” said Shenton.
had in hand a master’s degree and PhD in sta- Bowman’s life wasn’t all work, however. She also
tistics. She completed her education with a doc- found time to be a strong advocate for those with
“One
torate in mathematical engineering from Tokyo disabilities. She served on the National Science
University, but it was during her time at Virginia Foundation Equal Opportunities in Science and must not
Tech that her research really took hold. Engineering advisory committee and Committee
Bowman met L. R. Shenton, her thesis on People with Disabilities. She wrote a report
lose sight of
advisor, at Virginia Tech and—for the next while chairing the latter that resulted in grants what we want
45 years—worked with him to provide a for- being set aside to provide accommodation or
mula for approximating the distribution of special equipment for people with disabilities— to accomplish.
maximum likelihood estimators. They suc- which helped promote their participation in sci- So, I tried
ceeded by developing formulae for skewness ence and engineering at NSF.
and kurtosis statistics of maximum likelihood Bowman also chaired the Statistical Tracking of not to worry
estimators with respect to sample size and by Employment of People with Disabilities task force about slights
developing an approximation formula for the for the President’s Committee on Employment of
People with Disabilities. Her main duty was to or insignificant
”
percentage points of Pearson distributions.
By combining the two results, Bowman and monitor the questionnaire for Census 2000 so things.
Shenton were able to approximate the distri- relevant questions were asked about people with
bution of maximum likelihood estimators. disabilities and accurate results were obtained.
Another important accomplishment was Perhaps her interest in being a champion for
successfully implementing a divergent series people with disabilities stems from the physical
algorithm for large computers. “In sampling, hardships she, herself, has overcome. In her youth,
statistics series like this often occur, and the Bowman suffered from polio that paralyzed
algorithm for the coefficients is extremely her from the neck down. Doctors told her she
complicated and reaches four-dimensional would never walk again, but after two years of

Kimiko Bowman,
at Virginia Tech,
advised a num-
ber of gradu-
ate students
and published
three books and
approximately
200 papers.
rehabilitation, she was walking. Gladys Reynolds, In an effort to do more for minorities in
a statistician and good friend of Bowman’s from the statistics profession, Bowman also became
her Virginia Tech years, said, “I want to highlight a contributing editor to the Current Index to
her determination and perseverance. … In 1980, Statistics in 1977 and continued in this position
she was having polio-like symptoms and was for more than 10 years. During that time, she
diagnosed with Post Polio Syndrome. In spite translated articles from Japanese journals into
of this, she has continued to be as active and English, added keywords and abstracts, and
proliferate as always.” included them in the index. She also was asked
Bowman considers herself a “triple minority,” to go to Japan as a liaison scientist for the Office
being an Asian and a woman, as well as someone of Naval Research. She initially went for three
who suffers the effects of polio. “There were months, but continued to visit throughout her
many struggles I had to overcome,” Bowman career. In 1987, she was invited to attend the
said. “However, one must not lose sight of what International Statistical Institute meeting in
we want to accomplish. So, I tried not to worry Tokyo and to give a keynote address at the satellite
about slights or insignificant things.” meeting at Mt. Fujiyama. During the meetings,
Instead, she became one of the first women she received an audience with the Crown Prince
to be elected Fellow of the American Statistical and Princess of Japan—an unforgettable memory
Association in 1976, worked hard to advance sci- for her.
entific research using statistics, advised a number of In 1994, Bowman retired from Oak Ridge
graduate students, and published three books and National Laboratory as a senior research sci-
approximately 200 papers. One of those papers, entist. Retirement has not slowed her down,
“Tables for Determining Statistical Significance however. She continues to be a guest scientist
of Mutation Frequencies,” coauthored with M. A. in the Computational Sciences and Engineering
Kastenbaum, received a Citation Classic in 1989 Division at Oak Ridge, working on research in
for being the fourth most-cited paper in the his- distributional properties of estimators and test
statistics under non-normal sampling. She also
tory of Mutation Research, an international jour-
consults with colleagues within Oak Ridge on
nal. Additionally, Bowman is an elected fellow of
aspects of statistics such as procedures for validat-
the American Association for the Advancement ing computer models in economics and stochas-
of Science (1970), an elected member of the tic models in epidemiology and biology. In the
International Statistical Institute (1978), and an words of Reynolds, “As a woman, minority, and
elected fellow of the Institute of Mathematical a person with a disability, she has certainly been
Statistics (1987). an inspiration and mentor to many of us in the
scientific community.” ■

Interview with
Irene Hess
Sampling Is in
the Details
I
da Irene Hess was born in Muhlenberg County, Kentucky. Her father was a mining engineer in Central
City. Her mother, who graduated from Valparaiso College with her father, worked at home. After
graduating from Indiana University, Hess returned to Central City to teach math and English at the
local junior high school from 1932 to 1942. In 1940, she came to work with Leslie Kish at the Survey
Research Center. Under Kish, she trained many graduate students in the details of sampling. The following
is an interview with Hess, conducted in February 2008 by Fritz Scheuren, an ASA past-president.
[Scheuren] Irene, it is so good to see you well and still active pro-
fessionally at 97. Thanks for letting me interview you for all your
many friends and colleagues who want to catch up with all you
have done. Can we start with something about your family and
early life before you came to the University of Michigan?
[Hess] I grew up in Kentucky. In addition to my parents, I had
one sister, Beulah Marie. I was at Evansville College three years
and then went to Indiana University for one year for a bachelor’s
Fritz Scheuren degree, but that’s the only degree I have. I guess you want me to
talk about my start as a sampler?
[Scheuren] Well, of course, but a little more please about what
you did before that. Didn’t you teach in high school?
[Hess] No, I taught in junior high school in Central City,
Kentucky. I was really not happy doing something like that,
though. I was always interested in mathematics, and I would have
liked teaching mathematics at a higher level. And, when you are
teaching in public school, you first know that you’re to teach chil-
dren. And I really didn’t enjoy that. I couldn’t enjoy grade-school

Irene Hess (left and above) worked at the U.S. Census Bureau during the late ‘50s and stayed
there until her retirement in 1981.
Michigan, but his assistant was going to get married and move
away and Kish needed somebody else. I was recommended to
Kish and he wrote to me. In the meantime, at the [U.S.] Census
Bureau, I think at least twice, when I came to work in the morn-
ing, they told me I should leave. I had no retention points to
justify my continuing employment at the [U.S.] Census Bureau.
(When a person returned from military service, that person had
mathematics. I did not want to be responsible for teaching chil- to be re-employed.)
dren and encouraging them in what I would consider to be the [Scheuren] They were laying off people?
‘right way.’
So, I decided to try for statistical employment in the federal [Hess] Yes, that was because of the return of veterans from the
government. But to pass the civil service exam, you had to have war, and every time someone came back and had to have a job,
a minimum of six hours in statistics and I did not have that, so I they selected me to lay off. That happened twice. Joe Steinberg
went about getting it. I considered Indiana first, but there wasn’t worked hard to get me back each time. Anyway, on one Good
anything there that interested me at that time. I was aware of Iowa. Friday evening, I walked into my apartment and picked up my
Leslie [Kish] asked me once why I didn’t go to Iowa. I didn’t tell him mail and there was this communication from Leslie Kish out here
why, but my mother’s parents lived in Iowa. It was always so hot in Michigan.
in Iowa in the summer, so I just wouldn’t consider going to school When I got the letter, I felt it was going to be a job offer and
in Iowa because I just couldn’t take that heat. I ordered catalogues I knew that I was going to take it, but I didn’t want to. I liked
from several universities—Indiana, Michigan, and Kentucky, per- Washington and I always liked the [U.S.] Census Bureau. I
haps others. I came to Michigan two summers (1940, 1941). That responded to Leslie’s letter, and he invited me here to look around.
is how I got my six hours in statistics, and that was what I needed I’ve been here ever since, and that was 1954. In 1981, I retired. I’ve
for the civil service exam, which I passed. been retired for 27 years.
[Scheuren] And you came to Washington after that? [Scheuren] But you’ve been working here as a retired person,
every weekday ever since? That’s a real compliment to you
[Hess] Yes, it was right after the start of World War II in December and to the center.
and there was a civil service exam in May. The Bureau of Labor
Statistics sent out invitations, and that’s what I applied for and [Hess] Well it has been a long time. The way it began was I had
went first to the Bureau of Labor Statistics. Then, I was at the started this book [on sampling]. We had some very complex sam-
Bureau of Standards for a short time when they were developing ple designs for various projects and it bothered me that nobody else
the proximity fuse for bombs in the war. When the war was over, I around here knew anything about the sampling activities. So, I had
decided I better get away from the Bureau of Standards. Of course, written or was working on that book, Sampling for Social Survey
I had already been here, at Michigan, for two years because I was Research Surveys, 1947–1980. I decided to stay and finish it.
really interested in statistics. So, I went out to the [U.S.] Census
[Scheuren] So, that is how you are continuing to work after retire-
Bureau and talked to somebody out there. They were just organiz-
ment? What came next that kept you coming here every day?
ing the first sampling section at the [U.S.] Census Bureau.
Anyway, I joined the sampling section at the [U.S.] Census [Hess] Roe Goodman. I don’t know if you were ever acquainted
Bureau around 1944 or something like that. I was there until with Goodman.
I came here in 1954. Kish was in charge of sampling here in
[Scheuren] No, I wasn’t, but I know of his work.

Irene Hess worked with Leslie Kish, but
also worked with Mildred “Jean” Harter,
seen here with Hess (above and right) in
Hess’ office in the late ‘50s.
[Hess] Goodman was sold on controlled selection.

[Scheuren] Well, I am too. In fact, one of my former students,
Yan Liu, did her dissertation on balanced sampling with me—a
related idea.
Leslie Kish, ASA president in 1977;
[Hess] Goodman’s family was here in Michigan, so he came in one
he passed away in 2000.
day concerned about how well-controlled selection by computer
is compared with hand-drawn controlled selection. So, that’s how
we happened to have the book Controlled Selection Continued. As
I remember it, I think we were already well into the book before I
had to retire.
[Scheuren] You had to retire because of age? What was the age
you had to retire? [Hess] Yes, several. The last one of these was published in
September 2007, I think.
[Hess] Yes, I had to retire at 70. That was in 1981.
Anyway, Goodman was very interested in controlled selection and [Hess, changing the subject] As you know, the center has done a lot
the material we set up and worked on in the book. The last time of work on telephone surveys. But, I never got very deeply into the
I saw Goodman was one afternoon before Easter. Goodman’s wife telephone field. All of these things I worked on were face-to-face
was already in Kansas and he was going to go take the train and area probability surveys. I believed wholeheartedly in this approach.
join her. Anyway, he put in his final edits to the book. We didn’t And, hence, I never got involved in the telephone business.
know they were final edits, of course, at the time. Well, he went How it started at the center was Charlie Cannell and Bob Kahn
home to Kansas. He had a heart attack and died that night. That got a project to work on telephone sampling. One day, I don’t
made me determined to finish the book. know if it was Cannell or Kahn, asked me if there was somebody in
When my time came to retire, I probably still had something the sampling section who could work with them on this telephone
to do there. I don’t know if it was all done—probably not. I don’t project they had and I told them Bob Groves. And, I also told Bob
remember how long it took me to get these first two books finished Groves that Cannell and Kahn would be talking to him. And they
after I retired. did, and they worked on various projects, and Groves has been
with telephone surveys ever since. I was no admirer or interested in
[Scheuren] But there were more books after that weren’t there? the telephone. Anyway, I was getting to the point where I had to

[Hess] Early on, I did publish a number of papers with Kish. I
remember a paper in The American Statistician on nonresponse.
That paper was something I guess Kish shared with me, actually.
That was just shortly after I came here.
[Scheuren] That nonresponse paper that you and Kish did is a
wonderful piece of applied work. I have cited it many times. When
I used to do the History Corner in The American Statistician, I
republished it.
[Hess] After about 1960, Kish got out of the day-to-day sampling
activities. He was going to educate the world. So, we didn’t work
together much after that. For example, he did not have anything to
do with the projects that I was working on and what we were doing
with Roe Goodman. Kish was not involved with any of that. He
was just in a separate area of the center completely.
Irene Hess (left) celebrates her 97th birthday with Rhea Kish on [Scheuren] One last item? Can you talk about your involvement
August 26, 2007. with the Section on Survey Research Methods?
[Hess] Do you know how the section started? Within the Social
Statistics Section, there was a subsection, and I guess it existed for
retire, so it was not something for me to get into and I had enough maybe two or three years. But anyway, at one time, I was nomi-
of these other things [talked about earlier] to finish up. nated for chair of the subsection. There was an election within the
[Scheuren] You sure did. Look at these accomplishments, all done Social Statistics Section and I happened to have been nominated
after you retired, too. and then elected for chair of the subsection. In the meantime,
the powers that be decided to have a separate section for survey
[Hess] Now, I’m completely out of any project and I don’t have research, and they said—because I had just won the election—
anything that I really want to write about. So, I figure I ought to that I would automatically be chair of the new Section on Survey
clean up everything that I have around here and get out. Research Methods. So, I was the first chair of the Section on Survey
[Scheuren] Don’t do that. You do have more to do. At least help Research Methods, 1977.
me edit this interview, please? Do you remember this [showing the engraved silver plate given
to her by the Section on Survey Research Methods in 1998]?
[Hess] Bob Groves says he just likes to have me around.
[Scheuren] You still have this? I remember it well. I think I may
[Scheuren, changing the subject] Let me ask you about your work even have given it to you.
with Kish, and then we will finish up with your stint as the first
chair of the Section on Survey Research Methods. [Hess] That was my last annual meeting. I remember it very well.
I was surprised, overwhelmed.
[Scheuren] Irene, you have done such wonderful things for us all.
I’m going to read this inscription. The inscription says, “Irene Hess
for distinguished service and unstinting efforts in the furtherance
Books by Irene Hess of survey research methods.”
Probability Sampling of Hospitals and Patients (1961) [Hess] In July 1999, I had extensive surgery on my right leg, and
Irene Hess, Donald C. Reidel, and Thomas Fitzpatrick I have never been free of some kind of health struggle since that
time. And I have never been to another meeting of the American
Sampling for Social Research Surveys (1995) Irene Hess
Statistical Association, and I guess I’ll never get there again. But
Controlled Selection Continued (2002) Irene Hess and anyway, 1998 was very special. That was a real graduation for me.
Steven G. Heeringa [Scheuren] Oh my, you’ve done so many things, Irene, and most
The Practice of Survey Research at the Survey Research Center, of your major publications were finished after you retired. Earlier,
1947–1980 (1985) Irene Hess you were so busy day-to-day that you didn’t have time to write
these books. But, you did write them, eventually. How can we all
thank you? ■

Jack Chao-sheng Lee
The following excerpt is based on the 2007 article “In Memory

of Professor Jack Chao-sheng Lee (1941–2007),” which
appeared in the Journal of Data Science, Vol. 5, 143–150. For
a complete version, visit www.sinica.edu.tw/~jds/In-memory-of-
Jack-Lee.pdf.
C
hao-sheng Lee can be described as one of the most out-
standing people in Taiwan’s present statistical community.
His 40-year career of teaching and research has resulted in
nearly 100 academic papers in well-known international periodi-
cals. The National Chiao Tung University (NCTU) Graduate
Institute of Statistics, where Lee established himself when he
returned from the United States, has become a cornerstone for the
development of Taiwan’s growing statistical academic circles. The Prepared by the
NCTU Graduate Institute of Finance, which Lee helped estab- Institute of Statistics,
lish, has provided Taiwan’s financial sector with the talent needed National Chiao Tung University
for growth and innovation.
Lee graduated from National Taiwan University’s Department
of Business in 1964 and received his master’s in economics from
the University of Rochester in 1969. He studied under Seymour
Geisser at The State University of New York, Buffalo, and received
his PhD in statistics in 1972. Upon completing his education, he
began his teaching career at the University of Minnesota, and then
taught at Wright State University, as well as other universities, for
many years. He then went to Bell Labs and devoted himself to
research in economics and statistics.

In 1992, Lee was invited by Chi-Fu Den, the president of In order to adapt to the trend of global financial development,
NCTU, and Tsang-Hai Kuo, the dean of the College of Science, Lee participated in the establishment of the Institute of Finance
to return to Taiwan and take up the post of chair of the newly at NCTU in 2002 and took on the responsibility of head of the
established Institute of Statistics. Until stepping down from this institute for two years. During his tenure in charge, he was elected
post in 1999, he persuaded many top statistical researchers who University Chair Professor of NCTU in recognition of his dedica-
had graduated from well-known schools to return to Taiwan tion and accomplishments, as well as his reputation within aca-
and teach. He also encouraged the now common practice of demic circles.
research and acquiring software, as well as hardware. His efforts Lee had published many papers, and his research subjects
enabled the Institute of Statistics at NCTU to quickly rise included many fields in statistics, including multivariate analysis,
to the level of other established institutes of statistics, name- time series analysis, growth curves, Bayesian inference, and classi-
ly those of National Tsing-Hua University and National fication and pattern recognition. He also published in other areas,
Central University. such as portfolio management and option evaluation, with particu-
In August 1995, Lee was the first academic in Taiwan to lar application to finance.
receive the ASA Fellow honor from the American Statistical Lee conducted his academic studies conscientiously. Since his
Association. Lee had already been selected as a member of the return to Taiwan, he has supervised numerous exceptional master’s
International Statistical Institute in 1990 and elected as the and PhD graduates. He has four PhD graduates who are current-
president of the International Chinese Statistical Association in ly teaching in well-known universities and colleges. His master’s
1992. He was invited on multiple occasions to make speech- graduates have also held pivotal positions in numerous areas of
es around the world in many renowned schools, including business and academia. His pupils owe much to him, and have the
Harvard, Columbia, Cornell, and [the] University of Minnesota capability to build on his success and to further glorify what Lee
in the United States; Oxford and [the] University of London in has started.
England; Hitotsubashi University and Kobe University in Japan; Throughout his life, Lee lived plainly and simply, and his pas-
[the] University of British Columbia in Canada; and INSEAD sion for learning never diminished. He, now and always, will be a
in France, to name a few. great role model for scholars to admire and learn from. ■

Paul Dixon Minton
Building a
Department
P
aul Dixon Minton was born in Dallas, at SMU in the Department of Mathematics. It
Texas, the third of four sons in the family was at that point that Minton was introduced to
of William M. and Addie Evelyn Croft probability and statistics by Edwin Mouzon, who
Minton. Having grown up in the economically had, himself, written a dissertation on statistics
difficult times of the Great Depression, he was at the University of Illinois. After completing
able to attend college only because he received an his master’s degree, Minton was encouraged by
“emergency scholarship” to attend Southern Mouzon to continue his graduate studies in the
Methodist University (SMU). The university was new program in statistics at The University of
experiencing its own difficulties, struggling to fill North Carolina.
Dwight B. Brock,
Westat classes, and decided to offer such scholarships to Minton earned a PhD in statistics from
young Dallasites who had done well in high North Carolina State University under the
school, were recommended by their principals, tutelage of Gertrude Cox, during the time the
and could not otherwise afford to go to school. Institute of Statistics of the Greater University
Minton earned a bachelor’s and master’s of North Carolina consisted of the Department
degree from SMU, a school to which he would of Mathematical Statistics at UNC Chapel Hill
return later to found and direct a department of and the Department of Experimental Statistics
statistics. His studies were interrupted by World at NC State in Raleigh. Many famous statisti-
War II, during which time he worked as a crypta- cians were either faculty members or visitors
nalyst for the FBI. Following the war, he returned when Minton was a student there, including
to Dallas as an instructor and graduate student R. A. Fisher, Harold Hotelling, William Cochran,

