November 2014 • Issue #449 AMSTATNEWS The Membership Magazine of the American Statistical Association • http://magazine.amstat.org

ASA Gears Up for CSP ALSO: Using the Classroom to Bring Big Data to Statistical Agencies

Sirken Endows Fund to Recognize Survey Researchers

AMSTATNEWS NOVEMBER 2014 • ISSUE #449 Executive Director Ron Wasserstein: [email protected]

Associate Executive Director and Director of Operations features Stephen Porzio: [email protected] 3 President’s Corner Director of Science Policy Steve Pierson: [email protected] 5 This Month in ASA's History

Director of Education 5 ASA in Search of Editors Rebecca Nichols: [email protected] 6 Highlights of the August 2014 ASA Board of Directors Meeting Managing Editor Megan Murphy: [email protected] 8 Nominations Wanted for ASA President-Elect, Vice President, International Representative Candidates Production Coordinators/Graphic Designers Sara Davidson: [email protected] 9 SAMSI Offers Two New Research Programs for 2014–2015 Kathryn Wright: [email protected] 9 AAPOR Releases Report on Survey Refusals Publications Coordinator Val Nirala: [email protected] 10 Meet Joseph Reilly, Associate Administrator of NASS Advertising Manager 12 Using the Classroom to Bring Big Data to Statistical Agencies Claudine Donovan: [email protected]

Contributing Staff Members Pamela Craven • Amy Farris • Rebecca Nichols Steve Pierson • Kathleen Wert Amstat News welcomes news items and letters from readers on matters columns of interest to the association and the profession. Address correspondence to Managing Editor, Amstat News, American Statistical Association, 732 North Washington Street, Alexandria VA 22314-1943 USA, or email amstat@ 14 STATtr@k amstat.org. Items must be received by the first day of the preceding month Applying to Statistics PhD Programs 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. STATtr@k is a column in Amstat News and a website geared toward people who Articles will be edited for space. Accompanying artwork will be accepted are in a statistics program, recently graduated from a statistics program, or recently in graphics file formats only (.jpg, etc.), minimum 300 dpi. No material in entered the job world. To read more articles like this one, visit the website WordPerfect will be accepted. www.stattrack.amstat.org. If you have suggestions for future articles, or would like Amstat News (ISSN 0163-9617) is published monthly by the American to submit an article, please email Megan Murphy, Amstat News managing editor, Statistical Association, 732 North Washington Street, Alexandria VA 22314- at [email protected]. 1943 USA. Periodicals postage paid at Alexandria, Virginia, and additional mailing offices. POSTMASTER: Send address changes to Amstat News, 732 North Washington Street, Alexandria VA 22314-1943 USA. Send Canadian address changes to APC, PO Box 503, RPO West Beaver Creek, Rich Hill, Contributing Editor ON L4B 4R6. Annual subscriptions are $50 per year for nonmembers. Amstat Lee Richardson just graduated from the University of Washington News is the member publication of the ASA. For annual membership rates, with a BS in mathematics and statistics. He worked a year see www.amstat.org/join or contact ASA Member Services at (888) 231-3473. for the Institute of Health Metrics and Evaluation and is now a American Statistical Association PhD student in the department of statistics at Carnegie Mellon. 732 North Washington Street Alexandria, VA 22314–1943 USA (703) 684–1221 • FAX: (703) 684-2037 Richardson ASA GENERAL: [email protected] ADDRESS CHANGES: [email protected] AMSTAT EDITORIAL: [email protected] 18 175 ADVERTISING: [email protected] Celebrating the Impact of Our Profession WEBSITE: http://magazine.amstat.org Printed in USA © 2014 This year marks the ASA's 175th birthday. To celebrate, the column "175"—written American Statistical Association by members of the ASA's 175th Anniversary Steering Committee and other ASA members—will chronicle the theme chosen for the celebration, status of prepara- tions, activities to take place, and—best yet—how you can get involved.

Contributing Editor Promoting the Practice and Profession of Statistics® Fred Hulting is the director of Global Knowledge Services at General Mills, Inc. He holds a PhD in statistics from Iowa State University and has served the ASA in a variety of roles, including The American Statistical Association is the world’s largest section, chapter, and committee chair. community of statisticians. The ASA supports excellence in the development, application, and dissemination of statistical science through meetings, publications, membership services, Hulting 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. departments 19 education Online Article 'Getting Much Further, Much Faster' The following article in this issue can be found 20 meetings online at http://magazine.amstat.org. ASA Gears Up for CSP

25 To increase student participation in statistical statistician's view conferences and other professional activities, the Statistics Losing Ground to Computer Science ASA’s Quality and Productivity Section (Q&P) initiated a student scholarship program to attend the Joint Statistical Meetings. Q&P will offer up to three travel awards of $400 each for students enrolled in a graduate program with a concentration in applied statistics and/or quality member news management to attend JSM in Seattle, Washington, 27 People News from August 8–13. Complete information about the award and how to apply is posted at http:// 29 Awards and Deadlines community.amstat.org/QP/ScholarshipsAwards/ JSMStudentTravelAwards. Sirken Lecture From left: Nat Schenker, Ron Congratulations to Yuqi Chen, the winner Wasserstein, and of the ASA’s Trivia Challenge for the third and final Monroe Sirken. Sirken provided the quarter! association with an Thanks to all who played and helped us celebrate endowment fund our 175th anniversary. to recognize a dis- tinguished survey researcher. Make the most of your ASA membership Visit the ASA Members Only site: www.amstat. 34 Professional Opportunities org/membersonly.

Visit the ASA Calendar of Events, an online database of statistical happenings across the globe. Announcements are accepted from educational Follow us on Twitter @AmstatNews and not-for-profit organizations. To view the complete list of statistics meetings and workshops, Join the ASA Community visit www.amstat.org/dateline. http://community.amstat.org/Home

Like us on Facebook www.facebook.com/AmstatNews

2 amstat news november 2014 president's corner

The ASA and Big Data: What’s Been Happening Lately?

uring the past year, the The statisticians engaged with several dozen people, mostly ASA has engaged in a in interdisciplinary research young people, working in Big number of activities to involving Big Data will need Data enterprises. Dstrengthen the role of our profes- to be computationally savvy, The voices we heard comple- sion in the areas of Big Data and possessing expertise in ment the messages of the ASA data science. This column sum- statistical principles and an white paper with a business per- marizes some of those activities, understanding of algorithmic spective on the skills needed by with an emphasis on important complexity, computational statisticians. Participants in the steps we have taken in education cost, basic computer meetings noted that businesses and professional development. architecture, and the basics need people who can “jump in Three teams, each composed of both software engineering and swim with data” at a time of about a dozen ASA members principles and handling/ when both the volumes and com- and facilitated by ASA Science management of large-scale plexity of their data are increas- Policy Director Steve Pierson, data. Just as critical—for both ing rapidly. Statisticians need have created a series of white the training and recruitment— the skills to understand business papers (http://bit.ly/1DEBY0d). as the computational skills are problems and find solutions using interpersonal skills for the next These papers are aimed at data, but—even more impor- Nat Schenker research funding agencies, mak- generation of statisticians to tantly—the leadership ability to ing clear that statisticians are be effective communicators, “make it to the middle,” serving vital partners in advancing sci- leaders, and team members. as a bridge between highly techni- ence with their expertise in study cal employees and business man- design, inference, and quantify- To broaden our thinking, agement–oriented employees. In ing uncertainty. The papers artic- the ASA is conducting meetings summary, desirable traits include ulate the distinctive input statisti- around the country with leaders in strong skills in statistical methods, cians can provide to help tackle data science and users of Big Data. data collection and management, our nation’s critical research pri- It is extremely important to listen and computing; understanding orities. One of the papers (http:// to voices outside of our profession, important applied problems and bit.ly/1mBUuRD) discusses how so we have intentionally sought out how to use data to solve them; statistics can contribute to solving people in different business sectors. communication; and leadership. problems involving Big Data in In Cincinnati, Ohio, we met with What has the ASA been doing applications to business, govern- leaders in large companies who to help statisticians enhance their ment, health, science, and societal hire master’s-level statisticians for skills in these important areas? issues. This paper emphasizes that business analytics. In Palo Alto, We introduced short courses in work on such problems is most California, we heard from lead- the analysis of unstructured tex- effective if it is multidisciplinary, ers in high-tech companies who tual data at the 2014 Conference involving statisticians, computer hire doctoral-level statisticians for on Statistical Practice (CSP) scientists, mathematicians, and specialized problem solving. In (http://bit.ly/1rqpXTD) and Joint relevant domain scientists. The New York City, ASA Executive Statistical Meetings (JSM) (http:// authors conclude the following: Director Ron Wasserstein met bit.ly/12bXhZN), in addition to

november 2014 amstat news 3 Our profession has a great deal to offer to solving problems involving Big Data,

and the ASA will continue its efforts to scientific method and its relation to the statistical problem-solving cycle,” “real applications,” “focus attract data scientists to the Big Tent on problem solving,” and “the importance of data science,” are strongly related to skills empha- for Statistics and prepare statistical sized in our meetings with busi- ness leaders and in the white paper cited above. scientists for success in these areas. The ASA has taken on the responsibility of coordinating DataFest (http://bit.ly/1wfKwdj) activities around the United States at the request of the orga- the many other member-initiated 2016. We also revised our course nizers. DataFest is a data analysis short courses and sessions in Big on giving effective presentations competition for teams of under- Data, data mining, etc. We plan (http://bit.ly/1woq8TW) and graduates, who devote a weekend to offer more of these kinds of rolled it out at the 2014 Fall to analyzing a large and complex courses in the future. Such activi- Technical Conference (http:// data set, looking for insight and ties will help attendees strengthen bit.ly/1sPnDcA). The offerings learning to communicate what their ability to “swim with” com- in personal skills development they learn. This exciting program plex data. will help attendees improve their encourages collaboration among We launched a personal skills problem solving and communi- students from different disciplines development program (http://bit. cations skills and to and helps students ly/1tbWu6p). This program was “get to the middle,” develop skills they’ll designed to enhance the ASA’s thus enhancing their need in graduate offerings in the area of profes- influence. school and in the sional development and will We are updat- work force. For stu- include courses, workshops, and ing the ASA’s Guidelines for dent perspectives on DataFest, see other training in effective com- Undergraduate Programs in the September 2014 President’s munication, collaboration, lead- Statistical Science (http://bit. Corner (http://bit.ly/1D1RMIq). ership, and career planning. In ly/12jrh6b), the current ver- Our profession has a great particular, we developed training sion (http://bit.ly/1rfUFj6) of deal to offer to solving problems in statistical leadership (http:// which was written in 2000. involving Big Data, and the bit.ly/ZJtea5) and presented The draft updated guidelines, ASA will continue its efforts to a JSM course titled Preparing available from www.amstat.org/ attract data scientists to the Big Statisticians for Leadership: education/curriculumguidelines. Tent for Statistics and prepare How to See the Big Picture and cfm, list guiding principles. In statistical scientists for success Have More Influence (http://bit. particular, the first four, “the in these areas. ly/1DAHUra). Based on the suc- cess of this course at JSM 2014, future leadership training will be pursued for JSM 2015 and CSP

