October 2012 • Issue #424 AMSTATNEWS The Membership Magazine of the American Statistical Association • http://magazine.amstat.org

Recollections Thoughts AboutAbout JSMJSM inin& & SanSan DiegoDiego Many Honored at Presidential Address, Awards Ceremony

ALSO: NSF Graduate Research Fellowship

Statistician Greg Ridgeway New Deputy Director of National Institute of Justice

AmstatNews October 2012 • Issue #424 features Executive Director Ron Wasserstein: [email protected] 3 President’s Corner Associate Executive Director and Director of Operations 5 Member Spotlight Stephen Porzio: [email protected] Susan Gruber Director of Science Policy 6 ASA Commends NSF for Initiative, Commitment to Increasing Steve Pierson: [email protected] Statistics Profile Director of Education 6 ISS-2012 on Longitudinal Data Analysis a Success Rebecca Nichols [email protected] 7 Statistics for Policymakers Managing Editor Megan Murphy: [email protected] Statistical Modeling to Address the Problem of Illegal Immigration Production Coordinators/Graphic Designers Melissa Muko Gotherman: [email protected] 8 Statistician Greg Ridgeway New Deputy Director of National Kathryn Wright: [email protected] Institute of Justice Publications Coordinator 10 NSF Renews SAMSI for Five Years Val Nirala: [email protected] 11 ASA Records Viewable Online Advertising Manager Claudine Donovan: [email protected] 12 Meet NCHS Director, Edward Sondik

Contributing Staff Members 14 JASA Highlights Kathleen Wert • Rebecca Nichols • Eric Sampson Studies of Immune Response and False Discovery Rate Amstat News welcomes news items and letters from readers on matters Featured in September Issue of interest to the association and the profession. Address correspondence to Managing Editor, Amstat News, American Statistical Association, 732 North 16 Statistician's View Washington Street, Alexandria VA 22314-1943 USA, or email amstat@ A Tale of Two Researchers amstat.org. Items must be received by the first day of the preceding 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 artwork will be accepted in graphics file formats only (.jpg, etc.), minimum 300 dpi. No material in WordPerfect will be accepted. Amstat News (ISSN 0163-9617) is published monthly by the American columns Statistical Association, 732 North Washington Street, Alexandria VA 22314- 1943 USA. Periodicals postage paid at Alexandria, Virginia, and additional mailing offices. POSTMASTER: Send address changes to Amstat News, 732 18 Master’s Notebook North Washington Street, Alexandria VA 22314-1943 USA. Send Canadian NSF Graduate Research Fellowship address changes to APC, PO Box 503, RPO West Beaver Creek, Rich Hill, ON L4B 4R6. Annual subscriptions are $50 per year for nonmembers. Amstat This column is written for statisticians with master's degrees and highlights areas of News is the member publication of the ASA. For annual membership rates, employment that will benefit statisticians at the master's level. Comments and sug- see www.amstat.org/join or contact ASA Member Services at (888) 231-3473. gestions should be sent to Megan Murphy, Amstat News managing editor, at megan@ amstat.org. American Statistical Association 732 North Washington Street Alexandria, VA 22314–1943 USA Contributing Editors (703) 684–1221 • FAX: (703) 684-2037 Meredith Berthelson is pursuing her PhD in interdisciplinary studies at the University of Montana. She earned her master’s of interdisci- ASA GENERAL: [email protected] plinary studies at the University of Montana, and her undergradu- ADDRESS CHANGES: [email protected] ate career was spent at Montana State University, where she earned her BS in mathematics. AMSTAT EDITORIAL: [email protected] ADVERTISING: [email protected] Berthelson WebSite: http://magazine.amstat.org Printed in USA © 2012 American Statistical Association Jennifer Slimowitz Pearl serves as a program director in the Division of Mathematical Sciences at the National Science Foundation. She earned her PhD in mathematics, specializing in symplectic geom- etry from the State University of New York at Stony Brook, and her BS in mathematics from Duke University.

Promoting the Practice and Profession of Statistics

The American Statistical Association is the world’s largest Pearl community of statisticians. The ASA supports excellence in the development, application, and dissemination of statistical science through meetings, publications, membership services, education, accreditation, and advocacy. Our members serve in industry, government, and academia in more than 90 countries, advancing research and promoting sound statistical practice to inform public policy and improve human welfare. columns 20 STATtr@k Service-Oriented Statistics: What Can Students Do? Glen DePalma is a PhD student in the department of statistics at Purdue University and the Online Articles director of StatCom. His research interests include applied, Bayesian, and computational statistics, as well as distributed computing. The following articles in this issue can be found online Doug Baumann is a PhD student in the department of statistics at Purdue University. His at http://magazine.amstat.org. research focuses on the annotation-informed integration of “omic” data in next-genera- tion sequencing. Samuel Kou received this year’s Committee of Presidents Jeff Nisen is a fourth-year PhD student in the department of statistics at Purdue University. of Statistical Societies (COPSS) President’s Award for his His research focuses on the development of novel estimation and calibration procedures outstanding research contributions and service to the for stochastic models used in the financial engineering, risk management, and economet- field of statistics. Kou spoke to Bhramar Mukherjee about rics fields. winning the award. Read what he had to say online at Libo Wang is a PhD student in the department of statistics at Purdue University. Her http://magazine.amstat.org. research focuses on the application of linear mixed models in genome-wide associa- tion studies. The American Statistical Association/National Council of Teachers of Mathematics Joint Committee on Curriculum 21 175 in Statistics and Probability sponsored a Beyond AP The Future of Statistical Publications Statistics (BAPS) workshop during the annual Joint The ASA will celebrate its 175th anniversary in 2014. In preparation, column “175”— Statistical Meetings. Read the details at http://magazine. written by members of the ASA’s 175th Anniversary Steering Committee and other ASA amstat.org. members—will chronicle the theme chosen for the celebration, status of preparations, activities to take place, and, best yet, how you can get involved in propelling the ASA toward its bicentennial. The ASA sponsored a two-day Meeting Within a Meeting (MWM) statistics workshop for middle- and high-school Contributing Editor mathematics and science teachers in conjunction with David Banks is a professor of the practice of statistics at Duke the 2012 Joint Statistical Meetings. To find out more, visit University. He was coordinating editor of the Journal of the American Amstat News online at http://magazine.amstat.org. Statistical Association; editor of the Journal of Transportation Statistics; and associate editor for Statistical Methodology, American Mathematical Monthly, STAT, The Electronic Journal of Statistics, The department of statistics at Rajshahi University in and Environmetrics. Bangladesh will host the International Conference Banks on Statistical Data Mining for Bioinformatics, Health, Agriculture, and Environment from December 22–24 in Rajshahi, Bangladesh. The registration deadline is November 20. For more information, visit Amstat News online at http://magazine.amstat.org. departments

The editors of the peer-reviewed Journal of Research 22 meetings in Economics and International Finance (JREIF) are Recollections and Thoughts About JSM in San Diego accepting original articles in basic and applied research Many Honored at Presidential Address, Awards Ceremony and case studies and critical reviews, surveys, opinions, Committee of Presidents of Statistical Societies commentaries, and essays. Submit your manuscript(s) to COPSS Honors Statisticians [email protected]. A guide for authors and other details are available at http://interesjournals.org/JREIF/ 28 education Guide%20to%20Authors.htm. Planning Begins, Proposals Sought for USCOTS ’13 International Census at School Workshop Held in The Division of Mathematical Sciences (DMS) at the National Science Foundation launched a program Conjunction with JSM 2012 last October called Computational and Data-Enabled 2012 Educational Ambassador from Botswana Attends JSM Science and Engineering in Mathematical and Statistical Sciences (CDS&E-MSS). Awards have been made for the first round of competition. The submission window for member news the current year is November 25 to December 10. To find 31 Awards and Deadlines out more, visit Amstat News online at http://magazine. amstat.org. To read the program description, visit www. 32 People News nsf.gov/funding/pgm_summ.jsp?pims_id=504687. 36 Section • Chapter • Committee News Sara Murphy, series acquisitions editor for the ASA-SIAM 38 Calendar of Events Series on Statistics and Applied Probability, helps authors 41 Professional Opportunities determine the publishing route they want to pursue by outlining what the series has to offer. Visit Amstat News online2 amstatat http://magazine.amstat.org news october 2012 . president's corner Career Success Training for Statisticians: A Progress Update

From left: Rebecca Nichols, Rebecca Brafman, Bob Starbuck, and Ofer Harel listen to Jeanine Buchanich during the Train the Trainer work- shop at the 2012 Joint Statistical Meetings in San Diego, California.

ne of the presidential initiatives I described in Ballman, Jeanine Buchanich, Janet Buckingham, my January column was the formation of a Ron McRoberts, Gary Sullivan, Erin Tanenbaum, workgroup to develop career-building cours- and Jennifer Van Mullekom. For an update on their Oes that enable statisticians to become more successful. progress, here are Bob’s responses to some questions I In this column, the workgroup chair, Bob Starbuck, recently asked him. explains how this effort has evolved, how you can con- tribute, and how you will benefit. What courses are you pursuing? The workgroup first met at the 2011 Joint Statistical The Need for Career Success Training Meetings (JSM) in Miami Beach, Florida. We created Early in 2011, I invited a group of business execu- a list of 15 topics we thought had an important impact tives to meet at ASA headquarters for a discussion of on the career of a statistician. Knowing we could not the professional development needs of their statistical simultaneously tackle all these topics, we narrowed the staffs. When these executives were asked how the ASA list to the following four courses: might meet those needs, their top request was train- • Presentation Skills ing in career success factors such as communication, leadership, and teamwork. All of them said, “Our stat- • Influence and Leadership Skills Robert Rodriguez isticians need the ability to explain the relevance and • Personality Training and Team Building impact of their work to the rest of the organization.” In recent years, many of our members have • Career Planning expressed similar needs for career success training designed specifically for statisticians. University pro- Who is the audience? grams in statistics are the ideal place for this training The target audience is ASA members who are in the to begin. I encourage statistics departments to consid- early to middle stages of their careers and who stand er offering the types of courses described in “Preparing to gain the most from these courses. When a course Biostatisticians for Leadership Opportunities” is conducted on a university campus, we also expect at http://magazine.amstat.org/blog/2012/02/01/ students in the audience. And when a course is con- leadershipncarolin and “Scientific Course Strengthens ducted at a business facility, we expect statisticians who Students’ Communication Skills” at http://magazine. are not necessarily ASA members. amstat.org/blog/2012/02/01/pittsburgleaders. Today, however, our association has the opportunity to offer How do you plan to deliver the courses? training that will benefit the most members. The course on presentation skills has been developed To develop this training, I asked Bob Starbuck to for a classroom setting. For other topics, a webinar lead a workgroup whose current members are Karla or webcast might be a good medium.

october 2012 amstat news 3 train-the-trainer course conducted periodically to cre- ate a cadre of volunteer instructors. The first train-the-trainer course, delivered at JSM 2012, focused on teaching presentation skills. Twenty future trainers participated in this one-day course, which was led by Buchanich and Van Mullekom (DuPont). Participants were highly enthusiastic about the value of the material they learned to pres- ent. One described the content as “well prepared, very informative, with a large amount of useful information.” Because the participants, them- selves, were highly experienced presenters, they contributed to the training by suggesting ways to improve teaching presentation skills. Jennifer Van Mullekom during the Train the Trainer event at the In addition to preparing participants to be 2012 Joint Statistical Meetings in San Diego, California volunteer instructors, the training will pay off in their own work. One attendee concluded, “This Will the courses be affordable? session has given me excellent ideas for how to We are planning with ASA staff to make the courses teach my MS and PhD students in statistics to be affordable and accessible to the target audience. To better communicators.” minimize the cost of courses taught in classrooms, we are training a cadre of ASA volunteer instructors What is the process for hosting a course? who are located in regions where many ASA mem- ASA staff is preparing a web page that will indicate bers live and work. That will allow one-day courses geographical areas in which course instructors are to be offered locally, and it will reduce registration available. The web page will provide steps that ASA fees when instructor travel is not needed. chapters, sections, and other groups—such as compa- Because the courses will be taught by ASA vol- nies and university departments—can follow to host unteers, they will cost substantially less than corre- a course in their area. It also will announce train-the- sponding commercially available courses and deliver trainer opportunities. content that is tailored for professional statisticians. Host groups will be responsible for finding train- ing rooms, covering the cost of refreshment breaks and How are you creating course content? lunches, and publicizing courses to their members. Our original workgroup realized that, collectively, we ASA staff will provide support for online registration, did not possess enough material to create the courses downloading course materials, and publicizing the on our own. We decided to look outside of our group course outside the local area. for people who could contribute course content. After If you would like to attend a course in your area, we became aware of a course on presentation skills contact the leaders of your local group or chapter taught at the University of Pittsburgh by Jeanine (www.amstat.org/chapters/chapterofficers.cfm) and ask Buchanich, we enlisted her participation. them to host the course. We also became aware of a leadership develop- ment program created at Eli Lilly for their Statistics Will courses be available at conferences? and Advanced Analytics group. Gary Sullivan of Eli A course on presentation skills will be offered at the Lilly joined the workgroup and is providing significant Conference on Statistical Practice in New Orleans this input for the influence skills course. coming February. We are considering ways to offer We are still searching for material for courses in courses at future conferences. personality training, team building, and career plan- ning. If you have experience in these areas and would A Successful Start like to contribute, please contact me (Bob Starbuck) I am grateful to Bob Starbuck, the members of the at [email protected]. workgroup, and the initial volunteer instructors for launching our career success training. Please sign up How are instructors being trained? for this training and encourage others to participate. Course instructors will come primarily from the It will make a difference in your career and the careers ASA membership. For each course, we will prepare a of many others! n

4 amstat news october 2012 MEMBER SPOTLIGHT Susan Gruber

or many, the path to becoming a statistician is a circuitous one, and that was certainly the case for me. At the age of 45, with a history of job titlesF including chemist, computer scientist, teacher, and mom, I entered the MPH program in epidemiol- ogy and biostatistics and, from there, the PhD pro- gram in biostatistics at the University of California at Berkeley School of Public Health. The defining moment actually occurred more than a decade earlier. When my oldest son was born, we were told in no uncertain terms that the baby must never be put to sleep on his back because it would increase the risk of death due to aspiration. Fast-forward three years to the birth of my second child, when we were unequivocally Gruber advised to always place the baby on his back to reduce the risk of SIDS. How could medical decisionmakers In 2009, I had the privilege of visiting a group of have reversed course so quickly on an activity we have U.S. Food and Drug Administration (FDA) employ- been doing since the dawn of humanity? And today, ees with my PhD adviser, Mark van der Laan. The 20 years later, are we confident that the current recom- people in the room were the ones who approve or deny mendation is the correct one? use of drugs, write the warning labels, and study post- These questions speak to the heart of what we do market data on adverse events. Their determinations as statisticians and highlight a debate most of us have rely on statistical analyses of incomplete data, and any had with others or ourselves at some point. Decisions decision—even one to postpone a final decision—has must be made and actions must be taken in the face health consequences for the population. I’m not sure of incomplete information. As statisticians, are we how many of us outside government understand that bound to always deliver an estimate, or are there times each of these individuals shoulders a weighty public when the appropriate course of action is to conclude service burden. I certainly had not, and I found it the data do not contain sufficient information to reli- heartening to witness. ably estimate the desired quantity? To what extent are In 2012, I once again had the privilege of going we responsible for the conclusions drawn from esti- to a meeting of a consortium of FDA, industry, and mates we report? From a layman’s perspective, in the academic researchers, this time with Jamie Robins, absence of reliable data on all-cause mortality stratified who along with Miguel Hernán, is mentoring me in by sleeping position, should I have been admonished my current role as a postdoctoral research fellow in the to keep my oldest son from sleeping on his back? Can department of epidemiology at the Harvard School of a child aspirating vomit and dying in his sleep teach Public Health. Scientists, economists, policymakers, us anything about caution in translating statistical evi- and medical practitioners are increasingly turning to dence into a one-size-fits-all recommendation? What large observational data sets to find answers to causal biostatisticians do has real implications for real people. questions. There are enormous computational, philo- I greet the recent trend toward personalized medi- sophical, statistical, and mathematical challenges, and cine with enthusiasm. Treatment recommendations interdisciplinary teams are rising up to meet them. I that are conditional on covariates are likely to be see this as another cause for optimism and an indica- more effective. Babies who have acid reflux are pre- tion we are on the right path. sumably at higher risk of mortality when laid on their The turns my life has taken since beginning this backs than babies without. Should that be mentioned journey in 2005 amaze me. When I first dreamed up in instructions given to new parents? Or, in reality, the idea of going back to school, I almost didn’t do might that make matters worse by piling information it, because I thought I had waited too long. But it overload on top of parental sleep deprivation? The turns out I hadn’t. Given my background, it is not problems are complex and perhaps best addressed surprising that my research focuses on methodological through a process of asking clear questions, collect- advances in causal effect estimation. I am thrilled to ing high-quality data, and applying causal inference find myself part of the burgeoning community of bio- methodology to find answers where there is support statisticians working on causal inference. In the words in the data. of poet Louise Driscoll, “Hold fast your dreams!” n

october 2012 amstat news 5 ASA Commends NSF for Initiative, Commitment to Increasing Statistics Profile

uring the August meeting of the cooperation with members of other NSF advisory Mathematical and Physical Sciences committees and community members. Advisory Committee (MPSAC), National ASA Executive Director Ron Wasserstein DScience Foundation (NSF) Assistant Director Ed responded by commending the NSF leadership for Seidel announced in an initiative to “examine fund- its commitment to recognizing the statistical sci- ing for statistical sciences research at NSF including ences and the contributions statisticians are mak- organizational alternatives and new initiatives.” He ing to NSF’s mission. Wasserstein also said that, also affirmed the NSF’s “strong commitment to the “raising the profile of statistics within NSF is an statistical sciences.” important step for our scientific discipline to fulfill To demonstrate the commitment, Seidel noted its potential.” that, whenever appropriate, NSF would specifically Seidel’s announcement came in response mention statistics alongside mathematics in budget to a proposal last fall to rename the Division of requests and form a subgroup of the MPSAC in Mathematical Sciences (DSM) to the division of mathematical and statistical sciences. Seidel said the name will stay the same, but noted in his initia- tive that “the recent vigorous discussions in the rel- evant communities, with the MPSAC, and within ISS-2012 on Longitudinal NSF, have led to a better understanding of the criti- cal and expanding role of the statistical sciences.” Data Analysis a Success A detailed report on the name change—written Brajendra C. Sutradhar, Memorial University by a subcommittee of MPSAC—also was presented at the meeting. The report summarizes the com- The International Symposium in Statistics (ISS) on Longitudinal ments received from the mathematical and statistical Data Analysis Subject to Measurement Error, Missing Values, societies and individuals in the science community. and/or Outliers took place at Memorial University from July Participating in the meeting by phone, DMS 16–18. Fifty-one delegates from countries including Brazil, Director Sastry Pantula thanked the committee for Switzerland, Spain, The Netherlands, Mauritius, the United the report and said the name change was originally States, and Canada attended the meeting, which covered three proposed to recognize the distinct discipline of sta- specialized research areas for longitudinal data analysis. tistics. “I am happy that the process opened worth- Included were an academic program and various social events, while discussions within the communities,” he said. including a barbeque, symposium banquet, and local scenic bus “I look forward to working with both mathematical tour. The welcome address was given by Ray Gosine, associate and statistical sciences communities to look for col- vice president research at Memorial University, followed by a ple- laborative opportunities for future. There is strength nary talk given by Brajendra Sutradhar on successes and further in the union, and there is strength in numbers.” challenges in longitudinal data analysis when data are subject to To read Seidel’s memo and the report on the measurement error. name change, visit the NSF Advisory Committee Five invited speakers covered the three areas of the meeting, and meeting’s web page at www.nsf.gov/events/event_summ. contributed papers were presented by 10 speakers, including four jsp?cntn_id=124926&org=NSF. To read the ASA graduate students. Select papers will be published as lecture notes statement, visit www.amstat.org/about/pressreleases.cfm. in Springer’s lecture note series. For a detailed report on the meeting, read the blog Sutradhar organized the symposium, which was sponsored post written by ASA Science Policy Director Steve n by Memorial University, the Statistical Society of Canada, the Pierson at http://tinyurl.com/9owo2s2. Institute of Mathematical Statistics, and the Atlantic Association for Research in the Mathematical Sciences. More information can be found at www.iss-2012-stjohns.ca.