Jacob Wolfowitz, Herbert Ellis Robbins, Richard was organized in his honor, attended by many
L. Anderson, Raj Chandra Bose, and others too former students and colleagues.
numerous to mention. Minton was the recipient of numerous academic
Minton once related a story in which Fisher was and professional awards, including being named a
sitting in the front row of the audience when he Fellow of the American Statistical Association and
(Minton) gave his first paper as a graduate student. being an early recipient of the ASA Founders Award
As if that fact alone was not enough to frighten for service to the association and the profession. In
an already nervous graduate student, the situa- his honor, the SMU Statistics Department created
tion became “infinitely worse” (Minton’s words) the Paul Minton Award for the student who scores
when Fisher left the room in the middle of the talk. highest on the basic qualifying examination. Also,
Minton told another story about his final PhD oral the ASA created the Paul Minton Service Award,
examination, in which he had a bad case of laryn- which has been given annually since 1992. He was a
R. A. Fisher
gitis that he attributed to nerves. To his knowledge, longtime member of the ASA, IMS, and Biometric
he had the world’s only silent orals. Society. He founded the North Texas Chapter of
Minton returned once again to SMU and the ASA and was active in a variety of roles dur-
began to build a set of courses in mathematical and ing the years he was in Dallas. Later, he was very
applied statistics for students from a wide range of active with the Virginia Academy of Sciences (the
subject-matter departments in the university. These Virginia Chapter of the ASA) and in organizing the
offerings gradually evolved into the formation of Southern Regional Conferences on Statistics. He
the Department of Statistics at SMU, now known participated in numerous committees, task forces,
as the Department of Statistical Science. At the councils, and boards, and he served one term as
same time, Minton began to recognize the impor- vice president of the ASA.
tance of computing in statistics, and, as a result of Minton had an active life outside the statis-
his expressed opinions, he was assigned to direct the tics profession that included a love of classical
first computer center at SMU, which housed the music and the ability to create terrible puns. He
Univac 1103, one of the few large scientific com- combined those interests with his knowledge of
Harold Hotelling puters available at the time. statistics in the following way. As part of the
Funded primarily by a training grant from the ASA’s 150 celebration in 1989 in Washington,
National Institutes of Health for training biostat- DC, he and some friends created a “statisti-
isticians, Minton established a new Department cal songbook,” from which they performed at
of Statistics at SMU in 1962. It was significant a banquet held on that occasion. In the hands
in obtaining faculty approval of the new depart- of these statistical punsters, the tune “Bali
ment that he had provided research consultation in Hai” from the musical “South Pacific” became
either statistics or computing—or both—to every the results for a test of statistical significance:
department in the university. The new department “Barely High.” These and other rewritten song
offered master’s and doctoral degrees following the lyrics can be found in the songbook, which was
North Carolina model and received consultation subsequently published in the August 1990
and assistance from Cox. The department subse- issue of The American Statistician.
quently expanded to offer degrees at all levels, to In a sense, entertaining audiences with humor-
provide consultative assistance to faculty research ous statistical lyrics to well-known songs became
and outside clients, and to conduct research in sta- another way for Minton to be an advocate for
tistical theory and methods. statistics and promote the profession to a wider
Jacob Wolfowitz
After 10 years of developing and administering world. His life and work as a statistician were
the SMU department, Minton moved to summarized at his memorial service in the words
Richmond, Virginia, to take the position of of his son, Roland: “He recruited students to
dean of the School of Arts and Sciences at the newly developing field of statistics, he found
Virginia Commonwealth University in 1972. money to support them during their student
There, in addition to his duties as dean, Minton days, and he helped them find jobs after they
formed an Institute of Statistics—a form of graduated. Others have said of him that he was
liaison office between the Department of a gentle person, with a gentle temperament, but
Mathematical Sciences in the School of Arts and he changed the lives of hundreds, if not thou-
Sciences and the Department of Biostatistics in sands of statistical students, just as he changed
the Medical College of Virginia. He also was the lives of many others who knew him outside
active in statistical consulting in local industry, the profession.”
including paper manufacturing, polymer plastics A memorial session for Minton was held at the
processors, pharmaceuticals, a national baking 2008 Joint Statistical Meetings in Denver. Papers
William Cochran company, and a federal agency. He retired from from that session will be included in the ASA
VCU in 1988. That year, a special symposium Archives Collection at Iowa State University. ■

Mollie Orshansky:
Courtesy of the Social Security Administration Archives

Author of
the Poverty
Thresholds
Editor’s Note: Views expressed in this article are those of the author, and should not be construed as representing the
policy of the U.S. Department of Health and Human Services. This article is condensed from an unpublished draft
paper with references. That paper is based on extensive research involving published articles and documents, unpub-
lished documents, and conversations with Mollie Orshansky.
M
ollie Orshansky was born on January 9, A daughter from a poor family graduating from
1915, in the Bronx in New York City. college during the Great Depression, Orshansky did
She was the daughter of Ukrainian- not have the luxury of attending graduate school
Jewish immigrants who spoke limited English. before she started working. Instead, she found a job
Although her father worked hard at a number of as a statistical clerk in the New York Department of
jobs, Orshansky and her sisters grew up poor—in Health’s Bureau of Nursing, where she worked on
her words, the family could “barely … make ends infant mortality and other subjects for a year.
Gordon M. Fisher meet.” The girls slept two to a bed and wore hand- In 1936, the U.S. Children’s Bureau (then part of
me-down clothing. Orshansky remembered going the U.S. Department of Labor) offered Orshansky a
with her mother to stand in relief lines for surplus job as a junior statistical clerk. She accepted the job,
food. As she was to say later, “If I write about the moving from New York to Washington, DC. Her
poor, I don’t need a good imagination—I have a first job assignment involved logarithmic equations
good memory.” for 600 infants who had been in a study. In July
Orshansky was both the first high-school 1939, the bureau promoted her to research clerk, a
graduate and first college graduate in her family. job in which she stayed until January 1942, work-
She graduated from Hunter College High School ing on biometric studies of child health, growth,
in Manhattan (then a school for gifted young and nutrition.
women) in 1931. Because she received two schol- It was while she was working at the U.S.
arships from the college, she was able to attend Children’s Bureau that Orshansky began taking
Hunter College (at that time a women’s college). graduate courses. At various times between 1936–
She graduated from Hunter in 1935 with an AB 1937 and 1960, she took courses in economics and
in mathematics and statistics. She was a statistician statistics at the U.S. Department of Agriculture
by training and profession, although she has some- Graduate School and American University. From
times been referred to as an economist. January 1942 to March 1943, Orshansky took a job

as a statistician for the New York City Department of Health, working
on a survey of the incidence and therapy of pneumonia.
How Mollie Orshansky Talking about her government career, Orshansky once comment-
ed, “I basically always worked with women, except when I was in the
Developed the Poverty Thresholds war agencies [the National War Labor Board during World War II
When Orshansky developed her poverty thresholds, she and the Wage Stabilization Board during the Korean War].” During
made use of information she had worked with at the U.S. World War II, in particular, large numbers of men left civilian jobs to
serve in the military. As a result, a number of female workers were able
Department of Agriculture (USDA). She based her thresholds
to get jobs that would not have been open to them under ‘normal’
on the economy food plan, which was the cheapest of four
circumstances. Describing the situation for female federal workers,
food plans (hypothetical food budgets providing nutritious
Dorothy Rice, a colleague of Orshansky’s, later said, “Any female that
diets at different cost levels) developed by USDA. had anything on the ball really did very well during the war. All the
From a finding of USDA’s 1955 Household Food Consumption men went to the war and we had to carry on.”
Survey (the latest such survey available during the early In March 1943, Orshansky secured a job at the U.S. National War
1960s), she knew families of three or more persons had spent Labor Board as chief of the Program Statistics Division of the Research
approximately one-third of their after-tax money income on
and Statistics Branch. She planned and executed the collection and
analysis of data required for the board’s decisions on wage adjustments
food in 1955. The way in which she used this survey finding
and the effects of the board’s stabilization activities. She stayed in this
was by considering a hypothetical average family that was
job until September 1945, when the operations of the board were
spending one-third of its income on food and by assuming
being terminated.
the family had to cut back on its expenditures sharply. She From September 1945 to March 1951, Orshansky worked as
assumed expenditures for food and non-food would be cut a family economist at the Bureau of Human Nutrition and Home
back at the same rate so the family would continue to spend Economics of the U.S. Department of Agriculture (USDA). She con-
a third of its income for food. ducted research in family consumption and levels of living, carrying
out a variety of assignments. She was in charge of preparing data on
When the food expenditures of the hypothetical family
the estimated value of household furnishings and equipment on farms
reached the cost of the economy food plan, she assumed
for the department’s balance sheets of agriculture for 1947–1950.
the amount the family would spend on non-food items
In 1948, Orshansky and a colleague were responsible for respond-
would also be minimal, but adequate. (Her procedure did ing to letters from members of the public asking how they could make
not assume specific dollar amounts for any budget category ends meet on their existing income in the face of the severe (by U.S.
besides food.) Following this logic, she calculated poverty standards) post–World War II inflation. Orshansky and her colleague
thresholds for families of various sizes by taking the dollar would send the letter writers pamphlets about preparing a family bud-
costs of the economy food plan for families of those sizes and get and planning low-cost and moderate-cost meals using USDA’s
multiplying the costs by a factor of three—the “multiplier.” food plans—hypothetical food budgets providing nutritious diets at
(She followed somewhat different procedures to develop different cost levels. (This shows Orshansky was working with USDA’s
thresholds for two-person and one-person units.) food plans at least 15 years before she was to use them to develop her
poverty thresholds.)
She differentiated her thresholds not only by family size, but About 1949, Orshansky carried out an assignment to update a stan-
by farm/nonfarm status, by the gender of the family head, by dard budget (an estimate of necessary living costs) for a single work-
the number of family members who were children, and (for ing woman in New York. In June 1950, she presented a paper titled
one- and two-person units only) by aged/non-aged status. “Equivalent Levels of Living: Farm and City” at the annual meeting
The result was a detailed matrix of 124 poverty thresholds, later of the Conference on Research in Income and Wealth (CRIW). Her
reduced to 48. paper and the other papers presented at the meeting (including one
by Milton Friedman) were published in volume 15 of CRIW’s annual
To avoid confusion, the preceding explanation has been
(and still continuing) series, Studies in Income and Wealth.
phrased in terms of the economy food plan. However, In March 1951, during the Korean War, Orshansky secured
Orshansky actually developed and discussed two sets of a job at the U.S. Wage Stabilization Board as the director of the
poverty thresholds, one derived from the economy food plan Program Statistics Division of the Office of Economic Analysis. She
and one derived from the somewhat less stringent low-cost planned and directed the board’s statistical program until August
food plan. (The latter set was the one she preferred.) It was the 1953, when the operations of the board were being terminated due
lower of the two sets of poverty thresholds—the set derived to the end of the war.
from the economy food plan—that the Office of Economic In 1952, while Orshansky was working at the U.S. Wage
Opportunity adopted as a working definition of poverty in Stabilization Board, she was elected a member of the Econometric
May 1965. One probable reason for the adoption of the lower Society (an international society for the advancement of economic
set of thresholds was that the lower set yielded approximately theory in its relation to statistics and mathematics). Of the 73 can-
the same number of persons in poverty as the Council of didates for election to membership in 1952, Orshansky was the
Economic Advisers’ $3,000/$1,500 poverty line. only woman.

From August 1953 to February 1958, Orshansky worked as a food
economist for USDA in a successor office to the Bureau of Human Orshansky told several interesting stories about
Nutrition and Home Economics, where she had worked before. She events during the development of her thresholds:
planned and directed the collection and analysis of data on food con- ➤ As noted in the sidebar, “How Mollie Orshansky Developed the
sumption and expenditures of American households. Poverty Thresholds,” Orshansky based her thresholds on USDA’s
In 1956 and 1957, Orshansky was the senior coauthor of two economy food plan. She had been working with the USDA food
reports on family food expenditures and food consumption based on plans at least as early as 1948, and so was familiar with them. In
a food consumption survey of rural families in the North Central the context of developing poverty thresholds for families, she
[Midwest] region. She was one of a number of people who gave tech- became concerned about the economy food plan not allowing
nical assistance in the preparation of a series of reports on USDA’s
for purchases of food away from home, either at work or school.
1955 Household Food Consumption Survey, and she wrote a
For the purpose of developing poverty thresholds, she wanted
major section of a summary report on the same survey. This 1955
to modify the cost of the food plan by adding $0.15 a day per
Household Food Consumption Survey was the source from which
Orshansky would calculate the “multiplier” she later used to develop person to it to allow for the husband in a family to buy coffee
her poverty thresholds. at work or for children to buy snacks. However, her supervisor
In February 1958, Orshansky went to work for the Social Security would not allow her to do so.
Administration (SSA) in an office that later became the Office of ➤ One major source for Orshansky’s July 1963 article was a
Research and Statistics (ORS). ORS seems to have been one of a
special tabulation of Current Population Survey (CPS) data that
small number of federal offices that provided significant work oppor-
the Social Security Administration (SSA) purchased from the
tunities for women professionals at this time; during Orshansky’s first
U.S. Census Bureau for $2,500. The results showed the median
decade there, both the director and the deputy director were women.
Orshansky had several titles at SSA, but can best be described as a annual income of nonfarm, female-headed families with
social science research analyst. children was $2,340. Orshansky was horrified when she realized
Orshansky performed a number of assignments during her early half of these families lived for a year on less than SSA paid for
years at SSA. Her first was to prepare an article on standard budgets one statistical tabulation. She later commented, “I determined I
(family budgets) and practices in setting fee scales in 21 large cities. was going to get my $2,500 worth.”
She also prepared several annual updates of an analysis of the income
➤In the CPS, thousands of sample cases represent millions
sources of “young survivors” (widows under age 65, particularly those
of families and persons in the general population. Published
with minor children). She prepared a medical care standard for the
U.S. Census Bureau reports based on survey data always
Budget for an Elderly Couple, of which the Bureau of Labor Statistics
(BLS) was preparing an interim revision. She also prepared data for show figures that relate to the national totals of families and
16 charts on financial resources of the aged for the Chart Book for the unrelated individuals. However, when Orshansky got the results
1961 White House Conference on Aging. of her $2,500 tabulation of CPS data, they gave her only the
While Orshansky’s development of the poverty thresholds was a unweighted counts of sample households in various poverty
major milestone in both social policy history and statistical history, it and nonpoverty categories; SSA’s payment to the bureau had
grew out of ordinary work activities—an “answer for the record” for not been enough to pay for computing weighted national
a congressional hearing and an in-house research project. During a totals for her, so she had to do the work herself, “by hand,” to
1960 congressional hearing, a senator asked HEW Secretary Arthur calculate weighted national totals from the unweighted sample
Flemming if he had figures on how much it costs a retired couple to count. She also calculated the poverty gap “by hand” for her
live. Flemming said HEW would provide an answer for the record, January 1965 article. She didn’t even use a calculator.
and Orshansky was the civil servant who prepared an unattributed
submission for the record. She mentioned the Budget for an Elderly
Couple, which BLS was then revising, and a similar budget prepared In January 1964—only six months after the publication of
by a group in New York. In addition, she provided two rough mea- Orshansky’s obscure article—President Lyndon Johnson declared a
sures of income inadequacy for an elderly couple that she developed war on poverty. In a chapter on the problem of poverty in its 1964
by applying multipliers derived from USDA’s 1955 Household Food annual report, the president’s Council of Economic Advisers (CEA)
Consumption Survey to the cost of USDA’s low-cost food plan (at that put forward its own rough measure of poverty: $3,000 for families
time, the cheapest of USDA’s three food plans)—almost exactly as she of all sizes and $1,500 for unrelated individuals. The CEA’s $3,000
was to do several years later in her 1963 and 1965 poverty articles. figure was not derived in any way from Orshansky’s work; however,
In early 1963, Orshansky was assigned to do an in-house research the CEA report did cite two dollar figures from Orshansky’s July 1963
project on poverty as it affects children. At that time (the year before article (without giving her name as the author) to show that the CEA’s
the War on Poverty was declared), there was no generally accepted $3,000 figure was a reasonable level for a poverty line.
measure of poverty, so to carry out this research project, Orshansky When Orshansky saw the January 1964 CEA report (including the
developed one (see “How Mollie Orshansky Developed the Poverty reference to her dollar figures), she was disturbed by the CEA’s failure
Thresholds” on the previous page for her methodology). In July 1963, to vary its $3,000 family poverty line by family size, as this resulted
she published results of her research project in a Social Security Bulletin in understating the number of children in poverty relative to aged
article, “Children of the Poor,” in which she also described the initial persons. The CEA figure “led to the odd result that an elderly couple
version of her poverty thresholds.

Mollie Orshansky (third from left) at the first Conference on Women in the War on Poverty on May 8, 1967. With her are Mary Keyserling, Josephine
Weiner, and Hyman Bookbinder. Near the end of her speech, Orshansky said, “…our statistics, imperfect though they may be, show us where prob-
lems are even if they cannot always reveal exact dimensions…[C]alculations … relating to poverty … exist only to help make them disappear, and so
if we can think bold solutions and dream the big dream, we can wipe out the scourge of poverty before we all agree on how to measure it.”
with $2,900 income … would be considered poor, but a family with Orshansky was a member of this group. The technical working group
a husband, wife, and four little children with $3,100 income would reviewed and contributed to the plan for what became the Income
not be.” In addition, the president’s declaration of a war on poverty Survey Development Program—the research and development phase
evidently led SSA to give a higher priority to Orshansky’s poverty for the Survey of Income and Program Participation.
work. As a result, Orshansky’s supervisors asked her to do an analysis In 1982, Orshansky retired from SSA after a government career
extending her families-with-children poverty thresholds to the whole that lasted for more than 40 years. She died on December 18, 2006,
population. She completed this analysis in late 1964 and it was pub- in New York City.
lished in the Social Security Bulletin in January 1965 as “Counting the Orshansky received a number of honors for her achievements.
Poor: Another Look at the Poverty Profile.” She received a Commissioner’s Citation from the Social Security
The publication of Orshansky’s January 1965 article came when Administration in 1965 for her creative research and analytical work
the Office of Economic Opportunity (OEO)—the lead agency for and the Distinguished Service Award (the department’s highest recog-
the War on Poverty—was being set up. OEO officials were enthusi- nition of civilian employees) from HEW in 1976 for her “leadership
astic about Orshansky’s poverty thresholds, considering them to be in creating the first nationally accepted measures of income adequa-
an advance over the CEA’s $3,000-for-all-family-sizes figure. OEO cy and applying them diligently and skillfully to public policy.” In
research chief Joseph Kershaw commented, “Mollie Orshansky says 1974, she was elected a Fellow of the ASA for her leadership in the
that when you have more people in the family, you need more money. development of statistics for the measurement of poverty. After her
Isn’t that sensible?” In May 1965, OEO adopted Orshansky’s thresh- retirement, she received a national award from the Children’s Defense
olds as a working definition of poverty for statistical, planning, and Fund in 1989 and an Award for Distinguished Contribution from the
budget purposes, and, in August 1969, her thresholds were made the American Political Science Association in 1993.
federal government’s official statistical definition of poverty. Orshansky’s achievements also were recognized in a very different
When she developed the poverty thresholds, Orshansky was “an setting. She may be the only statistician to have been discussed on a
obscure civil servant” who worked “[d]own a dimly lit hall, among major television show. One subplot of “The Indians in the Lobby,”
stacks of computer printouts [at] a paper-covered desk …” However, an episode of “The West Wing” originally broadcast in November
after her thresholds were adopted as the federal government’s poverty 2001, involved the adoption of a new poverty measure, and one char-
line, she became much more well known. Because of frequent citations acter alluded to how Orshansky developed the current poverty mea-
of her work in academic articles and books, someone once referred to sure. While the discussion of issues relating to a new poverty measure
her as “the ubiquitous footnote.” Besides presenting papers at a num- sounded plausible, the episode grossly mischaracterized the rationale
ber of professional meetings and publishing a number of articles, she for Orshansky’s methodology for developing the poverty thresholds.
testified and/or provided written documents to congressional commit- Of the contributions to American public policy that Orshansky
tees on 10 occasions between December 1967 and 1990. made during her career, the greatest by far was her development of the
On five occasions between 1968 and 1980, Orshansky participated poverty thresholds. The poverty line has become a major feature of
in federal interagency committees that reviewed the poverty thresh- the architecture of American social policy. Although the measure may
olds. The 1968–1969 committee made two modest revisions in have its shortcomings, the poverty line gives us a means of identifying
the thresholds, and it was the thresholds with these revisions that and analyzing the makeup of the groups in our society with the least
were made the official federal statistical definition of poverty. In resources. Orshansky’s thoughtful analyses of the poverty population
1981, several minor changes recommended by the 1979–1980 began a tradition, and there are now numerous researchers and advo-
committee were made. cates who conduct such analyses and draw policy implications from
Following up on a 1973 subcommittee’s recommendation for a them. Even though there may not be consensus on answers, the ques-
new income survey vehicle, the HEW Technical Working Group on tion “How does it affect the poor?” has become a test for proposed
income data and models proposed that a new survey be developed policies and programs. And a simplified version of the poverty line is
to provide better information on the income and related characteris- used to determine eligibility not only for certain federal programs, but
tics of the population and on participation in government programs. for a number of state, local, and private programs, as well. ■