4 amstat news november 2014 ASA in Search This month in ASA’s history... NOVEMBER of Editors

he American Statistical Association and 1839 its publishing partners are searching The ASA was formed at a meeting in for editors for the following journals. the rooms of the American Education TPlease go to the link for each journal for more Society in Boston and was chartered by the information about nomination and application Commonwealth of Massachusetts. Present deadlines, as well as for more information about at the organizing meeting were William each editorship. Cogswell, teacher, fundraiser for the ministry, and genealogist; Richard Fletcher, lawyer and U.S. representative; John Dix Fisher, physi- • Journal of the American Statistical Association, cian and pioneer in medical reform; Oliver Applications and Case Studies http://bit.ly/ Peabody, lawyer, clergyman, poet, and editor; ZqW4M8 and Lemuel Shattuck, statistician, genealogist, publisher, and author of perhaps the most • Journal of Agricultural, Biological, and significant single document in the history of Environmental Statistics (published by Springer public health to that date. for the International Biometric Society and the ASA) http://bit.ly/1Dad2xq 1964 • Journal of Business & Economic Statistics (Co-Editors) http://bit.ly/1z8o08c On November 27, 1964, the American Statistical Association celebrated its 125th anniversary in • Journal of Computational and Graphical Boston. The one-day meeting started with after- Statistics (published with the Institute of noon sessions, including a talk by John Tukey: “The Technical Tools of Statistics: Past, Present, Mathematical Statistics and the Interface and Future.” An after-dinner session was chaired by Foundation) http://bit.ly/1trIZA8 Frederick Mosteller, and the cost to attend the entire one-day event was $15. • Journal of Nonparametric Statistics (published by Taylor & Francis) http://bit.ly/1Cil0SS • Journal of Statistics Education, www.amstat.org/ 1989 publications During the March 1988 board • Statistical Analysis and Data Mining meeting, the ASA Board of (published by Wiley) http://bit.ly/1CLfqZA Directors passed a motion to unanimously designate November • Technometrics (published with the American 1989 as National Statistics Month. Famous Society for Quality). Please note: The deadline To commemorate the sesqui- November for nominations for Technometrics is November centennial and in conjunction with National Statistics Month, a Birthdays 15, 2014. The deadline for applications is symposium called “Statistics—A William Farr, November 30. http://bit.ly/1Fr3LDf Guide to Assessing Societal Risk” George Gallup, was held on November 29, 1989, Erich Leo • Journal of Survey Statistics and Methodology at the National Academy (published by Oxford University Press for of Sciences complex Lehmann, and the American Association for Public Opinion in Washington, DC. Leonard Jimmie Research and the ASA) http://bit.ly/1zhVzVx Paul Meier, James Savage Ware, Lester Lave, and Peter Preuss Nominations and applications for all editorships were present. should be sent to [email protected]. With the exception of Technometrics (see above), the deadline for nominations is December 9. Applications are due no later than February 15, 2015.

november 2014 amstat news 5 ASA President Nat Schenker (left) and ASA Executive Director Ron Wasserstein Highlights of the August 2014 ASA Board of Directors Meeting

SA President Nat Schenker led the board the 175th anniversary, will help make high- through a 175th anniversary–oriented agenda school and undergraduate students, and those during its annual meeting at JSM in Boston who influence them, aware of the great oppor- AAugust 1–2, 2014. tunities presented by a career in statistics. Here are the highlights of the meeting. The board: • Welcomed the past president of the Royal • Approved the ASA budget for 2015. Statistical Society (RSS), John Pullinger, who is the National Statistician of the UK Statistics • Received the report of the ASA treasurer, Authority. Pullinger updated the ASA on the Mingxiu Hu, on the status of ASA invest- strategic planning and subsequent activities of ments, including observations about the RSS, which is a major partner of the ASA. investment policy, and the report of ASA While the ASA celebrates its 175th anniver- Associate Executive Director and Director sary, RSS is celebrating number 180. of Operations Steve Porzio on the mid-year financial status of the ASA. Porzio reported • Discussed the site for JSM 2021. A final deci- that our balance sheet is good and that we sion will be reached in the fall. The ASA plans should end in the black for 2014. JSM sites seven years out to be able to obtain the kinds of facilities we need in the time frame • Was briefed by ASA PR Coordinator we need them. Jeff Myers on the status of the ASA’s national campaign, “This Is Statistics” • Approved creation of the Monroe G. Sirken (http://bit.ly/ZGGD2y). This campaign, Award in Interdisciplinary Survey Research launched as one of the premier initiatives of Methods (http://bit.ly/1D1SRQq)

6 amstat news november 2014 • Approved the creation of an advisory com- mittee to guide the ASA’s new Personal Skills 2014 ASA Board of Directors Development Program (http://bit.ly/1tbWu6p). Nat Schenker (President) • Approved taking on the responsibility of serving as the umbrella organization for DataFest, an David Morganstein (President-elect) undergraduate Big Data analysis competition. Marie Davidian (Past President) • Heard a report on the plans for more fully Martha Gardner (3rd-Year Vice President, creating an ASA development program, replacing David Morganstein) including hiring a development officer. Jim Rosenberger (2nd-Year Vice • Greeted Alan Karr for the final time in his President) capacity as National Institute of Statistical Sciences (NISS) director. As he has done Jeri Mulrow (1st-Year Vice President) annually for two decades, Karr reported on David Banks (Publications the status of NISS activities. Representative) • Heard news of substantial progress being Ming-Yen Cheng (International made on the various strategic initiatives for Representative) 2014, including the development of a statis- Nick Horton (3rd-Year Council of tical leadership course, the creation of a “sta- tistical commons” (http://bit.ly/1pASw19), Chapters Representative) and the updating of the undergraduate Mary Kwasny (2nd-Year Council of curriculum in statistics. In addition, Chapters Representative) President-elect David Morganstein reported on four initiatives he has developed in con- Dan Jeske (1st-Year Council of Chapters sultation with the board. Details of these Representative) initiatives will be provided in future issues Janet Buckingham (3rd-Year Council of Amstat News. of Sections Representative) • Was updated by ASA Director of Science Policy Dick De Veaux (2nd-Year Council Steve Pierson on the wide range of advocacy of Sections Representative) efforts by the ASA and similarly updated by ASA Director of Education Rebecca Nichols Cyndy Long (1st-Year Council of Sections on the gamut of education efforts in which our Representative) association engages. Mingxiu Hu (Treasurer) • Discussed the status and activities of the Ron Wasserstein (Executive Director) committees in the Professional Issues and Visibility Council. These discussions help keep the board and ASA committees contact for members with their society and operating in sync. provide rich networking and professional development opportunities. • As always, heard the important reports of activities of the ASA’s sections and chapters • The board’s final meeting of 2014 is from representatives of their respective coun- November 14–15 at the ASA office in cils. These entities represent major points of Alexandria, Virginia. n

november 2014 amstat news 7 Committee on Nominations Nominations Wanted for ASA President-Elect, Vice President, International Representative Candidates Marilyn M. Seastrom, Committee on Nominations Chair

he ASA needs your help. at the table where appropriate, elect and vice president by email- Nominations are being whether it is academia; business ing Committee on Nominations sought for ASA presi- or industry; research firms; fed- Chair Marilyn Seastrom at Tdent-elect, vice president, and eral, state, or local government; [email protected] or ASA Questions & international representative can- or nonprofit organizations. This Executive Secretary Pamela Craven suggestions didates for the 2016 election year. is why we need your input. at [email protected]. Forward as may be Although the 2015 elections have The president-elect will be much information about your directed not yet been held, the Committee nominated from industry and the nominee as possible (though this to Marilyn on Nominations needs time to vice president from government is optional). We will research each Seastrom evaluate nominations to propose (federal, state, local, or foreign). candidate thoroughly and discrete- at Marilyn. the best possible slate of candi- Think about your colleagues and ly. The deadline for the nomina- Seastrom@ dates in 2016. associates who are members of tions for these two ASA positions ed.gov. As a member of the ASA, the ASA and would make good is February 1, 2015. you recognize the importance of candidates for these positions. Nominees for the Inter­ leadership in our diverse, com- Think about members who have national Representative position plex, and multidisciplinary field. helped run a conference or are must reside outside the United You and all fellow ASA members active in your section or chapter. States. Deadline for nominees deserve visionary leaders who can Then, nominate your choices for for this position is also February ensure our discipline has a voice the 2016 elections of president- 1, 2015. n

8 amstat news november 2014 SAMSI Offers Two New Research Programs for 2014–2015

he Statistical and Applied is fundamentally based upon forensic scientists about the Mathematical Sciences statistical comparisons of the mathematical sciences. Institute (SAMSI) has characteristics of a material Financial support is available Ttwo new programs for 2015– left at a crime scene to charac- for visiting researchers to be resi- 2016. They are Challenges in teristics of a source or suspect. dent at SAMSI for periods of one Computational Neuroscience These comparisons are often month to one year. Postdoctoral (CCNS) and Statistics and acknowledged by forensic sci- positions are available in both More online Applied Mathematics in Forensic entists to be highly subjec- programs and give many oppor- To find out Science (Forensics). ASA mem- tive. A series of reports by the tunities for fellows to collaborate more about bers who join the programs National Research Council with senior-level researchers. either of these will have the opportunity to (NRC) has raised deep ques- Workshops and working groups research collaborate with people in dis- tions about major forms of give many people the opportu- programs, or ciplines such as neurobiology, forensic evidence and has made nity to collaborate with others to apply, go applied mathematics, and foren- a clear case for heeding statis- on research projects and network to www.samsi. sic science. tical underpinning for foren- with their peers. info. sic procedures. These include SAMSI offers workshops to CCNS fingerprints, patterns and graduate and upper-level under- The CCNS program will devel- impressions (footprints and tire graduate students who want to op mathematical and statisti- tracks), toolmarks and firearms, learn about the latest research cal methods for neuroscience hair, fibers, documents, paints and applications in the statistical applications. These will be used and coatings, bloodstains, and and mathematical sciences. All to understand the underlying fire debris. Working groups will involved researchers will have mechanisms that bridge multi- focus on statistical issues for chances to broaden their inter- ple spatial and temporal scales, pattern evidence, bias, imaging, ests and skill sets, participate in linking the activity of individu- and quality control for forensics cutting-edge interdisciplinary al components (e.g., molecular laboratories. Crosscutting chal- projects, and make new con- biology, genetics, and neuron lenges are identifying where sta- nections. New researchers and networks), and their interac- tistics can have a quick impact members of under-represented tions to the complex dynamic and educating mathematical groups are especially encouraged behavior of the brain and ner- scientists about forensics and to participate. n vous system. Brain theory, modeling, and statistics will be essential to turn data into better AAPOR Releases Report understanding of the brain. The CCNS program will address the on Survey Refusals underlying methodological, In response to the steady decrease in the public’s participation theoretical, and computational in surveys over the past 25 years, the American Association for challenges. Probability and sta- Public Opinion Research (AAPOR) has released a study called tistics, dynamic systems, geom- Current Knowledge and Considerations Regarding Survey etry, and computer science will Refusals. It is a comprehensive report on current knowledge be combined with respect to about refusals in survey research—the impact, techniques to theory and in applications. minimize refusals, and ethical considerations for the rights of respondents with regard to survey refusals. The report is a Forensics resource for survey researchers, institutional review boards, SAMSI’s program on foren- consumers of research, journalists, and the public. sics is focused on strengthen- ing the statistical and applied More online: Download a PDF of mathematical bases of foren- the report at http://bit.ly/11kC1AJ. sic science. Forensic science

november 2014 amstat news 9 Meet Joseph Reilly, Administrator of NASS Amstat News invited the administrator of the U.S. and standardized processes that Department of Agriculture’s National Agricultural Statistics will make us more systematic and Service, Joseph T. Reilly, to respond to the following ques- efficient in our work. We are still in transition, so our entire leader- tions so readers could learn more about him and the agency ship team is evaluating where we he leads. Look for other statistical agency head interviews in are in the process and prioritiz- past and forthcoming issues. ing our next steps. Also, from an internal perspective, I mentioned What about this position ing the way we do our statistical the need to focus on the needs of appealed to you? work. When complete, we will be our staff—not only providing the Having worked with the U.S. much more efficient and flexible tools they need, but also the work- Department of Agriculture’s in meeting the needs of a wide ing environment that is conducive National Agricultural Statistics variety of program delivery. For to them. This involves a different Service (NASS) since 1997 in vari- example, we are starting to col- type of transformation. Newer ous leadership roles, I have devel- lect data or to expand existing efforts include more telework Prior to his tenure oped a true passion for the agency, programs on bees, conservation and flexible work arrangements, at the National our staff, and the service we pro- practices, urban agriculture and enhanced career path and advance- Agricultural vide. As the source of official U.S. current agricultural industrial ment opportunities, and more. Statistics Service, Joseph Reilly was agricultural production statistics reports to keep up with changing Externally, we always have an at the U.S. Census and a part of the federal statistical policy and information needs. In eye to the needs of those who Bureau for more system, we provide high-quality, a practical sense, this means pro- provide data by responding to than 21 years. objective data as a public service to viding the leadership that gives our surveys and those who use While there in everyone interested in agriculture, NASS staff the tools and work- our data. In the big picture, we 1997, he received the department’s while always keeping the identity of ing environment they need to do are working on modernizing highest award for farmers and ranchers who provide their jobs. At the same time, it and streamlining how respon- his outstanding their information confidential. means being as proactive as pos- dents provide information so it management At the same time, we are unique sible in anticipating needs of the is easier for them and the result- of the Census because many of our employees people who use the data to serve ing statistics are more complete of Agriculture program. He is a come from farms and we work agriculture, whether they be agri- and accurate. In particular, we 1975 graduate of directly with America’s farmers, business, land grant universities, are focusing on various forms of The Pennsylvania ranchers, and producers, which Congress, USDA agencies, or electronic data response, from State University keeps us well grounded. It is a huge producers themselves. using tablets to collect data with degrees and exciting responsibility, and I Taking on the responsibili- in the field to simplifying our in statistics and marketing. think everyone at NASS is proud ties of administrator of this great online response mechanism, and to be in service to U.S. agriculture. agency and working with such tal- looking at the use of previously It is also an exciting time of ented staff to modernize the work reported data and administrative change for the agency and for we do in service to America’s agri- information to reduce the bur- U.S. agriculture. It is incumbent culture sector holds great appeal den on respondents. On the data upon us to keep up to date with to me and is an honor. user’s side, we want to make our best statistical practices, trends in data as accessible as possible. To agriculture, and the needs of data Describe the top 2–3 priori- that end, we’ve recently added a users within budget and regulato- ties you have for the National simplified query tool as part of ry frameworks provided to us. To Agricultural Statistics Service. our Quick Stats database and those ends, former Administrator Our top internal priority is to are providing some documents Cynthia Clark and I, as associate complete our reorganization. In that summarize large amounts of administrator, began a reorgani- our new regionalized structure, data, such as the recent Census of zation in 2012 that is transform- we are using more centralized Agriculture results, into smaller