6 amstat news october 2012 Statistics for Policymakers Statistical Modeling to Address the Problem of Illegal Immigration David Salsburg

he problem of immigration reform resembles We would never resort to this in the United States, a situation that those of us who have engaged but the number of people who managed to escape in in statistical consulting often find. One mem- spite of these structures provides us with one basic bit Tber of the team working on the problem insists one of information. This is the probability that someone aspect be addressed first when that aspect is ill defined will cross even the most tightly controlled border. and often impossible to resolve. In the public discus- The Israeli fence blocking off the West Bank pro- In this inaugural article sion of immigration reform, there are politicians who vides another way to estimate minimum probability. for the series “Statistics insist we must first “secure our borders.” Over the Newspaper reports suggest that 2–3 would-be terror- for Policymakers,” David years, different attempts by the federal government to ists manage to cross each week in spite of the high-tech Salsburg provides an improve on border security have failed to meet this nature of the fence. illustration of how statistical thinking can address the ill-defined objective in the eyes of those who insist We can use statistical methods to get a handle on U.S. immigration stalemate. upon it. the probability that the border will be crossed for any Amstat News is inviting Statistical thinking can help defuse such a stale- given amount of effort put into “securing” that bor- articles by ASA members mate. A simple model can help focus the arguments der. We start with an estimate of how many would to share what they think around the parameters of that model and can be cross if there were no controls. Then, we can model policymakers should resolved with adequate data—data that are often the process so we have the decrease in probability of know about statistics. View details about this series at already on hand. crossing as a function of the amount of money put http://magazine.amstat. Immigrants who are illegally in the United States into the effort. org/blog/2012/07/01/ arrive in several ways. Some cross the Mexican bor- We have data from the last few years, when statthinking. der, some arrive legally but overstay their visas, and attempts were made to strengthen the border cross- some are put ashore without having gone through ings, that would enable us to make those calcula- customs. The U.S. Census Bureau has estimates of tions. Like any other statistical approach, we would the proportion that belong in each category. So, any be able to estimate the most likely relationship discussion of “securing the borders” has to deal with between money expended and reduction in prob- these three aspects. If, as I gather from reading news- ability of crossing and we would have a measure of paper accounts, the vast majority is about equally split the uncertainty involved. For instance, we might between the first two, the problem becomes one of learn that an increase in enforcement cost by 10% constructing two models. will decrease the probability of crossing from 80% In each case, suppose we model the probability of to between 68% and 52% when enforcement is first entering illegally as a function of the effort expended started, but will decrease a probability of crossing of to prevent it. It is usually convenient to quantify the 40% only down to something between 27% and effort in terms of monetary cost. Once we have some 39%. These hypothetical numbers describe a func- decent bounds on the probability of entrance as a tion with diminishing return, a situation usually function of cost, we can multiply those probabilities found in circumstances like this. by expected numbers of those seeking to cross and the Arguments will still be there, but the arguments discussion can become one of how much we are will- will now be over the validity of the statistical estimates, ing to spend to get to a minimal specific number of the level of uncertainty that exists about them, and illegal entries. the amount of effort the nation can afford. If there are Consider the question of crossing the Mexican- arguments over the validity of the data used to esti- U.S. border. The Communist governments of Eastern mate the parameters, the model can be expanded by Europe have provided us with one end of that prob- the use of Bayesian methods that force the contenders ability function. They erected high walls with barbed to quantify their arguments. wire in cities. Elsewhere, a no-man’s land was plowed The problem of expired visas could be modeled the up and sowed with land mines. There were high same way. It would be necessary to model each type of cyclone fences on either side and guard towers ever visa separately, and there may be a problem in find- couple of kilometers with orders to shoot to kill any- ing enough reliable data, but the basic methodology one who attempted to cross. would be the same. n

october 2012 amstat news 7 Statistician Greg Ridgeway New Deputy Director of National Institute of Justice

mstat News invited new National Institute I thought my analyses of gun violence in East of Justice (NIJ) Deputy Director Greg Los Angeles and racial profiling were having an Ridgeway to respond to the following impact. And that led to new opportunities and Aquestions about his role in this position. Ridgeway, more criminal justice research questions. The an ASA member, also speaks to what he, as a stat- justice system continues to be a topic with a lot istician, brings to NIJ’s strategy to strengthen its of room for more statisticians. science mission. How did you get involved in criminal What about this position appealed to justice research? you? I started at the RAND Corporation pretty much There were three primary reasons I took the job. fresh out of graduate school. I found that many First, NIJ is the lead federal agency for crimi- statisticians were already doing good work in nal justice research. It has a lot of influence on public policy areas such as health, education, which direction the field goes. The idea of having and national security. However, I found a near a prominent role in that process was very appeal- vacuum in criminal justice research and a lot ing to me. Second, I think the criminal justice of opportunities for statistical work. Early on, system is a fascinating world in which to conduct research and, while I could have continued my research at RAND, I saw moving to NIJ as an opportunity to be exposed to new people and new ideas. Even my RAND colleagues said that spend- Greg Ridgeway ing some time in government is essential to really earned his PhD understand public policy. Already, I have learned a in statistics from lot. Last, both current and former NIJ staff who I the University of talked to thought I had a lot to contribute to NIJ. Washington and It is always nice to feel needed. his BS in statistics from California Polytechnic State The 2010 National Academies (NAS) University, San Report, “Strengthening the National Luis Obispo. Before Institute of Justice,” made many recom- joining the National mendations to strengthen the science Institute of Justice, mission and research infrastructure of Ridgeway was a statistician at the RAND NIJ, noting resource, autonomy, and Corporation and directed RAND’s criminal authority challenges. Did that report’s justice research program. He specializes in findings play into your decision to the analysis of criminal justice issues, most accept the position? prominently policing, gun violence prevention, The NAS report was released just before John and drug policy. Laub, NIJ’s current director joined and provid- ed a good starting place for him to examine and

8 amstat news october 2012 map out NIJ’s future. The creation of my position constantly tune into the input of practitioners, cul- was part of his strategy to strengthen NIJ’s sci- tivate the next generation of criminal justice schol- ence mission, a way of securing a senior scientific ars, maintain a fair and transparent review process, leader. I had been an NIJ grantee for many years and be good stewards of public dollars. and thought I knew NIJ quite well. However, the NAS report showed the complicated environ- ment in which NIJ and its staff work. Besides Describe some of your specific goals and the resource constraints, which will always exist, challenges as you begin your tenure. NIJ is unique among federal science agencies in NIJ’s greatest strategic asset is its ability to make that it also has a mission to serve the practitio- investments in ideas. Therefore, almost all of our ner community. That was very much the kind of challenges revolve around how to get the best organization I wanted to join. return on those investments. We need to have a good process for synthesizing the key concerns Why do you think they hired a statisti- that practitioners face in police, courts, and cor- rections so we’re investing in the right ideas. cian? What about your background and We need to make sure our review process is fair, experience appealed to NIJ? transparent, and efficient. We need to take some I’m sure they did not set out to find a statisti- risks on ideas that might take a decade to mature. cian. That is just an added bonus! For me, being And last, we need to make sure the results of our a statistician was a pathway. At RAND, I worked investments are having an impact on the field. on dozens of projects on gangs, guns, drugs, polic- There are several examples of NIJ successes in ing, but data analysis was the theme across all of each of these steps, and I will be working toward these. While some of these analyses were ending improvements in each. up in statistical journals, I was also testifying at city council meetings about my findings; advis- ing major city police chiefs; and exchanging ideas Other than you and the statisticians at with judges, attorneys, and advocates. In addition, the Bureau of Justice Statistics, we’re not for the last five years, I directed RAND’s crimi- aware of many other statisticians within nal justice research program and gained a lot of the DOJ science-related units. Does your experience in managing personnel, budgets, and strategies in a research organization. I think the hiring signal an increasing appreciation combination of my academic research and practi- for statisticians across the DOJ? cal and management experience made me a good I have never had the sense that there is a lack of fit for the job. appreciation for statisticians, but rather that there are few statisticians thinking about justice issues. Admittedly, numerous fields such as health care What is your role as deputy director? and pharmaceuticals, finance, and environmental I am responsible for NIJ’s scientific offices cov- sciences compete to attract statisticians. I would ering social science, technology and physical sci- like to attract more statisticians to justice system ences, and forensic science. I work closely with research. Forensic science, for example, is a key the office directors and their staffs on issues big area in which there is much room for new statisti- and small. The most challenging questions are cal analysis and research. The NAS report recom- about how NIJ can affect and improve our nation’s mended that NIJ “nurture and grow the pool of criminal justice system. What issues need greater researchers involved in criminal justice research.” attention? What kinds of projects should we invest I will make sure the statistical community is part in? Which projects show the greatest promise? of that pool. n Where do we need to take more risks? Along the way toward answering these questions, we need to

october 2012 amstat news 9 NSF Renews SAMSI for Five Years

he National Science Foundation (NSF) our research over the years,” said Richard Smith, recently renewed the Statistical and SAMSI director. Applied Mathematical Sciences Institute’s Over the 10 years, SAMSI’s programs have grown, T(SAMSI) grant for five years, making this SAMSI’s adding summer programs in 2006 and increasing its second renewal. education and outreach programs to undergraduate SAMSI is one of eight mathematical institutes and graduate students. Last year, the institute started funded by the NSF’s Division of Mathematical holding some of its workshops on the West Coast in Sciences, but this is the only one that focuses on addition to the workshops held in Research Triangle statistics and applied mathematics. It was origi- Park, North Carolina. nally founded in 2002 and is now celebrating its SAMSI has mentored more than 80 postdoctoral tenth anniversary. fellows and associates, many of whom now hold aca- The grant is a collaboration of Duke University, demic positions as well as positions in industry and North Carolina State University (NCSU), The government. SAMSI has hosted hundreds of visitors University of North Carolina at Chapel Hill (UNC), each year to participate in multidisciplinary work- and the National Institute of Statistical Sciences shops that push the boundaries of data-driven and (NISS) in conjunction with the William Kenan Jr. model-driven research. Institute for Engineering, Technology, and Science. In 2013–2014, SAMSI will hold two major pro- “The renewal of our funding is a tribute to the grams, one focusing on computational methods in the incredible commitment to SAMSI by the whole com- social sciences and the other focusing on low dimen- munity, our partner institutions, the national leader- sional structure in high-dimensional systems. It also ship in statistics and applied mathematics, everyone will host a summer program on neuroimaging data who has served SAMSI as a directorate member or analysis. For more information about SAMSI and program leader, and—more broadly—all the peo- its programs, visit www.samsi.info. n ple who have visited SAMSI and participated in

10 amstat news october 2012 ASA Records Viewable Online Laura Sullivan, Assistant Archivist

selection of records from the American Statistical Association Records— heldA by the Special Collections Department in the Iowa State University Library—is now available online at http:// cdm16001.contentdm.oclc.org/ cdm/search/collection/ p15031coll20. The records document the his- tory of the ASA and include the constitution and bylaws, meeting minutes, correspondence, let- ters of acceptance of early mem- bers, resolutions, a report on the American Statistical Association Conference on Transfer of Methodology Between Academic and Government Statistics, and a report on the first 40 years (1967–2007) of the archival activities done by the ASA. Many of the digitized records document the formation and early history of the ASA. Included are the the documents were digitized, as Edward Jarvis, Francis Lieber, minutes from the first meeting they were rebound or rehoused J. M. Peck, and William Prescott; on November 27, 1839, which as necessary. In addition to now handwritten minutes of the first was held “for the purpose of con- being available online for any- two organizational meetings; sidering the expediency of form- one with an Internet connection and the original constitution. ing a statistical society.” to read, the records document- The Organization and As part of the digitiza- ing the early years of the ASA Proceedings of the American tion process, the early records are now back in the Special Statistical Association (1839– were sent to the Conservation Collections Department in bet- 1872) contains records docu- Laboratory to be stabilized and ter shape than they were when menting the creation of the prepared for digitization. The they departed. association; the constitution and old bindings were beginning to The early records in the digi- bylaws; meeting minutes and fail, dirt was adhered to pages, tized collection include adminis- proceedings; names of Fellows; pressure-sensitive tape had been trative records, correspondence, names of honorary, correspond- applied to pages in an effort proceedings, board of direc- ing, and foreign members; and to hold them together, and a tors materials, and founding account information. number of pages were torn or documents. The administrative The Board of Directors records folded. Iowa State’s conserva- records (1838–1860) docu- (1839–1872) include handwritten tor cleaned pages, unbound ment the earliest activities of the minutes for meetings, including the fragile volumes, removed American Statistical Association. attendance, officer and committee brittle tape and adhesive, and Included are letters of acceptance lists, presentations, proposals for mended tears and creases. After from early ASA members such membership, and resolutions. n

october 2012 amstat news 11 Meet NCHS Director, Edward Sondik

mstat News invited Edward J. Sondik of the National Center for Health Statistics (NCHS) to respond to the following questions so readers could learn more about him and the agency he leads. Look for other statistical agency head Ainterviews in past and forthcoming issues.

What have you enjoyed most about being our principal focus is national, we need to sup- head of NCHS? port measures at the state and local levels, too. Perhaps most challenging will be incorporating I’ve especially enjoyed balancing the center’s new data-collection technologies. Whether it’s dual role in the Department of Health and [using] the web to collect data or figuring out Human Services and our role as a federal sta- how to make sense of crowd technologies, I’m tistical agency. These complementary roles con- convinced that these technologies and electronic trast with the single statistical agency structure health records are in our future. The challenge? of other countries. We serve the core policy How to use them to give quality information. and research needs of government as well as the Edward Sondik started private sector. We’re responsible for birth and out post-PhD life as Describe your top two or three priorities for death data and for measures of health status— an electrical engineer, NCHS. teaching systems and both self-reported and measured directly—and operations research– a variety of health care data, both from doctors’ Preparing the agency for those new modes of data related courses at offices and from hospitals. collection while at the same time building a staff Stanford, but became I’d also include in my list reporting on trends with the expertise we need to maintain our mis- interested in health as sion to serve as the gold standard for health data. an application, especially in how we live (and are born and die); trends in of modeling health as a health status and health risks; meeting the moni- It’s also essential that our program meet the very Markov process. toring needs of the decades-long Healthy People diverse and changing needs of the Department of program important to so many state efforts; Health and Human Services. Our knowledge of and, perhaps most, I’ve enjoyed supporting the factors affecting health and the very measures the changes we’ve made in data collection and of health itself are changing, and it is our respon- reporting technology and methodology. sibility to respond to these changes to give our country the data it needs for policy decision and What do you see as the biggest challenge(s) research. It’s critical, therefore, that we have open for NCHS, and have they changed significantly and frequent communication with the policy and since you started in this position? research staffs throughout our large department. Several challenges come to mind: Anticipating What do you see as the role for the broader data collection needs and changing our programs statistical community in supporting NCHS? to collect new information while supporting crit- ical trends; adapting to new modes for collecting I see this as a matter of mutual support in pro- data such as the web; combining data from dif- viding the multi-domain information that’s ferent modes since we know that mode makes needed to address issues of well-being—not only a difference in answers; understanding the rela- health, but also other aspects of well-being such as tionship between self-reported and clinical mea- income, education, participation in civic affairs, sures; moving into a world of electronic heath housing, and freedom from crime and violence. records and perhaps even crowd sourcing; and I also feel that joint research on topics in survey online tools for data analysis. A continuing chal- research, cognitive studies, use of administrative lenge is information on health disparities. While records, and collaborative work with academe, the

12 amstat news october 2012 National Science Foundation, and the National Academies would be helpful to all of us. Our knowledge of the factors affecting What do you see as the biggest accomplish- ment of the agency during your tenure? health and the very measures of health Our biggest accomplishment is how we’ve changed. Our staff has greatly reduced the time itself are changing ... from data collection to publication in all cases from literally years to months. We’ve anticipat- ed a wide variety of data needs crucial to policy and research to improve Americans’ health. As a couple of examples, we collect data on exposures to environmental substances particularly impor- and we’ve not suffered a cutback, which I think tant to environmental research and have begun a reflects that we are recognized as a critical part of follow-back component on heart disease in our the department and health infrastructure. To date, ambulatory care survey. Plus, we’ve expanded we’ve maintained our independence in accord our data dissemination through new reports and with the National Academy of Sciences’ principles the web. I might add that our budget has grown and practices. n

NOMINATIONS SOUGHT FOR THE MARVIN ZELEN LEADERSHIP AWARD IN STATISTICAL SCIENCE

The Department of Biostatistics at the Harvard School of Public Health named Nicholas P. Jewell, PhD as the recipient of the 2012 Marvin Zelen Leadership Award in Statistical Science. Dr. Jewell, Professor of Biostatistics and Statistics, School of Public Health and Department of Statistics, University of California, Berkeley, delivered a lecture entitled “Counting Civilian Casualties” on June 1 at Harvard University.