The ASA
Brings the Library
to Your Desktop
Did you know your ASA membership includes online access to the
Journal of the American Statistical Association
Journal of Business & Economic Statistics
and
The American Statistician?
Log in at MEMBERS ONLY today to access your journals!

www.amstat.org/membersonly

Day in the
LIFE
1
9

Life as a Stochastic Modeler
Alan Gelfand, Duke University
F
irst, let me say I am flattered to be invited to write about my
career in statistics. I do not know if many readers will find it
interesting, but I do know it has evolved along a path that
may be viewed as a reflection of the evolution of our field.
What I have become is a stochastic modeler. In particular,
a modeler whose primary focus is in the area of environmental
processes, including ecological systems, exposure assessment, and
climate processes. A common ingredient of this analysis is the col-
lection of data across space and, often, across time. I am usually
studying a complex process with different types of information—
theoretical results drawn from physical principles, mechanistic
insights based on knowledge of aspects of process function, and
empirical knowledge as a result of previous data collection and
relevant laboratory and field experiments.
I imagine the process is described at multiple levels, with the
foregoing information entering in at different places and lev-
els. I typically represent the process through an acyclic-directed
graph with some nodes observed and others unknown, and then
I infer about the unknown nodes given the observed nodes. I
formulate the joint model in a hierarchical fashion driven by
the graph and fill in the stochastic details needed to complete
the model specification. In essence, a stochastic modeler seeks
the posterior distribution of what we don’t know given what we
have observed, so we usually fit these graphical models within
the Bayesian framework.
It has emerged, more clearly than ever, that such modeling is
my greatest strength as a statistician. Moreover, being part of a
team of researchers assembled to ‘brainstorm’ a complex prob-
lem is an exceptionally stimulating and rewarding activity. In
this setting, the modeler becomes a central player in synthesizing
inputs from team members, shaping progress on the problem,
and becoming a richer scientist as a result.
I think this research view serves as a contemporary perspective
of our field. The team research concept, which it presumes, dra-
matically revises the role of the statistician from someone brought
in at the end to carry out data analysis and create ‘pretty’ pictures.
Rather, the statistician is able to illuminate what we can learn with
what we have, as well as what we need to collect to learn about
what we want. In the midst of all this, it is essential that the mod-
eler retain technical rigor, attention to detail, and appreciation of
the properties and features encompassed by the modeling.

How Did I Get Here? which led me to empirical Bayes and, eventually, research within
So, that is the story of what I do. But, how did I get here? I fol- the Bayesian paradigm. Finally, Adrian Smith was most influen-
lowed a fairly typical course for my time, starting as an under- tial to me. Following conversations with him in the mid- to late
graduate major in mathematics. In my senior year, I took an ’80s, I arranged a sabbatical to work with him at Nottingham
introduction to mathematical statistics course from the book by University. I arrived seeking to use his numerical integration
the same name written by Robert Hogg and Allen Craig (embar- software for certain empirical Bayes problems and left with the
rassingly, using the first edition, spanning all of 240 pages). The Gibbs sampler!
material was so elegant, I was smitten, and I applied to graduate In this regard, I was fortunate to be involved in a seminal contri-
programs in statistics, moving to Stanford—as far from New bution to our field. Indeed, working with Smith, it was a remark-
York City as I could go. able feeling to come upon the Gibbs sampler from the 1984 paper
I thoroughly enjoyed my four years at Stanford. (The late of Stuart Geman and Donald Geman, “Stochastic Relaxation,
1960s was a remarkable time for a statistics department in the Gibbs Distributions, and the Bayesian Restoration of Images,” and
San Francisco Bay area—and for the country.) I emerged with recognize it was better suited for handling Bayesian computation
training as a mathematical statistician, really the only path avail- than it was for its original purpose—sampling Markov random
able at that time. As I have told many people, I had a ‘wasted fields. In fact, for the first half of the 1990s, I spent almost all
youth’ in some ways, not discovering until the late 1980s that I my time writing as much about the use of Gibbs sampling and
was born to be a Bayesian (and, with it, to be a modeler). Markov chain Monte Carlo algorithms for Bayesian model fitting
People who influenced me include my adviser from Stanford, and model determination as I could. These were heady times for
Herbert Solomon, a wonderful, generous man who was particu- those of us involved, as we realized we were elaborating on a tech-
larly good at recognizing how people could best contribute to nology to successfully address the problem of Bayesian compu-
projects. He was one of the earliest ‘team builders’ in our field. tation—to replace high-dimensional integration with simulation
Charles Stein, who was also from Stanford, influenced me in from high-dimensional distributions.
a distant way. I believe I hold the record for the most courses In the mid 1990s, I started moving from research confined to
ever taken from Stein by a single student. His beautiful, deep computational matters to the analysis of spatial and spatial-temporal
thoughts in decision theory led me down that research path, data. Again, I was fortunate. Spatial statistics had struggled as
LATEX
LATEX
LATEX
MacKichan
SOFTWARE, INC.

a field, functioning at the periphery of mainstream statistical
work, and was particularly in need of inexpensive, high-speed
computation to enable tools vital for visualization. Fortunately,
this computing power arrived and, with it, the advent of geo-
graphic information systems (GIS) software to create effective
Welcome to ASA
maps that tell a broad range of stories. However, GIS software
had an obvious limitation from a statistician’s perspective—it was
descriptive, but not formally inferential. A wonderful opportu-
nity for stochastic modeling revealed itself, an opportunity with
few players at the time and a growing need in the applied com-
munity as more data with spatial content were being collected.
Here, again, there was more good fortune for me, because
advantages to hierarchical modeling for analyzing spatial data
became evident—the advantages we attach to working in gen-
eral in the Bayesian framework: fully model-based inference with
accurate assessment of uncertainty. Additionally, the customary
asymptotics, which are employed in time series, so-called expand-
ing domain asymptotics, may be inappropriate for spatial data
where infill asymptotics are perhaps more relevant. However,
that the latter asymptotics typically reveal information about
unknowns is bounded in terms of inference. In this regard, the
Bayesian framework provides exact inference, avoiding possibly
inappropriate asymptotics. Of course, the implied caveat is that
the data never overwhelm the prior; we must be more attentive
to prior sensitivity than in other areas.
Today, my good fortune persists through continuing research
relationships with talented former students. In particular,
Sudipto Banerjee and Brad Carlin, both at the University of
Minnesota, invaluably helped to shape my spatial data research
agenda. During my career, I also found wonderful interdisciplin-
ary collaborators. When I was at the University of Connecticut,
I worked seriously at building bridges with ecology and evolu-
tionary biology, yielding productive research relationships with
John Silander and Kent Holsinger. Moving to Duke University Make the most of your
made things even better, as Duke is a naturally collaborative
institution; research teams develop over the campus with strong
ASA membership.
encouragement from the university. What program on campus
is more naturally interdisciplinary than statistics? Now at Duke,
Visit the ASA
I have productive relationships with researchers in the Nicholas Members Only site!
School of the Environment, particularly Jim Clark and Marie
Lynn Miranda.
In summary, it seems my career has been one of ongoing good -Personalize your login ID
luck. To some extent, I believe this is true, but I also believe -Access all your ASA journal, CIS, and
you can make your own luck. Scientific curiosity, receptivity to JSTOR subscriptions with one easy login
reading in other fields, availability for collaboration, willingness
to listen, and attention to effective communication all facilitate -Partner discounts from Wiley, CRC Press,
creating opportunities. In addition, developing the full toolkit— and others
strong theoretical training, stochastic modeling expertise, mod-
ern computing skills, and data analysis experience—enables one -ASA member forums
to take advantage of these opportunities. Some of us work better -Enews archive
with certain parts of the toolkit and enjoy certain parts more
than others; however, as statisticians in the 21st century, we can
each find our own way to contribute and, in that sense, we are
all fortunate. ■

DAY IN THE LIFE
On Becoming a Teacher
Daren Starnes, The Lawrenceville School
A
s a fledgling mathematician, I was blessed to learn from sev-
eral outstanding teachers. First and foremost, there was
Sammi Yopp, my second-grade teacher at the North Carolina
Charlotte Country Day School. Yopp sparked my mathematical
curiosity and cajoled me into playing chess as a means of honing my
analytic thinking skills. Several years later, my mathematical abilities
were challenged by Sue Sams, my ninth-grade honors algebra II/
trigonometry teacher at East Mecklenburg High School. Sams was,
to put it mildly, a disciple of Mary P. Dolciani. The following year,
I landed in the classroom of another Dolciani disciple, Roger
Bruwhel at West Charlotte High School. Bruwhel was the one who
encouraged me to leave my comfort zone by applying to the North
Carolina School of Science and Mathematics (NCSSM). Off to
NCSSM I went, where I was surrounded by a host of talented
“mathletes,” budding scientists, and computer programmers. As a
junior sitting in Dan Teague’s “Calculus with Topics” class one day,
I made the decision to become a high-school math teacher. Why?
Because I witnessed firsthand the tremendous passion, thirst for
1
understanding, and knack for asking just the right question at the
right moment that allows great teachers such as Teague to inspire
students to do extraordinary things.
From NCSSM, I went on to pursue a bachelor’s of science in
mathematics with secondary teaching certification at the University
of North Carolina, Charlotte (UNCC). At the time, UNCC was
a teaching university, which enabled me to take all my classes with
professors from the outset. At the urging of my professors, I par-
ticipated in Budapest Semesters in Mathematics, where I fell in
love with combinatorics and probability. Having never left the
southern United States until that moment, I admit to experiencing
a healthy dose of culture shock as a westerner in what was then
socialist Budapest.
From Eastern Europe, I accepted an offer to join a graduate
program in pure mathematics at the University of Michigan. As I
was nearing the end of my master’s degree requirements, it became
SEPTEMBER
SEPTEMBER
20082008
AMSTAT
AMSTAT
NEWS
NEWS2525

“
who were not bound for AP Calculus. When AP Statistics kicked
If I have become a teacher leader, it is off in 1996–1997, I happily agreed to play a leadership role. Better
still, CCDS science colleague Wright Robinson convinced me to
because I have followed in the footsteps team teach AP Statistics with AP Environmental Science, to col-
of the many great mentors and role mod- laborate on a field and lab manual, and to co-present at several
”
conferences. Robinson got me started as a writer, having himself
els who have shaped my thinking. authored several children’s books about ecological topics. More
than anything, Robinson convinced me to “get out there” as a con-
tributing member of my profession.
When our youngest son graduated from CCDS, my wife and
clear to me that my interests had shifted dramatically away from I headed west to The Webb Schools in California to begin the
classical pure mathematics in the direction of research methods, next chapter in our lives. I assumed leadership of the mathematics
combinatorics and graph theory, and statistics. Ironically, my only department, and with the support of some exceptional colleagues,
statistical coursework to that point was a calculus-based probability began to tweak the existing curriculum. We dissolved our existing
and statistics course as a sophomore at UNCC. Having reached an geometry course and created an integrated geometry with algebra
important fork in the road, I opted to end my relationship with course that continues to evolve today. In addition, we added a cap-
pure math and begin my high-school teaching career. stone course for our most talented math students titled “Calculus-
Charlotte Country Day School (CCDS) was my proving Based Probability and Statistics.”
ground, the place where I learned the craft of teaching from sup- Externally, my role with the College Board expanded dramati-
portive colleagues such as Tim Timson, Sheila McGrail, and Sue cally. I led four or five one-day workshops for AP Statistics teachers
Schwartz. My interest in curriculum development blossomed. every year and a similar number of week-long summer institutes.
CCDS launched the International Baccalaureate (IB) program, and These opportunities to help coach other teachers gave me a chance
I volunteered to take on the IB Higher Level Math course. In my to share some of the lessons I learned in the trenches. At the same
fifth year there, McGrail and I designed a course called “Probability, time, interacting with other highly motivated professionals encour-
Statistics, and Finite Math” to better serve the needs of students aged me to further refine my own teaching.
In the summer of 1998, I began reading AP Statistics exams.
Every year since, I have attended this annual gathering of high-
Environmental Section school and college/university statistics teachers, which some of us
refer to fondly as “stats summer camp.” For the past eight years,
Upcoming ENVR Workshop! I have been privileged to serve as part of the reading leadership
team, first as a table leader and, more recently, as a question leader.
Needless to say, reading hundreds of exam papers every day is not
What: Statistical Issues in Monitoring the Environment
the highlight of the experience for most of us. Instead, it’s the tre-
When: October 22–24, 2008 mendous opportunity to network with like-minded professionals
Where: National Center for Atmospheric Research who have amassed an abundance of content knowledge and a host
(NCAR), Boulder, Colorado of resources for teaching introductory statistics. There are abundant
opportunities for discourse during breaks, at lunch, in the evening
This workshop covers state-of-the-art applications and social lounge, and at various nightly professional events. Truth be
statistical methods in environmental monitoring. Sessions told, some of my best conversations about statistics teaching have
on applications include monitoring in ecology, monitor- taken place at AAA baseball games.
ing in air quality, monitoring of aquatic resources, and Shortly after my wife and I moved west, Dan Yates contacted
monitoring of climate change and its impacts. The spa- me about assisting him with some of the calculator features in the
tio-temporal data collected in environmental monitoring second edition of The Practice of Statistics, the successful high-
present interesting and challenging statistical problems, school AP Statistics textbook he coauthored with David Moore.
such as modeling of space-time correlation, analysis Having spent three years co-editing the Technology Tips column in
of the huge amount of correlated data, and analysis of the National Council of Teachers of Mathematics’ (NCTM) maga-
high-frequency monitoring data. Technical sessions will zine, Mathematics Teacher, I felt comfortable accepting this role as
cover recent developments in statistical methods for envi- a minor contributor in the writing process. Somewhere along the
ronmental data. A one-day short course on the analysis of way, I mentioned to Yates that I had a few suggested enhance-
spatial and spatio-temporal data will be offered and taught ments to the text based on my experiences as a user. To my surprise,
by Doug Nychka. There also will be a poster session. Poster he called my bluff and asked me to compose a revised version of
abstracts need to be submitted by September 23, 2008. Chapter 1 that incorporated my recommendations. Doing so took
For more information, contact Hao Zhang, Department of far longer than I had expected, with many false starts along the
Statistics, Purdue University, West Lafayette, IN 47906; (765) way. The net result of my efforts was a 120-page tome disguised
496-9548; zhanghao@purdue.edu. as a single textbook chapter. With a typical dose of his southern
charm, Yates quipped, “You know, if chapters were this long, we’d
have to package the book with a wheelbarrow.” In the end, I was

DAY IN THE LIFE
instructed to “trim” the chapter to 60 pages. That was probably the After a brief stint in independent school administration, I accept-
most painful editing task I have ever undertaken! (There was one ed an endowed chair position as Master Teacher in Mathematics at
consolation: Nearly all the trimmed material ended up in a supple- The Lawrenceville School this past fall. This unique position allows
ment somewhere.) me to team teach with a different member of the department each
So, my second career as a writer was launched. I teamed with Yates trimester. By planning, instructing, and assessing collaboratively,
and Moore on a second venture—Statistics Through Applications, a we are able to improve our individual teaching practices. Our stu-
high-school textbook for non–AP Statistics courses—that was com- dents also benefit from the distinct approaches and perspectives of
pleted in 2004. With both of my coauthors now retired, I’ll have my their co-teachers. Other responsibilities as master teacher involve
hands full with future revisions. curriculum design, technology training, and professional develop-
In 2004, I was appointed to the ASA/NCTM Joint Committee ment. Much of my time outside the classroom is spent in pro-
on Curriculum in Statistics and Probability by then NCTM president fessional growth activities involving other teachers, either here at
Johnny Lott. Over the years, the joint committee has been respon- Lawrenceville or across the country at meetings, conferences, and
sible for several significant projects to develop materials that promote workshops. Distilling that collective wisdom into better learning
K–12 statistics education, including the Quantitative Literacy Series, for students is at the heart of what I do on a daily basis.
the Elementary Quantitative Literacy Series, the Science Education If I have become a teacher leader, it is because I have followed in
and Quantitative Literacy Series, and Data-Driven Mathematics. the footsteps of the many great mentors and role models who have
The joint committee also has assumed responsibility for designing shaped my thinking. As my career has evolved, I have drifted from
teacher workshops to support effective classroom use of these materi- my origins as a pure mathematician into the realms of data analysis
als. With the ASA’s recent publication of Guidelines for Assessment and and statistics. All the while, my primary focus has remained square-
Instruction in Statistics Education (GAISE), the committee has begun ly on how to help students learn mathematics (including statistics)
developing materials and training to help teachers implement the cur- in ways that will ensure their ability to use what they have learned
ricular framework described therein. Roxy Peck and I just completed to solve important problems today, tomorrow, and in the distant
a capstone module for high-school students, titled “Making Sense of future. After all, if students see no potential to do anything produc-
Statistical Studies,” which began as an ASA Member Initiative sev- tive with what they’re learning, why should they bother? ■
eral years ago when Peck, Henry Kranendonk, and June Morita were
members of the joint committee.