10 amstat news november 2014 Taking on the responsibilities of administrator of topical publications. We’d like to hear more about how we can make the agricultural statistics this great agency and working with such talented we produce even more useful. What do you see as your biggest staff to modernize the work we do ... is an honor. challenge for NASS? I think there are a number of chal- lenges we face. Each has the poten- tial to be the most important at any given time. One is the uncer- and graduates. The same may be Census of Agriculture, which is tain budget process in recent years. said for labor, energy, or other one of our largest data-collection As an organization that produces fields. And there seem to be fewer and publication efforts. I com- statistics based on a calendar driv- opportunities to learn this subject mend NASS staff for handling en by sometimes-long data-collec- matter in an office setting. this major undertaking while tion lead times and crop-growing also tackling restructuring, push- cycles, we have difficulty conduct- Prior to your tenure as ing forward our standardiza- ing surveys and producing the administrator, what do you tion efforts, and continuing to public data that U.S. agriculture see as the biggest recent produce more than 400 of our and other data users expect when accomplishment of the agency? regularly scheduled surveys and we don’t get a budget on time. reports. It is a tremendous success Another challenge is the This is an easy one. We just spent story for our agency that we can decreasing understanding of the five years working on the 2012 build on going forward. n importance of responding to sur- veys among our potential respon- dents. We have a great oppor- tunity to raise this awareness by showing the value of the data and how it is and can be used to their benefit, as well as dispelling mis- information about negative uses of data. Declining response rates is a problem across the statistical community and something we as a field are exploring. What kind of support from the statistical community do you look for? Overall, I think we have really strong networks and a profes- sional community to advance our field. If I were to point to one area, it would probably be to see more emphasis placed on applied statistics in colleges. For example, in the case of agricultur- al statistics, there aren’t as many students with farm backgrounds coming into statistical programs so necessary subject matter exper- tise is lacking in statistics majors

november 2014 amstat news 11 Using the Classroom to Bring Big Data to Statistical Agencies Ron Jarmin, Julia Lane, and Alan Marco, with Ian Foster

new statistical measures of innovation. The task is interesting and challenging. The innovation system is complex and requires thinking hard about how to measure the creation, transmission, and adoption of knowledge using new types of transactional data on innovation activity. Since human beings are key fac- tors in innovation—J. Robert Oppenheimer point- ed out that the best way to send knowledge was to wrap it up in a human being—statistical agencies need new and better data on the supply chain feed- ing innovation, on developing measures of networks as potential new units of analysis, and on linking those networks to economic outcomes. Participants in the class used data from research universities to make progress on these issues. The particular focus was on the way in which scientific ideas generated from funded research projects move to the economic marketplace. This involved using Figure 1: Schematic of course syllabus several aspects of Big Data, including:

ig Data have the potential to fundamentally (A) Describing the subject areas of research change the way federal statistics are produced funding using natural language processing and disseminated. Most obviously, data types and machine-learning techniques, rather than Band coverage, timeliness, usability, and, of course, manual coding cost. If statistical agencies are to retain their unique role as the trusted source of statistics, they must (B) Describing the production function reconsider the way data are being generated and of scientific innovation, using complex used, lest they produce data that fail to meet user administrative records from the UMETRICS expectations and/or are too costly. Fortunately, stat- program (www.cic.net/projects/umetrics) to isticians are responding to this call to action. describe the networks of scientists and the In June 2013, Amstat News published a position purchases of scientific inputs for each project paper from the leaders of the American Statistical Association and the American Association for (C) Associating research funding received by Public Opinion Research has established a task principal investigators with their subsequent force to produce recommendations for the survey activities using disambiguation techniques and industry’s use of Big Data. There is also substantial investigating the quality of information that funding for research across federal agencies (e.g., could be scraped from the web or generated the White House Big Data initiative, http://1.usa. from existing data (e.g., web-scraped CV data, gov/1tbZ0K3). However, it is difficult to make use API data from the Public Library of Science of Big Data in the federal statistical context. Staff about Facebook postings and Twitter feeds, in federal statistical agencies are stretched thin pro- USPTO databases, and Census business data) ducing existing products, and concepts developed by computer scientists are often not easily translated The work was preliminary, was for statistical into a social science framework. purposes only, and all confidentiality protocols To meet these challenges, we developed a class were observed. for the U.S. Census Bureau and U.S. Patent and The analysis of university financial transaction Trademark Office (USPTO) to bring Big Data tools data provided new insights into the supply chain to bear on a concrete measurement task: developing for scientific research, and into the Census Bureau’s

12 amstat news november 2014 seasonality assumptions. One team (Grant Degler, combined with social science. For example, Figure 3: Brian Dumbacher, and Sharon O’Donnell (Census); although patents are often seen as a tangible mea- Visualizing the Evangelos Xydis (USPTO)) mapped the location of sure of innovation resulting from federal fund- time between suppliers for one large research university and found ing, the direction of causality may not be clear. patents and they were distributed across one-third of all U.S. Does funding go to researchers who have previ- federal funding counties. The team also exploited the fact that the ously patented, or does funding create research- Source: UMETRICS transaction data were logged daily to evaluate various ers who subsequently patent? The team (Richard time series assumptions used in the Census Bureau’s Watson, Joshua Tokle, Daniel Whitehead, and seasonal adjustments. Lindsay Longsine (Census Bureau); Amanda The students found it was possible to characterize Myers (USPTO)) examined the timing of the link dynamic scientific networks—and use them as a unit between funding to all scientists on funded proj- of analysis. One group of students (Chandra Erdman, ects and patents; they found mixed results in estab- Star Ying, and Mary Pritts (Census); Gerard Torres and lishing causality. The green lines in Figure 3 show Richard Miller (USPTO)) examined how National events where awards precede patents; orange where Science Foundation–funded research centers fostered patents precede awards (the visualization has the the development of such networks. They used machine interesting property that the steeper the line, the learning techniques to examine questions like: How closer the timeline between patenting and award). do the networks of research centers change over time? Research funding is also anecdotally tied to What kinds of people—particularly graduate students, business startups. (Kimberly Wortmann, Nick postdoctoral fellows, and undergraduate students— Zolas, Nathan Goldschlag, and Quentin Brummet enter and leave the network? What do the collabora- (Census Bureau); Chris Leithiser (USPTO)) used tions inside and outside the center look like? How do employment data from 62 nanotechnology grants the associated vendors change over time? They found to three large research universities—with more than that collaborations drop significantly after the cen- 1,000 faculty, students, and postdocs—and linked ter funding ends (Figure 2, found online at http://bit. those data to firm-level data collected by the Census ly/1wosfXI). Another team (Tyler Fox, Kelly Trageser, Bureau and USPTO. They found more than 250 and Randall Knol (Census); Michael Shaver (USPTO)) researchers who either started up non-employer also zoomed in on the networks resulting from National businesses or were involved in start-ups. The major- Science Foundation Engineering Grants. They found ity of nanotech researchers appeared to be graduate that although the research topics and total number of students who started consulting firms. awards varied widely by institution, the academic net- In sum, the Big Data class provided a number of works that existed resulted from relatively few employ- Census Bureau and USPTO employees with new ees who worked on multiple grants. ideas and techniques that can be brought to bear on Another set of students examined the link trenchant measurement issues—in this case, the study between research funding and important innovation of innovation. Our hope is that the approach will outcomes—both patents and business startups— energize changing the way federal statistics are col- not by mechanical methods, but by tying them to lected by empowering and educating the very people scientific networks using Big Data techniques who collect them. n

november 2014 amstat news 13 columns

STATtr@k Applying to Statistics PhD Programs Lee Richardson uring the last several months, I’ve spent the Applying to Statistics PhD Programs large majority of my free time applying to There are four main components graduate admissions statistics PhD programs. After accumulating committees will use to evaluate your application: Dso much information on the topic, I felt it would be • GRE scores a shame not to pass it on to applicants next year and beyond. This post is an attempt to pass on everything • Letters of recommendation I’ve learned about the application process, with hope it will be of value to future applicants. • Grades/transcripts Statistics has risen in popularity over the past sev- • Statement of purpose/CV eral years (http://bit.ly/1oaXSou). Clearly, this is due to the meteoric rise of Nate Silver (just kidding). I Below is an attempt to go through these with would actually say it’s a multitude of factors: finan- as much helpful information as possible. This cial prospects, media coverage, and, of course, sexi- will include tangents of other things I thought ness (http://bit.ly/1qwY7FD). Whatever the reasons, were important. Editor’s Note: the implication for aspiring PhD students is that Reposted with the competition is growing. I remember first asking Statement of Purpose permission from graduate students last year about my chances, and Something I found controversial was the debate Lee Richardson’s blog, Statistical the consensus was, “Of course you’ll get in every- over the statement of purpose (SOP). See the snip- Signal (http:// where!” Most were genuinely surprised by how pet of a Twitter conversation I had with Julian statistical many departments accepted them. Wolfson, a professor at the University of Minnesota signal.com/?p=5) While this may have been true in the past, a Department of Biostatistics. deep look into the epic Grad Cafe admissions This isn’t just shameless self-promotion of my thread (http://bit.ly/1DdlImP) tells a different story. Twitter handle, there’s serious debate (see http://bit. There are qualified students being rejected from the ly/1DdlKLv) as to how useful the statement of pur- majority of schools they apply to, and this trend pose is. I’m a little biased here, because I wasn’t aware does not appear to be reversing anytime soon. of this until I spent a lot of time on the SOP, but I do The first question to ask yourself is: Do you want think it’s important! This is the only portion of your to pursue a master’s, one of the newly developed one- application where you can explain things that didn’t year analytic programs (http://bit.ly/1zbuaEF), or a fit anywhere else. The SOP is also an opportunity for PhD? This question could spawn an entirely new dis- committees to assess the applicant’s English, so non- cussion of the pros and cons of all three, but it really native English-speaking applicants should seek out a depends on what you want. native speaker to read their draft before submitting. The key argument against the SOP is that gradu- Attributes of a PhD ate committees are primarily concerned with math- Generally fully funded, 4–5 years ematical ability and research potential, and the SOP Attributes of a Master’s isn’t a great indicator for this. I think there’s merit Generally not funded, 1–2 years to both views, but since the SOP is the most labo- rious process that forces you to think about your I’m dramatically simplifying, but these are two strengths, why you’re applying, etc., there is value pretty significant things to consider. Since I only beyond just increasing the likelihood of acceptance. applied to PhD programs, that’s the realm in which My intuition tells me that some schools my advice will be the most useful. look closer at the SOP than others do. I have