This annual award, supported by colleagues, friends and family, was established to honor Dr. Marvin Zelen’s long and distinguished career as a statistician and his major role in shaping the field of biostatistics.

The award recognizes an individual in government, industry, or academia, who by virtue of his/her outstanding leadership has greatly impacted the theory and practice of statistical science. While individual accomplishments are considered, the most distinguishing criterion is the awardees contribution to the creation of an environment in which statistical science and its applications have flourished. The award recipient will deliver a public lecture on statistical science at the Harvard School of Public Health and will be presented with a citation and an honorarium.

Nominations for next year’s award, to be given in May/June 2013, should be sent to the Marvin Zelen Leadership Award Committee, Department of Biostatistics, Harvard School of Public Health, 655 Huntington Avenue, Boston, MA 02115 or via email to [email protected]. Nominations should include a letter describing the contributions of the candidate, specifically highlighting the criteria for the award, and a curriculum vitae. Supporting letters and materials would be extremely helpful to the committee.

All nominations must be received by December 1, 2012.

october 2012 amstat news 13 JASA HIGHLIGHTS Studies of Immune Response and False Discovery Rate Featured in September Issue Hal Stern, JASA Applications and Case Studies and Coordinating Editor

ew technologies producing large data sets are A second application paper in the September a major force in modern statistical science, issue presents another example of high-dimensional motivating the development of new theory data motivating new methodology. In this case, andN methods and the application of these methods to network data that characterize the interactions of a important scientific problems. The increased atten- large number of individual units (e.g., predator-prey tion to methodology for large data sets is illustrated in relations among animal species) present the chal- the two JASA invited papers that were presented dur- lenge. Network data exhibit a number of complex ing the 2012 Joint Statistical Meetings in San Diego, phenomenon that are not easily accommodated California. The articles and the ensuing discussions by standard models, including a latent hierarchical make for interesting reading in the September issue of organization of the species, different types of interac- the Journal of the American Statistical Association. tions, and different network topologies (e.g., varying tendencies for within-subcommunity and between- Applications and Case Studies subcommunity interactions). The immune system in humans and other vertebrates Qirong Ho, Ankur Parikh, and Eric Xing pro- provides an adaptive and remarkably effective response pose “A Multiscale Community Blockmodel for to infections or vaccines. The response is determined Network Exploration” that allows investigators to largely by the spatio-temporal motion of lymphocyte infer these phenomena from a set of observed network cells. These cells move in response to unobservable interactions. Ho et al. develop a stochastic model for gradient fields; learning about these fields is critical to partitioning the units in the network, say species, in a understanding the basic biology behind infection and hierarchically organized tree. Each species’ interactions vaccine response. New technology measuring single- are governed by a multiscale membership vector that cell motion in real time provides an opportunity for describes the species likelihood of interacting with spe- investigators to infer the underlying gradient field by cies at different levels of the hierarchical tree. carefully modeling the motion of lymphocyte cells. Finally, a probability model that links the hierarchi- New types of data require new statistical approaches. cal tree and the membership vectors to observed net- One approach to the spatio-temporal lympho- work connections can be used to infer the parameters cyte motion data is described in “Bayesian Spatio- of the model. The authors demonstrate the approach Dynamic Modeling in Cell Motility Studies: on a network describing the predator-prey relation- Learning Nonlinear Taxic Fields Guiding the ships among a collection of 75 species of grass-feeding Immune Response” by Ioanna Manolopoulou, wasps and their parasites. Melanie Matheu, Michael Cahalan, Mike West, and Thomas Kepler. Manolopoulou and colleagues Theory and Methods develop a flexible statistical modeling framework by Multiple hypothesis testing is a fundamental problem building on a continuous-time stochastic differen- in high-dimensional statistical problems. For example, tial equation model for cell motion under a gradi- in genome-wide association studies, tens (or even hun- ent field. Markov chain Monte Carlo computational dreds) of thousands of tests are performed simultane- techniques are used to learn about the parameters that ously to determine which, if any, genetic markers are govern individual cell motion and to infer the under- associated with a given disease or trait. Researchers in lying gradient field. The approach they develop works such settings increasingly rely on procedures that con- extremely well on a simulated data set and provides trol the false discovery rate (FDR), the proportion of insight for experimental data from the lymph nodes rejected null hypotheses for which the null hypoth- of mice. Invited discussions by Edward Ionides, esis of no effect is actually true. Procedures have been Samuel Kou, and John Fricks (with colleagues Le developed that control the FDR in large problems Bao and Murali Haran) provide additional insight with independent test statistics. When test statistics are into the modeling and computational choices made correlated, false discovery control becomes challeng- by the authors. ing, especially if we wish to allow for arbitrary forms of dependence. 14 amstat news october 2012 The featured Theory and Methods paper— response for an individual for each treatment and then “Estimating False Discovery Proportion Under propose to assign patients to the treatment that yields Arbitrary Covariance Dependence” by Jianqing Fan, the highest expected response. This can work poorly if Xu Han, and Weijie Gu—proposes a novel method the first-stage models are overfit to the data. for controlling the false discovery rate based on a prin- The authors show that estimating an optimal treat- cipal factor approximation of the covariance matrix ment rule is equivalent to a classification problem (a of the test statistics. The approximation subtracts the patient with a bad outcome on the assigned treatment primary sources of dependence and this significantly is considered a misclassification). Some misclassifica- weakens the remaining correlation structure. This tions are more significant errors than others are; the allows for the development of an approximate expres- authors introduce differential weighting based on sion for the false discovery proportion. Discussants the patient outcome to address this issue. A machine Larry Wasserman, Peter Hall, Armin Schwartzman, learning approach, support vector machines, is used and Jiashun Jin provide additional insights and raise to find the optimal decision rule that minimizes the challenging questions about the proposed approach. expected weighted misclassification rate without esti- The potential for personalized medicine is explored mating expected responses separately for each treat- in the article “Estimating Individualized Treatment ment. The resulting estimator for the optimal treat- Rules Using Outcome Weighted Learning” by Yingqi ment rule has good statistical properties and performs Zhao, Donglin Zeng, A. John Rush, and Michael well in simulation studies and in an analysis of chronic Kosorok. Physicians note heterogeneous responses to depression data. treatment in many diseases; a drug that works well There are many other informative articles in for some individuals may not work at all for others. both sections of the September issue, as well as a Zhao et al. propose an optimal approach for using set of book reviews. The full list of articles and a list randomized trial results and individual prognostic of the books under review can be viewed at http:// factors (which may include genetic and other factors) amstat.tandfonline.com. ASA should log on through to develop optimal rules for assigning individuals to the Members Only link at www.amstat.org to access treatments. Standard approaches to this challenging their free online access to JASA. n problem first use the data to estimate the expected

october 2012 amstat news 15 statistician's view A Tale of Two Researchers

ere is a tale of two researchers living in side- further into the network. Two years later, there are by-side universes. They are working on the 211 follow-up papers. Compare this story to Jane 1.0. same research project. However, one of themH lives with a new form of academic publishing. The Current Universe Both academics are named Jane. Jane 1.0 lives in the present. When Jane 1.0 wants to publish her research, she emails the research to her The New Universe colleagues. They respond individually, unaware of the Jane 2.0 has discovered something important and others’ comments. They all think additional work is she publishes her research on a social network dedi- needed for publication. However, because the com- cated to academic publishing. The network doesn’t menters are not aware of the others’ comments, no have to be like Facebook, MySpace, LinkedIn, or individual colleague can identify the controversy that Twitter. Several websites have found the advantages Jane 2.0 was able to address quickly. of linking users without making the network the Jane 1.0 also posts her research in several places: central focus of the site. For example, YouTube and (1) on her web page, (2) as a technical report on her Yelp both host user-generated content. department’s page, and (3) on arXiv. Additionally, she The network automatically alerts her col- submits her research to a prestigious journal. Three leagues to her new publication, and it automati- months later, she receives a rejection from the journal. cally allows for a discussion thread beneath the Fortunately, the referee reports lead her to discover the link to her paper. Her research is controversial. It reason people are so frustrated. She runs the addition- sparks a discussion among the researchers who fol- al experiment (that Jane 2.0 ran within a few days of low her work. After discussing the problem for a first posting her research) and resubmits to the pres- couple days, the handful of commenters (who are tigious journal with a long letter explaining why she not anonymous, but are well-known researchers is resubmitting after being rejected. This letter sits on to Jane) agree she needs additional evidence. They the editor’s desk for at least a week before the editor suggest an additional simulation experiment. decides to send it out again. After sitting on the refer- Since Jane’s research has attracted quick attention, ees’ desk for another three weeks, they uniformly agree the network starts to suggest her paper to others who that this is groundbreaking research. After worrying might be interested. At the same time, Jane runs the about typesetting and other typographical issues, the other experiment and links the results as an adden- paper is printed in the journal six months later. Since dum to her paper. This link appears directly next not all researchers read through the abstracts of every to the original paper; everyone who has previously single journal, several researchers do not see the paper downloaded her paper is alerted to the addendum. until four months later, when one of their graduate (Two days later, she posts a revised version of her students mentions it. (In Jane 2.0’s universe, content paper that incorporates this addendum.) This addi- recommendation software alerts researchers to inter- tional evidence addresses the controversy and several esting/relevant papers.) of the researchers who joined the discussion “repub- Follow-up research begins roughly one year after lish” Jane’s research on their own pages. The people Jane 1.0 initially published her research. This research who follow these other researchers will be notified of is not published for an additional year. So, two years Jane’s research. after Jane originally published her research, there are Jane’s paper is thought to open up several interest- only 22 follow-up papers. ing lines of inquiry. In the two weeks after Jane 2.0 first published her research, she has gone through a What We Should Expect from a Web- round of revisions and her paper has gone viral in her Enabled Publishing Environment community. One year later, there are 43 other papers The pace of research in the current universe stalls for that follow up on her research. They have been pub- several reasons. Some of these are emphasized in the lished on the network and are all linked in a com- above stories. Others are not. mon thread of papers. In that time, the editors of two journals have sent a message to Jane 2.0 asking (1) There is no public forum to discuss a paper. if they may conduct a formal peer review. Jane selects Sometimes, there are discussions at conferences; which journal she prefers, the article is reviewed and however, it is difficult to invite other people into accepted, and then linked on the journal’s page inside these discussions. It is difficult to make these dis- the network. The people who follow these journals cussions public for others to observe. are alerted to her research. This propels her research

16 amstat news october 2012 statistician's view

(2) For-profit journals used to serve a purpose looks like. Some paths are suboptimal. Some paths of printing ink on paper. This used to speed up will be designed for other disciplines. We need to research. They no longer add this value. Now, design a system that works for our community. they slow down research by putting up pay walls. At the same time, a social network without any Editors, referees, and researchers create the value people is worthless; the value dramatically increases and donate their time to the community. They as more people buy in. Making a separate, second are often paid with taxpayers’ dollars. Their efforts most–popular social network is not worthwhile (see should not be hidden behind a pay wall. MySpace). If we make the second or third site, we might be forced to follow the path others blaze. For (3) The current system recommends content these reasons, we need to build a network that others by sorting research into journals. However, our will want to join, and we need to move fast. research interests are much more varied than the Having a network supported by a large organiza- number of journals allow. A new, unified system tion such as the ASA or IMS would give credence to would enable “content recommendation,” like a site. This is important. However, for a site to spark, Netflix or Amazon, to identify research of interest. it needs much more. It needs to be easy. It needs to be effective. It needs to add value to people’s lives. (4) The current system is bulky. Papers are often The first step forward should be fast and easy. published in at least four locations. There is no One option is to build a site on top of arXiv. At first, systematic way for someone to stay up to date on the site does not need user profiles. It merely links to the current research. The new system would pro- articles (hosted elsewhere). This puts all the research vide the opportunity for a high-quality user inter- in one spot. As a second step, there should be a com- face for researchers to read about developments ment thread associated with each linked article. when they first happen. Additionally, there should be a “republish” button (5) The current paradigm accepts research that that “pushes” the good research to the top of the front is roughly 10–50 pages. In most journals, the page. These are extremely valuable steps forward. effective page restrictions are much tighter. Later, a more formal system can give more authen- However, suppose a graduate student reads a ticity by giving users profiles and letting them connect paper and runs a follow-up simulation that gives with fellow researchers. Whether this system should some small insight into a certain method. Unless be built in concert with arXiv is not clear. A huge this graduate student can expand this idea into a upside is that it would lead to immediate buy-in from full paper, no one sees the simulation. The new a wide range of researchers. One potential downside is paradigm would allow for users to post follow- that if arXiv owns the system, then the statistics com- up simulations with some code for others to try. munity loses control. In addition to expediting follow-up research, it Our current system of publishing was designed could also lead to rapid identification of coding for a world with different constraints. We have errors. Additionally, this site could host academic begun to realize a new system, through academic blogs, which have proved to be a popular medi- blogs and homepages, arXiv, and online journals. um for several researchers. On the upper page However, these sources lack unity and are therefore limit, any grand research that exceeds 50 pages more cumbersome than the limitations of current often must be split into several papers or pub- technology. A central electronic clearinghouse with lished in a monograph. Web-enabled publishing links, alerts, and discussion boards would reduce would not have such restrictions. some of the burden of finding research and provide a unified place to discuss research. Ultimately, this (6) This system also could allow conferences to would lead to an efficient, expedited, and accessi- have a central web page. Conference speakers ble interface for publishing and discussing research could provide links to papers/slides/etc. that would actively propel lines of inquiry forward. Academia is characterized by creative thought, inno- The Path Forward vation, and timely applications that benefit society, The ASA has convened a panel to discuss the future and the irony of being hamstrung by arcane modali- of academic publishing. There are several researchers ties is striking. It is time to move forward. n with deep concerns as to how a new system might disrupt the current system. However, we do not get Karl Rohe, Assistant Professor to choose if we follow a new path; a step forward is University of Wisconsin-Madison Statistics inevitable. We might get to decide what the new path Department

october 2012 amstat news 17 columns

Master’s Notebook NSF Graduate Research Fellowship Meredith Berthelson and Jennifer Slimowitz Pearl

statistics graduate students is the National Science Foundation’s (NSF) Graduate Research Fellowship Tips for Students Program (GRFP). To enter the competition, you need to submit a complete Director of the NSF Division of Mathematical application via NSF FastLane at www.fastlane.nsf.gov/grfp. The Sciences Sastry Pantula stated, “[The] NSF Graduate application consists of a personal statement, description of previous Research Fellowship (or an honorable mention in the research experience, proposed plan of research, and transcripts. Part competition) is certainly a feather in any future scien- of the application also includes three letters of reference, submitted tist’s cap! There are many well-qualified mathematics and statistics students in this country, and I would separately via FastLane by the reference writers. love to see many, many more of them take advantage Reviewers evaluate the applications based on intellectual merit and of this excellent opportunity.” broader impacts. For intellectual merit, you will need to demonstrate In 2012, the GRFP awarded 2,000 fellowships; your academic capability and other conventional requisites for only 75 of those were to students in mathematics and scholarly, scientific study. Details such as the ability to plan and statistics (3.75%). conduct research, work in a team as well as independently, and What are the key elements of the fellowship? It is a five-year award worth $126,000. The NSF gradu- interpret and communicate research are useful. ate fellow receives three years of support (useable over To demonstrate broader impacts, convey how your research a five-year period). For each of these three years, the will contribute on a larger scale to society and the breadth of its fellow receives a $30,000 stipend and the graduate audience. Will it encourage diversity, broaden opportunities, and institution receives a $12,000 educational allowance allow participation of all citizens in science and research? If so, this to cover tuition and all required fees. The fellow also should be evident to the reviewer. has access to international research opportunities and supercomputing resources. When preparing your application, you should be clear and specific, Eligible applicants must be either a U.S. citizen or so the reviewer doesn’t struggle as he or she is reading the national or permanent resident and an early-career application. Describe your experiences—whether they are personal, graduate student pursuing a research-based master’s professional, or educational—that have been factors in your or doctoral degree in an NSF-supported field. In preparation and that have driven you to pursue graduate study. Be mathematical and statistical sciences, the following detailed about your involvement in any scientific research activities categories are included: and what you learned from those experiences. If you have not Algebra, number theory, and combinatorics been involved with any direct research, then describe any activities Analysis you think have prepared you to start research. Also don’t let the reader try to glean from your writing that you “could” be a leader in Applied mathematics some capacity. Instead, describe your leadership potential directly. Biostatistics How do you see yourself contributing to research, education, and innovation? Provide the reviewers with your career aspirations and Computational and data-enabled science specific goals you hope to accomplish. You need to sell yourself in Computational mathematics your application. Computational statistics Geometric analysis Editor’s Note: This any students getting ready to graduate with article originally their baccalaureate degrees contemplate Logic or foundations of mathematics appeared in the September 2012 issue graduate studies or plan to continue their education.M One of the major obstacles can be funding: Mathematical biology of IMS Bulletin. Students who have just finished their undergraduate Probability education may not want to add more tuition bills to the pile. If only there were a way to help them con- Statistics tinue their education and execute some of the research they wish to do. Ah, but there is! One of the most Topology valuable funding mechanisms for mathematics and Other (related fields)