Harold Nelson
Office of Financial Management
Where Are Your
Olympia, Washington
“My group does a lot of purely

descriptive work, a lot of it
Colleagues …
WASHINGTON centering on the lack of health
care insurance. ”
MONTANA NORTH DAKOTA
M
OREGON Nancy Wang

MDS Pharma Services
Lincoln, Nebraska
IDAHO SOUTH DAKOTA

“[I] use statistics in the
design and analysis of phase I
WYOMING and IIa clinical trials.”
Kristi Epke NEBRASKA

San Domenico School
NEVADA
San Anselmo, California
“I am a high-school Geoffrey Bohling

mathematics teacher, not a UTAH Kansas Geological Survey
statistician, but I teach COLORADO
Lawrence, Kansas
AP Statistics. ” “My work is aimed at fusing KANSAS
information to reduce the
non-uniqueness in subsurface
CALIFORNIA characterization problems. ”
OKLAHOMA
ARIZONA
NEW MEXICO
Willis Jensen
W.L. Gore & Associates
Flagstaff, Arizona
Stephanie Lopez Cano
“ I’m the primary statistician for a The University of Texas
portion of a medical device San Antonio, Texas
TEXAS
Graphic by Val Snider
manufacturing company. ” “Aside from teaching statistics in
the classroom, I am director of the
Statistical Consulting Center. ”

Liam O’Brien
Colby College
Waterville, Maine
“I teach statistics to
undergraduates and research
MAINE
statistical applications in
Robert public health. ”
Kabacoff
Management Research Group
Luke Davulis
VERM
Portland, Maine
Maine Department of Labor
“ I use statistics to develop Augusta, Maine
NEW
ONT
predictive models of leadership
MINNESOTA “ I work at gathering the data HAMPSHIRE
success in various industries and
that form estimates of
”
job functions.
employment and wages by MASSACHUSETTS
WISCONSIN
occupation.
NEW YORK
” RHODE
ISLAND
MICHIGAN
CONNECTICUT
PENNSYLVANIA
NEW JERSEY
IOWA
OHIO
INDI
DELAWARE
ANA
Cindy Ammons
ILLINOIS WEST Schools
Chapel Hill-Carrboro MARYLAND
VIRGINIA
Chapel Hill, North Carolina
VIRGINIA
“I teach AP Statistics to
high-school students.”
MISSOURI
KENTUCKY WASHINGTON, DC
NORTH
CAROLINA
TENNESSEE
James Hardin SOUTH
OMA ARKANSAS CAROLINA
University of South Carolina
Columbia, South Carolina
“I use statistics to analyze

MISS
data collected in various

GEORGIA
ISSIP
health studies. ” Malay Ghosh

University of Florida
ALABAMA
PI
LOUISIANA Gainesville, Florida
“I teach statistical theory and do

research in both statistical theory
”
And What Are
and its applications.
FLO
They Doing?
RID
A

DAY IN THE LIFE
What Does a Teaching

Associate Professor Do?
Pam Arroway, North Carolina State University
F
or about a decade now, the Department of Statistics at North
Carolina State University has been hiring faculty in career
path, non-tenure-track teaching positions. These positions
were first called clinical assistant professor positions, but have since
been renamed teaching assistant professors. These positions are
unusual in that they are not tenure-track, but they do have a career
track, unlike instructor, lecturer, or adjunct positions found at
many universities in the United States. Thus, a teaching assistant
professor has the opportunity to be promoted to a teaching associ-
ate professor and a teaching professor. For simplicity, I’ll refer to a
person in any of these positions as a TAP and to tenure-track or
tenured faculty as TT faculty.
What Do TAPs Do?

Currently, the department has four TAPs, all of whom began at
the assistant level and were promoted to the associate level. All
currently have five-year contracts and PhDs in statistics or math-
ematics. Their responsibilities include a nominal teaching load of
four courses per semester, though most TAPs begin with a “3+”
contract. That is, the teaching load is three courses per semes-
ter, plus some scholarly activity. That may include research, but
doesn’t have to.
The four TAPs (including me) have evolved into very differ-
ent roles within the department. The ways we contribute to the
department, university, and profession include teaching gradu-
ate and undergraduate courses for statistics majors and as service
courses; developing curricular and teaching materials to share
with colleagues and/or publish; writing research papers; serving in
administrative roles such as assistant department head, co-director
of graduate programs, and director of undergraduate programs;
recruiting, advising, and mentoring undergraduate and graduate
students, including directing dissertation research; writing grant
proposals; providing professional service by reviewing journal sub-
missions; being an associate editor or grading AP Statistics exams;
serving on committees… Sounds a lot like what a TT faculty mem-
ber might do, right?
So what does a TAP not do? In my department, there’s very little
we can’t do. We have the opportunity to achieve our full potential.
The biggest difference between a TAP and TT position is we allo-
cate our time differently to the realms of teaching, research, and
service. TAPs are not expected to publish in top statistical journals,
advise a regular stream of graduate students, or get research grants,
though we can do those things.
We are expected to do a lot of outstanding teaching, and the
flexibility of these positions has allowed us opportunities junior
30 30 AMSTAT
AMSTAT NEWS
NEWS SEPTEMBER
SEPTEMBER 20082008

Teacher Associate Professor (TAP) and Director of Undergraduate Programs Roger Woodard
(center) mentors North Carolina State graduate students.
faculty at a Research I university can rarely afford to consider. For

example, we spend time on large course redesigns; organize con-
ferences; develop teaching materials that are worth sharing and
publishing; take on administrative roles; and go after grants for
REUs, conferences, undergraduate and graduate support, enhanc-
ing diversity, and teaching activities.
I want to note that a TAP is not the same as a TT person doing
research in statistics education. We have recently hired someone in
this type of TT position to begin in January 2009. She will have
the same research, teaching, and service expectations as any other
TT faculty, with her research focused on statistics education. A few
other statistics departments have, or at least have advertised, similar
positions. While some TAPs are active in the national statistics edu-
cation community, there is not a requirement of education research
for reappointment or promotion. Teacher Associate Professor (TAP) Pam Arroway serves
as assistant department head and codirector of grad-
Why Would Someone Want uate programs at North Carolina State University.
to Be a TAP?
A TAP is typically someone who is passionate about teaching and
mentoring students and who wants to be in a research-intensive
environment. A TAP position may look similar to a TT position at a big effect on the profession and in training critical thinkers. A
an institution that places more emphasis on teaching and under- variety of seminars (Bayesian, bioinformatics, biostatistics, environ-
graduates. However, there are advantages to being in a Research I mental statistics, statistical genetics, regular departmental seminars,
school and NC State, in particular. and seminars by a number of interview candidates) and special
There are resources that may not be available at smaller institu- topics courses in the department and at the nearby Statistical and
tions, such as for travel and professional development. In addition Applied Mathematical Sciences Institute provide opportunities for
to supporting NC State’s well-established undergraduate program, continuous professional development.
we have opportunities to mentor graduate students in teaching,
research, or in their graduate careers in one of the oldest, largest, Why Does the Department
and best graduate programs in the United States. In a large depart- Want TAPs?
ment, we have a variety of undergraduate and graduate classes to The existence of TAPs at NC State does not mean TT faculty
teach and are surrounded by colleagues with various expertise. members are not excellent teachers. In fact, many have received
Opportunities also exist to provide statistical consulting for stu- university and national teaching awards. TAPs can make valuable
dents and postdoctoral fellows in other departments. We can have contributions to the long-term goals of the department and the

Teaching Associate Professor Kim Weems (left), David Blackwell, and
Department Head Sastry Pantula
Teaching Associate Professor Kim Weems at the presentation of an

honorary doctorate from North Carolina State to David Blackwell mission of the university. The creation of TAPs incorporates fac-
ulty members with different interests and skill sets and helps keep
teaching loads for research-active faculty at a level where they can
continue to be successful in their research productivity.
Unlike short-term adjuncts or instructors, TAPs have a vested
interest in the future of the department. So, while it is more of a
commitment than hiring an instructor, the benefit to the depart-
ment tends to be much greater. TAPs have the time for (and even
Teaching Associate Professor expectation of ) improving teaching and curriculum across the
Roger Woodard joined whole department. NC State TAPs have also taken on administra-
the department in 2003 and
tive and service roles that lessen the non-research workload of TT
currently serves as director of
faculty, make use of the professional strengths of those in TAP
undergraduate programs.
positions, and contribute to the long-term growth of the depart-
ment. TAPs also play an important role in training graduate stu-
dents, especially for academic positions.
And About That Pesky Tenure …
I don’t miss it. Is my position less secure than my colleagues? My
contract has to be renewed every five years. With reappointment,
promotion, and post-tenure review every three to five years, I’m
Teaching Associate Professor
not sure TT faculty members have it any easier. The university,
Kim Weems joined the which regards us as “special ranks” faculty, has explicitly stated that
department in 2001. the review process is the same for TAPs as for TT faculty, though
contributions in different realms must obviously be weighted dif-
ferently. Industry isn’t handing out five-year contracts either. Job
security is not a bigger concern for me than for anyone else.
Some professional organizations have been outspoken against
increase in non-TT positions in academia. I agree that non-TT
positions should not increase at the expense of TT positions. They
should be used to complement TT faculty members and support
the mission of the university. Universities must be ethical and
respectful in how non-TT faculty members are treated. A TAP is
Teaching Associate Professor not a person the department gets to abuse and then dump when
Jeff Thompson joined the funding gets tight. Neither are TAPs short-term, hired-as-needed
department in 2003. teachers for low-level courses. In my department, the TAPs are very
much respected by colleagues and the administration. We are not
second-class citizens, and we participate in nearly all activities. For
better or worse, we are expected to contribute as much as TT fac-
ulty, and we do in the realms of our passion and expertise. I love
my job! (Okay, it is pay raise time, and I want my department head
to count this as one of my publications.) ■


CAREER GUIDE
has a lot of value for people who are beginning an
T
his article is based on a talk I gave at the
JSM 2007 meeting for the ASA Committee academic career. So, I refer new faculty members
David L. Banks,
on Career Development. But, I should to that, and, in this article, shall focus on topics
Department of Statistical
Science, Duke University confess at the outset that I have no particular qual- that apply to everyone, not just recent PhDs and
ifications or any expertise on this topic, aside from not just academics.
having held a lot of jobs (which ought to raise Although statisticians are relatively homoge-
questions about my suitability in the first place). neous in our training, we have the usual range of
Years ago, I was involved in drafting the New talents, personalities, and utility functions. This
Researchers’ Survival Guide (available at www. creates many career paths and many ways to be
imstat.org/publications). Reading over it again, we successful. It also means you can be miserable if
were awfully earnest and a bit naïve, but I think it you get caught on a path that doesn’t fit your per-
sonal strengths and values.

All careers have a stochastic component, so are a key component of almost any suc-
we should look to dynamic programming as a cess story, and whatever you can do to
model for continual reappraisal of our situations build strength in this area will repay your
and ways that may better them. This implies a efforts. If English is not your native lan-
portfolio analysis perspective: We each have a guage, that can be an obstacle, but per-
different mix of strengths and weaknesses, and sonality and style are much more impor-
we should try to adaptively invest our energy in tant. Feng Liang of Duke University and
combinations that seem most likely to pay off. Xiao-Li Meng of Harvard, whose native
Some skills that apply to employment in all sec- language is not English, give wonderful
tors are the following: talks, and there are many others.
Technical Strength. This is the Writing. It is crucial to be able to write

foundation when you are starting out, clearly, correctly, and briefly. This is a
but it often becomes less important as lifelong learning process—anyone who
you advance. Especially in business and writes well is constantly studying how to
government, one needs breadth more write and attending to their prose.
than depth at the higher levels.
Social Networking. This is crucially
Computational Ability. Anyone important, and it sometimes seems statis-
who can do solid statistical program- ticians study it more, while learning less,
ming will never miss a meal. It is a blue than those in any other field. You need a
chip skill and a way of thinking that has diverse network; having a lot of friends
unique value. But, it is hard to become who work on local asymptotic minimaxity
rich or famous on this alone. is not as helpful as having friends with
Public Speaking. Every member of

the ASA has survived at least 1.5 decades
of dull lectures in school, which is why it
amazes me that so many of us have not
learned enough from that experience to
avoid giving bad talks. Good presentations

complementary strengths. In particular, unconsciously and with benign intent, hijack a
bring your secretary flowers, talk to the meeting. Probably no constructive meeting lasts
system administrator about D&D, and more than 50 minutes, and no participant can
go to a conference where there aren’t make more than one key point per meeting (but
many statisticians. you should generally try to make a point, just to
prove you are relevant). Except when brainstorm-
Organization. This sounds mundane, ing over beer, you should never think out loud,
but it is very hard for a manager to pro- unless you are the smartest person in the room, and
mote you if you are sloppy or slow about even then it is perilous. A good rule of thumb for a
paperwork. And the discipline of quick meeting is to pretend you are one of the characters
“hold
turnaround on such items (phone calls, on the television show “West Wing.”
Most of us will emails, appointments, referee reports) Most of us will hold many jobs. Our chief assets
helps in other aspects of one’s career. are reputation, capability, and social capital—not
many jobs. seniority. So, when one moves on, try to leave only
Our chief assets Time Management. When I was friends behind—it isn’t always easy, as changing jobs
a young faculty member at Carnegie is usually prompted by dissatisfaction, but, after a
are reputation, Mellon, Jay Kadane shared with me this few years, things will seem less intense. My sense is
advice: Carry a pocket diary and mark that no organization is more than two bad manag-
capability, and down everything you need to do. But, also ers (in time or hierarchy) away from meltdown, so
social capital— realize everyone goofs off some (except it is good to have an exit plan. And, each time you
”
maybe Peter Hall and David Dunson). change jobs, you get the chance to learn new skills
not seniority. So, don’t waste time feeling guilty about and make new friends; most of us can advance faster
wasting time—just be efficient when you by moving than by staying. In terms of long-range
actually get down to work. planning, you need to think at least two moves
ahead to achieve significant career growth, and you
Someone else would probably generate a should choose those moves to ensure an ironclad
slightly different list, but these are all key areas résumé for the position you ultimately want.
to cultivate. For those who need to stick in their job, there
One situation where many people damage are still ways to advance. Personality counts for a
their careers is at internal business meetings. lot. Try to pretend to be happy and productive.
Always be concise. Try to read the interper- Read the newspaper so you have a wealth of con-
sonal dynamics—you need to hear the sub- versation topics and aren’t stereotypically dull or
text, read the body language, and notice what narrow. You should avoid doomed projects, those
doesn’t get said. Do not pursue pet projects or that do not build new professional assets, and those
raise dead issues—the way to lobby for those is for which you are not central. I’d recommend look-
through one-on-one talks, and it can be a slow ing for projects that cross division boundaries—it
process to build support. helps to have a broad base of good opinion, and
When I joined the National Institute of you can build unique collaborations the organi-
Standards and Technology, Lynne Hare advised zation needs. When David Blackwell introduced
me to “avoid other people’s nonsense.” Help keep Cuthbert Daniel to the Berkeley statistics depart-
meetings focused; people loathe those who, often ment, he said “This man is worth 10 of us—not
because he is better, but because he is different.”
So, try to differentiate yourself. Think of at least
one new idea a week, but be properly skeptical
of its value.

CAREER GUIDE
Some people can climb high by sucking up. Sections, committees, and chapters that are
And this is not evil, nor is it uncouth if done ladders for leadership within the ASA, offer
with panache and a degree of dignity. It crosses chances to network with like-minded people;
a moral line (I think) if one takes advantage of and help build your résumé
such favoritism to torpedo others in the office.
More broadly, statisticians may face special ethi- The opportunity to practice public speaking
cal challenges; we often have to find honest ways at every JSM and a 20-minute forum at which
to let the data speak on behalf of our employers’ to advertise yourself
interests, and sometimes that just isn’t possible. In
such cases, you must use all your powers of diplo- Journals, a directory of members, online
macy to head off direct conflict. But, if such arise, job listings, short courses, a conference
then doing the right thing is not going to protect calendar, and other ways to keep up with
you. So, keep an updated résumé handy and don’t our profession and colleagues.
wait around when the weather changes.
The ASA was created to be an engine for career These ASA resources are a huge benefit of
growth for all statisticians. Among its many other membership, and their relative advantage is
services, it offers the following: especially great for master’s statisticians, isolated
statisticians, and those working in industries or
Salary surveys for statisticians, broken out agencies whose professional staff includes many
by useful covariates such as sector and rank— nonstatisticians. ■
this is essential for evaluating job offers and
negotiating raises
Professional job fairs at every JSM—the

last day is open to all, and it never hurts to
look at your options and keep your search
skills sharp
Continuing education opportunities, both

as a student and as a teacher

Top Five
Favorites
We asked a few statisticians what
their top five favorite statistical
books were and why they chose
them as important added compo-
nents to their particular job. Here is
what they had to say.

Dalene Stangl
Director, Professor of the Practice of Statistics
and Public Policy, Duke University
Editor, Reviews of Books and Teaching
Materials, The American Statistician
1. Bayesian Statistics 8, published by Oxford

Press, edited by J. M. Bernardo,
M. J. Bayarri, J. O. Berger, A. P. Dawid, D.
Heckerman, A. F. M. Smith, and M. West
2. Statistical Decision Theory and Bayesian

Analysis by James O. Berger
3. Statistics by David Freedman, Robert Pisani,

and Roger Purves
4. Theory of Statistics by M. J. Schervish
5. Bayesian Approaches to Clinical Trials

and Health-Care Evaluation by David J.
Spiegelhalter, Keith R. Abrams, and
Jonathan P. Myles

Barry Nussbaum
Chief Statistician, U.S. Environmental
find compelling, real-to-life examples for
any level of course I am teaching.
Protection Agency 2. Statistical Sleuth by Fred Ramsey and

Daniel Schafer—a difficult book for a
first course for undergrads, but one we
1. The Cartoon Guide to Statistics by Larry
refer to often over the course of students’
Gonick and Woollcott Smith
time here.
Despite its name (or, perhaps because of
it), it is a good primer on statistics. Its
3. Generalized Linear Models by P.
catchy illustrations and succinct prose
McCullagh and J. A. Nelder provides a
provide just the explanation of statistical
definitive, unified, treatment of methods
concepts a statistician needs when describ-
for the analysis of diverse types of data.
ing some basic work to nonstatistician
decisionmakers.
I realize the following two books do not
exactly make for leisure time reading, but
The next three are classics in terms of
they come in handy as references.
established books, which display and
explain statistical methodology.
4. The R Book by Michael J. Crawley
2. Statistical Methods by George Snedecor
5. Handbook for Statistics Using Stata by
and William Cochran
Sophia Rabe-Hesketh and Brian S. Everitt
or A Handbook of Statistical Analyses Using
3. Applied Regression Analysis by Norman
R by Brian S. Everitt
Draper and Harry Smith
4. Nonparametric Statistics for the

Behavioral Sciences by Sidney Siegel and
N. John Castellan Jr.
5. Finally, I rely on an old favorite: Darrell

Chris Olsen
AP Statistics Teacher,
Huff ’s How to Lie With Statistics. Far from Thomas Jefferson High School
being an instruction manual for statisti-
cal lies, it has a wealth of illustrations 1. Statistics by David Freedman, Robert
describing how statistics can be misused. Pisani, and Roger Purves
Frequently, in complex analyses, we fail to
realize we are making some of the mistakes 2. Statistics for Experimenters: An
Huff explains so well. So, I gladly use his Introduction to Design, Data Analysis, and
examples to explain the fallacies we may Model Building by George Box, William
be getting into. Hunter, J. Stuart Hunter, and William
Gordon Hunter

Julie Legler
Director, Statistics Program,
3. Introduction to Design and Analysis of
Experiments by George Cobb
4. Elementary Survey Sampling by William

Lyman Ott, Richard Scheaffer, and
St. Olaf College William Mendenhall
1. Statistics: A Guide to the Unknown by 5. Statistics: A Guide to the Unknown by
Roxy Peck, George Casella, George Cobb, Roxy Peck, George Casella, George Cobb,
Roger Hoerl, and Deborah Nolan. I first Roger Hoerl, and Deborah Nolan
became enthralled with this book many
years ago when it was edited by Judith
Tanur. In its current form, it is possible to