14 amstat news november 2014 columns

Snippet of a Twitter Conversation

my department-specific hunches based on how quickly some schools responded, but this could vary from year to year based on things like com- mittee members, number of students leaving, etc. There’s also the argument that since the number of applications has risen, there’s less time for com- mittees to spend with individual applications. A lot to consider for sure. NSF-GRFP Fellowship Spoiler: Here’s what I think is one of my best pieces of advice. Although it was time consuming, I think the best decision I made was applying for the National Science Foundation Graduate Research Fellowship (www.nsfgrfp.org). The biggest upside is the deadline is much earlier than departmental application deadlines, so it forces you to think hard beforehand. Visit http://bit.ly/1w8Eh7t for a good about why you’re applying to graduate school and summary of best practices if you decide to do it from what type of research you would like to be doing. a related field (yes, that’s the same person who teaches Then, you can write what can easily be adapted into Udacity’s CS101). I invested an insane amount of a SOP. Even if you don’t get the fellowship (I didn’t), time researching professors, reading their papers, and you will have a huge jump-start on finishing your looking closely at the structures of many departments. applications. I came away with a polished, general I found this to be very useful, but it’s my no means template that worked as an SOP for nearly all my necessary; there are indeed people who got into great applications (only Michigan and Berkeley required programs without emailing anybody. The reason this personal/diversity statements beyond the SOP). is related to the NSF GRFP fellowship is that when Another significant benefit of the NSF GRFP is you’re putting together your research statement, you’ll that it requires a research statement, which means be reading papers of professors, and, naturally, those you have to think deeply about what type of statisti- are the one’s you’re most likely to contact. cal research you will be doing in graduate school. You might be thinking, “What the hell is statis- CV/Research Experience tics research?!” Don’t worry, you’re not alone, and Another requirement is to submit a CV, which has it’s perfectly all right if your research statement is no real defined structure, but a vague Google search bad. In fact, one of the most retrospectively funny can lead you to many templates/examples. The key moments of my application process was meeting part for me was being able to talk about research with one of my professors to discuss my application experience and ASA membership and deliberately a couple days after it was due. He remarked, “Your highlighting the best parts of my application. There’s personal statement is good, but your research state- not a lot of unique information in the CV that isn’t ment needs a lot of work!” Oh well, another plus covered elsewhere, but it gives an opportunity to is that you’ll have two more years to apply. (Note: frame you as attractively as possible. There are more fellowships than the NSFGRFP This leads to another conversation regarding fellowship, Such as the National Defense Science research experience. It can be comforting to know & Engineering Graduate Fellowship (https://ndseg. that some say having real, statistical/mathemati- asee.org) and the Hertz Foundation (www.hertz cal research experience before attending graduate foundation.org). school is rare and not strictly necessary. While The research statement ties into a related debate it’s perfectly acceptable to enter without research regarding the importance of contacting professors experience, it’s obviously a plus if you do. If you’re

november 2014 amstat news 15 columns

in your early years of undergraduate study, I high- as possible and get the highest grades you can (see ly encourage seeking out research experiences at http://bit.ly/1w9qqzF). It’s not strictly necessary to your institution or elsewhere through an REU (see be a math/stat/CS major, as Kristin Linn famous- http://1.usa.gov/1wJGwOH) or job. The real ben- ly majored in music before getting into the PhD efit of doing research before applying to graduate statistics program at NC State (see www4.ncsu. school is trying it out and seeing if it’s something edu/~kalinn), but it will boost your chances. you enjoy. You don’t want to plunge into a PhD The most common classes you see as strictly program not knowing whether you enjoy research, required are multivariate calculus and linear algebra, because if you don’t, it’s going to be a long ride. and some departments strongly recommend taking a real analysis sequence. If you’re still an undergrad- GRE uate, I recommend taking as many mathematical If you look at the number of applications Stanford proofs courses as possible. In fact, Peter Guttorp receives, you might think “Hey, something isn’t right recommended that I should take the real analysis here.” Well, the reason is that they require the math- sequence in my gap year. ematics GRE test, which weeds out lots of potential applicants. If you’re hell-bent on going to Stanford, Letters of Recommendation then you have to take the mathematics GRE. (Some Letters of recommendation (LOR) are a tricky other schools—like UW, Chicago, and Columbia— topic. It seems you have to be strategically fortunate strongly recommend it.) I didn’t take the exam, to stumble upon good LOR writers. Obviously, the but I’ve heard it is pretty hard and requires a lot of younger you are, the more time you have to develop studying and deep coursework in pure mathemat- a relationship with a professor or boss through a ics. There’s uncertainty around what particular score research project and lock down a good letter testify- would hurt/help you if you include it in your applica- ing you will be successful carrying out research in tion. Stanford posts an average of the 82nd percentile graduate school. That’s the golden standard for an (thanks, wine in coffee cups, http://bit.ly/1EZu3fM), LOC writer. Another thing to do while you’re still but it’s unclear what the distribution of scores are, in the early years of undergrad is participate in an and whether the scores at Stanford are similar to REU. I’ve heard many people get their LORs from scores at other schools that strongly require it. REU supervisors. The most annoying part of applying was the GRE You will hear this ad nauseam, but I’ll reiterate test. You’ll be quizzed on vocabulary words you can that it’s better to get an LOR from someone less effortlessly look up, write timed essays on strange top- well known who knows you well than someone well ics, and answer short quantitative questions. For our known who barely knows you. If you’re a senior and purposes, the most important part is by far the quan- you don’t have anyone who can write a good let- titative section, a significant piece of information ter, my suggestion is to go all out next quarter, be a graduate committees use to assess mathematical abil- totally overachieving student (going to all the office ity. I can understand this; as a statistics department hours), and see if you can connect with a current you probably don’t want to invest in someone who professor. Better yet, it would be good to seek out has evidence of struggling quantitatively. However, advice from people in the department about work- if you’re coming from a relatively quantitative back- ing on an undergraduate research project. Chances ground and take some practice tests, you should be are there are professors in your university looking able to score in the mid-high 160s. The GRE won’t to have some data collected/cleaned/analyzed. It’s get you in anywhere, but it can disqualify you. So the never too late to establish a good relationship with takeaway message is to not bomb it. a professor; however, it does take time and effort. After you take the GRE test, you can send your The LORs are tricky, but they’re also a criti- scores to four departments for free. If you want to cal piece of your application. It can certainly be send your scores to anyone after this, it costs you awkward asking a professor to write one for you, $25 per department. That’s right; the ETS is charg- as professors are very busy people. However, you ing $25 to send about 8 bits of information. The should keep in mind that they are asked for LORs Singularity Is Near. all the time as part of their job. Try to establish as many good relationships as possible, but if it’s not Grades/Transcripts feasible, it’s not the end of the world to default to This is pretty straightforward. Basically, you want professors you’ve simply taken a course with. One to take as many math/stat/computer science classes piece of advice is that, instead of asking whether a

16 amstat news november 2014 columns

professor can write a letter, ask if they can write a A reason PhD programs are so selective is that good letter. By asking directly, you can avoid send- they’re investing money into your potential as a stu- ing unflattering letters to admissions committees. dent, and it’s a waste for them to invest resources in someone who won’t be successful. Odds are certainly Where to Apply higher applying to master’s programs; however, you At a certain point, you’ll have to live with your per- will most likely be footing the bill (although I’ve formance in these categories and start filling out heard of some master’s programs funding students). applications. The natural question to ask is, “Where should I apply?” There are some basic places to start TheGradCafe looking—the most recent graduate department Chances are the majority of your social circle rankings (see http://bit.ly/1jwiQvc) and this insanely will not be simultaneously applying to statistics comprehensive list of all statistics programs (http:// PhD programs. Due to this, you may be seeking bit.ly/1veVZbl)—but I think the most important people eager to talk about the application process. thing is to meet with someone in your department TheGradCafe is a terrific community where fel- (if you’re still an undergraduate) and solicit sugges- low applicants can discuss the application process. tions on where to apply. If you’re not an undergrad- I found it a useful place to discuss various topics uate, an alternative to this is posting your profile relating to application season (as well as express my on TheGradCafe (http://bit.ly/1sRajqQ), and you’ll fears of being rejected everywhere). Also, it’s extraor- probably garner some candid responses. dinarily entertaining to meet people on visits whose One of the biggest mistakes I made while usernames you recognize. applying was never getting an honest assessment from someone inside the process about how com- Costs/Suggested Scholarships petitive my application would be. Per sources at Applying to graduate school is quite costly. You can TheGradCafe (biostat_prof, cyberwulf), the tier get a rough estimate from the following: system is a realistic way to view schools. This basi- cally means that if you’re accepted into one highly 200*(# of GRE Tests) + (25[GRE]+75 ranked school, chances are you will be accepted [App Fee]) *(Number of Schools) into others—the converse is also true (IFF!). I’m somewhat skeptical about this. I do think some Plus countless hours and -5 years of life expec- schools look more in depth at different factors tancy due to stress (sort of kidding). So, if you take than others do, and that they’re looking for stu- the general GRE once and apply to 10 schools, dents who fit their culture. There are applicants you’re looking at roughly $1,200. One thing I’ve who get into one of their reach schools, despite always felt would be a good idea is for undergradu- being rejected by others reaches (yours truly). ate institutions to foot the bill for their students’ The bottom line: You’re not going to get into application fees with something like an “applying anywhere you don’t apply, so I wouldn’t let rank- to a STEM PhD” scholarship. This could easily ings intimidate you if it’s a school you really want already be happening, so my apologies to anyone to attend. who’s already doing it. My personal criterion for applying was: “If you get rejected everywhere except this school, would Post-Application Stress you actually go?” This was retrospectively pretty I didn’t want to make this post about anything risky, and it’s certainly possible I could have been except the application process, but in case I rejected everywhere (everyone’s worst-case sce- don’t revisit this topic: It’s pretty common to nario). If your goal is to be in a PhD program feel as if you’re not going to be accepted every- no matter what, I would suggest having someone where directly after you’ve applied everywhere. give you an objective, candid assessment of which Relax, you probably will, especially if you’ve got- schools you could potentially be accepted into ten honest assessments of where you stand as (described above). an applicant. Applying to PhD programs is just Apply to some reaches, but also apply to schools extremely stressful. you’re fairly confident will accept you. Other If you have specific questions for Richardson than about $100 and some opportunity cost, about applying to a graduate statistics program, visit being rejected from schools doesn’t really carry his blog, Statistical Signal, at http://statisticalsignal. any downside. com/?p=5. n

november 2014 amstat news 17 columns

175 Celebrating the Impact of Our Profession Fred Hulting

he 175th anniversary of the ASA occurs Global Impact—Past, Present, and Future,” this month, and we are wrapping up a for JSM 2014. Many of the StatImpact-themed ses- year of commemoration and celebration sions described significant past accomplishments. Tof this tremendous milestone. The anniversary In addition, several sessions focused specifically theme was “Celebrate Our Past, Energize Our on the future impact of the profession, outlining Future,” and the ASA focused on three efforts in emerging opportunities in medicine, education, support of this theme: StatSharp, StatGrowth, public health, and the sciences. To learn more, visit and StatImpact. www.amstat.org/meetings/jsm/2014. The aim of StatImpact is to highlight how statis- Journals: Currently, The American Statistician ticians have successfully tackled the world’s biggest will publish articles commemorating the anniver- problems and inspire future efforts. The motiva- sary and celebrate the impact of statistics. The ASA tion for StatImpact came from Roger Hoerl’s call to 175th Committee continues to work with other action in his 2011 Deming Lecture. journals, as well. “Given the types of problems the world faces, Outreach: The ASA has been busy with out- such as AIDS … we might ask why statisticians are reach efforts, and several of these initiatives not working on these problems,” Hoerl said. "The focus on communicating our impact. First, the short answer, of course, is that they are. … They are ASA is building collaboration with other scien- doing precisely what I am suggesting. Unfortunately, tific societies. As an example, working with the I don’t think there are enough of them. We need to American Association for the Advancement of do more, and I think we can do more.” Science (AAAS) has led to renewed collaboration As an applied science, statistics is all about with Section U–Statistics, the addition of statis- impact, and the expertise of our profession is tics reviewers to raise the quality of published needed as the world confronts the enormous soci- statistical work, and Science columns focused on etal, medical, environmental, and technological statistics topics. Second, the ASA has published a challenges taking shape. The magnitude of this series of white papers to help drive research fund- opportunity was articulated again by Sallie Keller ing by demonstrating we are vital partners who at JSM 2014 when she asked that we “… [C] bring essential expertise. Finally, a nationwide onsider the health and welfare of the next billion campaign, “This Is Statistics,” has been launched, people on earth. How do we fully understand the centered on thisisstatistics.org. Aimed at students, human condition and accommodate the needs parents, and educators, the site’s main message this group?” is that statistics “is not what you think it is.” It In that spirit, StatImpact set forth several objec- is broader and deeper and more impactful than tives that leverage the ASA’s meetings, journals, many people realize. and public outreach efforts. First, infuse the 2014 All of these StatImpact activities are raising Joint Statistical Meetings with the ideas and inspi- the visibility of the power of statistics. ASA Past ration of StatImpact. Second, publish articles in President Bob Rodriguez often talks to students our 2015 ASA journals that highlight the impact about this power, saying that, as a statistician, “You of our profession. Finally, communicate the power can find the power to remain in demand, the power of our profession as a way to influence science, to become purpose driven, and the power to reach policy, and the growth of the ASA. higher. What could you do with this power?” As JSM 2014: ASA President Nat Schenker’s the StatImpact efforts have demonstrated, we can support for driving the impact of statistics make a tremendous difference in our society with was reflected in his chosen theme of “Statistics: that power. n