18 amstat news october 2012 columns

Applicants must be planning to enroll in an accredited institution in the United States by the fall following announcement of the award. Anyone who has already received a graduate degree is not eligible. Adam Kapelner and Gina-Maria Pomann, two current NSF graduate fellows, hammer home the importance of some of these requirements. Adam earned his bachelor’s degree in mathematics and com- puter science at Stanford University and is working on his PhD in statistics at The Wharton School of the University of Pennsylvania. His research involves machine learning and model selection. He attributes the GRFP for giving him the time to be able to immerse himself in his research and, as a result, sub- mitting his work to and publishing in various jour- nals. He is helping lead the charge in assisting interest- ed students in his department with their applications to the GRF. When asked what advice Adam could give students interested in applying to the GRF, he stated his best recommendation would be for candi- dates to describe their research experience. “Can you make an impact in science? You need to illustrate your potential in research,” he said. He also acknowledged that he heard about the fellowship through a friend who thought it might be beneficial when applying to graduate school. Gina-Maria Pomann is pursuing her PhD in statistics at North Carolina State University. Her research interests are functional data analysis with applications to magnetic resonance imaging and dynamic treatment regimens. She thinks the GRF, in combination with her AT&T Labs fellowship, has allowed her to work on an array of projects as well as with different mentors. Gina-Maria started out earning an AS degree from Middlesex County College and then transferred to The College of New Jersey, where she earned her bachelor’s in mathematics with a minor in statis- tics. Gina-Maria first learned about graduate school and the GRF at the Mathematical Science Research Institute Undergraduate Program (MSRI-UP). MSRI-UP also took Gina-Maria and her fellow par- ticipants to a Society for Advancement of Chicanos and Native Americans in Science (SACNAS) confer- ence, where the students were further informed about the GRF and other opportunities. Her advice to stu- dents seeking a GRF is, “Get as much undergradu- ate research experience as possible!” She said her early research experiences helped her focus her research interests and write her GRF application. For the NSF solicitation, more information, and tips from awardees and reviewers, visit www.nsfgrfp. org, call (866) 673-4737, or email info@nsfgradfellows. org. For access to online applications, user guides, and other official announcements, log on to the FastLane website at www.fastlane.nsf.gov/grfp. n

october 2012 amstat news 19 columns

STATtr@k Service-Oriented Statistics: What Can Students Do? Glen DePalma, Doug Baumann, Jeff Nisen, and Libo Wang

ervice-oriented statistical Many universities have benefits public school superinten- consulting is a natural way StatCom (or a similar organiza- dents receive in Indiana. for statisticians to help meet tion) to help nonprofit organi- A fast way to find projects such societalS needs. StatCom is a pro- zations in their communities. as these is to contact the organi- bono statistical consulting group These programs increase aware- zations directly or find someone that originated in 2001 in the ness among the general public of who knows the community well department of statistics at Purdue statistics and the potential that and ask them to help spread the University. The Purdue branch of statisticians have to make posi- word. Contact your local schools StatCom (www.stat.purdue.edu/ tive contributions. Furthermore, and governments. external_relations/statcom) is run nonprofits, local governments, and staffed by graduate students in schools, and other community The Future the department and includes statis- organizations realize the ben- There are constant challenges tics undergraduate students and efits of quantitative thinking and involved in any volunteer organi- graduate students from other assessment. And, finally students zation. For one, students must be departments (e.g., sociology and gain practical experience, acquire kept motivated. Every student is computer science). skills useful in any future career, already heavily burdened with class interact with professionals in work and maintaining some type the field, and develop a sense of of social life. Students may show volunteerism even before they interest at the beginning, but lose become practicing statisticians. interest as the semester wears on. It is important to find key individu- What Can Students Do? als who know the graduate student Many organizations need statisti- population well and can help keep cal help, from designing surveys students motivated. The success to developing predictive models. of volunteer efforts is the result of There are numerous opportuni- only a few individuals. ties for delivering service-oriented Societal awareness of quanti- statistics within local communi- tative thinking and the need for ties. Here, we highlight several assessment and analysis is growing recent projects. every day. As more data are made StatCom designed a survey widely available, the need for peo- for patrons of the local city sym- ple to extract meaningful informa- phony orchestra to inform board tion is heavily sought after. Most members of the potential areas of nonprofit organizations have to improvement. Initiatives are ongo- show community value to receive ing to improve the facilities and funding, and statistical graduate atmosphere at orchestra events as students can help fill this need. a result of our analysis. StatCom has worked with A Call to Action members of the Tippecanoe We encourage statistics gradu- County Assessor’s Office to devel- ate students everywhere to get op a predictive model for sale involved in their communities. prices of housing units throughout You will find the experience to be Tippecanoe County, Indiana. rewarding in many ways. We can StatCom has worked with the help you. Please contact us at Indiana Association of Public [email protected]. n Superintendents to assess the

20 amstat news october 2012 columns

175 The Future of Statistical Publications

n anniversary that ends in ‘0’ is an occasion for celebrating the past. When one ends in ‘5,’ it is an opportunity to plan the future. In thatA spirit, and responding to an invitation from the 175th Anniversary Committee of the American Statistical Association, I urge that we re-evaluate our publication processes. Electronic media are transforming access to information; it is time for the ASA to decide how to manage this change. I fear our current approach to publishing does not serve us well. It takes too long, so our best scien- tists are driven to other journals in faster disciplines. Refereeing is noisy and often achieves only minor in summary form, and mailed those summaries to gains. And the median quality of reviews is deterio- other members. But his hand grew weary, and he rating due to journal proliferation, pressure on junior began to send notes of the following kind: faculty to amass lengthy publication lists, and the slow burnout of conscientious reviewers. Our present paradigm has other structural prob- lems. Published articles are static—correction and improvement are impossible. Published research often does not replicate, which is hard to flag. Readers must reach too far to find the code behind the article, and the data are nearly impossible to obtain. Correct work that is not sufficiently novel is excluded and lost. And there is a large gray literature that cannot be easily accessed or assessed (e.g., fed- Given his technology, Oldenburg’s stringencies eral reports, weighting schemes for official surveys, were essential. Printing and distribution costs were lecture notes, classroom exams, code documenta- the limiting factors; pauca sed matura had to be the tion, data/metadata, PhD theses). standard. An entire economic ecology grew up around I am far from the first person to raise these con- those constraints: Publishers set type and sold vol- cerns. Karl Rohe has a parable that illustrates many umes; societies created editorships and referees; and of these issues on Page 16. Larry Wasserman, Jim libraries emerged. Authors and editors created content Pitman, Nick Jewell, Nick Fisher, and Roger Peng, for free and publishers made fortunes; this was the among others, have grappled with various features of best solution possible. Until the Internet. the problem. Since 2005, the ASA has formed three Now, we have fresh choices. Electronic articles committees to study the matter; most recently, Len can be living documents, as on arXiv; better versions Stefanski is chairing one, which will make recom- layer on top of the old. Articles may use color and mendations to the ASA Board in November. Among dynamic graphics and be as long as necessary, with young statisticians, almost all perceive the inefficien- detailed proofs and worked-out examples (while cies of traditional publication and share a common reader feedback enforces concision). Article quality sense of the improvements that are possible. can be signaled in multiple ways, either by conven- Change will happen. If we fail to plan ahead, the tional review or by ungameable rating systems, similar ASA will be forced to adopt whatever system Wiley or to page-ranking algorithms. Readers can use person- Springer or the American Mathematical Society estab- alized recommender systems to discover papers. And lishes as the standard, but their interests and needs data, code, and gray literature become easy to access. align imperfectly with ours. Space limitations prevent a full catalog of the possible Today’s publication process was essentially invent- features (and bugs), but I expect Stefanski’s report will ed by Henry Oldenburg, the first corresponding sec- be more comprehensive. retary of the Royal Society. He received letters from I invite readers to comment on this topic. Just go to members describing their research, copied them out http://magazine.amstat.org and post your thoughts. n

october 2012 amstat news 21 Recollections Thoughts About JSM in &San Diego

Steve MacEachern, JSM 2012 Program Chair

particular pleasure within the statistics com- Section for Statistical Programmers and Analysts munity—whether corporate, government, or (more than 1,800 members in just two years of academic—is the chance to travel to profes- existence) and the Section on Statistical Learning sionalA meetings. I find I am always energized by a and Data Mining. The former has attracted many meeting and also worn out by its end. The energy from outside academia; the latter includes many comes from the excitement of seeing novel ideas that with backgrounds in computer science and related give us a new perspective on old issues, from being fields. Sessions sponsored by these sections pro- introduced to new problems that have been opened vided a good opportunity to broaden one’s knowl- up by technological innovation, from seeing sharp, edge, and both fit well with ASA President Bob clear examples of terrific statistical practice, and from Rodriguez’s Big Tent initiative. seeing the importance of problems the community is The technical sessions are always a highlight addressing. Additional energy that may account for of JSM, and we each have our favorites. For me, being worn out by the end of JSM week comes from Thursday morning is a treat. The relaxed atmosphere the vibrant social scene—seeing old friends and meet- as JSM winds down imparts the feel of a smaller ing new ones. What follows are a few of my recollec- meeting, and the program is every bit as strong as tions and thoughts about JSM in San Diego. I’m sure the earlier days. This year, I saw excellent sessions on each of us has many fine memories of the event. Bayesian methods in the health sciences and design of First off, San Diego: What a wonderful location computer experiments. for JSM. As advertised, the weather was perfect, there The posters were enjoyable, with a good loca- was an abundance of hotels and eateries nearby, and tion in the exhibit hall and ample room for a the food was excellent—even at the street fair I wan- nice display. The new electronic posters were a dered into. The convention center was laid out well, success, opening up many dynamic possibili- with good room size and acoustics. The “tunnel” view ties. Anecdotally, there seem to be more students coming down the escalator was worth a photo or two. attending the meetings. A poster presentation is a I enjoyed the sessions being split between the conven- terrific way to get a start on one’s career. tion center and headquarter hotel, as it generated a Attendance at the meetings was strong, with more few short walks outside. I think it’s a fine tradition to than 6,300 attendees. Interestingly, I bumped into our circulate JSM among a large number of cities, but I divisional dean, a probabilist by trade, and learned that certainly hope for a return to San Diego. attendance at the big annual meeting in mathemat- The Introductory Overview Lectures (IOLs) were a ics runs in the mid-5,000s. The comparison shines a big hit once again, as they provide us all with a chance spotlight on the vigor and currency of our discipline. to learn about a new area from experts who lay out the I’m already looking forward to next year’s JSM in main ideas and a few pithy examples—all in an hour Montréal, where I’m told the meeting will open with and 50 minutes. This year, the IOLs covered causal “Bienvenue au Canada.” inference, personalized medicine, sparsity, and climate. I had a terrific time as program chair; my thanks go Thanks to Judea Pearl, Don Berry, Dave Donoho, and to Bob Rodriguez and the Committee on Meetings for Peter Guttorp for their efforts in putting together a selecting me for the position. The real work of orga- fine set of presentations. nizing the meeting falls on two groups—the ASA staff Big import sessions included the late-breaker on and the JSM Program Committee. During the year, statisticians’ roles in health care reform, organized by I gained a new appreciation for the difficulty of put- John Adams, and the session on statistics and human ting on JSM and the dedication of the ASA staff. My rights, organized by Jay Kadane. thanks to them for the long hours and late nights they Developments in professional organizations are put into making the meeting run so smoothly. The always of interest. JSM has grown dramatically since Program Committee was singularly well prepared. the first I attended, nearly 30 years ago. The meeting They put together a fine technical program and made is larger, the crowds are far more diverse, and good scheduling easy. My special thanks go to the general penmanship has been replaced by artistic page layout. methodology chairs, Mark Glickman and Yoonkyung New sections are forming, and their growth is expand- Lee, and the poster chair, Kristin Duncan, all of whom ing the ASA in varied directions. Two of note are the proved a remarkable sounding board. n meetings Many Honored at Presidential Address, Awards Ceremony

Batcher Stokes Chuang-Stein

ighlighting the ASA Fellows, as program chair of the Melissa Begg Presidential Address Biopharmaceutical Section, as Columbia University and Awards Ceremony a member of the JSM Program Hduring the Joint Statistical Committee, and as president F. DuBois Bowman Meetings in San Diego, of the Southwestern Michigan Emory University California, were the announce- ASA Chapter. ment of the Founders Award Lynne Stokes, Southern Amy J. Braverman winners and the official induc- Methodist University, for out- Jet Propulsion Laboratory tion of 48 ASA Fellows. standing service to ASA over the Congratulations to all. past 20 years through her leader- N. Rao Chaganty The Founders Award was ship in chapters and sections as Old Dominion University given to Mary Batcher, Christy chair of the Council of Chapters Chuang-Stein, and Lynne and Council of Sections; editor Dongseok Choi Stokes for extended, outstand- of The American Statistician; and Oregon Health and Science Visit http:// ing service to the statistics dedicated contributions on sev- University magazine. profession through a variety of eral committees, including the leadership roles. JSM Program Committee, the Mark R. Conaway amstat.org Mary Batcher, Ernst & Census Advisory Committee, University of Virginia Health to view more Young, for outstanding service to and the Committee on Privacy Systems Department of Public pictures from ASA chapters through her service and Confidentiality. Health Sciences JSM 2012. as president of the Washington Each year, ASA Fellows are Statistical Society and as chair nominated by the member- Susmita Datta of the Council of Chapters; for ship and selected by the ASA University of Louisville many years of diligent, steady Committee on Fellows, chaired leadership of and service to the this year by Ji Zhang. The num- Marc N. Elliott ASA’s accreditation program; and ber of Fellows named is limited RAND Corporation for dedicated service on many to no more than one-third of 1% ASA committees. of the active ASA member total. Scott Evans Christy Chuang-Stein, The following 48 ASA Fellows Harvard School of Public Pfizer, Inc., for outstand- were inducted this year: Health ing service in multiple roles, including vice president, group Mousumi Banerjee Debashis Ghosh leader of the Accreditation University of Michigan The Pennsylvania State Committee, and chair of the University 175th Anniversary Steering Sudipto Banerjee Committee; for dedicated ser- University of Minnesota Mithat Gönen vice on the Committee on Memorial Sloan-Kettering Cancer Center

october 2012 amstat news 23 meetings

Forty-eight ASA members received the honor of Fellow in 2012.

Robert L. Gould Jen-Pei Liu Christian P. Robert UCLA National Taiwan University Universite Paris Dauphine

Deborah H. Griffin Thomas Lumley Abdul J. Sankoh U.S. Census Bureau University of Auckland Vertex Pharmaceuticals

Jeffrey H. Hooper Clyde F. Martin Sanjay Shete BT Group US&C Texas Tech University MD Anderson Cancer Center

Nicholas J. Horton Nancy Mathiowetz Judith D. Singer Smith College University of Wisconsin– Harvard University Milwaukee Xiaoqiong Joan Hu Marc A. Suchard Simon Fraser University Bhramar Mukherjee University of California, Los University of Michigan Angeles Michael G. Hudgens The University of North Anna B. Nevius Thaddeus Tarpey Carolina at Chapel Hill U.S. Food and Drug Wright State University Administration Gareth James Ram C. Tripathi University of Southern Thomas E. Nichols The University of Texas at San California University of Warwick Antonio

Jae-Kwang Kim A. James O’Malley Andrea Troxel Iowa State University Harvard Medical School University of Pennsylvania

Michael D. Larsen R. Todd Ogden Roshan Joseph Vengazhiyil The George Washington Columbia University Georgia Institute of Technology University Liang Peng Colin O. Wu Lawrence I-Kuei Lin Georgia Institute of Technology National Heart, Lung, and Baxter Healthcare Corporation Blood Institute José C. Pinheiro Bo Henry Lindqvist Janssen R&D Daowen Zhang Norwegian University of North Carolina State Science and Technology University

24 amstat news october 2012 meetings

Kelly Hong Zou Pfizer Inc.