CAREER GUIDE
or Media Savvy?
Are You Media Shy…
Rosanne Desmone,
ASA Public Relations Specialist
W
hile many of the people I’ve quoted below, while others are incorporat-
worked with would sooner have ed into the text. One of the journalists who
several teeth pulled than do a uses our experts list fairly frequently is Carl
media interview, others take to the process Bialik, The Wall Street Journal’s Numbers
like a duck to water. The basic difference Guy. I asked Bialik what he was looking for
between the two attitudes seems to be the when he spoke with an expert and included
individual’s confidence in his or her ability his answers in the sections below. Bialik also
to control the interview. They would rather said he wanted an expert willing to give
run the other way than get involved in a an overall assessment of the statistical rela-
situation where they might be misquoted or tionship or statistical techniques used in a
misrepresented. The good news, however, is paper he’s covering, rather than addressing
that every media-shy person can become a few minor quibbles … the big picture, so
media savvy with a little knowledge and to speak. In general, this is a good point to
practice. You can learn how to be more in remember for every interview.
control of an interview and to understand Below are some guidelines to help you
what the interviewer wants and needs. become a media savvy interviewee.
Sir Laurence Olivier attributed his suc-
cess to two things: the confidence to per- Respect deadlines, but buy
form and the humility to prepare. some time to think.
Six months ago, the ASA unveiled its The first point on Bialik’s list of what he
Media Experts Program, and we have had looks for when he speaks to an expert is
good feedback from the media who have “awareness of the importance of deadlines.”
used it. Some of our experts have had a That said, however, if you’re asked to do an
great deal of media experience, and I asked interview, buy a little time. Ask what the
them for their best advice on doing media editor/writer wants to discuss, and don’t
interviews. Some of their comments are do the interview until you can think about

what you’re going
to say. News people
almost always work
on deadlines of a few
hours to a few days, but
that doesn’t mean you have
to jump in without knowing
what will be discussed. If you
tell the journalist you’re in the
middle of something and can talk
to them in 15 or 30 minutes, you are
respecting their need for speed, and
those few minutes will give you time
to think about the topic before you start
answering questions. If they have a longer
lead time, usually they will be happy to Keep it simple
schedule a time to talk later.
All our media experts agree you have to
“Make sure you understand what is
speak in plain English and not use techni-
being asked, and that you can contribute
cal jargon in interviews. You need to be as
something sensible on the point,” advised
accurate and complete as possible in your
David Peterson, who is semi-retired and
answers, but you shouldn’t go into so much
formerly of the Duke statistics faculty.
detail that the answer becomes confusing.
Using language your interviewer doesn’t
Remember that you may quite understand is one of the major rea-
be talking to someone sons experts are misquoted. The more tech-
with little or no knowledge nical you are, the more likely you are to be
misunderstood and/or misquoted.
of statistics. “Avoid statistical jargon. The audience
There are basically two types of news peo- won’t understand and may, in fact, take
ple: generalists and specialists. The gener- a different meaning from your intend-
alists are not subject-matter experts; they ed message.” This is sound advice from
may cover a strike in the morning, an auto C. Shane Reese of Brigham Young
accident in the afternoon, and a program University.
on nuclear fusion in the evening. Some
science writers have a lot of science back-
ground. In both cases, however, you may Know your audience
be dealing with people who know little or Yes, you’re talking to a journalist, but you’re
nothing about statistics. Don’t overestimate really talking to his audience. Framing
the interviewer’s expertise; you may need to your answers for the journalist’s audience is
provide enough background about a sub- really the key to giving a good interview.
ject so you can lead the interviewer to ask That’s also the key to getting your com-
the right questions. As one journalist once ments used—the more relevant they are
aptly put it, “If I don’t ask the right ques- to the interviewer’s audience, and the bet-
tion, make sure you give the right answer.” ter you are at explaining things succinctly,
“If a reporter’s question is unclear or not the more likely you are to be quoted—and
posed in such a way that one can answer it quoted correctly. So, if you’re not familiar
usefully, the most valuable thing one can do with the news outlet, ask who makes up the
is to first provide some background expla- audience. Journalists use their audience to
nation to help them understand the issues. filter what you say: What is the implica-
That is, lead the reporter to ask the right tion of the particular topic to the audience?
question, then answer it,” said Peter F. Thall Does it mean an additional congressional
of the M. D. Anderson Cancer Center. district in their state? More money for a

CAREER GUIDE
school district? Better health of Pennsylvania School of Medicine. “No Anything that can
care? If you cannot make your one is expected to know everything.” go wrong might.
comments relevant to the audience, Don’t say anything you don’t want to see Finally, it’s important to know things can
they won’t get used. in print or on the 6 p.m. news, and don’t let go wrong no matter how good a job you
Jessica Utts of the University of interviewers put words in your mouth. do. The story could turn out to be unfair
California-Davis, said she tries to “translate You can’t control the interviewer’s ques- or inaccurate. If it does, bite the bullet
technical information into something the tions, so focus on controlling your answers. and don’t call the journalist or editor to
public can understand.” Anything you say in an interview can be complain. It’s a bad idea to get into an
quoted, even if it originated with the inter- argument with someone who buys ink
viewer. For example, a journalist may ask by the barrel or videotape by the mile.
Use examples to clarify you if you have problems with the meth- Once the journalist files her story, it
and explain. odology used in a study. If you repeat what goes into the system, to the editor, the
Our media experts have varying opinions was said, or even deny it, you can be quoted copy editor, and so on. Any of those
about using examples or analogies in inter- as talking about questionable methodology. folks can change the story without
views, although most of them voted a quali- Instead, answer the question by making a checking with the journalist. The head-
fied yes to the use of these tools. Some said positive statement about the study or some line writer can change the whole focus
the wrong example can seem to trivialize an aspect of it. Additionally, it’s important to of a story with the headline he writes,
issue that isn’t trivial, while others feel talking remember that the interview is not over until or the producer can cut the story and
to the journalist’s audience is not much dif- it’s over—until the phone is hung up, the change its meaning.
ferent than teaching statistics to nonstatistics cameras or mikes are off, or you’ve left the Bottom line: There are no guaran-
majors. While journalists welcome the kinds building. You have to be disciplined enough tees, but your best chance of having a
of examples that help illustrate a difficult to not get pulled into discussion after the good outcome is to be prepared and fol-
point, you probably have to use your judg- fact because it could become the story. low the guidelines above. Good luck! ■
ment in each case. According to Bialik, he “If you wouldn’t write it and sign it,
wants experts he interviews to have “the abil- don’t say it,” said Earl Wilson, former Red
ity to put their findings in terms accessible Sox pitcher.
to a general audience, including providing
analogies and examples.”
“Yes, it is helpful to use an example,
which can clarify how the statistical issue/
result is directly relevant to people’s lives,”
said David Dunson of the National Institute
of Environmental Health Sciences. “It makes
things more concrete.”
If you don’t know, say so.

If the topic of the interview is out of the
scope of your experience or responsibility,
your best bet is to politely decline and, if
possible, direct the writer to someone who
may be able to help. Once you’re into the
interview, don’t speculate or talk beyond
your expertise, and remember that opinions
are not facts. Some journalists will try to get
you to answer a question even if you don’t
feel qualified, but the majority of good jour-
nalists will accept your refusal to answer. In
fact, Bialik cited the following as a desirable
quality in an expert: “Honesty when they
can’t speak with expertise about the issue”
being addressed.
“If you’re asked a question to which you
do not know the answer, don’t be afraid to say
so,” advised Susan Ellenberg of the University

Up-and-Coming
Statisticians on
the Verge of
Great Things
S
tatisticians work in so many fields—
from animal health to ecology to
agriculture to computer science—
and, every year, an untold number make a
name for themselves. Amstat News asked
around, and the names of the statisticians
below just kept coming up as those belong-
ing to people on the verge of doing great
things. Allow us to introduce some of the
ASA’s up-and-coming statisticians.

Age: 33
Title: Assistant
Professor
Employer:
Department of
Statistics, University
of Michigan
Alma Mater: UC
Berkeley (PhD 2002)
Research Interests:
Liza Levina’s interests center on inference for high-dimensional
data, including large p small n problems, covariance estimation,
dimension reduction, and networks.
Name a statistician you admire and What do you hate most about
tell us why. your profession?
If I’m supposed to name a historical figure here, The same things that everyone else does—papers
I would pick Fisher, but the most honest answer getting rejected, particularly by incompetent ref-
would have to be my advisor, Peter Bickel. Not erees. Dealing with bureaucracy. The time pres-
only because of his work, but because of the great sure of tenure track. Also, that every doctor and
positive influence he’s had on so many people’s nurse I’ve ever met felt obliged to tell me how
lives and careers. much they hated statistics in college.
What do you love most about What was the best career advice you
your profession? were given?
When something you’ve been struggling with Co-advising graduate students. We have all had
finally works, be it a proof or a piece of code. conversations with colleagues that end in “we
When you are collaborating with someone and should really talk more about this,” and then
know that, together, you are doing work neither nothing happens because no one has time. But,
of you could have produced on your own. When if you get a joint student, then the collaboration
a student just understood something you’ve been really will happen.
explaining and you can see it’s a fascinating revela-
tion for them. Of course, all these things are not What is one web site/blog you can’t
unique to statistics. As for statistics, itself, the fact go a day without visiting?
that almost everyone needs it is a nice bonus. None. After my daughter was born (she’s almost
three now), my free time has become much too
scarce to spend it on reading blogs.
What is something your friends

would be surprised to learn about you?
I think they’ll be quite surprised to see me do this
article…

CAREER G UIDE
Age: 42
Title: Biostatistician
What do you hate most about
Employer: USDA your profession?
Agricultural Research What I dislike most about the statistical profession
Service is the degree of mistrust and suspicion investigators
Alma Mater: University from other fields hold for statistics. I believe this sim-
of Louisiana at Lafayette ply results from a lack of understanding of statistics
and the bad press to which statistics is sometimes
subjected.
What is the best career advice you

were given?
The best career advice I was given actually came
from my brother, who is an electrical engineer. He
told me to never sell myself short and to view a job
What are your research interests? interview as a two-way street. That is, I should inter-
The research I do involves statistical and epidemio- view the prospective employer as much as he/she is
logic approaches to identify relationships among interviewing me.
dietary, physiological, genetic, and behavioral char-
acteristics that may be associated with overweight or What is one web site/blog you
obesity in humans. can’t go a day without visiting?
This is going to sound so nerdy, but one web site I
Name a statistician you admire and cannot go a day without visiting is the SAS docu-
tell us why. mentation web site that contains documentation
The name of a statistician that I admire is Don for all their procedures. I am constantly learning
Mercante, who is the director of the biostatistics (or relearning) the theory behind statistical meth-
program at the LSUHSC School of Public Health in ods simply by reading the documentation provided
New Orleans. While Don is not a ‘famous’ statisti- to correctly use these procedures. It’s such a great
cian, he helped me out tremendously when I started resource for me.
my first job as a statistician. I knew I could always
go to Don for help with a statistical issue, or even a What is something your friends would be
professional issue. He was (is) truly a great mentor— surprised to learn about you?
always giving good advice and making time for you, Something my friends might be surprised to learn
no matter how busy he happens to be. about me is that mathematics did not always come
easily to me. In fact, it was often the subject I strug-
What do you love most about gled with the most in high school and college.
your profession?
What I love most about the statistical profession is
the wide variety of scientific disciplines to which
I am exposed. My first love was biology, so I truly
enjoy seeing the relevance of statistics in biological
fields, especially nutrition and medicine.

where we have a very collegial atmosphere. Also, I do
not generally enjoy people pushing dogmatic views
Age: 32 on the different approaches to statistical inference.
Title: Assistant
Professor What is the best career advice you
were given?
Employer: University of I have been guided by a number of excellent advisors.
Chicago When enrolled in an applied mathematics program
Alma Mater: University back in Germany, it was Friedrich Pukelsheim who
of Washington first brought my attention to statistics. During my
PhD studies in Seattle, I received a lot of guidance
from my advisors, Thomas Richardson and Michael
Perlman. It was through Thomas that I met Bernd
Sturmfels, whose advice to spend a year at Berkeley
as a postdoc was probably the most influential pro-
fessional advice I have received. Based on my experi-
ence, I would recommend doing a postdoc to any
fresh PhD. It is a great way to finish off thesis papers
and explore new avenues for research.
Name a statistician you admire and What is one web site/blog you
tell us why. can’t go a day without visiting?
Looking beyond the circle of my doctoral and post- After checking an old email account, I usually take
doctoral advisors but keeping within my areas of a look at http://de.yahoo.com to pick up some news
interest, the name Steffen Lauritzen immediately about Germany. I probably use the site mostly to
comes to my mind. He has done groundbreaking read up on sports news. They also have a lot of live-
work on graphical models, and his book is the stan- tickers for soccer games and, of course, they always
dard reference in this area. Over the past few years, cover my favorite club: FC Bayern Munich.
Research Interests: I have had several opportunities to be at meetings
The focus of my research is on organized or attended by Steffen and, over and over What is something your friends would be
graphical models and algebraic again, I am impressed by his quick thinking and his surprised to learn about you?
statistics. Graphical models are truly constructive comments on the various speak- I am not sure there is anything I could do to surprise
multivariate statistical models ers’ presentations. I also admire Steffen’s supportive my friends. So, I will try to surprise some strangers
in which observed variables nature and positive attitude toward younger fac- by sharing two pieces of trivia:
are constrained to exhibit ulty. He is always encouraging, and I have never I have been playing soccer since I was a little
dependence patterns associated seen him pass on judgmental views. I think I am kid and, naturally, I racked up a number of inju-
with a graph. My work in not alone in having benefited and drawn motiva- ries. Sitting at a desk most of the day doesn’t help
algebraic statistics is primarily tion from his suggestions. with keeping in shape, and so the frequency of
concerned with statistical
injuries has increased over the last 7–8 years. In
inference in models that have What do you love most about fact, I am currently working from home while
parameter spaces with algebraic your profession? recovering from my fourth knee surgery. For sev-
structure. Many graphical Academic statistics is a small field when judged by eral years, someone’s Achilles tendon has been
models are, in fact, of this a headcount, but it is incredibly diverse. Research playing ACL in my left knee, and, since two
algebraic type. may address purely mathematical questions, solve weeks ago, someone’s patella tendon serves as the
algorithmic and computational problems, or be ACL in my right knee.
concerned with applied questions. Each one of these The second piece of trivia involves a prize. Very
three areas is challenging in its own way, and I very few will find it surprising that pronouncing my last
much enjoy the exchange of ideas and problems name is a very painful task in virtually all of this plan-
between the three camps of mathematical, compu- et’s languages. However, some may find it surprising
tational, and applied statistics. that outside the circle of my (great-)grandparents,
parents, and three sisters, I have never met anyone
What do you hate most about with my last name. This makes me wonder whether
your profession? I am the world’s only male heir to this concatenation
I really dislike turf wars between mathematically of mostly consonants. For possible resolution of this
minded and applied statisticians about who does matter, I would like to offer a pint of beer at the next
‘good’ or ‘important’ research. I consider myself very Joint Statistical Meetings to the first few people who
fortunate to be part of the department at Chicago, know/have heard of a person with last name Drton
who is not related to me.

Age: 30
Title: Assistant Professor
Employer: The
University of North
Carolina at Chapel Hill
Alma Mater: Nankai
University (BS, 1999);
The Ohio State University
(MS, 2001; PhD, 2004)

your profession?
It is difficult for me to think of anything that I really
hate about my profession.
What was the best career advice

Research Interests: you were given?
My research interests are primarily in the areas To form strong collaborations with peers. Although I
of statistical learning and data mining, as well as enjoy doing independent research, I have greatly ben-
their applications in bioinformatics. In particular, efited from collaborations with many other people,
I have been developing new statistical techniques both in statistics and in bioinformatics. Besides the
for analyzing high-dimensional data and data with values of making contacts in the profession, stimulat-
complex structure. ing discussions with my collaborators inspire me to
think deeply about problems and develop relevant
Name a statistician you admire and new techniques, as well as identify future problems
tell us why. to work on.
R. A. Fisher, who proposed and developed many
fundamental statistical techniques and concepts, What is one web site/blog you
including maximum likelihood, analysis of variance, can’t go a day without visiting?
Fisher’s information, and Fisher linear discriminate On the professional side, I visit the Google web site
analysis. It is impossible to be a statistician without every day to find information. On the personal side,
being influenced by Fisher’s statistical philosophy. In I frequently visit www.mitbbs.com. It is a forum for
addition to his most celebrated career in statistics, Chinese students and scholars to exchange informa-
Fisher was also an accomplished geneticist. With tion, news, and views about study, jobs, and everyday
remarkable contributions in both theoretical and life in the United States.
applied statistics, Fisher is certainly a wonderful role
model for young statisticians. What is something your friends would be
surprised to learn about you?
What do you love most about I am now comfortable with public speaking, although
your profession? I used to be shy and sometimes tremble while speak-
As a faculty member, I enjoy the freedom of being ing in public. I love swimming and reading Chinese
able to work on research problems that interest me. martial arts novels.
I also enjoy the satisfaction of solving a challenging
problem after great effort.

CAREER GUIDE
Age: 38
Title: Assistant Professor
Employer: University of
Missouri-Columbia
Alma Mater: University
of Illinois at Chicago
Research Interests:
My main research area is optimal design of
experiments. Currently, I am working on opti-
mal design for generalized linear models, an
underdeveloped area.
Name a statistician you admire and

tell us why.
Jack Kiefer, because of his fundamental contribution
to the design of experiments.
What do you love most about

your profession?
The opportunity to tackle challenging research prob-
lems that may make a contribution to science.

your profession?
Because of what I love most about my profession,
when I focus on a problem, I cannot live a ‘normal’
life until I find an answer.
What was the best career advice you

were given?
Work on problems that you honestly believe will
have a big impact and don’t worry about whether
it is the fashion today.
What is one web site/blog you

can’t go a day without visiting?
Sports news web sites.
What is something your friends would be

surprised to learn about you?
I am doing what I love to do. But, if I could choose a
second career, I would be a tennis player.

CAREER GUIDE
Age: 28
Title: Assistant
Professor of Statistics
Employer:
University of
California, Berkeley
Alma Mater:
Indian Statistical
Institute, Calcutta

your profession?
A pessimistic feeling that the honesty and thorough-
ness mentioned in the answer to the previous ques-
Research Interests: tion is slowly being replaced by other, less sustain-
Probability theory, theoretical statistics, statistical able ideals borrowed from other, younger disciplines.
physics. I am more comfortable with theoretical
issues, rather than applied ones. Mathematically, What was the best career advice you
I’m better at analytical thinking than combinato- were given?
rics. Currently, I’m working on a number of top- “Everyone is insecure. If you don’t feel insecure,
ics in probability, including spin glasses, random look at Gauss, and you will.” — From my adviser,
matrices, and concentration of measure. Persi Diaconis
Name a statistician you admire and What is one web site/blog you
tell us why. can’t go a day without visiting?
Charles Stein. I have very little understanding or Wikipedia. It’s addictive.
intuition about applied statistics, and how can I
idolize someone whose work I cannot understand What is something your friends would be
at a deep level? So that precludes applied statisti- surprised to learn about you?
cians. If I have to choose a theoretical statistician, I have only a few friends, and they know almost
Charles Stein’s name is the first that comes to mind. everything that there is to know about me. So, no
First, because I am more familiar with his work big surprises. Still, here’s a small one: When I was
than that of other towering figures in theoreti- young, I used to spend hours and hours trying to
cal statistics. Second, because his work is extraor- figure out how to spin a cricket ball. Never played
dinarily deep and beautiful. Whether it had any cricket, though.
significant impact on data analysis is a moot point.
What do you love most about

your profession?
A degree of honesty and thoroughness unmatched
by almost any other branch of the applied math-
ematical sciences. I think that’s part of our legacy,
and I hope we don’t lose that.