18 amstat news november 2014 education

'Getting Much Further, Much Faster' Chris Wild, Data to Insight e all know the universe of data, big and derived from a large observational study (NHANES) otherwise, is growing explosively—in and tackle key questions of development and health volume, in the areas it reaches into, in using 30 country-level indicators over the last Whow it is constituted, and in what you can do with 50 years from Gapminder. For motivational reasons, it. By comparison, changes in what students expe- it catches students up in the potential of data analy- rience are glacial. Statistics education is simply far sis for discovery before being mired in the “limi- too slow. The window we open up on this rapidly tations” swamp. Serious consideration of system- expanding universe is a tiny porthole. We have to atic biases and confounding and random error (all find ways to get much further, much faster—to crucial topics) does not come until the fifth week, open our students’ eyes to an enormously wider which also forms the bridge to bootstrap confidence world of possibility, opportunity, and excitement. intervals (Week 6) and experimentation and ran- Editor’s Note: This is necessary to prepare our students for a domization tests (Week 7). The course concludes Though the rapidly changing world, but the self-interest of stat- (Week 8) with an unrelated, lighter bonus: explor- site is officially isticians also calls for it. We need to adapt to pre- ing sets of time series. closed, it will vent the data domain we have thought of as our stay open own being completely overrun by species that are Strategies Used for Acceleration “silently” until fleeter of foot (most notably computer scientists). The most novel strategies used are the following: December Otherwise, we may expend our last breaths picking • Being intensely visual and driving all argu- 20 for Amstat up a well-deserved Darwin Award. ment off things you can see supplemented by News readers. After signing This year, by agreeing to be my university’s metaphor MOOC guinea pig, I was given the space and pro- up and duction support to prototype an ambitious and • Building software solutions that prevent “how receiving a radical re-thinking of intro statistics—a much- do I get this out of the software?”, limiting the login, you will further-much-faster, more-data-more-quickly, speed at which students can encounter new have access for introductory statistics course that emphasizes situations and new ideas many months data visualization. This new approach was cre- (instructions ated as much for the statistical community as for • Finding some powerful, conceptually unde- at http://tinyurl. the MOOC audience. We have called it “Data to manding “extender-capabilities” that immedi- com/4Stat Insight: A First Course in Data Analysis” (www. ately open much wider horizons Educators). futurelearn.com/courses/data-to-insight). Most of the content is delivered in 42 five-min- Other strategies are more obvious: limiting mes- ute videos. The received wisdom from MOOC sages to just those most critical for real-world learn- analytics is that viewership really starts to drop off ing from data; stripping concepts back to their bar- at about five minutes. It is a daunting discipline (I est bones; and exploiting, feeding, and reshaping am guilty of minor lapses), but I have been amazed primary intuition. Additionally, we use vivid images by what you can get into five minutes if you “don’t (verbal and visual) to make key messages linger. sweat the small stuff.” It really puts the pressure on to convey just the essence of ideas efficiently. Can It Work? The course has an eight-week, “three hours a What can you pack into a total of four hours of week” structure and features just 30 minutes each movies? Can it possibly work? Come and see. A page week of instructional video. But it is not passive. It has been set up at http://tinyurl.com/4StatEducators is a hands-on data analysis course that takes begin- specifically for statisticians and educators. It lets you ners and quickly gets them working with up to five jump straight from a linked course outline to par- numeric and categorical variables simultaneously ticular movies. You are encouraged to sample and in the same analysis (over the first four weeks and “steal” ideas, because this MOOC is a prototype just two “teaching” hours). Students extract mean- built for this community and a course designed to ing from a 10,000-observation, 70-variable data set benefit its students. n

november 2014 amstat news 19 he fourth annual ASA Conference on Statistical Practice will provide attendees opportunities to learn new Tstatistical methodologies and best practices in statistical analysis, design, consulting, and sta- tistical programming. The CSP also will provide opportunities for attendees to further their career development and strengthen relationships in the statistical community. The conference is designed to aid applied statis- ASA ticians in improving their abilities to consult with and help customers and organizations solve real- Gears Up world problems. Look forward to learning the following: for • Statistical techniques that apply to your job as an applied statistician • How to better communicate with customers CSP • How to have a positive impact on your organization

New Orleans, Louisiana February 19-212015

20 amstat news november 2014 Two attendees participate in a poster discussion during last year's conference in Tampa, Florida.

attend are able to research current job openings and contact employers they are interested in pursuing via our online system in advance of the conference. The close-knit atmosphere of CSP is ideal for fostering a mentoring program (www.amstat.org/ meetings/csp/2015/mentoring). Registrants will be The registration given the opportunity to express interest in partici- form and a list of pating as either a mentor or mentee. Just register for short courses, the conference and the ASA will automatically send tutorials, and you the program registration link. There is no cost practical to participate in this program. computing ASA President Nathaniel Schenker gives his CSP keynote Early registration for the conference closes on demonstrations address in 2013 in Tampa, Florida. January 5, 2015. The housing deadline is January can be found on 19, 2015, and the conference registration deadline is the following Once again, the steering committee has orga- February 5, 2015. Our goal is to maintain a cozy, inti- pages. nized a terrific program, with talks, courses, tutori- mate, and personal feel to the conference, so register als, and posters focused on the needs of applied and before all the conference spots and courses are filled. consulting statisticians concerned with urgent prob- It is worth reminding everybody that this confer- lems, issues, and systems for clients and organiza- ence is for you. When you are in a course or attend- tions to improve processes, products, and decisions. ing a session, do not shy away from asking how the ASA President-elect David Morganstein will deliver work presented is applied and how it can help you the keynote address, “Communication: A Two-way be a better statistical practitioner. When you attend Street.” You can check out the program for details the closing feedback session on February 21, speak at www.amstat.org/meetings/csp/2015/onlineprogram. up and let us know how the conference has served Better yet, register for the conference and sign up you, as well as offer suggestions for improvements. for the courses. Visit www.amstat.org/meetings/csp/2015 to The CSP career placement service will con- learn more about CSP 2015. For conference nect employers and applicants so they can pursue updates, follow @AmstatNews on Twitter and use informal meetings and interviews. Applicants who #CSP2015. n

november 2014 amstat news 21 CSP REGISTRATION FORM INSTRUCTIONS: 1. Print or type all information and retain a copy for your records. 2. Use a separate form for each registrant. 3. Mail form with payment to CSP Registration, 732 N. Washington Street, Alexandria, VA 22314. Fax form (credit card only) to (703) 684-2037. Registration form must be received by February 5, 2015. www.amstat.org/meetings/csp/2015 Forms received without payment will not be processed. Purchase orders will not be accepted. No exceptions. ASA Federal ID #53-0204661

ATTENDEE INFORMATION REGISTRATION FEES (required) By Jan. 5 Jan. 6–Feb. 5 ❑ ASA Member $375 $435 $______❑ ASA Member ❑ SSC Member ASA ID # (if known) ❑ New* Member $530 $590 $______❑ Nonmember $570 $630 $______Name ❑ Student $180 $180 $______❑ ASA/SSC Accredited Member ® Preferred First Name for Badge (PStat /GStat/A.Stat.) $300 $300 $______*Includes discounted first-year ASA dues; not available to renewing or recently lapsed members. Organization The ASA reserves the right to adjust member and/or accreditation registration type to an eligible type and to charge the difference if stated membership and/or accreditation is not currently active. In such an event, you will be notified first and given the opportunity to update your membership.

Address ADDITIONAL FEES (optional) Space is limited. Full-Day Short Courses—February 19, 8:00 a.m.–5:30 p.m. Oct. 1–Jan. 5—$330 members and students; $265 PStat®/GStat/A.Stat.; $380 nonmembers Jan. 6–Feb. 5—$375 members and students; $265 PStat®/GStat/A.Stat.; $425 nonmembers City State/Province ZIP/Postal Code ❑ SC1: Practical Data Mining: Challenges and Solutions $______❑ SC2: From Statistical Consultant to Effective Leader $______Country (non-U.S.) Half-Day Short Courses—February 19 Oct. 1–Jan 5—$215 members and students; $170 PStat®/GStat/A.Stat.; $265 nonmembers Jan. 6–Feb. 5—$250 members and students; $170 PStat®/GStat/A.Stat.; $300 nonmembers Phone 8:00 a.m.–12:00 p.m. ❑ SC3: What Can We Learn from Software Engineers? $______Email ❑ SC4: How to Start and Run an Independent Statistical Consulting Business $______❑ SC5: An Overview of Clustering: Finding and Extracting Group Structure in In case of emergency, list the name and phone number of the person we should High-Dimensional Data $______contact (remains confidential). 1:30 p.m.–5:30 p.m. Emergency Contact ______❑ SC6: Building Your Professional Brand $______❑ Update my ASA customer contact information with this meeting contact information. ❑ SC7: Design of Not-Simple Graphs $______❑ Exclude my information from contact lists managed by the ASA for use by outside ❑ SC8: Text Analytics: Integrating Topic, Opinion, and Sentiment Analysis $______entities, including offers for onsite receptions or activities and booth giveaways. ❑ Exclude my name from the conference attendee roster that will appear on the Tutorials—February 21, 2:00 p.m.–4:00 p.m. conference website. Oct. 1–Jan. 5—$65 members and students; $60 PStat®/GStat/A.Stat.; $75 nonmembers Jan. 6–Feb. 5—$80 members and students; $60 PStat®/GStat/A.Stat.; $90 nonmembers This meeting is ADA accessible. ❑ T1: A Case Study in Big Data Analytics $______❑ Please check here if you need special services due to a disability and attach a ❑ T2: An Introductory Tutorial on Mixed Models $______statement regarding your needs. ❑ T3: Speak & Connect: Harnessing PowerPoint $______❑ T4: Tutorial on Parallel Programming in R $______PAYMENT Practical Computing Demos—February 21, 2:00 p.m.–4:00 p.m. ❑ Check/money order enclosed payable to the American Statistical Association (in U.S. dollars Included in registration fee. Pre-registration is requested to ensure proper preparation. on a U.S. bank) ❑ PCD1: Interactive Predictive Modeling with JMP 12 Pro: Keeping It in the Flow Credit Card ❑ American Express ❑ Discover ❑ MasterCard ❑ VISA ❑ PCD2: Tessera: Open Source Tools for Big Data Analysis in R ❑ PCD3: Mathematica and Statistical Computing ❑ PCD4: Rating College Football Teams: A Case Study on Integrating Minitab with Statistical Card Number Programming Languages CAREER SERVICE (optional) Expiration Date Security Code Post your résumé for viewing by interested employers. This is an online-only service. Visit www.amstat.org/meetings/csp/2015 for details. Name of Cardholder ASA Member Nonmember Student ❑ $50 ❑ $75 $______❑ $75 ❑ $150 $______Authorizing Signature Nonstudent GUEST (optional) CANCELLATION POLICY Guest (name______) $95 $______Cancellations received by January 5, 2015, will be refunded, less 20% all items. Requests for refunds received after January 5 will not be honored. All cancellations must be made in writing to [email protected], via fax to Allows admission to the opening mixer, poster sessions/refreshments, and continental breakfast for both (703) 684-2037, or mailed to CSP Registration, 732 N. Washington Street, Alexandria, VA 22314. Friday and Saturday mornings. Admission to other conference sessions and events is not included. MEETING CONDUCT POLICY TOTAL FEES: $______Rules and regulations apply. See www.amstat.org/meetings/csp/2015/conductpolicy for details.