Rebecca Zwick Educational Testing Service

Many more people were honored for their contribu- tions to various causes that advance the field of statistics. Following is a list of awards and recipients: Samuel S. Wilks Memorial Award The Samuel S. Wilks Memorial Award was established in 1964 to honor the memory and dis- tinguished career of Sam Wilks 2012 Edward C. Bryant Scholarship Award winner Mark Dahlke by recognizing outstanding contributions to statistics that carry on the spirit of his work. The 2012 Wilks award winner standing contributions to the with spatially dependent mark is Peter Hall of the University theory, applications, and teaching distributions and using them to of Melbourne for his pioneering of nonparametric statistics. assess fingerprint individuality. and influential contributions to This research was published in a wide variety of areas of statis- Statistics in Chemistry their paper, titled “Assessing tics and probability and for his Award Fingerprint Individuality outstanding service to the pro- The Statistics in Chemistry Using EPIC: A Case Study fession, with an extensive record Award recognizes outstanding in the Analysis of Spatially of editorial roles, collaborative collaborative endeavors between Dependent Marked Processes,” work, and promotion of statisti- statisticians and chemists. While published in the May 2011 cal science to the wider scientific this is a section award, the board issue of Technometrics. and educational communities. of directors grandfathered it in Edward C. Bryant Gottfried E. Noether 1995 for presentation at the ASA Presidential Address and Awards Scholarship Award Awards Ceremony. The 2012 Statistics The Bryant scholarship trust is The Noether awards were estab- in Chemistry Award winners are a permanent scholarship fund lished in 1999 by the wife and Bradley Jones, JMP Division, endowed by Westat to honor its daughter of the late Gottfried SAS Institute, and Scott Allen, cofounder and longtime leader, Emanuel Noether of the Novomer, Inc., for outstanding Edward C. Bryant. The award University of Connecticut as a collaborative work in developing honors an outstanding graduate tribute to his memory. They rec- a new catalyst for CO2-based student who is studying survey ognize distinguished researchers polymers that sequester CO2. statistics. The 2012 scholarship and teachers and support research recipient is Mark Dahlke of in nonparametric statistics. The Outstanding Statistical Colorado State University for Gottfried E. Noether Young Application Award outstanding academic achieve- Researcher Award winner for Each year, the ASA recognizes ment in a statistical program. 2012 is Guang Cheng, Purdue a paper that is an outstanding University, for outstanding early application of statistics in the W. J. Dixon Award for career contributions to nonpara- physical, biological, or medi- Excellence in Statistical metric statistics. The Gottfried cal sciences. This year’s win- Consulting E. Noether Senior Scholar ners are Chae Young Lim and Award winner for 2012 is Joseph Established through a gift from Sarat C. Dass of Michigan the family of Wilfrid Dixon, L. Gastwirth, The George State University for develop- Washington University, for out- this award recognizes outstand- ing novel point process models ing contributions to the practice

october 2012 amstat news 25 meetings

that recognizes organizations. This year’s SPAIG award was given to Baylor University and Eli Lilly & Company in recog- nition of the vision and support provided by Eli Lilly to Baylor University in the creation of a partnership for the enhance- ment of pharmaceutical research and development. Gertrude M. Cox Scholarships Jessica Orth of the University of Minnesota, Morris, and Christine Ho of the University of California at Berkeley are the winners of the Getrude M. Cox Scholarship in Statistics Award. Since 1989, the schol- 2012 SPAIG Award winners arship has been awarded by the ASA Committee on Women in Statistics and the Caucus of statistical consulting. The authors developed a simple but for Women in Statistics to 2012 award was presented flexible hierarchical model to encourage women to enter sta- to Gary Grove Koch of The address multi-analyte, multi- tistically oriented professions. University of North Carolina laboratory studies with outly- The following women were at Chapel Hill for outstanding ing observations and tested the awarded honorable mentions: contributions to the develop- model using both simulated and Lucy D’Agostino, Naomi ment and review of pharma- real interlaboratory study data. Brownstein, Julia Shin-Jung ceutical licensing applications, Lee, and Diana Marie Liley. to the research on and software Waller Education Award development of categorical Retired ASA Executive Director Karl E. Peace Award data analysis, and for advanc- Ray Waller and his wife, The Karl E. Peace Award ing the science and art of sta- Carolyn, established the Waller for Outstanding Statistical tistical consulting. Education Award in 2002 to Contributions for the Betterment recognize a statistics teacher W. J. Youden Award in of Society recognizes statisti- early in his/her career for excel- cians who have made substantial Interlaboratory Testing lence and innovation in teach- contributions to the statistical The W. J. Youden award was ing introductory statistics at profession and society in gen- established in 1985 to recog- the undergraduate level. The eral. The award—established nize the authors of publications 2012 Waller Award winner is by Christopher K. Peace, son of that make outstanding contri- Michael A. Posner of Villanova Karl Peace, on behalf of the Peace butions to the design and/or University in recognition of family to honor the life work of analysis of interlaboratory tests outstanding contributions to his father—was offered for the or describe ingenious approach- and innovation in the teaching first time this year. The 2012 es to the planning and evalu- of elementary statistics. Peace award winners are Fritz ation of data from such tests. Scheuren for an exemplary career The 2012 W. J. Youden Award The SPAIG Award that has translated impressive winners are David Dunson of The ASA established the SPAIG statistical contributions into sup- Duke University and Garritt L. Award in 2002 to recognize out- port of humankind and Marvin Page of Pontifica Universidad standing partnerships between Zelen for outstanding statistical Catolica de Chile in recogni- academe and business, industry, contributions and dedication to tion of their paper, “Bayesian and government organizations the establishment of collaborative Local Contamination Models and to promote new partner- partnerships between biostatisti- for Multivariate Outliers.” The ships. It is the only ASA award cal and clinical sciences. n

26 amstat news october 2012 meetings

Committee of Presidents of Statistical Societies COPSS Honors Statisticians

Visit Amstat News online at http:// magazine. amstat.org to Kou Gray Little read an in- terview with embers of the outstanding statistical research professional and academic COPSS Presi- Committee of in the modeling and evalua- leadership contributions. His dents’ Award Presidents of Statistical tion of missing data, sample lecture was titled “In Praise of winner Sam SocietiesM (COPSS) are pleased to survey, and causal inference; Simplicity, not Mathematistry! Kou. announce the 2012 awards, which for the clear and comprehen- Ten Simple, Powerful Ideas for were presented to the winners at sive application of these and the Statistical Scientist.” JSM in San Diego, California, by other methodologies in sci- Award criteria and nominating COPSS President Xihong Lin. ence and public policy arenas; procedures are available on Page The winner of the Presidents’ and for diverse and effective 31 and at www.niss.org/copss. n Award is Samuel S. Kou of Harvard University for ground- breaking contributions to sto- chastic modeling and statistical inference in single molecule biophysics; for pioneering the equi-energy sampler; for fun- damental contributions to Bayesian, empirical Bayes, and nonparametric methods; and for outstanding service to the statistical profession and contri- bution to statistical education. The Elizabeth L. Scott Award was presented to Mary Gray of American University for her lifelong efforts to fos- ter opportunities in statistics for women and to further the careers of academic women; for creating a forum for discussing the role of women in math- ematics; for exposing discrimi- nation; and for exchanging strategies, encouraging political action, and promoting affirma- tive action. The 2012 Fisher Lecturer was Roderick J. Little of the University of Michigan for

october 2012 amstat news 27 education

Planning Begins, Proposals Sought for USCOTS ’13

lanning has begun for the next United States To propose a breakout session, send a descrip- Conference on Teaching Statistics tion to Allan Rossman, program chair, at arossman@ (USCOTS), which will be held in Raleigh, calpoly.edu by November 10. Your proposal of no NorthP Carolina, from May 16–18, 2013, and host- more than 1,000 words must include the following: ed by the Consortium for the Advancement of • Title for proposed session Undergraduate Statistics Education (CAUSE). Members of the USCOTS ’13 program committee • Names, email addresses, and brief biographical are seeking ideas for active, participant-focused sketches for all leaders of the session breakout sessions addressing the conference theme, “Making Change Happen.” • Description of how the session relates to the Change can be difficult, but also exhilarat- conference theme ing, scary, but also liberating. Previous USCOTS • Explanation of how the session will actively sessions have advocated various types of changes engage participants in teaching statistics likely to produce enhanced student learning. Conference participants often • Discussion of how participants will be able to become excited about visionary ideas presented, implement ideas presented in the session but struggle to follow through and affect sustained changes at their home institutions. USCOTS ’13 Proposals will be reviewed by members of the will address the goal of making change happen. USCOTS ’13 program committee and notifica- This conference is designed to model good tions will be made by January 15, 2013. Proposals teaching in its sessions, social activities, and hall- for “poster and beyond” sessions will be solicited at ways. As with previous USCOTS, it will consist of a later date; those proposals will be due on February plenary sessions, breakout sessions, and “poster and 1, 2013. beyond” sessions. These will address how to make change happen in four main areas: curriculum, ped- Conference sessions will consider many types agogy, resources, and educational research. and aspects of change, including the following: Thus, members of the USCOTS program com- mittee are requesting proposals for 80-minute • Changing an individual class, course, or breakout sessions that relate directly to the confer- program in terms of content, technology, or ence theme and focus on actively engaging partici- assessments pants. It is not appropriate for a breakout session to • Changing pedagogical methods used to pro- consist primarily of a presentation. mote student learning • Changing resources or delivery systems used to teach statistics The deadline for listing an internship opportunity in the December 2012 issue of Amstat News is October • Collaborating to make change happen 20. Any listings received after October 20 will be post- • Overcoming resistance to change ed on the ASA website only. Visit www.amstat.org/ • Evaluating the impact of change education/internshipopportunities.cfm for details. More information can be found at www. CAUSEweb.org/uscots. n

28 amstat news october 2012 education

International Census at School Workshop Held in Conjunction with JSM 2012 Rebecca Nichols, ASA Director of Education

he American Statistical Association spon- Statisticians involved or interested in get- sored the 2012 International Census at ting involved with U.S. Census at School and School Workshop at the end of the Joint representatives from the U.S. Census Bureau TStatistical Meetings in San Diego, California, on and their Statistics in Schools program also par- August 2 and 3. ticipated in the meeting. ASA President Bob Census at School is a free, international class- Rodriguez and ASA Executive Director Ron room project that engages students in grades Wasserstein welcomed participants. 4–12 in statistical problemsolving using their August 2 sessions included an overview of own real data. Under the direction of their teach- the project and a look to the future by Neville ers, students involved in the program anony- Davies of Census at School UK and creator of mously complete an online survey, analyze their the International Census at School program. class census data, and then compare those results Country representatives then gave updates with results from random samples of participat- from their countries (Rebecca Nichols, United ing students throughout the world. The Census States; Kate Richards, United Kingdom; at School project began in the United Kingdom Angela McCanny and John Brewster, Canada; in 2000 and now includes Australia, Canada, Chris Wild, New Zealand; Delia North, South New Zealand, South Africa, Ireland, Japan, Africa; and Michiko Watanaba and Kanzunor Korea, and the United States. Statistics educa- Yamaguchi, Japan). There also was time spent tion leaders from other countries also are investi- remembering Martha Aliaga, former ASA direc- gating bringing the project to their country. tor of education, who was instrumental in bring- The two-day meeting provided an opportuni- ing Census at School to the United States. ty for international Census at School leaders and Additionally, Stephen Miller taught a sepa- U.S. Census at School champions to coordinate rate workshop in the morning to introduce the international Census at School project, share middle- and high-school teachers and statis- hands-on curriculum materials, achieve common ticians interested in getting involved in U.S. understanding of the international project, acquire Census at School to the program. Workshop experience with country-specific data-handing participants learned about the Census at School activities and resources for teachers and students, program, how it aligns with the Common Core and increase awareness of international efforts to State Standards in Mathematics, how to regis- improve statistical literacy in school children. ter a class, and what resources are available for Roxy Peck of Cal Poly and Rob Gould of the teachers using Census at School. Additionally, University of California at Los Angeles chaired participants actively collected data, learned how the workshop program committee, which includ- to enter it into the Census at School database, ed international Census at School leaders and and selected random samples of data entered U.S. Census at School leaders and champions. by students. Participants generated statistical Workshop presenters and participants included questions that can be answered using Census international leaders and representatives from the at School data, used software to make numeri- United Kingdom, South Africa, Canada, New cal and graphical summaries of the data, and Zealand, Japan, and Paraguay. Although they answered the statistical questions. could not attend in person, leaders from Australia U.S. teachers and statisticians participating in sent a written report to update the international the morning workshop joined the international leaders regarding the project in Australia. group in the afternoon, when international leaders

october 2012 amstat news 29 education

demonstrated international Census at School resources, identifying common concerns, volun- curriculum resources, statistics software, and teer support, and strategies to address concerns. activities. All attendees were invited to a dinner that evening and a presentation given by Eric Getting Involved Newburger of the U.S. Census Bureau. As of August 2012, there were more than 8,500 August 3 workshop sessions for international students from 41 states plus the District of leaders included discussions about evaluating Columbia who had participated in the U.S. the effectiveness of Census at School programs, Census at School program. Teachers who are growing support for Census at School, build- comfortable with statistical problemsolving and ing relationships with official census organiza- data analysis can begin using the program in their tions, expanding the program, funding, sharing classes at any time. There are detailed instruc- tions, five instructional webinars, a PowerPoint presentation, lesson plans, and other resources on the website. The ASA is seeking champions to expand the 2012 Educational U.S. Census at School program. Champions can be teachers who use the program in their classes Ambassador from or statisticians and statistics educators who assist teachers who are not yet comfortable with statistics Botswana Attends JSM and statistical problemsolving. There is a variety of ways to get involved, including sharing informa- tion about the program with local schools, writ- s the 2012 ASA ing lesson plans, and teaching local workshops Educational for teachers. For those interested in teaching local Ambassador, workshops, the ASA will provide materials. KeamogetseA Setlhare of the The ASA also is building online Census at University of Botswana attend- School resources and seeking those interested in ed the Joint Statistical Meetings writing new lesson plans or adapting internation- (JSM) in San Diego, California, al Census at School lesson plans for U.S. data. to participate in Continuing Those teaching grades 4–12 pre-service teachers Education (CE) courses. might consider encouraging them to create les- The Educational Ambassador son plans using U.S. Census at School data and Program is an ASA out- submit them to the STatistics Education Web reach effort launched by the (STEW), an online bank of peer-reviewed les- Committee on International Relations in Statistics to fos- son plans for K–12 teachers. STEW lesson plans ter international collaboration and enhance statistics educa- relating to Census at School also will be pub- tion worldwide. The program subsidizes an ambassador from a lished on the Census at School website in the developing country to attend JSM and take CE courses. It also resources area. provides a one-year ASA membership. Educators teaching or advising undergraduate or Candidates are required to have a PhD in statistics, graduate statistics students might consider encour- an interest in teaching, and be open to study in new areas aging or requiring them to get involved in service of research. After attending CE courses in an emerging area learning by working with grades 4–12 teachers of research, the educational ambassador returns to his or and students to incorporate Census at School and her country and teaches the subject matter learned in the CE enhance their statistical problemsolving skills. course(s) within the next year to at least 10 students. For more information about the U.S. Census While at JSM, Setlhare took courses in Bayesian meth- at School program, see the February 2012 ods and computation, simulation and sampling of data, and Amstat News article at http://magazine.amstat.org/ analysis of univariate and multivariate extremes, as well as sev- blog/2012/02/01/censusatschool-2 and the U.S. eral computer technology workshops. Census at School website at www.amstat.org/ Since the program launch in 2005, the Committee on censusatschool. Other ideas to enhance and expand International Relations in Statistics has chosen educational the program are welcome. Contact Rebecca ambassadors from Argentina, Ethiopia, Vietnam, Morocco, Nichols, ASA director of education, at rebecca@ Armenia, Costa Rica, and Botswana. amstat.org about these or any efforts regarding ser- vice learning or other activities. n

30 amstat news october 2012 awards and deadlines

Nominations Sought for years of the date of the award. is sponsored jointly by COPSS COPSS Awards Starting in 1991, this award has and the Caucus for Women in been given in odd years and con- Statistics. No member of the Nominations are being sought for sists of a plaque and cash award. award committee or officer of the the following awards presented by Nominations of an individual Caucus of Women in Statistics is the Committee of Presidents of and an associated publication or eligible to receive the award during Statistical Societies (COPSS). publications should be sent by his or her term of service. Eligible January 15, 2013, to Nilanjan nominations shall be based on a The Fisher Lectureship is Chatterjee, committee chair, at nomination letter, letters of sup- awarded for outstanding con- [email protected]. port, curriculum vitae, and other tributions to aspects of statis- appropriate documentation as tics and probability that closely The Florence Nightingale David requested by the award commit- relate to the scientific collec- Award is presented biennially to tee. Nominations should be sent tion and interpretation of data. recognize a female statistician who by January 15, 2013, to Nancy The award exists to recognize exemplifies the contributions of Reid, committee chair, at reid@ the importance of statistical Florence Nightingale David, an utstat.utoronto.ca. methods for scientific investiga- accomplished statistician in com- These awards are jointly tions. The hour-long lecture is binatorial probability theory and sponsored by the American delivered during JSM. Eligible the first recipient of the Elizabeth Statistical Association, Institute nominations should be sent L. Scott Award. The criteria for of Mathematical Statistics, in PDF format by December the award are excellence as a role International Biometric Society 15 to Kathryn Roeder, model to women, excellence in (ENAR and WNAR), and committee chair, at kathryn. statistical research, leadership of Statistical Society of Canada. [email protected]. multidisciplinary collaborative Detailed award criteria and nomi- groups, statistics education, and nating procedures are available at The Presidents’ Award is pre- service to the profession. The www.niss.org/copss. n sented in recognition of outstand- award was established in 2001 and ing contributions to the statistics profession. It is typically granted to an individual who has not yet reached his or her 41st birthday. In the special case of an indi- vidual who has received his or her statistically related terminal degree fewer than 12 years prior to the nomination deadline, the individual will be eligible if he or she has not yet reached his or her 46th birthday during the year of the award. Eligible nominations should include a current curricu- lum vitae, the nominee’s date of birth, a nomination letter (up to three pages), and up to five sup- porting letters. Nominations should be sent in PDF format by January 15, 2013, to Raymond J. Carroll, committee chair, at [email protected].