SPECIAL MATERIAL
MathStatistics
Is Music ;
Is Literature
(Or, Why Are There No Six-Year-Old Novelists?)
Richard D. De Veaux,
Williams College, and
Paul F. Velleman,
Cornell University
A
lmost 30 years ago, something happened Council of Teachers of Mathematics and the ASA
that made introductory statistics harder to that produced the Guidelines for Assessment and
teach. Students didn’t suddenly become less Instruction in Statistics Education (GAISE) Report—
teachable, nor did professors forget their craft. It was have pushed us all to change our teaching. And a
that we began to switch from teaching statistics as a new generation of texts has appeared following the
mathematics course to teaching the art and craft of advice of these reports, challenging statistics teach-
statistics as its own discipline. When statistics was ers to use this new approach.
viewed as a subspecialty of mathematics, students But why is it more difficult to teach this way?
were taught to manipulate formulas and calculate And why is it so important that we do?
the ‘correct’ answer to rote exercises. Life for the By comparison, let’s look at mathematics. Much
teacher, both as instructor and grader, was easy. of the beauty of mathematics stems from its axiom-
That started changing in the early 1980s. The atic structure and logical development. That same
video series “Against All Odds” appeared, and David structure facilitates—in fact dictates—the order
Moore and George McCabe published Introduction in which the material is taught. It also ensures the
to the Practice of Statistics. Since then, two pioneering course is self-contained, so there are no surprises.
committees—one for the Mathematical Association But, modern statistics courses are not like that, and
of America and the ASA and one for the National that can frustrate students who were expecting a

math class. As a student of one of us once wrote on remarkable that some of the music Mozart wrote at
the course evaluation form, “This course should be age 5 is still in the repertoire.
more like a math course, with everything you need Also, chess prodigies continue to appear.
laid out beforehand.” Sergey Karjakin is the youngest grandmaster ever
Mathematics has a long history of prodigies and at 12 years, 7 months. The infamous late Bobby
geniuses, with many of the most famous luminar- Fischer—who was youngest in 1958 when he
ies showing their genius at remarkably early ages. became a grand master at 15 years, 6 months, and
We’ve all heard at least one version of the famous 1 day—is now only 19th on that list.
story of young Carl Friedrich Gauss. A web search But there are only a few fields that develop
finds more than 100 retellings of the story, but prodigies, and all seem to be self-contained. For
an article by Brian Hayes in American Scientist’s example, as professor of English at the University
“Gauss’s Day of Reckoning” identifies a version of Connecticut, Thomas Dulack observed, “There
actually recounted at Gauss’ funeral. In that ver- are no child prodigies in literature.” Although one
sion, Gauss—age 7 and the youngest in the class— might argue that William Cullen Bryant, Thomas
summed the numbers from 1 to 100 in seconds, Chatterton, H. P. Lovecraft, or Mattie Stepanek
wrote the answer on his slate, and then threw it qualifies as a literary prodigy, that list doesn’t have Carl Friedrich Gauss, a
down on the table mumbling “there it lies” in the quite the same panache as the others we’ve cited. It’s math prodigy at the age of 7
local dialect. It was perhaps an hour later that the no easier to find prodigies in art, poetry, philosophy,
teacher discovered that his answer was, in fact, the or other endeavors that require life experience.
only correct one in the room. What does any of this have to do with statistics
Prodigies in math can develop at remarkably and how can it help us understand why introduc-
early ages because math creates its own self-con- tory statistics is so hard to teach? The challenge for
sistent and isolated world. Pascal had worked out the student (and teacher) of introductory statistics
the first 23 propositions of Euclid by age 12 when is that, as literature and art, navigating through
his parents, who wanted him to concentrate on and making sense of it requires not just rules and
religion, finally relented and presented him with a axioms, but life experience and “common sense.”
copy of Euclid’s Elements. Galois wrote down the Although working with elementary statistics
essentials of what later became Galois Theory the requires some mathematical skills, we ask so much
night before a fateful duel when he was 20, or so the more of the intro stats student than is required by,
legend has it. In the modern era, Norbert Weiner for example, a student in his or her first calculus
entered Tufts at age 11; Charles Pfefferman of course. A student in calculus I is not asked to com-
Princeton was, at 22, the youngest full professor in ment on whether a question makes sense, whether Blaise Pascal was a French
American history; and Ruth Lawrence of Hebrew the assumptions are satisfied (e.g., Is the reservoir mathematician, and a
University passed her A-levels in pure math at age from which the water pouring really a cone?), to prodigy by the age of 12.
9 and became the youngest student ever to enroll at evaluate the consequences of the result, or to write
Oxford two years later. a sentence or two to communicate the answer to
Of course, mathematics isn’t the only field others. But, that’s exactly what the modern intro
that shows prodigies. Mozart, Schumann, and stats course demands.
Mendelssohn, among others, were young musical The challenge we face is that, unlike calculus I,
prodigies. Even though his music matured, it is we have a wide variety of skills to teach, and most of
them require judgment in addition to mathematical
manipulation. Judgment is best taught by example
and experience, which takes time. But, we’re sup-
posed to produce a student capable of these skills
in one term. It would be challenging enough
to teach the definitions, formulas, and skills
in the standard first course. To convey in
addition the grounds for sound judg-
ment is even more difficult. It should Évariste Galois was a French
be no wonder that the first course in mathematician and prodigy
statistics is widely acknowledged to who finished the Galois Theory
by the time he turned 20.
be one of the most difficult courses
to teach in the university.
It is not merely that we hope to
teach judgment to sophomores; we
are actually asking our students to
change the way they reason about

We thank the following JSM 2008 Sponsors for their financial support:
Platinum
Gold
Silver
the knowledge to act™

SPECIAL MATERIAL
the real world. P. F. Velleman in his 2003 key- communicate their results in plain language, not
note address to the Beyond the Formula con- only in statistical jargon.
ference called the skills they must acquire the As GAISE emphasized, we must place more
seven unnatural acts of statistical thinking: emphasis on the plan and communicate steps. The
emphasis of the traditional mathematical course,
➊ Think critically. on the do step can be largely replaced by relying
Challenge the data’s credentials; look on technology for the calculations and graphics.
for biases and lurking variables. In teaching students to think through the prob-
lem, plan their attack, and communicate results,
➋ Be skeptical. we bring students face-to-face with their real-world
Question authority and the current knowledge and experience—with the literature
theory. (Well, okay, sophomores do side of their maturing intellect. We owe them an
find this natural.) acknowledgement that we’ve done this. It isn’t fair
to emphasize the simplicity of the calculations or to
just provide a bunch of definitions in little boxes.
➌ Think about variation, rather than No comparative literature or philosophy teacher
about center.
would do that, and neither should we.
What guidance should we offer? First, we can
➍ Focus on what we don’t know. note that the judgment often called for in statistics
For example, a confidence interval is one that invites students to state their personal
exhibits how much we don’t know views. (After all, they are the ones who must be
about the parameter. 95% confident in their interval.) But, we can offer
guidance for their judgments; they must be guided
➎ Perfect the process. by the ethical goal of discovering, describing, mod-
Our best conclusion is often a refined eling, and understanding truth about the world.
question, but that means a student Second, we can remind students their introduc-
can’t memorize the ‘answer.’ tory statistics course is related to every other course
➏ Think about conditional

probabilities and rare events.
Humans just don’t do this well. Ask
any gambler. But, without this, the
student can’t understand a p-value.
Embrace vague concepts. Symmetry, center,
outlier, linear … the list of concepts fundamen-
tal to statistics but left without firm definitions
is quite long. What diligent student wanting to
learn the ‘right answer’ wouldn’t be dismayed?
How can we help students navigate these
woods? We don’t have definitive answers to the
question, in spite of our more than 50 years
(combined) teaching of introductory statistics.
But, we’d like to identify some themes that
might help us as a community to start a conver-
sation about some of the challenges.
We can help students by giving them a
structure for problemsolving that incorpo-
rates the requirement that they exercise their
judgment. In our books, we’ve recommended
that students follow the steps W. E. Deming
created more than 50 years ago in his advice
to industry: plan, do, check, act. We’ve sub-
stituted communicate for act to underscore
the importance of communicating to oth-
ers the results we see. Students must learn to

SPECIAL MATERIAL
Resources
statistics can be taught—or even less plausible, can
Guidelines for Assessment and Instruction in Statistics Education be learned—in a single term. Any objective consid-
(GAISE) Report, www.amstat.org/education/gaise eration of the breadth and depth of the concepts
and methods covered shows this to be absurdly
Hayes, Brian. (2006) “Gauss’s Day of Reckoning.” American Scientist, optimistic. Yet, few academic programs require
94(3):200. more than one course, and many of those that
require two are cutting back. We need to argue as a
Velleman, P.F. (2008) “Truth, Damn Truth, and Statistics.” Journal of discipline that an introductory statistics course must
Statistics Education, 16(2): www.amstat.org/publications/jse/v16n2/ cover more than an introduction to inference for
velleman.html. means if it is to teach the reasoning of statistics—
and that teaching that reasoning must be its goal
Velleman, P.F. (2003) “Thinking With Data: Seven Unnatural Acts
(not just teaching definitions and formulas.) But,
and Ten 400-Year-Old Aphorisms.” Keynote address to the Beyond
a more complete course that covers techniques that
the Formula conference, Rochester, New York. require more than rudimentary sophistication, such
Weiss, Cindy. (2006) “New York Philharmonic Selects as inference for regression and multiple regression,
UConn Prof to Revive Concert Series.” http://advance.uconn. is unlikely to have time to teach judgment, plan-
edu/2006/060424/06042412.htm ning, and communication. It will most likely be
pared down to a collection of equations and rules.
As a community, we need to make it clear that
the subject of statistics deserves both more respect
they may study. The reason they are taking statis- and more time, not because it covers so many
tics (or perhaps, the reason it’s required) is that they methods, but because it should teach the founda-
are accumulating the kind of knowledge about the tions of reasoning when we have data. Part of the
real world that will help them write literature and argument might be that, unlike students in subjects
read philosophy, and that kind of knowledge makes that exhibit prodigies, our students must summon
them qualified to make statistical judgments. Of their real-world knowledge to learn to think statisti-
course, by asking students to call upon what they’ve cally. And, that the effort by statistics teachers and
learned in other courses, we are encouraging them students will pay back correspondingly in all our
to solidify their knowledge from those courses. students do. Math is sometimes said to be the lan-
Third, we must actually require students to guage of science (and much social science), but sta-
demonstrate all the steps of a statistical analysis, tistics should teach students the structure for what
from problem formulation to communicating the it communicates.
results to making real-world recommendations on Is the effort to teach the modern course
what they find. Unfortunately, homework and worth it? We believe it is. Rather than a col-
exam problems that carry these requirements are lection of techniques or a ‘cookbook’ of situa-
harder to write and harder to grade. Training teach- tions and formulas, a modern course in statis-
ing assistants to reliably grade these efforts can be tics must teach students to reason about the
problematic. Moreover, many statistics instructors world. Although that makes the course more
are not trained in statistics, and they, too, can find difficult to teach and assess, it will make a dif-
this approach challenging. But, the results of teach- ference in students’ lives and serve them for the
ing a modern course reward both the student and rest of their academic careers and beyond. ■
teacher, in spite of its challenges.
We should also face outward to the academic Editor’s Note: This paper is based on several
community. There is a wide- talks given by the authors at the United States
spread impression Conference on Teaching Statistics (USCOTS).
that introductory

Shop
ASA
Purchase T-shirts,
books, JSM Proceedings,
and gift items!
Visit the ASA’s online
marketplace at
www.amstat.org/asastore
SAVE 10%
on your first purchase by entering ASASTORE at checkout.

PROFESSIONAL OPPORTUNITIES
P rofessional Opportunity Listings may not exceed 65 words, plus equal oppor-
tunity information. Ads must be received by the first of the preceding month
to ensure appearance in the next issue (i.e., September 1 for the October
issue). Ads received after the deadline will be held until the following issue.
Listings are shown alphabetically by state, followed by international listings.
Illinois
■ Full-time assistant professor, tenure-
track position in the Department of
Preventive Medicine, Northwestern
Vacancy listings may include the institutional name and address or be identified University. Requires a doctoral degree in
by number, as desired.
biostatistics or statistics for methodology
Professional Opportunities Vacancies also will be published on the ASA’s web
research and epidemiologic studies in the
site (www.amstat.org). Vacancy listings will appear on the web site for the entire
calendar month. Ads may not be placed for publication in the magazine only: area of cardiovascular disease. Excellent
All ads will be published both electronically and in print. communication skills are essential. To
Rates: $320 for nonprofit organizations (with proof of nonprofit status), $475 apply, send CV and references to: Marie
for all others. Member discounts are not given. Display advertising rates are at Lee, Department of Preventive Medicine,
www.amstat.org/advertising. NU Feinberg School of Medicine, 680
Listings will be invoiced following publication. All payments should be made N. Lake Shore Suite 1102, Chicago, IL
to the American Statistical Association. All material should be sent to Amstat News, 60611. Northwestern University is an
732 North Washington Street, Alexandria, VA 22314-1943; fax (703) 684- affirmative action/equal opportunity
2036; email advertise@amstat.org. employer.
Employers are expected to acknowledge all responses resulting from publica-
tion of the ad. Personnel advertising is accepted only with the understanding that ■ Faculty biostatistician, Department of
the advertiser does not discriminate among applicants on the basis of race, sex, Internal Medicine. PhD in biostatistics or
religion, age, color, national origin, handicap, or sexual orientation. related field required. Background in lon-
Also look for job ads on the ASA web site at www.amstat.org/jobweb.
gitudinal analysis and field epidemiologic
studies highly desirable. Excellent com-
munication and computing skills required.
Level of appointment commensurate with
experience. Letter, curriculum vitae, three
references to Carlos F. Mendes de Leon,
PhD, Rush Institute for Healthy Aging,
Rush University Medical Center, 1645
W. Jackson Blvd., Suite 675, Chicago, IL
60612. Rush University Medical Center
is an EOE.
■ Two, non-tenure-track full/associ-

ate/assistant professor positions in the
Biostatistical Collaboration Center.
Requires a doctoral degree in biostatistics/
statistics for collaboration, consultation,
teaching, and research in biomedical field.
Excellent communication skills are essen-
tial. At least one year post-doctoral expe-
rience. To apply, send CV and references
to: Thongsy Singvongsa, Department of
Preventive Medicine, NU Feinberg School
of Medicine, 680 N. Lake Shore Suite
1102, Chicago, IL 60611. Northwestern
University is an affirmative action/equal
opportunity employer.

Department Head Maryland

■ Seeking PhD/experienced master’s
The University of Texas at Dallas statisticians for Center for Devices
and Radiological Health, FDA, HHS.
Department of Mathematical Sciences Grapple with rich array of statistical
issues in clinical trials for new technolo-
gies, from LASIK and artificial hearts to
The University of Texas at Dallas seeks a distinguished scientist to head its Department
genetic tests and robotic surgery. Review
of Mathematical Sciences and enhance its national and international stature. A Ph.D. in
Mathematics, Applied Mathematics, Statistics, or related field is required together with a statistical designs/analyses issues in medi-
proven record of scholarly research commensurate with tenure at the full professor level. cal devices from invention to postmar-
The successful candidate will have a multidisciplinary research emphasis and will present a ket. Send CV to Greg Campbell, 1350
vision for the future of the department and clear evidence of leadership potential. The position
represents an unusual opportunity to shape a mathematics department for the future in an
Piccard Drive, HFZ-550, Rockville MD
expanding, research-oriented university with strong science, engineering, and management 20850, greg.campbell@fda.hhs.gov. FDA
emphases. Salary will be commensurate with qualifications is a smoke-free environment and an equal
and experience. opportunity employer. Permanent resi-
The department maintains strong undergraduate and graduate degree programs in
dency required.
mathematics, applied mathematics, statistics, and bioinformatics and computational
biology, and expects to add M.S. and Ph.D. degree programs in Actuarial Mathematics and
Biostatistics in the near future. Massachusetts
The university is located in the rapidly growing and increasingly cosmopolitan suburbs of ■ The Center for Health Policy and
North Dallas amid one of the largest and most vibrant concentrations of high technology, Research at UMass Medical School seeks
multi-national corporations in the nation, including a substantial energy sector. For more a senior biostatistician for appointment
information, visit http://www.utdallas.edu.
at the associate professor level or above.
Inquiries may be addressed to: Responsibilities: health services/policy
Dr. Raimund Ober, Professor, Chair of the Search Committee, ober@utdallas.edu study design and consultation to faculty
Dr. Michael Baron, Professor, mbaron@utdallas.edu developing scientific proposals and pub-
Dr. Mieczyslaw K Dabkowski, Assistant Professor, mdab@utdallas.edu
Dr. Sam Efromovich, Endowed Professor, efrom@utdallas.edu
lications. Qualifications: PhD in statistics
Dr. Robert Hilborn, Professor & Department Head, rhilborn@utdallas.edu or closely related discipline, 5+ years post-
Dr. David Lewis, Senior Lecturer II, dlewis@utdallas.edu doctoral experience, including applied
Dr. Viswanath Ramakrishna, Professor, vish@utdallas.edu research design and preparation of propos-
Review of applications will begin November 17, 2008, and will continue until the position is als for competitive funding, good inter-
filled. The successful candidate will fill the position effective August 2009, although other personal skills. www.umassmed.edu/chpr/
arrangements are negotiable. index.aspx. University of Massachusetts
Medical School is an AA/EOE.
Candidates should submit a complete resume or curriculum vitae, a statement of research
interests, a letter describing his/her vision for the development of the department, and the
contact information (names, addresses, telephone numbers, and email) for five professional ■ Massachusetts Institute of Technology
references. Electronic applications may be submitted via http://go.utdallas.edu/ Mathematics Department seeking to fill
facultyjobs and are highly encouraged. The electronic process will allow candidates to
combined teaching and research posi-
upload documents into a secure space directly from their computers. Emailed materials
will not be accepted. Indication of gender and ethnic origin for affirmative action statistical tions at instructor, assistant professor,
purposes is requested as part of the application process but is not required for consideration. or higher levels in statistics or applied
probability beginning September 2009.
Alternatively, application materials may be mailed to:
Academic Search #20096
Appointments are mainly based on
The University of Texas at Dallas exceptional research qualifications. PhD
Mail Station AD 42, Room MP 2.228 required by employment start date.
800 West Campbell Road Submit online at www.mathjobs.org: CV,
Richardson, Texas 75080-3021
U.S.A.
research description, three recommenda-
tion letters. Applications should be com-
plete by 1/1/2009. (See full classified text
at mathjobs for more information.). MIT
is an equal opportunity, affirmative action
The University of Texas at Dallas is an equal opportunity/affirmative action university and
encourages applicants from candidates who would enhance the diversity of the employer.
university’s faculty and administration.