22 amstat news november 2014 Courses

Conference registration is required. For additional information, visit www.amstat.org/meetings/CSP/2015.

SHORT COURSES known outside the discipline. This course will and statistical approaches to clustering with an explain the techniques of software engineering emphasis on how to choose an approach and and how they apply to managing your software. its related parameter. Note that while we use Full-Day Short Courses Topics range from code-level practices to design the statistical software package R, these meth- 8:00 a.m. – 5:30 p.m. issues and project control. The course will focus ods are available on other platforms. on software engineering in the context of R, SC1—Practical Data Mining: which provides a rich environment for statisti- Challenges and Solutions cal programming. Participants are expected to Afternoon Half-Day Short Courses Instructor: Richard De Veaux, Williams College arrive with R and RStudio installed on their lap- 1:30 p.m. – 5:30 p.m. tops. Some familiarity with R is required. Large data sets (or Big Data) are becoming more SC6—Building Your Professional common as our ability to collect and store data increases. Many new tools and methods are SC4—How to Start and Run an Brand now available to both the experienced analyst Instructor: Bill Williams, Organizational Learning and casual user. Unfortunately, there is a strong Independent Statistical Consulting Consultant Business belief—due in large part to a series of popular Big The world of work is full of people with ambi- Data books—that good results are guaranteed Instructor: Stephen Simon, P. Mean Consulting tion and aspirations to do bigger things as with just powerful algorithms and a lot of data. An independent statistical consulting job is both their careers progress. While the rules for suc- Instead, success is dependent on the skill and rewarding and challenging. If you follow this cess—most of which are unwritten—vary from domain knowledge of the analyst and the quality career path, you will need to learn many business organization to organization, two ingredients are and relevance of the data. However, by using prin- skills. This course will review practical issues you always essential: 1) your current performance ciples of statistical engineering and sound statistical will face in setting up an independent consulting on the job and 2) the potential other people knowledge, the chance of success in these prob- business. Should you set up a limited liability cor- see in you. How people view your performance lems is significantly increased. Through a series of poration or a subchapter S corporation? Should and potential is derived only in part by what case studies, we will show how to be successful in you bill by the hour or the project? What insur- you know and the functional expertise you pos- Big Data problems. We will show applications of ance do you need? Should you have a standard sess. The rest is based on the image you project many current and popular algorithms, and when contract in place prior to any consulting work? and the exposure to other people your job and where they are most successful. In addition to these legal and accounting affords you. In this session, we’ll examine both requirements, there are human issues that you the impression you want others to have of you SC2—From Statistical Consultant as an independent consultant will have to face. as a professional—your “brand”—and how your to Effective Leader Your most important job is finding new clients. communications can influence the impressions The best method, by far, is “word of mouth,” and of others. You will define the brand you would Instructors: Ronald Snee, Snee Associates, LLC; there are several strategies you can adopt to most like people to associate with you and con- Roger Hoerl, Union College enhance your visibility and increase the number sider how to manage your behavior to support This workshop is designed to enhance the leader- of referrals you receive. You also need to know your brand, particularly when communicating ship skills of statisticians working in business, indus- how to keep your current clients happy. This with senior managers and leaders. try, and government. The goal is to help statisticians class will include several small-group exercises during which you will share your thoughts and transition from being viewed as passive consultants SC7—Design of Not-Simple Graphs to being viewed as proactive leaders within their experiences on how to handle specific cases organizations. Issues addressed include under- involving independent statistical consulting. No Instructor: Richard Heiberger, Temple University standing what statistical leadership is and how specific knowledge about business models, accounting, or legal issues will be assumed. Complex data analyses may require complex it differs from consulting, why it is important to graphs to place the full information of the analysis be viewed as leaders, and critical leadership skills into a form the intended client will be able to required. As part of the course, each participant SC5—An Overview of Clustering: read. In our opinion, graphs are the heart of most will develop a personal action plan to enhance Finding and Extracting Group statistical analyses; the corresponding tabular their leadership in their own work environment. Structure in High-Dimensional Data results are formal confirmations of our visual impressions. Data analysts are responsible for the Instructors: Rebecca Nugent, Carnegie Mellon display of data with graphs and tables that sum- Morning Half-Day Short Courses University; Samuel Ventura, Carnegie Mellon 8:00 a.m. – 12:00 p.m. University marize and represent the data and the analysis. Clustering is the search for similar or homo- The graphs are often the best means of com- SC3—What Can We Learn from geneous subgroups in a population, say, of munication between the data analyst and client. This course will emphasize the design of graphical Software Engineers? consumers, patients, genes, images, text docu- ments, anything that can possibly contain group displays that best represent the message of an Instructor: Paul Teetor, Quant Development LLC structure. For example, consumers might be analysis. Do any of the following problems sound familiar? divided into market segments based on their Your organization is swimming in SAS or R code. preferences and spending habits. In public health, SC8—Text Analytics: You’ve saved numerous versions because you we might be interested in predicting which Integrating Topic, Opinion, and can’t afford to lose anything. People are unsure of outcome group a patient is likely to be in given Sentiment Analysis which version is best. Testing your code is difficult. their symptoms, past history, and current treat- You’ve cut-and-pasted your code so often you’re Instructor: Edward Jones, Texas A&M ment. In document clustering, the goal is to Statistical Service, LP seeing the same parts over and over. Everyone group similar pieces of text (e.g., blogs, emails, does their work differently, and people can’t posts, letters, articles, etc.) based on the words This workshop discusses current statistical share code easily. The code is now so convo- used, the frequency, and other text features. approaches to conducting a linked analysis of luted that newcomers cannot understand it. The In all cases, the goal is to extract structure reviewer comments, sentiments, and rating. thought of major changes makes your head hurt. from potentially high-dimensional data. The Today, statisticians have powerful tools avail- Software engineers have spent decades difficulty, however, often lies in which cluster- able for integrating the analysis of structured dealing with these problems, and the result is a ing approach to adopt, particularly given that and unstructured data. Reviewer and customer body of best practices for managing software. results are rarely independent of approach. This comments can be used with their ratings and These best practices are an art and not well tutorial will give an overview of algorithmic other background information to build models

november 2014 amstat news 23 Courses

Conference registration is required. For additional information, visit www.amstat.org/meetings/CSP/2015.

linking ratings, opinions, and emotions. Done questions, you should attend this tutorial. Have scientists to explore and analyze large, complex well, this provides a more complete picture of some of your slides been loaded with text, bullet data. The Tessera computational environment what people think and feel about services and points, or complex data? Have there been times is powered by divide and recombine (D&R), products. when you’ve read off your slides? Has your audi- an approach for dividing data into subsets and ence ever looked bored, inattentive, or asleep? computing on them in parallel. At the front end, Learn how to share data and information and the analyst programs in R. At the back end is a TUTORIALS create better decks. distributed parallel computation environment such as Hadoop. In between are three Tessera 2:00 p.m. – 4:00 p.m. T4—Tutorial on Parallel packages: DataDR, Trelliscope, and RHIPE. The DataDR R package provides a high- T1—A Case Study in Big Data Programming in R level interface to D&R operations, making Analytics Instructors: Josh Hewitt, Colorado State Uni- specification of divisions, analytic methods, and versity; Henry Scharf, Colorado State University; recombinations easy. The interface is designed Instructor: Patrick Hall, SAS Enterprise Miner and Miranda Fix, Colorado State University to be back-end agnostic, so it can harness new So what exactly do you do when faced with This tutorial introduces participants to high-per- distributed computing technologies as needed. a huge data set from which you are to derive formance computing in R for analyzing research Trelliscope is a scalable visualization tool in insights? This happens in banks, insurance com- data and developing practical analytics. R is a which data sets are divided into subsets and a panies, government agencies, manufacturing cen- free, open source programming language that visualization method is applied to each subset ters, and other institutions all the time. This tuto- concisely supports a wide range of statistical and shown in a multi-panel trellis display. This rial illustrates best practices for mining large data computing and machine learning needs. Modern framework has proven to be a powerful mecha- sets in the context of a case study. Participants data sets are large and computational proce- nism for all data, large and small. will learn real-world techniques to explore and dures can be intense, which may become pro- RHIPE is the R and Hadoop Integrated preprocess data; to select, extract, and engineer hibitive to practical data analysis projects. This Programming Environment. RHIPE allows an the most predictive features; to build the best tutorial introduces participants to workflows analyst to run Hadoop MapReduce jobs from predictive model for the job at hand; and to and packages that let practitioners use R to take within R. RHIPE is used by DataDR when the leverage predictive analytics to make decisions advantage of the power of modern comput- back end is Hadoop. for their organization. Instructors also will point ing resources like multicore architectures and out common pitfalls and trade-offs inherent cloud technologies. Applications and examples PCD3—Mathematica and ßto contemporary Big Data approaches. SAS include demonstrating parallel forms of popular Enterprise Miner will be used for the analyses, classic and machine learning methods, using Statistical Computing but the focus will be on the methods and not bootstrapping and cross validation to estimate Michael Kelly, Wolfram Research Inc. the software. Participants will have access to the uncertainty and accuracy, simulating data to ana- Mathematica is the world’s leading symbolic example data for further study. lyze “what if” scenarios, and discussing related and numerical software, pioneering the use of topics. Demonstrations will be presented with symbolic functional programming for the repre- T2—An Introductory Tutorial on R. Attendees are encouraged to bring laptops sentation of mathematical, statistical, and com- Mixed Models with R installed so they may follow along and putational objects in a universal consistent high- experiment with these tools. level language that has allowed for a systematic Instructor: Funda Gunes, SAS Institute Inc. treatment of the entire area of statistical analysis. Mixed model analysis is one of the cornerstones Unlike other statistical programs that are mainly of modern statistics. It extends the general linear PRACTICAL COMPUTING numerical, Mathematica combines the many model for independent and equivariant data by advantages of symbolic representation of math- allowing a more flexible covariance for the error DEMONSTRATIONS ematical statistics with the numerical capabilities term. Using mixed models, you can fit models to a 2:00 p.m. – 4:00 p.m. of advanced and novel algorithms. See www. variety of data that follow the normal distribution, wolfram.com/solutions/industry/statistics. including repeated measurements and data from PCD1—Interactive Predictive a randomized block design. This tutorial introduces the basics of mixed model methodology and illus- Modeling with JMP 12 Pro: Keeping It PCD4—Rating College trates the analysis of linear mixed models in typi- in the Flow Football Teams: A Case Study on cal applications, with numerous examples using Mia Stephens and Scott Wise, JMP Division of SAS Integrating Minitab with Statistical the MIXED procedure in SAS/STAT® software. This tutorial also includes an overview of other Interactive predictive modeling in JMP Statistical Programming Languages mixed modeling procedures in SAS®, giving a Software from SAS is more than building mod- Daniel Griffith and Joel Smith, Minitab els. It allows you to take advantage of interactive brief introduction to analyzing generalized linear College football is a sport with highly variable and dynamic graphs and advanced analytic tools, models by using the GLIMMIX procedure and outcomes and teams that play highly unbalanced keeping data visualization, analysis, and modeling discussing the scenarios in which you would use schedules due to conference affiliation, a large in the flow. In this talk, we will use case studies the nonlinear mixed models and NLMIXED pro- pool of potential opponents, and incentives that to see how to explore and prepare data using cedure. Prerequisites are a working knowledge of disfavor competitive balance. Despite these dif- the Column Switcher, Data Filter, Recode, and the general linear model and basic matrix algebra. ficulties, it is highly desirable for fans, media, and Graph Builder. We will use the Partition plat- the playoff selection committee to rate teams form, Fit Model, and Generalized Regression as accurately as possible. Utilizing an unconven- T3—Speak & Connect: Harnessing platforms, as well as tools such as the Prediction tional method, the case study demonstrates how Profiler and the Solution Path in JMP Pro 12, PowerPoint teams can be rated with minimal effect from to interactively explore parameters and select Instructor: Andrew Causley, Ballistic Fish Studios uncontrollable aspects of the game. The method potential models. Finally, we’ll compare a variety is performed using a combination of Minitab Data-heavy presentations can quickly overload an of competing models using Model Comparison. audience with information, causing them to tune Statistical Software for its ease of use and graphi- out. Learn how to create and deliver PowerPoint cal capabilities integrated with a statistical pro- presentations that are interesting, effective, and PCD2 —Tessera: Open Source Tools gramming language for complex routines. memorable! It’s a fresh approach, one that com- for Big Data Analysis in R bines information with effective visuals and per- sonal engagement to connect with an audience Ryan Hafen and Amanda White, Pacific Northwest in a credible captivating manner. National Laboratory If you can answer YES to any of the following Tessera is a set of R-based tools to enable data