The George W. Snedecor Award, established in 1976, honors an individual who has been instru- mental in the development of statistical theory in biometry. The award is for a noteworthy publi- cation in biometry within three

october 2012 amstat news 31 people news The Jeanne E. Griffith Mentoring Award Stephanie Shipp and Clyde Tucker, Jeanne E. Griffith Committee

performance and team awards covering work on the redesign of the Mass Layoff Statistics Program and use of the Internet for data collection. As head of the Behavioral Science Research Center, Mockovak has successfully mentored a genera- tion of younger staff members over almost 15 years. John Eltinge, associate commissioner of the Office of Survey Methods Research, nominated Mockovak for this award. He said he has “developed very deep respect for Bill’s extraordinary style of quiet leader- ship of the Behavioral Science Research Laboratory and his mentoring of junior and mid-level colleagues. For example, he has a remarkable talent for carefully calibrated integration of relatively junior colleagues into very complex research projects. He does this in a way that leads to individual work assignments that are challenging (but not overwhelming) and that lead Recent Jeanne E. Griffith awardees—Deborah Griffin, Jenise Swall, and Kevin to increasing levels of professional growth and career Cecco—with this year’s awardee, William Mockovak (third from left) recognition for his colleagues.” Those who have worked for Mockovak over the illiam P. Mockovak was recently years echoed Elting’s words in their supporting let- announced winner of this year’s Jeanne ters, describing him as a wonderful mentor who has E. Griffith Mentoring Award for being helped staff transition from junior to mid-level, teach- Wa champion role model and mentor. ing them how to work with program staff in present- Mockovak is chief of the Behavioral Science ing new ideas and critiquing existing approaches, and Research Center in the Bureau of Labor Statistics acting as a role model by his modesty and generosity (BLS). His professional experience covers a wide range and his dedication to his staff and work. of areas, including program evaluation, interviewer training, computer-assisted information collection Nominations for 2013 Award (CASIC), questionnaire design, usability testing, and The Jeanne E. Griffith Mentoring Award is intended cognitive interviewing. to encourage mentoring of junior staff in the federal, After service in the U.S. Air Force, Mockovak state, or local government statistical community. It is joined the U.S. Census Bureau’s Center for Social presented annually to a supervisor, technical director, Science Research in September 1978. Over the next team coordinator, or staff member who is nominat- 20 years, he held a number of positions of ever- ed by coworkers and/or supervisors and selected by increasing responsibility at the Census Bureau. In members of the Jeanne E. Griffith Mentoring Award 1998, he moved to his current position at BLS. In this Selection Committee. role, he supervises behavioral scientists whose primary Jeanne E. Griffith worked for more than 25 years responsibilities are to identify, investigate, and reduce in the federal statistical system. Throughout her career, measurement error in BLS surveys. His responsibilities and especially in her senior management positions include cognitive testing of survey questions, question- at the National Center for Education Statistics and naire design, usability testing, nonresponse bias analy- National Science Foundation, one of Griffith’s high- ses, program evaluation, and basic research. est priorities was to mentor and encourage junior Mockovak has degrees from the U.S. Air Force staff to learn, grow, and recognize and seize career Academy (BS, psychology), The Ohio State University opportunities as they came along. After her death (MA, human performance psychology), and The from breast cancer in 2000, the Jeanne E. Griffith Pennsylvania State University (PhD, educational psy- Mentoring Award was set up in her honor. chology). He received a Bronze Medal award from the To help sponsor the award and increase the Census Bureau for significant contributions to the amount of the honorarium, contact Stephanie development of major new approaches in training, Shipp at [email protected]. managing, and supporting field personnel. He also To view more photos and a list of sponsors, visit received BLS awards for outstanding management http://[email protected]. n

32 amstat news october 2012 people news

The Pakistan Journal of Statistics and Operation Gibbons has written more than 200 peer- Research (PJSOR) recently published a special volume reviewed journal articles in various areas of statistical in honor of Mir Masoom Ali, an ASA Fellow and theory and practice and is an author of Longitudinal George and Frances Ball Distinguished Professor of Data Analysis. He also wrote two environmental Statistics Emeritus at Ball State University, on the statistics texts: Statistical Methods for Ground Water occasion of his 75th birthday. The papers can be Monitoring and Statistical Methods for Detection and viewed at www.pjsor.com/index.php/pjsor/issue/current/ Quantification of Environmental Contamination. Read about showToc. He is a member of the Institute of Medicine of your colleagues the National Academy of Sciences, a Fellow of the Jianqing Fan of Princeton and Ker-Chau American Statistical Association, and a member of and friends in Li of Academia Sinica (AS) and the University the American College of Neuropsychopharmacology. the news. Go of California at Los Angeles were elected to He has received numerous awards for his research, to www.amstat. Academicians of AS during the July 2012 convoca- including two Youden awards for contributions org and click on tion in Taipei, China. Becoming an AS Academician to interlaboratory calibration, the Harvard Award “Statisticians in is a major honor for scholars of Chinese origin. in Psychiatric Epidemiology and Biostatistics, and the American Statistical Association Award for the News.” Nancy Flournoy, professor in the department Outstanding Statistical Applications. of statistics at the University of Missouri’s College of Arts and Sciences, is the recipient of the 11th James Landwehr and Linda Young are this annual Janet L. Norwood Award for Outstanding year’s recipients of the distinguished service awards Achievement by a Woman in the Statistical Sciences. from the National Institute of Statistical Sciences Spanning four decades, her rich diversification in (NISS). The awards were presented by NISS director academic appointments range from directorships at Alan Karr during the 2012 Joint Statistical Meetings. the Fred Hutchinson Cancer Research Center and The NISS distinguished service awards were National Science Foundation’s Program in Statistics established by the board of trustees in 2005 to recog- and Probability to chairs with American University nize individuals who have given extraordinary service and the University of Missouri (MU). Having that significantly advances NISS and its mission. As recently stepped down as MU’s chair of statistics, Karr noted, the recipients “didn’t have to do what Flournoy is returning from sabbatical as professor. they did for NISS, but did it because they believe in Her research interests in theoretical and applied NISS and what it does for the statistics community.” statistics include clinical trials, adaptive sequential Landwehr, who works at Avaya Laboratories, was designs, transplantation biology and infectious recognized for his many years of service on the NISS disease, specifically cytomegalovirus at a key point prior to the AIDS epidemic. Her initial collaborative efforts on transplantation research with the team of E. D. Thomas led to that team receiving the Nobel Prize in Medicine in 1990. Has Your Email

Robert D. Gibbons was recently announced the Address Changed? Rema Lapouse Award recipient for 2012. The Rema Lapouse Award is granted annually to an outstanding recipient for excellence in psychiatric epidemiology. Let Us Know! Gibbons is an internationally renowned biostatistician and psychometrician with expertise • Log in to Members Only in broad areas of mental health and environmental at www.amstat.org/ research. Since 2010, he has been the director of the membersonly and update Center for Health Statistics and professor of biosta- your member record tistics in the departments of medicine, health studies, and psychiatry at The University of Chicago. Prior • Email your new address to to that, he directed the Center for Health Statistics at [email protected] the University of Illinois at Chicago. Major themes in his work include development of linear and non- • Call us toll-free at linear mixed-effects regression models for the analy- (888) 231-3473 sis of longitudinal data, analysis of environmental monitoring data and interlaboratory calibration, item response theory and computerized adaptive testing, and statistical methods in pharmacoepidemiology and drug safety.

october 2012 amstat news 33 people news

Board of Trustees, which includes chairing the board for three years and serving this past year as chair of Obituaries the board’s Affiliates Committee. Young, professor of statistics at the University Sung C. Choi of Florida, was acknowledged for her service on the board of trustees and leadership role in the Sung C. Choi, 81, passed away on July 21, 2012. NISS-NASS research program, which ran from Born in South Korea, Choi came to the United 2009 to 2011. States for his education, earning a BS in mathemat- For more information about NISS, visit www. ics and an MA in statistics from the University of niss.org. Washington. He went on to earn a PhD in biosta- tistics from UCLA in 1966. He was supported by a The Johns Hopkins University professor Thomas predoctoral fellowship in biometry from the National Institutes of Health (NIH). A. Louis will join the U.S. Census Bureau through Choi began a career in academia in the depart- an interagency personnel agreement as the new asso- ment of applied mathematics at Washington ciate director for research and methodology and chief University in St. Louis, Missouri. He served as chair scientist effective January 7, 2013. of the program in statistics from 1969–1974 in the Former Census Bureau director Robert Groves school of engineering and applied sciences and, in said, “I am overjoyed that we were able to attract a 1978, moved to Richmond, Virginia, to join the world-class scientist to lead the research director- department of biostatistics in the Medical College ate. Dr. Louis will add to the innovation Dr. of Virginia (MCV) of Virginia Commonwealth Roderick Little brought to the position over his University (VCU) as a professor. In 1989, he was successful leadership. He will, no doubt, increase given a joint appointment as professor in the the rate of scientific discoveries, which permits the Division of Neurosurgery at MCV/VCU. He Census Bureau to do its work more efficiently.” served as the primary adviser for two postdoctoral The National Institute of Statistical Sciences trainees, seven doctoral students, and four MS stu- (NISS) recently presented the 2012 Jerome dents over the course of his career. Sacks Award for Cross-Disciplinary Research to Choi was a prolific writer, having published two books, three book chapters, five government publica- William Q. Meeker of Iowa State University. tions, and more than 100 peer-reviewed publications. Susan Ellenberg, chair of the board of trustees, His methodological research focused on design and announced the award during the 2012 Joint statistical issues related to clinical studies of severe Statistical Meetings. The annual award, named head injury, including misclassification of primary in honor of the founding director of NISS, was outcome measures, interim analyses with delayed established in 2000 to recognize “sustained, high- observations, and sequential methods of estima- quality cross-disciplinary research involving the tion. He was a Fellow of the American Statistical statistical sciences.” Association, served as a referee for many biomedical Meeker received the award for outstanding and statistical journals, reviewed grants for the NIH, sustained research that develops, implements, and served as a consultant for multiple pharmaceuti- documents, communicates, and teaches statistics cal companies and health sciences centers. for the solution of relevant engineering and cross- Harold Young, professor and chair of neurosur- disciplinary problems, especially in reliability, gery at VCU, described Choi as “a great partner accelerated testing, reliability software, degradation of the original head injury program” at VCU and data analysis, and statistical methods for nonde- an “indispensible part of the team.” For more than structive evaluation. two decades, Choi was the principal investigator of the Head Injury Biostatistics Center, which sup- The board of directors of the Council of ported the clinical research conducted by the Head Professional Associations on Federal Statistics Injury Clinical Research Center at MCV, a NIH- (COPAFS) recently announced the appointment sponsored center. of Katherine R. Smith as executive direc- Choi’s colleagues remember him as a quiet, unas- tor. Smith comes to COPAFS from the American suming, and modest person, as well as a “solid citi- Farmland Trust (AFT), where she was the chief zen” in the department, willing to serve as needed. economist and vice president for programs. Prior His humor and encouragement to students and fac- to her position at AFT, she worked for 20 years in ulty will be missed. His family noted that the depart- progressively responsible positions at the Economic ment of biostatistics at VCU was an important part Research Service of the U.S. Department of of his life, as were the colleagues, staff, and students Agriculture, including as administrator for her last with whom he had the pleasure of working during five years. n his 25 years there.

34 amstat news october 2012 people news

recipient of the Lindback Foundation and Sigma Lester R. “Randy” Curtin Kappa Phi awards for distinguished teaching. de Lester R. “Randy” Curtin, who worked at the Cani was a Fulbright scholar at the Norwegian Centers for Disease Control and Prevention’s School of Economics. During World War II, he National Center for Health Statistics, died June 20 at completed 38 missions as a radio operator on his home in Huntingtown, Maryland. He was 60. B-24 bombers as a member of the U.S. Army Air Curtin joined the National Center for Health Corps. After the war, he earned a BS in mathe- Statistics in the late 1970s. He was an expert on matics from the University of Wisconsin, followed survey design and his methodologies helped improve by MBA and PhD degrees in statistics from the health statistics’ accuracy. He worked for many years Wharton School. He is survived by his beloved in the Mortality Statistics Branch and helped compile partner, Joanna Williams, professor at Teachers data that enhanced the detection of changes in mor- College, Columbia University. tality trends. He also helped design health surveys and studies. Curtin was born in Trenton, New Jersey. He was Bob Hodges a 1973 mathematics graduate of The University of Bob Hodges died on January 26, 2012, in Seneca, North Carolina at Chapel Hill and earned a doctor- South Carolina, of congestive heart failure. Born on ate in biostatistics from UNC in 1978. February 21, 1933, he served in the United States Army Reserve and began college at Georgia Institute of Technology, ultimately completing his bachelor’s John S. de Cani and master’s degrees at Georgia State University. He Jean Dickinson Gibbons and Samuel Litwin earned a PhD in marketing, statistics, and psychol- ogy from The University of Texas at Austin. Hodges spent most of his adulthood in John S. de Cani passed away on August 8, 2012, in Richmond, Virginia, teaching business statistics and New York City. Born on May 8, 1924, in Canton, marketing at the University of Richmond, Virginia Ohio, he spent most of his life as a member of Commonwealth University, Christopher Newport the statistics department of the Wharton School, University, and The George Washington University University of Pennsylvania, and served as department while also running Hodges & Associates, a marketing chair from 1971 to 1978. The breadth of his scientif- research firm. ic work illustrates the possible diversity of a career in A lifelong technology enthusiast, Hodges pro- statistics. He contributed to general statistical theory, grammed mainframes, flew airplanes, owned early especially the design of experiments; his work also versions of laptop computers and mobile phones, included applications to business, law, the military, and took pride in writing his own marketing software anthropology, psychology, economics, education, applications. An avid cook and traveler, he was active biology, and medicine. Besides statistics, de Cani in the Presbyterian church and YMCA. A loving contributed to econometrics, operations research, friend, father, and partner, he is survived by his wife, and mathematical programming. Marlane Fairleigh Hodges. He was active in civil liberties, testifying as an expert witness in racial discrimination in employ- ment litigation. He risked his life to testify about Steve Samuels the statistical significance of racial discrimination in jury selection in the South. He also collaborated Steve Samuels passed away July 26, 2012, after a with biologists and pharmaceutical companies in the long battle with lung cancer. Emeritus professor at design of clinical trials for medical research. Purdue University, Samuels was born in Brooklyn, de Cani’s other interests were many and varied. New York, and grew up on Long Island. He com- He was active in university administration and poli- pleted his undergraduate work at MIT and earned tics and supervised many PhD student dissertations. his PhD from Stanford. He joined the Purdue He loved listening to classical music, particularly faculty in 1963 as one of the original members of Beethoven quartets, and spent many pleasant hours the department of statistics. Samuels headed many after work over drinks with colleagues and other important committees both at Purdue and in the friends. These get-togethers were often held at the broader statistical community, including the Statistics University of Pennsylvania Faculty Club, where he in Sports Section of the ASA. Samuels retired in served as president. 2003. Read more about his life at www.legacy.com/ He was elected a Fellow of the American obituaries/jconline/obituary.aspx?pid=158764122 Statistical Association in 1978. He was also the #fbLoggedOut.

october 2012 amstat news 35 calendarsection • ofchapter events • committee news

sectionnews Fang Han, Ming Yuan, John Lafferty, and Larry Wasserman Biometrics Jennifer Sinnott, Harvard University, for The Biometrics Section held its annual business “Omnibus Risk Assessment via Accelerated committee meeting at this year’s Joint Statistical Failure Time Kernel Machine Modeling” with Meetings (JSM), during which they shared election co-author Tianxi Cai appointments, summarized past conferences, and Applications are invited for the 2013 Byar Young announced the winners of several section awards. Investigator Award and Biometrics Section travel The section appointed Wei Sun as the 2013 JSM awards; the deadline is December 1. Program chair, Doug Schaubel as the 2013 ENAR Finally, Section Chair Dianne Finkelstein raised Program chair, and Donglin Zeng as the 2013 the prospect of more active participation in the Continuing Education chair Also, Mike Daniels is the WNAR annual meeting. In the past, the section has new chair-elect-elect, Yu Shen is the section’s 2013– not participated formally, but there was considerable 2014 secretary/treasurer, and Limin Clegg is the interest in expanding the section’s role. Council of Sections representative for 2013–2015. Complete minutes of the meeting will be made Another highlight of the business meeting was available at www.bio.ri.ccf.org/Biometrics. a review of the ENAR meeting that took place in On another note, the section is sponsoring a sym- Washington, DC, from April 1–4. posium titled “Statistics in Biomedical Research: To view Also at the meeting, Barry Graubard gave a brief Making and Translating New Discoveries,” to take account of David Byar, after whom the section’s section news in place November 10 at the Natcher Conference young investigator award is named. This year’s winner its entirety, visit Center in Bethesda, Maryland. Registration is free, is Yang Ning, from The Johns Hopkins University. http://magazine. but required. For information, visit http://ncifrederick. This year’s travel award winners also were named cancer.gov/events/BiomedicalResearch. amstat.org. and include the following: Huaihou Chen, Columbia University, for “A Quality and Productivity Marginal Approach to Reduced-Rank Penalized Spline Smoothing with Application to The Section on Quality and Productivity (Q&P) is Multilevel Functional Data” with co-authors happy to welcome Di Michelson as chair-elect and Yuanjia Wang, Myunghee Cho Paik, and H. Ming Li as program chair-elect for 2013. If any sec- Alex Choi tion member would like to nominate a candidate for an elected office (chair or program chair) or Shuo Chen, Emory University, for “A Bayesian appointed office for 2013, please email the nomi- Hierarchical Framework for Modeling Brain nation to Teresa Utlaut at [email protected]. Connectivity of Neuroimaging Data” with co- Three members of the section were elected as ASA author F. DuBois Bowman Fellows this year: Jeffrey Hooper of BT Group US&C, Jeff Goldsmith, The Johns Hopkins University, Dongseok Choi of Oregon Health and Science for “Corrected Confidence Bands for Functional University, and Joan Hu of Simon Fraser University. Data Using Principal Components” with co- The 2013 Quality and Productivity Research authors Sonja Greven and Ciprian Crainiceanu Conference will be held June 5–7 at GE Global Research in Schenectady, New York. A short course Min Jin Ha, The University of North Carolina, will be presented by Stephan Steiner from the for “Testing and Estimation of Partial University of Waterloo on June 4. For details, visit Correlation Networks” with co-author Fred A. www.qprc2013.com or contact Martha Gardner at Wright [email protected]. Peisong Han, University of Michigan, for Members can submit JSM 2013 invited session “Conditional Empirical Likelihood Inference proposals to Willis Jensen at [email protected]. for Unbalanced Longitudinal Data” with co- authors Peter X.-K. Song and Lu Wang Statistics and the Environment The Section on Statistics and the Environment Yen-Tsung Huang, Harvard University, for (ENVR) is sponsoring a student paper competi- “Joint Analysis of SNP and Gene Expression Data in Genome-Wide Association Studies” tion on the topic of environmental statistics. Papers with co-authors Xihong Lin and Tyler can consist of novel approaches to the analysis of VanderWeele environmental data, new methodology with a clear application to an environmental statistical issue, or Han Liu, The Johns Hopkins University, for application of statistics to environmental problems. “The Nonparanormal Skeptic” with co-authors The selected winner will present his/her paper in a