■ Tufts University School of Dental
Medicine seeks a doctoral-level statisti-
Faculty Position
cian to collaborate with students and fac-
ulty on clinical and basic science research
Biostatistician
The Division of Public Health Sciences
and teach introductory statistics. PhD or
equivalent in statistics, biostatistics, or epi- The Division of Public Health Sciences of the Fred Hutchinson Cancer Research Center invites
demiology; research experience; familiar- applications for a faculty level biostatistician with an interest in the conduct and analysis of cancer
clinical trials and associated correlative studies to work primarily with the Southwest Oncology Group
ity with statistical software packages (SAS (SWOG) Statistical Center. SWOG is an NCI-funded national research network that conducts clinical
and SPSS); teaching experience; strong trials and related studies aimed at the prevention or cure of cancer. Experience with methodological
and collaborative research in clinical trials or high dimensional data analysis (genomics, proteomics) in
communication skills. Contact: Paul the context of clinical studies is desired. The position will be at the rank of either Assistant or Associate
Stark, MS, ScD, Director of Advanced Member (equivalent, respectively, to Assistant or Associate Professor at a university). An established
track record is required for a position at the Associate Member level. The incumbent will be expected
and Graduate Education, paul.stark@ to conduct an active program of independent and collaborative research pertinent to the mission of
tufts.edu. Tufts University is an AA/EOE the Fred Hutchinson Cancer Research Center and the Division of Public Health Sciences. The Fred
Hutchinson Cancer Research Center is a world-renowned research institution with large and active
employer and actively seeks candidates efforts in basic biological sciences, human biology, clinical research, epidemiology, biostatistics and
from diverse backgrounds. cancer prevention research. Its mission is the elimination of cancer as a cause of human suffering and
death. The Center conducts research of the highest standards to improve prevention and treatment of
cancer and related diseases. The Fred Hutchinson Cancer Research Center is an equal opportunity
■ The MIT Sloan School of Management employer. The Center has a culturally diverse faculty and strongly encourages applications from female
and minority candidates.
seeks applicants for a tenure-track faculty Deadline for applications is November 15, 2008. Please send applications including curriculum vitae,
position in statistics starting July 1, 2009. a letter describing research interests, and the names of four references to:
Applicants should possess a PhD in a rel- Search Committee Chair
Division of Public Health Sciences
evant field by date of appointment. Please Fred Hutchinson Cancer Research Center
submit your curriculum vitae, relevant 1100 Fairview Avenue North
M2-B500, Box 19024
Seattle, WA 98109-1024
The Fred Hutchinson Cancer Research Center is an equal opportunity employer.
Novartis Oncology –
Join an Industry Leader
At Novartis Oncology, our mission is to become the world’s premier oncology company by consistently discovering,
developing, and producing broadly available novel therapies that improve and extend the lives of cancer patients.
Our work is about harnessing the power of the imagination. It’s about applying our pooled experience in new ways, exploring new possibilities
and novel pathways, and finding better solutions for patients. Our primary goal is to research and develop new and more effective ways to treat
cancer—in essence, to bring our vision of research to life. Our passion is about believing in our ability to change the way cancer is treated—and
acting on this belief. Evidence of turning our thought into action can be seen in our array of innovative products designed to treat a variety of can-
cers. It is evident through our efforts in medical education, and through our unprecedented level of support, education, and financial assistance
for patients. Our portfolio provides a broad range of innovative therapies and practical solutions that enhance the lives of patients and their fami-
lies. Our efforts to discover and develop innovative approaches for the treatment of cancer have produced breakthrough medicines such as the
leukemia therapy Gleevec®/Glivec®; the breast cancer agent Femara® and Zometa® for the treatment of cancer-related bone complications.
We have approximately 3800 employees and operate in 50 countries. Our main locations include our global offices in Florham Park, New Jersey
and Basel, Switzerland.
Biostatistics and Statistical Reporting opportunities

At Novartis Oncology Global Development, we believe our people are our greatest strength. It is through their passion for our work that we can
provide novel therapies to patients in need. Within our Global Oncology Biostatistics and Statistical Reporting organization, we have several out-
standing opportunities available for statisticians and statistical programmers to be based in our global offices in New Jersey.
Skill and Experience Requirements

For both statisticians and statistical programmers, excellent quantitative abilities and communication skills (written, verbal and presentation) are
a must with a high level of analytical and conceptual ability in order to provide strategic focus to projects. Candidates must have a proven track
record of strong execution and results. Educationally, a PhD is preferred for our Biostatistics opportunities.
For our Programming opportunities a Bachelors degree is required and an advanced degree is strongly preferred. Education and experience
in Statistics is highly preferred. These positions require experience in the use of statistical software, particularly SAS. Experience analyzing and
reporting clinical research data is highly desired especially experience with Oncology data.
Email: Natalie.Wood@novartis.com • Linda.Finelli@novartis.com

Novartis Oncology • 180 Park Avenue • Bldg 104/2K37 • Florham Park, NJ 07932
Novartis Oncology is an EOE - MFVH

publications, a statement of objectives

and aspirations in research and education,
information about teaching experience
and performance, and three recommen-
dation letters by December 1, 2008 to:
stat-search@mit.edu. MIT is an equal
opportunity employer committed to
building a culturally diverse intellectual
community and strongly encourages
applications from women and minorities.
■ Amherst College, Assistant Professor

of Mathematics (Statistics). Tenure-track
position in the mathematics and com-
puter science department, beginning fall
2009. PhD in statistics, biostatistics, or
related field required with evidence of
excellence in both teaching and research.
Two courses per semester: statistics and
some mathematics. See www.amherst.
edu/~deanfac for details. Applications
(by December 15, 2008) and inquiries
to search@math.amherst.edu. Amherst
College is an AA/EOE.

New Jersey
■ Novo Nordisk Inc., U.S. affiliate of
Novo Nordisk A/S in Copenhagen,
Denmark, a world leader in diabetes care,
currently has an opening for an associate
director, biostatistics within our clinical
development division in our corporate
headquarters in Princeton, NJ. For more
information or to apply online, please visit
the careers section of our web site, www.
novonordisk-us.com, and reference position
number 1162BR. EOE.
North Carolina
■ Bioinformatician for the Section of
Statistical Genetics and Bioinformatics,
Biostatistical Sciences, Division of Public
Health Sciences, Wake Forest University
Health Sciences. Research collabora-
tion in genomics, proteomics, imaging,
bioinformatics, and related disciplines.
Provides bioinformatics solutions and
DEPARTMENT CHAIR AND ENDOWED ROLLINS PROFESSOR
DEPARTMENT OF BIOSTATISTICS AND BIOINFORMATICS

ROLLINS SCHOOL OF PUBLIC HEALTH and WOODRUFF HEALTH SCIENCES CENTER OF
EMORY UNIVERSITY
The Department of Biostatistics and Bioinformatics of the Rollins School of opportunities for research collaborations in clinical trials, the new Atlanta Clinical
Public Health and Woodruff Health Sciences Center of Emory University, Atlanta, and Translational Sciences Institute, computational life sciences, population based
Georgia, seeks a proven scholar and innovative leader to chair a growing and studies, emerging technologies and university-wide initiatives in global health,
dynamic department in a vibrant school and outstanding health sciences center. predictive health and neurosciences. The successful candidate will be expected
The Rollins Chair is supported by endowment funds. to capitalize on a wide range of collaborative opportunities at Emory and with
nearby institutions, including the Centers for Disease Control and Prevention, The
Candidates should possess a doctoral degree in an appropriate discipline; a strong Georgia Institute of Technology (e.g., the Emory-Georgia Tech Nanotechnology
record of academic research and teaching, particularly at the graduate level; Center for Personalized and Predictive Oncology) and other local universities.
a demonstrated capacity to secure external funding to support research; and a
compelling vision for an integrated faculty and programs in biostatistics and The Rollins School of Public Health is dedicated to teaching and research,
bioinformatics. The candidate should have a capacity for exceptional leadership currently employs 140 full-time faculty members and enrolls over 900 full and
in an academic setting and be able to foster multidisciplinary collaboration among part-time graduate students in its masters and doctoral programs http://www.sph.
academic departments and centers both within and outside Emory University. emory.edu/index.php. The department offers MPH, MSPH and PhD degrees and
a combined 5-year BS/MSPH degree with Emory College. The school is located
The current Department consists of 33 primary and joint faculty members as well on the Emory campus in close proximity to the Centers for Disease Control and
as 21 adjunct faculty members http://www.sph.emory.edu/bios/index.php. The Prevention, Emory’s Graduate Division of Biological and Biomedical Sciences,
Department has grown rapidly in recent years with active research programs in School of Nursing, School of Medicine, Yerkes National Primate Research Center
statistical modeling, spatial statistics, bioinformatics, imaging, survival analysis, and the Winship Cancer Institute and is near the American Cancer Society head-
misclassification models and multivariate survival models as well as health quarters. A major research university, Emory enrolls 11,350 students in under-
preparedness, infectious diseases, cancer, cardiovascular and renal diseases, and graduate, graduate and professional programs http://www.emory.edu.
mental health.
Applicants should send a letter indicating their interest and a curriculum vitae to:
The Chair will be expected to provide guidance and vision to the development of Melissa Sherrer, Rollins School of Public Health, Emory University, 1518 Clifton
an expanding department at the interface of biostatistics and bioinformatics while Road, N.E., Atlanta, GA 30322 USA, preferably via email to msherre@sph.emory.
performing as a scholar, a teacher in the department’s graduate academic pro- edu. Screening of applications will begin immediately and continue until the posi-
grams, a faculty colleague and mentor, and a leader in the school and university. tion is filled. Starting date is negotiable. Applications will be considered confiden-
Within the school and the Woodruff Health Sciences Center, there are substantial tial and references will not be contacted without the permission of the applicant.
Emory is an Equal Opportunity/Affirmative Action Employer

software development for publications

and research proposals. Email CV, research
statement, and three letters of reference to
dbsrecruit@wfubmc.edu. Wake Forest
University Health Sciences is an equal
opportunity/affirmative action employer.
Ohio
■ Associate/full professor, statistics, Miami
University, Oxford, Ohio. Visit www.
muohio.edu/mathstat/stat_ad08.html for
details. Send CV, transcripts, research
description, teaching philosophy, and 4
letters of recommendation to statsearch@
muohio.edu. At least one letter should
discuss teaching ability and potential.
Application screening begins 10/15/08
and continues until the position is filled.
Miami University is an EOE/AA employer
with smoke-free campuses. Campus Crime
and Safety Report—www.muohio.edu/
righttoknow. Hard copy upon request.
■ Cleveland Clinic: biostatisticians, out-

comes researchers, and statistical program-
mers; doctoral, masters, and bachelors
levels. The clinic is a nonprofit academic
medical center known globally as a leader
in patient care, research, and education.
Its Department of Quantitiative Health
Sciences has over 90 professionals, includ-
ing more than 50 biostatisticians and
statistical programmers. Openings are
posted at www.clevelandclinic.org/qhs. The
Cleveland Clinic is an equal opportunity
employer.
Pennsylvania
■ Assistant Director/Associate Director
Biostatistics (Clinical Neuroscience).
Wyeth Pharmaceuticals, Collegeville/
Philadelphia, PA. A senior biostatistician
opportunity is available in the CNS
group of Wyeth Pharmaceuticals, which
has one of the most robust Alzheimer’s
pipelines in the industry. Be respon-
sible for strategy and the overall statis-
tical support for key clinical projects.

Thinking of Your
Future?
Let the ASA help you realize
your professional goals.
Visit the ASA’s JobWeb—The JobWeb is a

targeted job database and résumé-posting service
www.amstat.org/jobweb

Details about this position and oth-

ers can be found at www.wyeth.com/
careers. (Requisition #21390). EOE,
M/F/D/V.
Texas
■ University of Texas Health Science
Center, Houston, is seeking a senior
biostatistician. The position offers exten-
sive opportunities for research leader-
ship/collaboration in the Coordinating
Center for Clinical Trials. Responsibilities
include holding leadership positions in
clinical trials, conducting biostatistical
research (methodological/collaborative),
teaching. Qualifications include earned
doctorate in biostatistics/related field,
commitment to teaching, evidence of
research/recognition in biostatistics/
clinical trials. Contact Barry Davis,
MD, PhD, barry.r.davis@uth.tmc.edu.
UTHSC is an EOE.
continued on page 71

BIOSTATISTICS and CLINICAL TRIALS
Faculty Position NORC conducts high quality social science
research in the public interest from its head-
quarters at the University of Chicago and
The University of Texas School of Public Health (UTSPH) invites applications from senior investiga- from its offices in Chicago, IL, Washington,
tors in biostatistics to fill a tenure track faculty position at the Associate or Full Professor level at DC, Bethesda, MD, and Berkeley, CA.
the UTSPH Houston Campus in the Texas Medical Center.
We conduct research in economics, demo-
The position offers extensive opportunities for research leadership and collaboration within graphics, education and child development,
the UTSPH Coordinating Center for Clinical Trials, which has made significant contributions to health, substance abuse, mental health,
cardiovascular disease and vision research by serving as a coordinating center for 17 nationwide justice, and survey quality both in the U.S.
multicenter clinical trials. Responsibilities include: holding leadership positions in clinical trials; and internationally. We offer full-service
conducting biostatistical research (methodological and collaborative); teaching biostatistics survey design and operations as well as
courses for students in MS, PhD, MPH, and DrPH degree programs; advising graduate students; strengths in analysis, information technol-
and participating in community service. ogy, and technical assistance. NORC sup-
ports the research needs of government in
Qualifications include: (1) earned doctorate in biostatistics or related field; (2) commitment to
the U.S. and abroad, international donor
excellence in teaching and advising graduate students; (3) evidence of both methodological and
agencies, foundations, academic research-
collaborative research accomplishments in clinical trials; (4) evidence of national/international
recognition in biostatistics and clinical trials; and (5) excellence in written and oral communica-
ers, and private organizations.
tion skills. NORC is actively seeking statisticians,
survey methodologists, statistical program-
Review of applications will begin immediately and continue until the position is filled. Academic
mers, data managers, survey directors,
rank will be determined by the qualifications of the candidate. Candidates should e-mail a letter
and social scientists with advanced training
describing their qualifications and interests along with their curriculum vitae, and contact infor-
or experience in survey research or survey
mation for three professional references to:
operations. New staff will be based in our
Barry R. Davis, M.D., Ph.D., Chair, Biostatistics Search Committee, The University of Texas Health
Chicago, IL or Washington, DC offices. To
Science Center at Houston, School of Public Health, 1200 Herman Pressler, E-809, Houston, Texas
learn more about NORC and to apply for
77030, FAX: 713-500-9530, email: Barry.R.Davis@uth.tmc.edu
employment, visit our website at:
The University of Texas Health Science Center at Houston is an EO/AA employer. M/F/D/V. Minorities and
women are strongly encouraged to apply. This is a security-sensitive position and thereby subject to Texas
http://www.norc.org/careers
Education code §51.215. A background check will be required for the final candidate. NORC is an affirmative action, equal opportunity employer
that values and actively seeks diversity in the workforce.


■ The University of Texas School of
Public Health is seeking a director of Instructor Position
the Division of Biostatistics. Leadership at the Virginia Commonwealth University School of Medicine Department
of the division requires the vision to of Social and Behavioral Health
develop new programs, to increase and
complement the success of exist- The Department of Social and Behavioral Health Virginia Commonwealth University School of
Medicine seeks applicants for an academic position at the Instructor level in the area of Biostatistics,
ing programs, and to mentor junior Social Epidemiology, or related field. The appointee’s responsibilities will include collaborative and
faculty and students. Candidates independent research related to social and behavioral sciences.
must have a PhD in biostatistics or Applicants must have graduate level training in Biostatistics, quantitative methodology and/
or Epidemiology, strong interest in collaborative work with faculty, and a record of peer-reviewed
statistics and have progressed to the publications in applied behavioral research. Candidates should have a working knowledge of
rank of full professor. Information statistical software packages (e.g., SPSS, SAS) for data analysis and statistical modeling. Candidates
contact: Robert.J.Hardy@uth.tmc. with an interest in cancer research are highly desirable. This position will provide support,
consultation and collaboration for researchers specializing in various aspects of social and behavioral
edu. The University of Texas Health health and cancer research. The candidate will provide consultation on statistical methods
Science Center at Houston is an EO/ commonly used in behavioral and cancer research. The candidate will be expected to work closely
with researchers in writing grant proposals and articles for publication. The successful candidate will
AA Employer. M/F/D/V. Minorities be encouraged to develop their own research program.
and women are strongly encouraged The Department of Social and Behavioral Health (http://www.behavioralhealth.vcu.edu/) at VCU
to apply. ■ has multiple externally funded research projects involving various aspects of health behavior and the
social context of healthcare outcomes. Faculty in the Department of Social and Behavioral Health are
closely linked to the Massey Cancer Center (http://www.massey.vcu.edu/) at VCU, which is one of the
nation’s NCI-designated cancer center offering outstanding opportunities for research.
Candidates should send a cover letter with their CV and the names of three references to:
Chair, Instructor Search Committee, Department of Social and Behavioral Health
Virginia Commonwealth University, McGuire Hall Annex
1112 East Clay Street, P.O. Box 980149
Richmond, VA 23298-0149
Expand Your Career Horizons! Consideration of application materials will continue until the position is filled.
Visit the ASA JobWeb at www. VCU is an urban, research intensive institution with a diverse university community and
commitment to multicultural opportunities. VCU is an equal opportunity/affirmative action
amstat.org/jobweb/index.cfm employer. Women, minorities, and persons with disabilities are encouraged to apply.
Open Faculty Positions
Biostatistics Support Unit Director and

RAHC Biostatistics Core Director
The Department of Epidemiology and Biostatistics at the University of Texas Health Science Center at San Antonio has two positions open. We seek a Director for
our new Biostatistics Support Unit (BSU). The BSU provides coordinated study design and statistical support for clinical and translational research within the univer-
sity community and with partner organizations. The BSU Director supervises and assigns projects to BSU personnel; monitors project flow and deadlines; maintains
contact with investigators to track project progress; tracks grant reviews and funding or publication status; and oversees the collection of web-based quality evalua-
tions of services. We also seek a Director for our new Biostatistics Unit that will be located in our Regional Academic Health Center (RAHC) in Harlingen, Texas. The
Biostatistics Unit will provide coordinated study design, statistical and clinical informatics support for clinical and translational researchers in our campuses located in
Harlingen and Edinburg, Texas in the Rio Grande Valley. The Unit Director will provide direct collaborative support for RAHC researchers.
These positions require a DrPH, a PhD in Biostatistics or Statistics, or a PhD in a related health science discipline. We strongly prefer 5 or more years of post-
doctoral experience in a collaborative research environment. Successful candidates are expected to have outstanding collaborative scientific research experience, as
well as excellent verbal and written communication skills. The candidates will be expected to help expand academic mission of the Department of Epidemiology and
Biostatistics in the School of Medicine.
Currently, the Department of Epidemiology and Biostatistics, consisting of 29 faculty and 25 staff, is a part of several large research programs such as the CTSA-
funded Institute for Integration of Medicine and Science (IIMS), the Cancer Therapy & Research Center, Greehey Children's Cancer Research Institute, and the
Barshop Institute for Longevity and Aging Studies. The department’s activities span basic, translational, clinical and population-based research. Several other new
departmental faculty will be located in Harlingen, Texas.
The University of Texas Health Science Center at San Antonio is located in suburban Northwest San Antonio.
Interested candidates should send a letter of interest, curriculum vitae along with names, addresses and phone numbers of three references to:
John E. Cornell, Ph.D., Chair, Search Committee
Department of Epidemiology and Biostatistics
The University of Texas Health Science Center at San Antonio
7703 Floyd Curl Drive, Mail Code 7933
San Antonio, TX 78229-3900
The University of Texas Health Science Center at San Antonio is an Equal Employment Opportunity/Affirmative Action Employer.
All faculty appointments are designated as security sensitive positions.

Associate/Assistant
Professor of Epidemiology
The Department of Epidemiology, Graduate School of
Public Health, University of Pittsburgh invites applica-
tions for a faculty position at the Associate or Assistant
Professor level in the Epidemiology Data Center
(EDC). A tenure stream or non tenure stream position
is available, with teaching required for a tenure stream
appointment. Candidates at the Associate Profes-
sor level must have a record of external funding and
peer-reviewed publications. Candidates at the Assistant
Professor level should have demonstrated potential for
external funding and peer-reviewed publications. The
faculty member will be expected to develop his/her
own research, publish in peer-reviewed journals and
collaborate with clinical investigators. The successful
candidate will have excellent communication skills. A
ASA’s Online
FORUM
doctoral degree in Statistics, Biostatistics, Epidemiology,
Bioinformatics or a related field, or a medical degree
with graduate training in public health is required. All
specialty areas will be considered, but bioinformatics,
clinical trials, epidemiological modeling and genetics
will be given high priority. For information about the
Epidemiology Data Center visit http://www.edc.gsph. is available to members.
pitt.edu. Salary and rank will be commensurate with
experience. Applications will be reviewed until position
is filled.
Applicants should send to Position #0124944, cur- The forum encourages general discussion related
riculum vitae, a letter indicating area(s) of expertise and
a list of three references to:
Steven Belle, PhD
to statistics. Access to the forum may be
Graduate School of Public Health
University of Pittsburgh
504 Parran Hall
Pittsburgh, PA 15261
found in theASA Members Only
The University of Pittsburgh is an Affirmative Action
area at www.amstat.org/membersonly.
Equal Opportunity Employer.