24 amstat news november 2014 statistician's view

STATISTICIAN’S VIEW Statistics Losing Ground to Computer Science Norman Matloff he American Statistical Association leader- by a prominent author on the use of mixed ship and many in statistics academia have labeled and unlabeled training data in clas- been undergoing a period of angst the last sification. Sadly, the paper cites nothing in the Tfew years. They worry that the field of statistics is extensive statistics literature on this topic, going headed for a future of reduced national influence back to 1977. and importance, with the feeling that the field is • The CS “engineering-style” research model to a large extent being eclipsed by other disciplines, often causes a cavalier attitude toward under- Norman Matloff is a professor notably computer science (CS). lying models and assumptions. Consider, for This is exemplified by the rise of a number of of computer example, a talk I attended by a machine learn- science at the new terms, largely in CS, such as data science, Big ing specialist who had just earned her PhD University of Data, and analytics, with the popularity of the term at one of the very top CS departments in the California, Davis, machine learning growing rapidly. To many of us, world. She had taken a Bayesian approach, and was formerly in the statistics this is “old wine in new bottles,” just statistics with and I asked why she had chosen that specific new terminology. department at prior distribution. She couldn’t answer—she that institution. I write this as both a computer scientist and stat- had just blindly used what her thesis adviser His research istician. I began my career in statistics, and though had given her. Moreover, she was baffled as interests my departmental affiliation later changed to CS, to why anyone would want to know why that include parallel processing, much of my research in CS has been statistical in prior was chosen. nature. And I submit that the problem goes beyond statistical • CS people tend to have grand—and some- computing, and the ASA’s understandable concerns about the well- regression and being of the statistics profession. The broader issue times starry-eyed—ambitions. On the one classification is not that CS people are doing statistics, but rather hand, this is a huge plus, leading to highly methodology. that they tend to do it poorly. impressive feats such as recognizing faces in This is not a problem of quality of the CS a large crowd. But this mentality leads to an researchers themselves; indeed, many are highly tal- oversimplified view, with everything being ented. Instead, there are a number of systemic rea- viewed as a paradigm shift. sons for this: • The CS research model is based on very rapid Neural networks epitomize this problem. publication, with the venue of choice being Enticing phrasing such as “neural networks work conferences rather than slow journals. The work like the human brain” blinds many CS research- is refereed, but just on a one-time basis, not ers to the fact that neural nets are not funda- with the back-and-forth interaction of jour- mentally different from other parametric and nals. There is also rapid change in fashionable nonparametric methods for regression and classi- research topics. Thus there is little time for deep, fication. Among CS folks, there is often a failure Editor’s Note: This article also long-term contemplation about the problems to understand that the celebrated accomplish- ments of “machine learning” have come mainly was published at hand. As a result, the work is less thoroughly on the StatsLife conducted and reviewed. from applying huge resources to a problem, rath- website at www. er than because fundamentally new technology statslife.org.uk. • Due in part to the pressure for rapid publica- has been invented. tion and the lack of long-term commitment to None of this is to say that people in CS should research topics, most CS researchers in statisti- stay out of statistics research. But the sad truth is that cal issues have little knowledge of the statistics the process of CS overshadowing statistics research- literature, and they seldom cite it. There is ers in their own field is causing precious resources— much “reinventing the wheel,” and many research funding, faculty slots, the best potential grad missed opportunities. students, attention from government policymakers— to go quite disproportionately to CS, even though the • For instance, consider a well-known CS paper

november 2014 amstat news 25 statistician's view

statistics community is arguably better equipped to “messy” problems, such as the face recognition make use of them. Statistics is important to the nation example cited earlier. and to the world, and if scarce resources aren’t being used well, it’s everyone’s loss. • Though many statisticians have first-rate com- What can be done? I offer the following as a puting skills, stat should reach out to CS for start: collaboration in advanced areas, such as the R project is doing with CS compiler experts. • There should be more joint faculty appoint- ments between CS and statistics departments. • Stat undergraduate and graduate curri- Teaching a course in the “other” department cula should be modernized (while retaining forces one to think more carefully about the mathematical rigor). Even math stat courses issues in that field and fosters interaction should involve computation. Emphasis on between fields. significance testing, well known to be under- informative at best and misleading at worst, • CS undergraduates should be encouraged to should be reduced. Modern tools, such as pursue a double major with statistics, and to cross-validation and nonparametric density/ go on for graduate work in statistics. There regression estimation, should be brought in are excellent precedents for the latter, such as early in the curriculum. Hadley Wickham and Michael Kane, both of them winners of the John Chambers Statistical The academic world is slow to change, but the Software Award. stakes here are high. There is an urgent need for the fields of CS and statistics to re-examine their • Statistics researchers should be much more roles in research, both individually and in relation aggressive in working on complex, large-scale, to each other. n

26 amstat news november 2014 people news

to strengthen forensic science “… outstanding service to through the identification and Bayesian statistics and the society.” development of new standards 2014 was the first year of and guidelines. n the award. The presentation was made at the latest in the Terry Speed ISBA biennial series of World Terry Speed, an ASA member Meetings in Mexico in July and leader of the bioinformat- of 2014. ics division of the Walter and West, who was one of the Eliza Hall Institute of Medical founders and a past president of Research in Melbourne, Australia, the society, received the 2014 was awarded one of that coun- inaugural award with a cita- Nathaniel Schenker try’s top science prizes. He was tion that noted, “A scientific presented the Commonwealth life spent always at the top, and Scientific and Industrial Research ASA President Nathaniel a vision of the future which Organisation-sponsored (CSIRO) Schenker is the recipient of became reality, from the work to Eureka Prize for Leadership in the 2014 Owen Thornberry establish ISBA as a society to its Science on September 10 for his Excellence in Leadership construction on sound bases.” superb leadership of the bioinfor- Award, a major award bestowed For more information about matics team at the Hall Institute annually by the National the medal, visit Duke University and his other contributions to the Center for Health Statistics website at https://stat.duke. science of bioinformatics. The (NCHS). He was presented edu/news/mike-west-wins Australian Museum presents the the honor September 17. The -zellner-medal. n Eureka awards, and CSIRO is award was established in 1996 Australia’s national science agency. to recognize the outstanding To watch Speed’s YouTube managerial leadership of an video and read more about NCHS employee. Its name- the award, visit the Australian sake, who served as the direc- Museum website at http:// tor of the Division of Health australianmuseum.net.au/media/ Interview Statistics, led efforts 2014-Eureka-Leadership. n to conduct the first nation- Nandini Kannan wide surveys of health promo- The National Institute of tion and disease prevention and ASA fellow Nandini Kannan Standards and Technology AIDS knowledge and attitudes. recently was named a Statistics (NIST) has appointed ASA Thornberry is recognized for Program director of the National members Hal Stern of the his contributions to data qual- Science Foundation’s Division of University of California, Irvine, ity in telephone surveys. n Mathematical Sciences (DMS), Bruce Weir of the University announced DMS Director of Washington, and William Mike West, the Arts & Michael Vogelius. Prior to join- Guthrie of NIST to Scientific Sciences Professor of Statistics ing DMS full-time, Kannan was Area Committees (SAC). Stern and Decision Sciences at Duke a professor of statistics at The was named to the physics/pat- University, was recognized University of Texas at San Antonio tern group, Weir to the biology/ recently by the International (UTSA). She brings significant DNA panel, and Guthrie to the Society for Bayesian Analysis administrative experience to her chemistry/instrumental analy- (ISBA) through the award of the new position, having served as a sis committee. These NIST- Zellner Medal. department chair at UTSA and administered committees pro- Named for the founding as president of the International vide the critical bridge between president of the society, the Indian Statistical Association. the Forensic Science Standards medal is awarded once every two Previously, she was a DMS “rotat- Board and the 23 discipline- years to (at most two) statisti- ing” Statistics Program director specific subcommittees working cians in recognition of their from 2011 to 2013. n

november 2014 amstat news 27 people news

research. Morton is professor and chair of the department of biostatistics at the University of Pittsburgh’s Graduate Sally Morton School of Public Health. C. R. Rao PCORI assists patients, clini- ASA Past President Sally cians, purchasers, and policy- The Polish Statistical Morton was appointed makers in making informed Association recently awarded to the Patient-Centered health care decisions by pro- C. R. Rao the Splawa-Neyman Outcomes Research Institute’s viding quality, relevant Medal. The citation read, “In (PCORI) Methodology evidence on how best to pre- appreciation and admiration Committee. This commit- vent, diagnose, treat, and mon- of your extraordinary scientific tee assists in developing and itor diseases and other health work and your invaluable com- updating methodological stan- conditions. Read more at www. mitment to statistical education dards and guidance for com- gao.gov/press/pcori_methodology all over the world.” parative clinical effectiveness _comm2014sep18.htm n The medal was established in 2012 to celebrate the 100th anniversary of the Polish Statistical Association and to OBITUARY appreciate the most outstanding statisticians, teachers of statistics, and all people of science that Robert "Bob" Newcomb by their work significantly con- Robert "Bob" Newcomb of the University of California at tributed to the development of Irvine (UCI), the founding president of the ASA’s Orange statistics or its education. County Long Beach Chapter (OCLB), passed a For information about the way on July 10 due to complications of a sudden illness. Polish Statistical Association, Newcomb was faculty at UCI for more than 40 years, was the visit pts.stat.gov.pl/en. n founding director of the UCI Center for Statistical Consulting, helped form the UCI Department of Statistics in 2002, and was a driving force in the creation of the UCI men’s volleyball program. A dedicated professor, Newcomb continued to teach statistics courses at UCI, completing his last course in the spring of 2014. Three regional ASA chapters in southern California—San Diego, Southern California, and OCLB—are organizing a memorial service to celebrate Newcomb’s successful life, which influenced the statistical community and lives of many Read about your statisticians. The service will be followed by a statistical semi- colleagues and friends nar with presentations by invited statisticians who had devel- in the news. Go to oped personal and professional relationships with Holcomb www.amstat.org and over many years. The service and seminar will take place click on “Statisticians November 15, 2014. in the News.” Details about the program are available at www.sc-asa.org/ calendar.html. To read more about Newcomb’s extraordinary life, visit www.legacy.com/obituaries/orangecounty/obituary.aspx?pid =171958185.