36 amstat news october 2012 section • chapter • committee news contributed session at the Joint Statistical Meetings was elected vice chair of the Council of Sections (JSM) in August of 2013 and receive a $1,000 sti- Governing Board. pend toward reimbursing the expenses for attending This year, the section contributed support toward JSM. The deadline for submission is December 15. the Mathematical Association of America’s Project To be eligible, applicants must be in a college pro- NeXT. Funding requests are due by February 1, gram. Also, applicants must submit an abstract to the 2013, to Deborah Nolan at [email protected]. ASA by February 1, 2013, and request a presentation The section newsletter publication schedule in an ENVR session. Applicants must follow all rules changed last year from semiannual to annual (early required for paper submission and attendance at JSM. fall) as an archival digest, with members and offi- This award will recognize an environmental sta- cers able to relay more timely communications and tistics contribution with a certificate and recogni- post announcements through the centralized com- tion during the ASA ENVR section business meet- munication tools of the Section’s ASA Community ing/mixer at JSM. For information about student site (go to http://community.amstat.org and select paper awards, visit www.amstat.org/sections/student- “My Communities”). paperawards.cfm. Application details can be viewed at http://magazine.amstat.org/?cat=17. Statistics in Epidemiology The Statistics in Epidemiology (SIE) section would Statistics Education like to congratulate the following winners of the Reported by Larry Lesser 2012 SIE Young Investigator Awards: The Statistics Education Section had a productive Xinxin Dong, University of Pittsburgh JSM 2012, sponsoring or co-sponsoring 11 invited Brianna Heggeseth, University of California at panels/sessions, 14 topic-contributed panels/ses- Berkeley sions, 10 contributed paper sessions, 12 roundtable Chen Hu, University of Michigan discussions, and a poster session. Also, many section members were recognized. Xiaobi (Shelby) Huang, Merck & Co. Michael Posner won the 2012 Waller Education Andrew Jaffe, The Johns Hopkins University Award; Thomas Willemain won the National Mu Sigma Rho’s 2012 William D. Warde Statistical Stephanie Kovalchik, National Cancer Institute Education Award; and Anna Jonsdottir, a student at Fu-Wen Liang, The University of Texas the University of Iceland, won the section’s best con- tributed paper of JSM 2011. Additionally, several Sandra Mohammed, University of California at members became Fellows. The student award will be Davis named for ASA executive director and longtime edu- Ming Wang, Emory University cation advocate Ron Wasserstein. Also at JSM 2012, Nathan Tintle, Jill VanderStoep, Hong Zhu, The Ohio State University Vicki-Lynn Holmes, Brooke Quisenberry, and Todd Cory Zigler, Harvard University Swanson won the Journal of Statistics Education (JSE) Best Paper Award for “Development and Assessment Jose Zubizarreta, University of Pennsylvania of a Preliminary Randomization-Based Introductory Details for the 2013 Young Investigator award Statistics Curriculum” in the March 2011 issue. competitions will be announced in November. Another winner was John Gabrosek, who received a “retiring editor certificate of appreciation” for his work Survey Research Methods as 2010–2012 JSE editor. The Survey Research Methods Section (SRMS) was At the section’s business meeting, committees were well represented at this year’s JSM in San Diego, formed in membership/fundraising, Common California. The section had five invited, 16 topic- Core, Fellows nominations, and GAISE revision. contributed, and 13 contributed sessions. There was Patty Frazer Lock gave a progress report on MAA one poster session with 17 posters and two round- guidelines for introductory statistics courses taken tables, one led by Lars Lyberg on total survey error by mathematical sciences majors. and one led by Joe Sakshaug on linking adminis- Along with the award winners, the 2012 sec- trative and survey data sets. The section also had a tion officers were announced. The officers include Continuing Education course, Paradata in Survey Amy Wagaman, 2013 JSM Program chair; Jim Research, taught by Frauke Kreuter and a technolo- Albert, chair-elect; Dexter Whittinghill, secretary/ gy workshop, SAS Procedures for Analyzing Survey treasurer; and Brigitte Baldi and Herle McGowan, Data, taught by Pushpal Mukhopadhyay. Executive Committee at large members. Also, Kreuter is incoming program chair-elect for SRMS. Dick De Veaux was elected as a Council of Sections Board representative and Tom Short

october 2012 amstat news 37 calendar of events

Burdwan University, Golapbag, Road, Melbourne, International 2012 Burdwan, International 713104; 3000, Australia; SusannaCramb@ 919433261178; [email protected]. cancerqld.org.au.

December 27—Eighth International ››21–23—American Statistical To view *2–7—68th Annual Deming Triennial Calcutta Symposium Association Conference on Statistical Practice, New the entire list Conference on Applied on Probability and Statistics, Statistics, Atlantic City, New Kolkata, India Orleans, Louisiana of statistics Jersey For more information, visit http:// For more information, visit www. meetings and For more informaion, visit www. triennial.calcuttastatisticalasso- amstat.org/meetings/csp/2013 or demingconference.com or contact ciation.org or contact Arindam contact ASA Meetings, 732 North workshops, visit Washington St., Alexandria, VA www.amstat. Walter Young, 16 Harrow Circle, Sengupta, 35 Ballygunge Circular Wayne, PA 19087; demingchair@ Road, Department of Statistics, 22314; (703) 684-1221; meetings@ org/dateline. gmail.com. University of Calcutta, Kolkata, amstat.org. International 700019, India; +91- 10–12—SAMSI-FODAVA 9433590336; [email protected]. March Workshop on Interactive Visualization and Analysis of 5–7—New Techniques and Massive Data, Research Triangle Technologies for Statistics Park, North Carolina 2013 (NTTS 2013), Brussels, Belgium For details, visit www.samsi.info/ For details, visit www.ntts2013. workshop/samsi-fodava-workshop- January eu or contact NTTS 2013 interactive-visualization-and-analysis- 2–5—IISA Conference: Statistics, Secretariat, Unit B1, Luxembourg, massive-data-december-10-12-20 Science, and Society: New International L-2920, Luxembourg, or contact Jamie Nunnelly, 19 T. W. Challenges and Opportunities, 00352430138327; ESTAT-NTTS@ Alexander Drive, Research Triangle Chennai, India ec.europa.eu. Park, NC 27709; (919) 685-9350; For more information, visit www. [email protected]. iisaconference.info or contact Subrata 13–15—IAENG International Kundu, Rome Hall 557, 801 22nd St. Conference on Data Mining and Applications, Hong Kong 15–17—Twenty-First NW, Washington, DC 20052; (202) 994-6355; [email protected]. For details, visit www.iaeng.org/ International Conference on IMECS2013/ICDMA2013.html or con- Interdisciplinary Mathematics, tact IAENG Secretariat, Unit 1, 1/F, 6–10—ISBA Regional Meeting Statistics, and Computational 37-39 Hung To Road, Hong Kong, and International Workshop/ Techniques (IMSCT 2012-FIM International HK, Hong Kong; (852) Conference on Bayesian Theory XXI), Chandigarh, India 3169-3427; [email protected]. For more information, visit imsct2012. and Applications (IWCBTA), Varanasi, India puchd.ac.in or contact Kalpana ››18–23—3rd Joint For details, visit www.bhu.ac.in/isba Mahajan, Department of Statistics, Statistical Meeting Deutsche or contact Satyanshu Upadhyay, Panjab University, Chandigarh, Arbeitsgemeinschaft Statistik, Department of Statistics, Banaras International 160014, India; 91-172- Freiburg, Germany Hindu University, Varanasi, 2541776; [email protected]. For more information, visit International 221005, India; +91- dagstat2013.uni-freiburg.de or 9918922333; [email protected]. 21–23—6th International contact Jan Beyersmann, Eckerstr. Conference of IMBIC on Salt Lake 1, Freiburg, International 79104, City, Kolkata, West Bengal, India ››18–19—Workshop on New Germany; dagstat2013@imbi. For details, visit www.imbic.org/ Directions in Monte Carlo uni-freiburg.de. forthcoming.html or contact Avishek Methods, Gainesville, Florida Adhikari, AH 317, Salt Lake City, For more information, visit www. Sector 2, Kolkata, WB, International stat.ufl.edu/symposium/2013/index. April 700091, India; 00913323598617; html or contact Robyn Crawford, 22–25—7th Meeting of the [email protected]. 103 Griffin-Floyd Hall, Gainesville, FL Eastern Mediterranean Region 32615; (352) 392-1942; robyn@stat. International Biometric Society, 24–25—Young Statisticians ufl.edu. Tel Aviv, Israel Meet: An International For details, visit event.pwizard.com/ Conference 2012, Burdwan, ims or contact Ilana Lobel, The India February Gertner Institute for Epidemiology For more information, visit ››7–8—The Young Statisticians and Health Policy Research, Tel www.buruniv.ac.in/Notices/ Conference 2013, Melbourne, Hashomer, International 52621, UBUR_2012032_NOT_WEBPAGE. Australia Israel; 972-3-5305390; ilanal@gertner. pdf or contact Arindam Gupta, For details, visit www.ysc2013.com health.gov.il. Department of Statistics, or contact Susanna Cramb, Clayton

38 amstat news october 2012 calendar of events

››26—Philosophy of 20–25—29th European Meeting Information: The Value of of Statisticians (EMS2013), The following events are the latest additions to Information, Washington, DC Budapest, Hungary the ASA’s online calendar of events. Announce- For more information, visit www. For more information, visit www. american.edu/cas/economics/ ems2013.eu or contact Eszter ments are accepted from education and not-for- info-metrics/workshop/workshop- Zsigmond, Szilágyi E. fs. 79., profit organizations only. To view the complete list 2013-spring.cfm or contact Amos Budapest, International 1026, of statistics meetings and workshops, visit www. Golan, 4400 Massachusetts Ave. Hungary; +36 1 212 00 56; amstat.org/dateline. NW, Kreeger 104, Washington, DC [email protected]. 20016; (202) 885-3770; info-metrics@ * Indicates events sponsored by the ASA or one of american.edu. its sections, chapters, or committees August ›› Indicates events posted since the previous issue *3–8—2013 Joint Statistical May Meetings, Montréal, Québec, ››13–17—SAMSI Undergraduate Canada Modeling Workshop, Research For more information, contact ASA Triangle Park, North Carolina Meetings, 732 North Washington Str. 155, Munich, International For details, visit www.samsi.info/ St., Alexandria, VA 22314-1943; (703) 80687, Germany; +49 8954823462; workshop/undergraduate-modeling- 684-1221; [email protected]. [email protected]. workshop-may-13-17-2013 or contact Jamie Nunnelly, 19 T. W. 24–31—59th ISI World Statistics Alexander Drive, Research Triangle Congress, Hong Kong, China Park, NC 27709; (919) 685-9350; For more information, visit www. September [email protected]. isi2013.hk or contact Hans Lucas, P.O. 9–12—Seminar of the Box 24070, Den Haag, International Austro-Swiss Region of the 2490 AB, The Netherlands; [email protected]. International Biometric Society, June Dornbirn, Austria For more information, contact 10–12—4th Nordic-Baltic ››25–29—34th Annual Hanno Ulmer, Innsbruck Medical Biometric Conference, NBBC13, Conference of the International University, Schöpfstraße 41, Stockholm, Sweden Society for Clinical Innsbruck, International 6020, For more information, visit nbbc13. Biostatistics, Munich, Germany Austria/Europe; +43512900370900; org or contact Marie Jansson, Box For details, visit www.iscb2013.info [email protected]. 281, Stockholm, International or contact Tanja Lypp, Landsberger SE-17177, Sweden; +46 8 52486150; [email protected].

››13–15—MedicReS World Congress on Good Medical Biopharmaceutical Applied Statistics Symposium Research, Vienna, Austria The 19th annual biopharmaceutical applied statistics sympo- For more information, contact sium, BASS XIX, will be held November 5–9 at the Mulberry Inn Jenny Knapp, Mariahilfer Strasse Suites in historic Savannah, Georgia. At least 16 one-hour tutori- 123/3, Vienna, International 1060, Austria; +43 1 599 99 8070; jenny. als on diverse topics pertinent to the research, clinical develop- [email protected]. ment, and regulation of pharmaceuticals will be presented from November 5–7 by speakers from academia, the pharmaceutical industry, and the U.S. Food and Drug Administration (FDA). July Four parallel two-day short courses will be presented November 7–9. Highlights of the symposium include the keynote address and reception on November 5 and the FDA biometrics session ››15–23—2013 Industrial and Mathematical Statistical on November 7. Modeling Workshop, Research BASS is a nonprofit entity, sponsored by the department Triangle Park, North Carolina of biostatistics at Virginia Commonwealth University and the For more information, visit www. Jiann-Ping Hsu College of Public Health at Georgia Southern samsi.info/workshop/2013-industrial- University. Its purpose is to raise funds for graduate fellowships mathstat-modeling-workshop- in biostatistics. graduate-students-july-15-23-2013 Registration is now open at www.bassconference.org. or contact Jamie Nunnelly, 19 T. W. For more information, contact Karl Peace at bassxix2012@ Alexander Drive, Research Triangle gmail.com or [email protected]. Park, NC 27709; (919) 685-9350; [email protected].

october 2012 amstat news 39 40 amstat news october 2012 professional opportunities

Alabama n Auburn University invites applications for the 12-month non-tenure track associ- Professional Opportunity listings may not exceed 65 words, plus equal oppor- ate research professor faculty position of tunity information. The deadline for their receipt is the 20th of the month two director of statistical consulting. Applicants months prior to when the ad is to be published (e.g., May 20 for the July issue). must possess a PhD in statistics; we seek Ads will be published in the next available issue following receipt. statistical consulting or postdoctoral expe- Listings are shown alphabetically by state, followed by international listings. rience. Further details available at www. Vacancy listings may include the institutional name and address or be identi- auburn.edu/cosam/news/director-of-statistical- fied by number, as desired. consulting1.htm. Review of applications Professional Opportunities vacancies also will be published on the ASA’s starts on January 15, 2013. For more infor- website (www.amstat.org). Vacancy listings will appear on the website for the mation about the department visit www. entire calendar month. Ads may not be placed for publication in the magazine math.auburn.edu/.Auburn University is only; all ads will be published both electronically and in print. an Affirmative Action/Equal Opportunity Rates: $320 for nonprofit organizations (with proof of nonprofit status), Employer. Women and minorities are $475 for all others. Member discounts are not given. For display and online encouraged to apply. advertising rates, go to www.amstat.org/ads. Listings will be invoiced following publication. All payments should be California made to the American Statistical Association. All material should be sent n University of California, Davis, to Amstat News, 732 North Washington Street, Alexandria, VA 22314- Department of Statistics. Apply now - one 1943; fax (703) 684-2036; email [email protected]. open-rank tenure-track faculty position Employers are expected to acknowledge all responses resulting from pub- beginning 7/1/2013. Requires PhD in lication of their ads. Personnel advertising is accepted with the understand- statistics or related field, research interest in ing that the advertiser does not discriminate among applicants on the basis of statistical methodology, theory or computing race, sex, religion, age, color, national origin, handicap, or sexual orientation. for problems involving large/massive and Also, look for job ads on the ASA website at www.amstat.org/jobweb. complex data. Interests in scientific applica- tions in bioinformatics, bio-medical imag- ing, or genomics are encouraged to apply. Application review begins 12/1/12 until position filled. See www.stat.ucdavis.edu/ employment/academic for more information. UC Davis is an affirmative action/equal for details or go to www.rand.org/statistics. tenure-track appointments in statistics employment opportunity employer and is Applications received by December 15, beginning August 2013. Applicants w/back- dedicated to recruiting a diverse faculty com- 2012, will receive priority. Applications must grounds in theoretical, applied, or bio-statis- munity. We welcome all qualified applicants be submitted online following the instruc- tics preferred. Candidates must demonstrate to apply, including women, minorities, indi- tions at www.rand.org/statistics/jobs.html commitment to, and excellence in, under- viduals with disabilities and veterans. (Job ID #3221). Send questions to Susan_ graduate teaching and scholarly achieve- n San Francisco VA Medical Center and [email protected]. EO/AA Employer. ment. PhD required. This position carries a 3-2 teaching load with generous leaves for NCIRE have an immediate opening for a Iowa statistician. Qualified applicant will have junior and senior faculty. http://academics. n Drake University Actuarial Science holycross.edu/mathcs/statsearch. College of the master’s degree in statistics, biostatistics, Program invites applications for two full- epidemiology, or equivalent, and 3+ years Holy Cross is an Affirmative Action/Equal time positions starting August 2013: (1) Opportunity Employer. experience in statistical analysis, data mod- tenure-track assistant professor of actuarial eling, and advanced programming skills. science (PhD preferred, ABD considered), n Professor or Associate Professor/Division SAS experience and ability to work with and (2) non-tenure-track Robb B. Kelly Chair of the Mathematics and Science large national administrative datasets, such Professor of Actuarial Science and Risk Division. The mathematics and science divi- as VA and CMS datasets, highly preferred. Management (master’s degree with actuarial sion of Babson College invites applications www.ncire.org Apply at referring to job credentials, or PhD/ABD required). View for September 1, 2013, appointment at the 2012-1888. EOE. details and upload materials at https://drake. rank of professor or associate professor if the n RAND Corporation is seeking PhD HireTouch.com. Reviews began Oct 1, 2012. candidate will soon be ready for promotion statisticians for exciting opportunities to col- Equal opportunity employer (EEO). to professor. This position will include an laborate on multidisciplinary public policy initial three year appointment as division research projects. Openings exist for recent Massachusetts chair with the possibility of a three year n graduates and experienced statisticians. The department of mathematics and extension. www.Click2Apply.net/gqgy77n. See our ad in the September Amstat News computer science at the College of the Holy EOE/AA. Cross invites applications for two full-time