Professor/Associate
Professor of Epidemiology
(Two Vacancies)
The Department of Epidemiology, Graduate School

of Public Health, University of Pittsburgh invites
applications for faculty positions at the Professor
or Associate Professor level in The Epidemiology
Data Center (EDC). Tenure stream or non tenure
stream positions are available, with teaching required
for a tenure stream appointment. Candidates must
have a record of external funding and peer-reviewed
publications as well as have excellent communication
skills. Faculty members will be expected to further
develop their own research, publish in peer-reviewed
journals and collaborate with clinical investigators
both nationally and within the University of Pitts-
burgh. A doctoral degree in Statistics, Biostatistics,
Epidemiology, Bioinformatics or a related field, or a
medical degree with graduate training in public health
is required. All specialty areas will be considered,
but bioinformatics, clinical trials, epidemiological
modeling and genetics will be given high priority.
For information about the Epidemiology Data Center
visit http://www.edc.gsph.pitt.edu. Salary and rank will be
commensurate with experience. Applications will be
reviewed until positions are filled.
Applicants should send to Position # 0124942,
curriculum vitae, a letter indicating area(s) of expertise
and a list of three references to:
Steven Belle, PhD
Graduate School of Public Health
University of Pittsburgh
504 Parran Hall
Pittsburgh, PA 15261
The University of Pittsburgh is an Affirmative Action

Equal Opportunity Employer.

Article
Reprints
Article reprints are
available from the ASA at http://
www.amstat.org/membership/
Individual_Articles.pdf
(single copies)
or
through the
Copyright Clearance Center,
www.copyright.com
(multiple copies).

ASA RESOURCES DIRECTORY
General Information Links

Information about any of the following may be Go to www.amstat.org/site/index.cfm?fuseaction
=linktous to add an ASA text or banner link to
found by linking to the web sites listed below or
your web site. From there, copy the text from the
by contacting Member Services. text box beside the option you select and paste it
into your web site. If you need help setting up the
Amstat Online link, please contact webmaster@amstat.org.
Access up-to-date information from Amstat

Online at www.amstat.org. Updated often, the Member Demographics
ASA’s web site gives you access to recent news; The ASA collects demographics data to create
the ASA’s products and services; and section, summary reports of the ASA membership and to
chapter, and committee homepages. develop new services for our members. All indi-
vidual demographic information remains confi-
ASA Membership dential. Please visit www.amstat.org/membersonly
and click “My Demographics” under “My Account”
ASA members receive Amstat News; enjoy a to participate.
variety of discounts on peer-reviewed journals,
magazines, newsletters, proceedings, and bro-
chures; build an invaluable network of more Member-Get-a-Member
than 18,000 members; expand career horizons; Nominate your colleagues for ASA membership!
and access up-to-date information from Amstat You depend on the ASA, and we count on you, too.
Online at www.amstat.org. To join, go to www. We know our current members are the best pos-
amstat.org/join. sible sources for new members who could benefit
from all the ASA has to offer. Please visit www.
Member Services amstat.org/membership/mgm.
Do you have an address change, membership
question, claim, or general inquiry? Please call the Sponsor an ASA Membership
ASA Member Service’s toll-free direct line, (888) Now you can sponsor the membership of a col-
231-3473, for all your ASA needs. If you prefer,
league, student, or K–12 school. You also can
email Member Services at asainfo@amstat.org,
Contacts give ASA membership as a gift! Please visit www.
or fax to (703) 684-2037.
amstat.org/membership/SAM.
Main
Campaigns for the ASA Student Membership Drive
American Statistical Association Students join the ASA at the special rate of $10.
732 North Washington Street Please visit www.amstat.org/membership/index.
Alexandria, VA 22314-1943
Giving to the ASA
cfm?fuseaction=student.
Phone: (703) 684-1221 Members and friends of the ASA may contribute
to the statistics profession by financially support-
Toll-free: (888) 231-3473 ing our mission and goals. The ASA is a 501(c) (3)
Chapter-Get-a-Member
Fax: (703) 684-2037 not-for-profit corporation. Contributions to the Chapters are the grassroots of the ASA, which
Email: asainfo@amstat.org ASA are tax-deductible. Please visit www.amstat. is why we are challenging our Chapters to help
org/giving for more information. increase membership with the Chapter-Get-a-
Member campaign. Please visit www.amstat.org/
ASA Departments membership/cgm.
Host a Membership Social
Advertising: advertise@amstat.org Host a membership social and the ASA will reim-
burse up to $100 to cover the cost of pizza, sand- Publications & Products
Amstat News: amstat@amstat.org
wiches, snacks, and soft drinks! Please visit www. Back Issues
Awards: awards@amstat.org amstat.org/membership/ChapterSchoolSocials.
pdf for more information. Back issues of ASA publications may be ordered at
Education: educinfo@amstat.org www.amstat.org/publications/index.
cfm?fuseaction=backissues.
Fellowships and Research Programs: K–12 Teachers, Join the ASA Today!
farp@amstat.org Join the ASA as a full individual member and Books
enroll your school in a K–12 Membership for only
Journals: journals@amstat.org $95! Visit www.amstat.org/membership/ ■ The ASA-SIAM Series on Statistics and
k12promotion. Applied Probability provides works of gen-
LINK: link@amstat.org eral interest to statisticians, applied mathemati-
cians, engineers, and scientists. The series is
Meetings: meetings@amstat.org
jointly published by the ASA and the Society
Membership/Member Services: for Industrial and Applied Mathematics.
Please visit www.amstat.org/comm/index.
asainfo@amstat.org
cfm?fuseaction=commdetails&txt
Web site: webmaster@amstat.org Comm=CBJPC06.

Ethical Guidelines
The purpose of this document is to encourage ethi-
cal and effective statistical work in morally conducive
working environments. Please visit www.amstat.org/
profession/index.cfm?fuseaction=ethicalstatistics or
contact Member Services to receive a copy.
Insurance
ASA members can take advantage of low group
insurance rates and a wide selection of insurance
coverage. For more information, contact ASA
Member Services to request a Group Insurance
brochure or visit www.amstat.org/services/index.
cfm?fuseaction=main#insurance.
Journals & Magazines

The ASA offers a variety of publications available
in print and online, including Amstat News; Journal
■ Teaching Statistics: Resources for Brochures of Agricultural, Biological, and Environmental
Undergraduate Instructors is a publication of The ASA publishes brochures on topics of inter- Statistics (JABES); Journal of the American Statistical
the ASA and the Mathematical Association of est to professional statisticians and those who Association (JASA); Journal of Business & Economic
America. For more information and to order, visit want to explore the field of statistics. Many bro- Statistics (JBES); Journal of Computational
www.amstat.org/publications/TeachStats.pdf. chures are available at no charge or for a nomi- and Graphical Statistics (JCGS); Journal of
nal fee. For more information and to order, visit Educational and Behavioral Statistics (JEBS);
■ Guidelines for Assessment and Instruction in www.amstat.org/ASAStore/Brochures_C104. Journal of Official Statistics (JOS); The American
Statistics Education (GAISE) Report: A Pre-K–12 cfm. Statistician (TAS); Technometrics (TECH); Survey
Curriculum Framework provides a conceptual Methodology; CHANCE; and STATS: The Magazine for
framework for K–12 statistics education and Students of Statistics. Please visit www.amstat.org/
CDs publications/index.cfm?fuseaction=main for more
a coherent picture of the overall curriculum.
■ 41 Years of Technometrics. Now, you can information.
For more information and to order, visit www.
have all articles published during the first
amstat.org/education/gaise.
41 years of Technometrics at your fingertips
Videotapes
with 41 Years of Technometrics on CD-ROM.
The mission of Technometrics has been to The Distinguished Statistician Video Series is now
contribute to the development and use of available for sale. The series offers the opportunity to
statistical methods in the physical, chemical, bring a lecture or discussion by a renowned statisti-
and engineering sciences. For more informa- cian to any ASA chapter meeting, university class-
tion and to order, visit www.amstat.org/ room, company archive library, or home collection.
publications/41yearstech.pdf. For more information, please contact educinfo@
amstat.org.
■ JSM Proceedings. Each year, the ASA pub-
lishes papers presented at JSM, as well as those
presented at other Section-sponsored con-
ferences. For more information about JSM
Member Benefits and Honors
Proceedings, please visit www.amstat.org/ Awards
publications.
Each year, a number of awards are presented in rec-
ognition of outstanding contributions or dedication
to the field of statistics, professional accomplish-
ments, and distinguished service. For more informa-
tion, please visit www.amstat.org/awards or email
awards@amstat.org.
ASA JobWeb/Career Center

The ASA JobWeb, a targeted job database and
résumé-posting service, will help you take advan-
tage of valuable career opportunities. Check out the
many services available at www.amstat.org/jobweb,
or email jobweb@amstat.org.

ASA RESOURCES DIRECTORY
Chapters & Sections
Network with thousands of colleagues through
ASA’s regional chapters and special-interest sec-
tions. For more information, visit www.amstat.org/
chapters or www.amstat.org/sections.
Current Index to Statistics (CIS)

ASA members enjoy free online access to the
Current Index to Statistics (CIS). To activate your
CIS access, log in to ASA “Members Only,” www.
amstat.org/membersonly, and select the CIS Web
Access tab at the top of the page for instructions.
Committees Other Resources

Since its founding in 1839, the ASA has depended JSTOR Advertising
on the invaluable service of dedicated volunteers,
ASA members can purchase access to the JSTOR Advertise in the ASA’s most popular publications or
working through committees, to achieve its goals
database of JASA, TAS, Technometrics, JABES, online. ASA corporate and institutional members
of furthering the statistical profession. Please visit
JBES and JCGS. JSTOR is a web archive of every receive significant discounts on display advertising.
www.amstat.org/com for information.
issue of these publications with the exception of Please visit www.amstat.org/advertising or email
the most recent five years. To purchase access to advertise@amstat.org.
Education these publications via JSTOR, visit www.amstat.
org/publications/JSTORSpecialOffer.pdf.
With the help of knowledgeable and well- ASA Calendar of Events
respected leaders in the profession, the ASA’s
The ASA Calendar of Events is a searchable online
Center for Statistics Education is able to provide Member Directory database of statistical events throughout the world.
quality educational opportunities to statisticians of
A full member directory is available at www. Promote your meetings and events through the
all levels of achievement. Please visit www.amstat.
amstat.org/membersearch/index.cfm or via ASA’s online calendar and Amstat News. Visit
org/education or email educinfo@amstat.org.
the ASA “Members Only” web site, which allows www.amstat.org/dateline.
a more detailed search. Only members who
E-newsletter choose to participate are listed.
Author Style Guidelines
ASA members also receive our monthly
Author style guidelines are available at
e-newsletter, ASA Member News, an update full of “Members Only” www.amstat.org/publications/index.
and special announcements. View the latest issue
As a member of the ASA, you have access to cfm?fuseaction=style-guide.
at www.amstat.org/newsletters, or view the entire
many additional features and options, including
archive at www.amstat.org/membersonly.
an enhanced searchable member directory, the
Editorial Calendar
E-newsletter archive, CIS, and “Members Only”
Fellows discounts. You also have the ability to manage The Amstat News editorial calendar is available at
your account online. Please visit www.amstat. www.amstat.org/publications/amsn/Media/05.pdf.
Annually, full members of established reputation
who have made outstanding contributions in org/membersonly.
some respect of statistical work are recognized by Mailing Lists
selection for the honorary title of Fellow. Please Reciprocal Societies of the ASA The ASA membership list is available for rental
visit www.amstat.org/fellows for more informa-
The ASA is happy to announce an additional from our list broker, MGI Lists. Detailed information
tion.
membership discount for members of reciprocal about the ASA list is online at www.mgilists.com/
statistical societies. If you reside in a developing datacards/ASA.pdf. Please contact Candy Brecht of
Grant Program country and are a member of one of the ASA’s MGI Lists at cbrecht@mgilists.com for more infor-
reciprocal societies, you are eligible to receive mation or to order the ASA mailing list. Only mem-
Six major research programs and one grant pro-
an additional $5 off your developing country bers who choose to participate are listed.
gram are administered by the ASA. Please visit
www.amstat.org/research_grants or email farp@ membership dues. ICSA members receive an
amstat.org for more information. alternate discount at www.amstat.org/member-
ship/ICSAAd.pdf. Please visit www.amstat.org/
membership/devcountries/app
Joint Statistical Meetings (JSM) for more information.
JSM is the largest annual gathering of statisti-
cians in North America. Sponsored by the ASA, Salary Report
International Biometric Society (ENAR and WNAR),
the Institute of Mathematical Statistics, and the An annual report of salaries of academic statisti-
Statistical Society of Canada, JSM is attended by cians for instructors, assistant professors, associ-
more than 5,000 individuals. Please visit www. ate professors, and full professors is available
amstat.org/meetings or email JSM@amstat.org. online at www.amstat.org/profession.

September 2008 • Issue #375
A MSTATNEWS
ADVERTISING DIRECTORY
Listed below are our display advertisements only. If you are looking for job-placement
ads, please see the professional opportunities section. For more job listings or more
information on advertising, please visit our web site at www.amstat.org.
MISC. PRODUCTS AND SERVICES University of Louisville......................................................p. 66

Birkhäuser .........................................................................p. 39
University of Pittsburgh ............................................. p. 72, 75
Cambridge University Press...............................................p. 44
The University of Texas at Dallas ......................................p. 62
CRC Press .........................................................................p. 27
The University of Texas Health Science Center .......... p. 69, 71
Springer.............................................................................p. 59
The University of Utah .....................................................p. 76
Virginia Commonwealth University ..................... p. 67, 71, 75

Westat ...............................................................................p. 61
astellas ...............................................................................p. 74
Cincinnati Children’s Hospital Medical Center .................p. 64

SOFTWARE
The Cambridge Group......................................................p. 64
JMP, a business unit of SAS ..............................................p. 24
Emory University ..............................................................p. 65
MacKichan Software .........................................................p. 22
Fannie Mae .......................................................................p. 68
Minitab, Inc. .............................................................. p. 40, 41
Fred Hutchinson Cancer Research Center.........................p. 63
NCSS ................................................................................p. 45
Mayo Clinic ............................................................... p. 69, 70
SAS ............................................................................... cover 3
NORC ..............................................................................p. 69
Salford Systems .................................................................p. 42
Novartis Oncology ............................................................p. 63
SPSS ............................................................................. cover 4
RAND Corporation ..........................................................p. 65
StataCorp ..........................................................................p. 53
SAMSI ..............................................................................p. 73
Stat-Ease, Inc. ...................................................................p. 12
Smith Hanley ....................................................................p. 76
StatPoint/statgraphics ........................................................p. 33
SUNY Upstate Medical University ....................................p. 66
StatSoft ......................................................................... cover 2
Takeda Pharmaceutical ......................................................p. 72
SYSTAT ............................................................................p. 57
U.S. Census Bureau ..........................................................p. 75
The University of Georgia .................................................p. 68

SEPTEMBER AMSTAT FINAL.indd i 8/20/08 2:27:11 PM
SEPTEMBER AMSTAT FINAL.indd ii 8/20/08 2:27:12 PM

Sept 08 Am Stat News

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Sept 08 Am Stat News

Uploaded by

Copyright:

Available Formats

SPECIAL CAREER ISSUE • SPECIAL CAREER ISSUE • SPECIAL CAREER ISSUE

September 2008 • Issue #375

The Membership Magazine of the American Statistical Association

Publications Agreement No. 41544521

SEPTEMBER AMSTAT FINAL.indd A 8/20/08 2:26:50 PM

SEPTEMBER AMSTAT FINAL.indd 1 8/20/08 2:26:53 PM

Statistics in Monitoring the Nation’s Health

SEPTEMBER AMSTAT FINAL.indd 2 8/20/08 2:26:54 PM

SEPTEMBER 2008 AMSTAT NEWS 3

SEPTEMBER AMSTAT FINAL.indd 3 8/20/08 2:26:54 PM

Never Say Never

4 AMSTAT NEWS SEPTEMBER 2008

SEPTEMBER AMSTAT FINAL.indd 4 8/20/08 2:26:54 PM

SEPTEMBER AMSTAT FINAL.indd 5 8/20/08 2:26:54 PM

SEPTEMBER AMSTAT FINAL.indd 6 8/20/08 2:26:54 PM

SEPTEMBER 2008 AMSTAT NEWS 7

SEPTEMBER AMSTAT FINAL.indd 7 8/20/08 2:26:54 PM

SEPTEMBER AMSTAT FINAL.indd 8 8/20/08 2:26:54 PM

[Hess] Goodman was sold on controlled selection.

SEPTEMBER 2008 AMSTAT NEWS 9

SEPTEMBER AMSTAT FINAL.indd 9 8/20/08 2:26:54 PM

10 AMSTAT NEWS SEPTEMBER 2008

SEPTEMBER AMSTAT FINAL.indd 10 8/20/08 2:26:55 PM

Jack Chao-sheng Lee

The following excerpt is based on the 2007 article “In Memory

SEPTEMBER 2008 AMSTAT NEWS 11

SEPTEMBER AMSTAT FINAL.indd 11 8/20/08 2:26:55 PM

12 AMSTAT NEWS SEPTEMBER 2008

SEPTEMBER AMSTAT FINAL.indd 12 8/20/08 2:26:55 PM

Paul Dixon Minton

SEPTEMBER 2008 AMSTAT NEWS 13

SEPTEMBER AMSTAT FINAL.indd 13 8/20/08 2:26:55 PM

SEPTEMBER AMSTAT FINAL.indd 14 8/20/08 2:26:55 PM

Courtesy of the Social Security Administration Archives

SEPTEMBER 2008 AMSTAT NEWS 15

SEPTEMBER AMSTAT FINAL.indd 15 8/20/08 2:26:55 PM

16 AMSTAT NEWS SEPTEMBER 2008

SEPTEMBER AMSTAT FINAL.indd 16 8/20/08 2:26:55 PM

SEPTEMBER AMSTAT FINAL.indd 17 8/20/08 2:26:55 PM

18 AMSTAT NEWS SEPTEMBER 2008

SEPTEMBER AMSTAT FINAL.indd 18 8/20/08 2:26:55 PM

Log in at MEMBERS ONLY today to access your journals!

SEPTEMBER AMSTAT FINAL.indd 19 8/20/08 2:26:55 PM

SEPTEMBER AMSTAT FINAL.indd 20 8/20/08 2:26:56 PM

SEPTEMBER 2008 AMSTAT NEWS 21

SEPTEMBER AMSTAT FINAL.indd 21 8/20/08 2:26:57 PM

22 AMSTAT NEWS SEPTEMBER 2008

SEPTEMBER AMSTAT FINAL.indd 22 8/20/08 2:26:57 PM

SEPTEMBER AMSTAT FINAL.indd 23 8/20/08 2:26:57 PM

SEPTEMBER AMSTAT FINAL.indd 25 8/20/08 2:26:58 PM

26 AMSTAT NEWS SEPTEMBER 2008

SEPTEMBER AMSTAT FINAL.indd 26 8/20/08 2:26:58 PM

SEPTEMBER 2008 AMSTAT NEWS 27

SEPTEMBER AMSTAT FINAL.indd 27 8/20/08 2:26:58 PM

“My group does a lot of purely

OREGON Nancy Wang

IDAHO SOUTH DAKOTA

Kristi Epke NEBRASKA

“I am a high-school Geoffrey Bohling

SEPTEMBER AMSTAT FINAL.indd 28 8/20/08 2:26:58 PM

“I use statistics to analyze

data collected in various

health studies. ” Malay Ghosh

LOUISIANA Gainesville, Florida