october 2012 amstat news 41 42 amstat news october 2012 n Massachusetts Institute of Technology Department of Mathematics. Mathematics department MIT seeking to fill com- bined teaching and research positions as instructor, assistant professor and higher, in statistics beginning September 2013. Appointments based mainly on exceptional research qualifications. PhD required by employment start date. Submit online, www.mathjobs.org: CV, research description, three recommendation let- ters. Applications should be complete by December 1, 2012. (See full classified text at mathjobs.) Massachusetts Institute of Technology is an Equal Opportunity Affirmative Action Employer. n The Babson College Mathematics and Science Division invites applicants for tenure-track assistant professor position starting September 1, 2013. The ideal candidate will possess a terminal degree in statistics, analytics, or a related field or expect to complete the requirements for the degree by September 1, 2013, a proven teaching record with the desire to deliver statistics and analytics courses, and a strong research agenda. www.click2apply. net/9b6488s. EOE/AA. Michigan n University of Michigan Biostatistics Department has a research faculty posi- tion in cancer research beginning prior to January 2013. Department seeks outstand- ing individuals with interests in the devel- opment of statistical methods, collaborative scientific research and teaching. Cancer research experience required. Information available at www.sph.umich.edu/biostat/jobs. Contact chair of Cancer Biostatistics Search Committee, Dept. of Biostatistics, SPHII, University of Michigan, 1415 Washington Heights, Ann Arbor, MI 48109-2029. Email: [email protected]. EOE/AA. n The Survey Research Center (www.src. isr.umich.edu) in the Institute for Social Research invites applications from out- standing candidates for faculty research fel- low appointments in any area of social sci- ence. These appointments are intended to lead directly into a research professor ten- ure-track career. Applicants should submit a cover letter, vita and one or two publica- tions. Three reference letters should be sent electronically to [email protected]. Reference position #72722. The University of Michigan is an Affirmative Action/Equal

october 2012 amstat news 43 Opportunity Employer and is responsive to to conduct epidemiological and method- summarizing findings for internal reports, the needs of dual career couples. Women ological research for the recently completed FDA submissions, and professional publica- and minority candidates are encouraged Two Sister Study, a family-based GWAS of tions. Experience developing trial protocols, to apply. genetic and environmental risk factors for sample size calculations, stat plans, and young-onset breast cancer. See www.sister- study reports strongly preferred. SAS profi- New Mexico study.org/English/index1.htm and www.niehs. ciency essential. Requirements: Master’s in n Assistant/Associate Professor of Applied nih.gov/research/atniehs/labs/bb/staff/index.cfm statistics/equivalent skills, PhD preferred. Statistics. The department of econom- for more information. Send CV, statement www.pinneyassociates.com AA/EOE. ics, applied statistics, and international of research interests and contact informa- n business at New Mexico State University tion for 3 references to Clarice Weinberg at Penn State Statistics Department - Faculty invites applications for a tenure-track assis- [email protected] in Research Triangle Positions. Multiple tenure-track positions tant/associate professor of applied statistics, Park, NC. www.sisterstudy.org/English/ at all levels available in statistics department 9-month position. Doctorate required. index1.htm The NIH is dedicated to build- beginning August 2013. PhD in statistics Specialty in time series and/or spatial sta- ing a diverse community in its training and or related field, excellent research and teach- tistics preferred. Other fields considered. employment programs. ing credentials; particulars at www.stat.psu. Position requires graduate and undergradu- edu. Please apply at mathjobs.org. Screening ate teaching, research, and service. For full Pennsylvania begins November 15. Employment will position announcement see www.nmsu. n Penn State University. Fixed-term faculty require successful completion of back- edu/~personel/postings/faculty/9265247. for innovative online applied statistics pro- ground check(s) in accordance with html EOE. gram. Excellence teaching applied statistics, University policies. www.stat.psu.edu. experience in consulting and collaborative Penn State is committed to affirmative New York research.Requirements: PhD in statistics or action, equal opportunity and the diver- n Tenure-track positions in Statistics at related quantitative field. See www.mathjobs. sity of its work force. Assistant/Associate Professorship Levels. org/jobs?joblist-514-3891 for more informa- n RAND Corporation is seeking PhD Statistics/CIS department, Zicklin School of tion. Submit cover letter, teaching statement, statisticians for exciting opportunities to Business, Baruch College (CUNY). PhD in resume, four references to www.mathjobs. collaborate on multidisciplinary public Statistics, good teaching ability and proven org. This is a fixed term position funded for policy research projects. Openings exist for record or potential for strong research. one year from date of hire with possibility of recent graduates and experienced statisti- Send a CV, three letters, of recommenda- refunding. www.stat.psu.edu. www.mathjobs. cians. See our ad in the September Amstat tion, research and teaching statements to: org/jobs?joblist-514-3891. Employment will News for details or go to www.rand.org/ Statistics Search Committee, Statistics/ require successful completion of background statistics. Applications received by December CIS Department,Box 11-220, Baruch check(s) in accordance with University poli- 15, 2012, will receive priority. Applications College, One Baruch Way, New York, NY cies. Penn State is committed to affirmative must be submitted online following the 10010. Deadline: January 20, 2013. Equal action, equal opportunity and the diversity instructions at www.rand.org/statistics/jobs. Opportunity /Affirmative Action/Equal of its work force. html (Job ID #3221). Send questions to Access Employer. [email protected]. www.rand.org/ n Possible tenure-track, lecturer, visiting statistics. EO/AA Employer. n New York University Stern School of positions. Collegial environment empha- sizing disciplinary and cross-disciplinary Business Statistics Group, tenure-track n The statistics department at Temple research and teaching. All areas of statistics assistant professor appointment in statis- University seeks a tenure-track, senior welcome. Joint appointments possible with tics. Candidates should have evidence of assistant/associate professor. Candidates in other units in the Pittsburgh area. See www. boundary-spanning interests across fields any area of statistics must have a PhD in stat.cmu.edu (email: [email protected]). that reflect significant interfaces of statistics statistics, publications in top-tier journals, Send CV, research papers, relevant tran- with areas of relevance in a business school. teaching excellence, and strong theory/appli- scripts, and three recommendation letters Expected that candidate will be produc- cation background. Apply electronically to to: Faculty Search Committee, Statistics, tive researcher and effective teacher at both Sanat Sarkar, [email protected], with Carnegie Mellon University, Pittsburgh, undergraduate and graduate levels. See www. cover letter, CV, teaching evidence and three PA 15213, USA. Application screening stern.nyu.edu/experience-stern/about/ recommendation letters. For more infor- begins immediately, continues until posi- departments-centers-initiatives/academic- mation, www.fox.temple.edu/dept/statistics. tions closed. www.stat.cmu.edu. Women departments/ioms/faculty-staff/position-open- Temple University is an Equal Opportunity/ and minorities are encouraged to apply. ings/statistics/index.htm for details, including Affirmative Action Employer. information on application procedure. www. AA/EOE. stern.nyu.edu/ioms. New York University is n Texas an Affirmative Action/Equal Opportunity Pinney Associates, Pittsburgh (www. n The department of mathematics and Institution. pinneyassociates.com) has an immediate statistics invites applications for a tenure- opening for a senior biostatistician/data ana- North Carolina lyst. We provide strong science-based con- track assistant professor position in statis- n sultation to pharmaceutical and consumer tics beginning fall 2013. A PhD degree NIEHS/NIH. Postdoc position for is required. Apply for position number biostatistician with interest in epidemiology healthcare clients. Responsibilities include conducting univariate/multivariate analyses; T96800 at http://jobs.texastech.edu. Include

44 amstat news october 2012 AMS standard cover sheet and vita. Have mission. We seek and welcome applications excellence in teaching and research or three letters of reference sent to Alex Wang, from individuals of all backgrounds, experi- evident potential. Review of applica- Hiring Committee Chair, Department ences, and perspectives. tions begins 11/15/2012 and continues of Mathematics and Statistics, Texas Tech until filled. Send letter, CV, and arrange University, Lubbock, TX 79409-1042. Canada to have three recommendation letters Review will begin immediately. jobs to Chair, Department of Statistics and texastech.edu EOE: Texas Tech is an AA/ Ontario Actuarial Science, University of Waterloo, EO employer. n One tenure-track or tenured posi- 200 University Avenue West, Waterloo tion in actuarial science. PhD in actu- ON N2L 3G1, CANADA. http://math. Wyoming arial science,and excellence in teaching and uwaterloo.ca/statistics-and-actuarial-science/ n Tenure-track Assistant Professor, research. Review of applications begins faculty-positions. The University of Waterloo Statistics, beginning August 2013. 11/15/2012 and continues until filled. Send encourages applications from all qualified Requirements include PhD in statistics or letter, CV, and arrange to have three recom- individuals, including women, members of related field and excellence in both research mendation letters to Chair, Department of visible minorities, native peoples, and per- and teaching at all levels. Seeking candidate Statistics and Actuarial Science, University sons with disabilities. with expertise in computational statistics. of Waterloo, 200 University Avenue West, Valuable complementary interests include Waterloo ON N2L 3G1, CANADA. http:// n University of Waterloo. One or more Bayesian statistics and interdisciplinary math.uwaterloo.ca/statistics-and-actuarial- tenure-track or tenured positions in statis- research. Collaborative opportunities exist science/faculty-positions. The University of tics. PhD in statistics, demonstrated excel- both on campus and with NCAR. Full Waterloo encourages applications from all lence in teaching and research or evident description and application instructions are qualified individuals, including women, potential. Review of applications begins at www.ezfacultysearch.com/uwyo/stats/10. members of visible minorities, native peo- 11/15/2012 and continues until filled. Send The University of Wyoming is committed to ples, and persons with disabilities. letter, CV, and arrange to have three recom- diversity and endorses principles of affirma- mendation letters to Chair, Department of tive action. We acknowledge that diversity n University of Waterloo. One or more Statistics and Actuarial Science, University enriches and sustains our scholarship and tenure-track or tenured positions in biosta- of Waterloo, 200 University Avenue West, promotes equal access to our educational tistics. PhD in biostatistics, demonstrated Waterloo ON N2L 3G1, CANADA. http://

BIOSTATISTICIAN ASSISTANT PROFESSOR or ASSOCIATE PROFESSOR Non Tenure Line (Research) PSYCHIATRY AND BEHAVIORAL SCIENCES The Department of Psychiatry and Behavioral Sciences seeks a biostatistician to join the Department as an Assistant or Associate Professor in the Non-Tenure Line (Research). The major criterion for appointment for faculty in the Non-Tenure Line (Research) is evidence of high-level performance as a researcher for whose special knowledge a programmatic need exists. Faculty rank will be determined by the qualifi cations and experience of the successful candidate. The successful applicant should have a doctoral degree in Biostatistics, Statistics, or a related fi eld, re- lated post-training work experience in an academic setting, and experience supervising statistical sup- port staff. We are particularly interested in candidates with a strong record of research accomplish- ments, including extramural research funding, in the fi eld of biostatistics that includes characterizing longitudinal mental health, neurocognitive, and neuroimaging trajectories across the lifespan and in aging. Candidates with at least 6-10 years of experience in psychiatric applications, signifi cant research training, and a demonstrated track record in empirical research are strongly desired. Stanford University is an equal opportunity employer and is committed to increasing the diversity of its faculty. It welcomes nominations of and applications from women and members of minority groups, as well as others who would bring additional dimensions to the university’s research, teaching and clinical missions. Interested candidates should send a copy of their curriculum vitae, a brief letter outlining their interests, and the names of three references via e-mail only to: Booil Jo, Ph.D. c/o Bevin DeMuth, [email protected] Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine 401 Quarry Road, Stanford, CA 94305 Phone: (650) 498-5107 / Fax: (650) 724-4794

october 2012 amstat news 45 math.uwaterloo.ca/statistics-and-actuarial- science/faculty-positions. The University of Waterloo encourages applications from all qualified individuals, including women, members of visible minorities, native peo- ples, and persons with disabilities. International Russia n The Skolkovo Institute of Science and Technology (Skolkovo Tech) seeks candidates for tenured and tenure-track faculty positions in science, technology and innovation to begin Fall 2013 or thereafter. Skolkovo Tech is an innovative, new, private university located just out- side of Moscow, Russia. Please visit http:// web.mit.edu/sktech/faculty-positions for more information and submit application materials to https://sktech-search.mit.edu by December 15, 2012. EOE. China n Non-Tenure-Track Teaching Position for Business Statistics in the Dept of ISOM. Applications will be accepted until the posi- tion is filled. Excellence in teaching, and

UNIVERSITY OF PENNSYLVANIA PERELMAN SCHOOL OF MEDICINE

The Division of Biostatistics in the Department of Biostatistics and Epidemiology at the University of Pennsylvania Perelman School of Medicine seeks highly qualifi ed candidates for standing faculty positions in both clinician educator (non-tenure) track and tenure track at the Assistant, Associate, or full Professor level. Academic rank will be commensurate with credentials and experience. A doctoral degree in Biostatistics, Statistics, or a related discipline is required. Review of applications will begin on September 15, 2012. Application will continue to be accepted after this date, until the positions are fi lled. The expected start date is July 2013 or later.

Clinician Educator track applicants will focus primarily on collaborative research as co-investigators, with secondary emphasis on methodological research. Applicants with collaborative research interests in translational science, health services, pediatrics, mental/behavioral health, comparative effectiveness research, and/or clinical trials are especially encouraged to apply. Applicants in other areas of research areas will also be considered.

Tenure track applicants will focus primarily on methodological research, with secondary emphasis on collaborative research projects within the School of Medicine. Ap- plicants with biostatistical research interests in meta-analysis, structural equation modeling, psychometrics (test/scale development, item response theory), causal models, mediation analysis, instrumental variables, nonparametric statistics, and specialized biostatistical methods for next generation sequence data are especially encouraged to apply. Applicants in other areas of research areas will also be considered. There is a rich mix of ongoing biomedical research projects in the Perelman School of Medicine to provide motivation and opportunities for the development of novel statistical methods on wide ranging topics.

Candidates for both tracks are expected to have a strong commitment to teaching and must demonstrate outstanding research productivity. Primary teaching responsibilities include participation in Penn’s Center for Clinical Epidemiology and Biostatistics academic programs.

The Graduate Group in Epidemiology and Biostatistics, jointly with the Department of Statistics in the Wharton School, offers degree programs leading to both the Doctor of Philosophy (PhD) and Master of Science (MS) in Biostatistics.

The University of Pennsylvania, founded by Benjamin Franklin, is a world-class research institution, located near the heart of Philadelphia. All of Penn’s 12 schools are located within walking distance of one another. The Penn Perelman School of Medicine is one of the top ranked medical schools in NIH funding.

The University of Pennsylvania is an affi rmative action/equal opportunity employer. Women and minorities are strongly encouraged to apply. Qualifi ed applicants should send a cover letter indicating the specifi c position to which they are applying, curriculum vitae, three letters of reference, and a statement of research interests to:

Clinician Educator: Apply for this position online at: http://www.med.upenn.edu/apps/faculty_ad/index.php/g303/d3026 Tenure Track: Apply for this position online at: http://www.med.upenn.edu/apps/faculty_ad/index.php/g303/d3027

46 amstat news october 2012 PhD required by employment start-date. WILLIAMS COLLEGE The successful applicant is expected to play Assistant Professor an important role in teaching and develop- ing business statistics courses for undergrad- The Williams College Department of Mathematics and Statistics invites applications for a uate and MBA programs of the business tenure-track position in statistics, beginning fall 2013, at the rank of assistant professor (in school. Submit CV and three referees to: an exceptional case, a more advanced appointment may be considered). We are seeking a highly qualifi ed candidate who has demonstrated excellence in teaching and research, [email protected]. http://jobs.amstat.org/hr/ and who will have a Ph.D. by the time of appointment. The candidate will become the jobdetail.cfm?job_id=4883315. The Hong fourth tenure-track statistician in the department, joining a vibrant and active statistics Kong University of Science and Technology group. Williams College is a private, residential, highly selective liberal arts college with is an Equal Opportunity Employer. an undergraduate enrollment of approximately 2,000 students. The teaching load is two courses per 12-week semester and a winter term course every other January. In addition to excellence in teaching, an active and successful research program is expected. Mexico n Visiting Positions Related to the To apply, please send a vita and have three letters of recommendation on teaching and re- International Year of Statistics and search sent to the Hiring Committee, Department of Mathematics and Statistics, Williams Mathematics of Planet Earth. CIMAT, College, 18 Hoxsey Street, Williamstown, MA 01267. Teaching and research statements Mexico. CIMAT calls for applicants for two are also welcome. Evaluations of applications will begin on or after November 15 and will continue until the position is fi lled. All offers of employment are contingent upon comple- 6- or 12-month visiting positions (January tion of a background check. Further information is available upon request. For more in- to December 2013). Candidates are sought formation on the Department of Mathematics and Statistics, visit http://math.williams.edu/. to conduct research and activities oriented by the aims of the 2013 celebrations with Williams College is a coeducational liberal arts institution located in the Berkshire Hills of impact on CIMAT academic programs. western Massachusetts with easy access to the culturally rich cities of Albany, Boston, and Expected monthly income $4,000 US Cy. New York City. The College is committed to building and supporting a diverse population Further information at www.cimat.mx/ of approximately 2,000 students, and to fostering an inclusive faculty, staff and curricu- n lum. Williams has built its reputation on outstanding teaching and scholarship and on the index.php?m=541 EOE. academic excellence of its students. Please visit the Williams College website http://www. williams.edu/. Beyond meeting fully its legal obligations for non-discrimination, Williams College is committed to building a diverse and inclusive community where members from all backgrounds can live, learn, and thrive.

october 2012 amstat news 47 October 2012 • Issue #424

AMSTATNEWS 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 about advertising, please visit www.amstat.org.

misc. products and services Cambridge University Press...... p. 19 Wiley-Blackwell...... p. 40

professional opportunities NORC...... p. 45 SAMSI...... p. 42 Stanford University ...... p. 45 U.S. Census Bureau...... p. 43 University of Pennsylvania ...... p. 46 University of Warwick...... p. 47 Williams College...... p. 47 Westat...... p. 46

software Advise Analytics ...... p. 27 Aptiv Solutions ...... centerfold Cytel Inc...... p. 10 JMP, a business unit of SAS...... cover 4 Minitab Inc...... centerfold NCSS...... p. 31 Salford Systems...... p. 20 SAS...... cover 3 Statistical Solutions...... p. 15 StatSoft...... cover 2

48 amstat news october 2012