28 amstat news november 2014 awards and deadlines

Sirken Endows Fund to Recognize Survey Researchers onroe Sirken, a long- time member of the ASA, recently provided Mthe association with an endowment fund to recognize a distinguished survey researcher. The honoree will give the first Sirken Lecture at this For more year’s Joint Statistical Meetings in information Seattle, Washington, and receive a about this $5,000 honorarium. award, visit “Monroe’s generosity opens www.amstat. an opportunity to recognize org/awards/ outstanding interdisciplinary monroe work by survey researchers,” sirkenaward. Ron Wasserstein, ASA Executive cfm. Director, said. “By shining a beacon on this work, the Sirken To read more Award will encourage and stimu- about Sirken, late future research in this impor- visit http:// tant area of statistical practice.” magazine. Sirken earned his PhD amstat.org/ in 1950 at the University of blog/2011/ Washington after majoring Sirken Lecture 09/01/ in sociology with a minor in From left: Nat Schenker, Ron Wasserstein, and Monroe Sirken. Sirken provided monroesirken. mathematics. Awarded a Social the association with an endowment fund to recognize a distinguished survey Science Research Council Post researcher. Doctoral Fellowship, he spent the 1950–1951 academic year intractable survey problems. He the award will be highly honored in the statistics laboratory at strongly advocated conduct- to be associated with Dr. Sirken the University of California, ing survey methods research at in this way.” Berkley, and the following sum- the intersection of the statistical, With the quality of Sirken’s mer at the U.S. Census Bureau. social, and cognitive sciences and work in mind, the Sirken Award Sirken was a charter member of information technology. Sirken Committee will select the Sirken the National Center for Health also introduced the first cogni- Lecturer based on outstanding Statistics (NCHS) and worked tive research laboratory dedicated contributions to interdisciplin- there for more than 50 years as to designing and testing survey ary survey research that improve chief mathematical statistician questionnaires and data presenta- methods of collecting, verifying, and associate director for research tion methods. processing, presenting, or analyz- and methodology. Wasserstein said it is fitting ing survey data. Interested in survey methods, that this award bears Sirken’s Nominations are due Sirken introduced network sam- name. “Monroe’s contributions to December 15 and should include pling to improve survey designs NCHS, to the monitoring of our a nominating letter, three sup- of rare and elusive populations national’s health, are legendary porting letters, and a curriculum and applied the method to a and still impact the work being vitae. Send nominations to Pam broad range of practical and often done at that agency. Recipients of Craven at [email protected]. n

november 2014 amstat news 29 awards and deadlines

2014 Causality in Statistics Education Prize Winners Announced

uring the 2014 Joint Statistical Meetings, the ASA awarded the second Causality 2014 Prize Committee in Statistics Education Prize, which was Members Destablished to “encourage the teaching of basic Dennis Pearl (The Ohio State causal inference methods in introductory statistics University, CAUSE, co-chair) courses.” The prize was awarded to Maya Petersen and Laura B. Balzer for developing the path-blaz- Judea Pearl (University of ing course “Introduction to Causal Inference” at California at Los Angeles, co-chair) the University of California at Berkeley. Felix Elwert (University With clear lectures, detailed discussion assign- ments, and innovative labs and homework using R, of Wisconsin-Madison) Petersen and Balzer prepared a new generation of Daniel Kaplan (Macalester College) scientists, who have acquired the tools of modern Michael Posner (Villanova University) causal analysis and are equipped to tackle each step of the causal roadmap. The course is publicly avail- Larry Wasserman (Carnegie able at www.ucbbiostat.com and thereby provides Mellon University) other institutes an educational resource for this foundational material. Peterson and Balzer’s course was chosen primar- is aiming to close a growing gap between research and ily on the basis of its “teachability” and appeal to a education in this field. “While researchers are swept in broad range of statistics-minded disciplines. an unprecedented excitement over new causal infer- Judea Pearl, who donated the prize and serves as co- ence tools that are unveiled before us almost daily, the chair of the prize selection committee, said the prize excitement is hardly seen among statistics educators and is totally absent from statistics textbooks.” In a recent interview (see http://magazine.amstat. Student Paper Competitions org/blog/2012/11/01/pearl), Pearl said to ASA Offered for JSM Travel Awards Executive Director Ron Wasserstein, “I hope this prize Many ASA sections sponsor student paper competitions will stimulate the generation of effective course mate- to reward gifted students and help them mitigate the costs rial. … And would convince every statistics instruc- of attending JSM. This is an opportunity for students to get tor that causation is easy (It is!) and that he/she too involved by being a participant at the meeting. Students can sub- can teach it for fun and profit. The fun comes from mit papers to up to two sections; however, they cannot accept showing students how simple mathematical tools can more than one prize. answer questions that Pearson-Fisher-Neyman could Papers are due to the sections’ award committees by December not begin to address (e.g., control of confounding, 15, 2014. Winners will be contacted by January 15, 2015. model diagnosis, Simpson’s paradox, mediation analy- Winners must submit their abstracts online at the JSM website sis), and the profit comes because most customers of by February 2. Some sections schedule their student paper award statistics ask causal, not associational, questions.” winners in one JSM session, while others match each winner’s The following prize criteria set by the selection topic with a similarly themed session. committee were pragmatic: Each section’s award program varies in the submission • The extent to which the material submitted requirements and what the winners receive. For a listing equips students with skills needed for effective of which sections sponsor student paper competitions and causal reasoning links for more information, visit www.amstat.org/sections/ studentpaperawards.cfm. • The extent to which the submitted material assists statistics instructors gain

30 amstat news november 2014 awards and deadlines

Deadlines and Contact Information for an understanding of the basics of causal infer- ASA National Awards, Special Lectureships, ence and prepares them to teach these basics in undergraduate and lower-division graduate and COPSS Awards classes in statistics www.amstat.org/awards

• The skills listed were, likewise, problem-orient- February 1, 2015 February 20, 2015 ed, and included the following: Karl E. Peace Award for ASA Statistics in the Physical Outstanding Statistical and Engineering Sciences Award • Ability to take a given causal problem Contributions for the Nominations and Questions: and articulate in some mathemati- Betterment of Society Philip J. Ramsey [email protected] Nominations: Pam Craven cal language (e.g., counterfactuals, [email protected] equations, or graphs) both the target Questions: G. David Williamson February 23, 2015 [email protected] ASA Gertrude M. Cox Scholarship causal quantity to be estimated and the Applications: Pam Craven assumptions one is prepared to make [email protected] (and defend) to facilitate a solution February 1, 2015 Questions: Eloise E. Kaizar ASA W. J. Dixon Award [email protected] • Ability to determine, in simple cases, for Excellence in Statistical Consulting how the target causal quantity can be Nominations: Pam Craven March 1, 2015 estimated using the observed data [email protected] ASA Edward C. Bryant Scholarship Questions: Michael H. Kutner Applications: Pam Craven • Ability to take a simple scenario (or a [email protected] [email protected] Questions: Tapabrata Maiti model), determine whether it has sta- [email protected] tistically testable implications, and then February 15, 2015 apply data to test the assumed scenario Causality in Statistics Education Award March 1, 2015 Nominations and Questions: ASA Excellence in Statistical This year, the committee received four nomi- [email protected] Reporting Award Nominations: Pam Craven nations. The selection was difficult, considering a [email protected] balance had to be struck between three competing February 15, 2015 Questions: Morteza Marzjarani objectives: mathematical rigor, breadth of topics, ASA Samuel S. Wilks [email protected] Memorial Medal and accessibility to a large audience of students Nominations: Pam Craven and instructors. [email protected] March 1, 2015 The inaugural Causality in Statistics Questions: Lynne Billard ASA Fellows [email protected] Nominations accepted online Education award in 2013 was given to Felix at www.amstat.org beginning Elwert of the department of sociology at the February 15, 2015 October 1, 2014 University of Wisconsin-Madison for his two- ASA Waller Distinguished Questions: Stephanie S. Shipp [email protected] day course, “Causal Inference with Directed Teaching Career Award Nominations: Pam Craven Acyclic Graphs.” Elwert received $5,000 and [email protected] March 1, 2015 a plaque at the 2013 Joint Statistical Meetings Questions: Bradley A. Hartlaub ASA Outstanding Statistical in Montréal, Québec, Canada. A gift from [email protected] Application Award Microsoft Research enabled the prize to double Nominations: Pam Craven in 2014 to $10,000. Slides covering about eight February 15, 2015 [email protected] ASA Waller Education Award Questions: DuBois Bowman lecture hours of Elwert’s short course and accom- Nominations: Pam Craven [email protected] panying publications are available at www.ssc. [email protected] wisc.edu/~felwert/causality. Questions: Bradley A. Hartlaub March 15, 2015 [email protected] ASA Founders Award Nominations Sought for 2015 Prize Nominations: Pam Craven [email protected] Nominations are wanted for the 2015 Causality February 15, 2015 ASA W. J. Youden Award Questions: Nathaniel Schenker in Statistics Education Award, which, due to a gift in Interlaboratory Testing [email protected] from Microsoft Research, will again be $10,000. Nominations: Pam Craven The nomination deadline is February 15, 2015. [email protected] Questions: Blaza Toman For additional information, see www.amstat.org/ [email protected] education/causalityprize. n

november 2014 amstat news 31 awards and deadlines

Student Paper Competition, John M. Chambers Statistical Software Award Competitions Announced he Statistical Computing and Statistical Graphics sections will review the applications. Graphics sections of the ASA are co- The selection criteria used by the committee will sponsoring a student paper competition include innovation and significance of the con- Ton the topics of statistical computing and statis- tribution as well as the professional quality of the tical graphics. Students are encouraged to sub- manuscript. Award announcements will be made mit a paper in one of these areas, which might by January 15, 2015. be original methodological research, some novel More information about the competition computing or graphical application in statistics, or can be found at www.amstat.org/sections/student any other suitable contribution (e.g., a software- paperawards.cfm. related project). The selected winners will present their papers John M. Chambers in a topic-contributed session at the 2015 Joint Statistical Software Award Statistical Meetings. The sections will pay regis- The Statistical Computing Section also announc- tration fees for the winners as well as a substan- es the competition for the John M. Chambers tial allowance for transportation to the meetings Statistical Software Award. In 1998, the Association and lodging. for Computing Machinery presented its Software Anyone who is a student (graduate or under- System Award to John Chambers for the design and graduate) on or after September 1, 2014, is eligible development of S. Chambers donated his award to to participate. An entry must include an abstract, the Statistical Computing Section to endow an a six-page manuscript (including figures, tables, annual prize for statistical software written by, or and references), blinded versions of the abstract in collaboration with, an undergraduate or gradu- and manuscript (with no authors or references ate student. The prize carries with it an award of that easily lead to identifying the authors), a CV, $1,000, plus a substantial allowance for travel to and a letter from a faculty member familiar with the annual Joint Statistical Meetings (JSM), where the student’s work. The applicant must be the first the award will be presented. author of the paper. The faculty letter must include Teams of up to three people can participate a verification of the applicant’s student status and, in the competition, with the cash award being in the case of joint authorship, should indicate split among team members. The travel allow- what fraction of the contribution is attributable to ance will be given to just one individual in the the applicant. Electronic submissions of papers in team, who will be presented the award at JSM. Postscript or PDF are preferred, and all materials To be eligible, the team must have designed and must be in English. implemented a piece of statistical software. The Students may submit papers to no more than individual within the team indicated to receive two sections and may accept only one section’s the travel allowance must have begun the devel- award. Students must inform both sections applied opment while a student and must either currently to when he or she wins and accepts an award, there- be a student or have completed all requirements by removing the student from the award competi- for her/his last degree after January 1, 2014. To tion for the second section. apply for the award, teams must provide the fol- All application materials must be received by lowing materials: 5:00 p.m. EST, December 14, 2014. Members • Current CVs for all team members. of the Student Paper Competition Award Committee of the Statistical Computing and • A letter from a faculty mentor at the

32 amstat news november 2014 awards and deadlines

Nominations Sought Spiegelman Award The Applied Public Health Statistics Section (http://bit.ly/1nz9QaS) of the American Public Health Association (APHA) invites nomi- nations for the 2015 Mortimer Spiegelman academic institution of the individual indi- Award, which honors a statistician below the cated to receive the travel award. The letter age of 40 in the calendar year of the award who should confirm that the individual had sub- has made outstanding contributions to health stantial participation in the development of the software, certify her/his student status statistics, especially public health statistics. The when the software began to be developed award was established in 1970 and is presented (and either the current student status or annually at the APHA meeting. the date of degree completion), and briefly The award serves the following three discuss the importance of the software to purposes: statistical practice. • To honor the outstanding achievements • A one- to two-page description of the soft- of both the recipient and Spiegelman ware, summarizing what it does, how it does • To encourage further involvement in pub- it, and why it is an important contribution. lic health by the finest young statisticians If the team member competing for the travel allowance has continued developing the • To increase awareness of APHA and the software after finishing her/his studies, the Applied Public Health Statistics Section description should indicate what was devel- in the academic statistical community oped when the individual was a student and what has been added since. To be eligible, a candidate must have been born in 1976 or later. Please email a nominat- • An installable software package with its source ing letter that states the candidate’s date of birth code for use by the award committee. It and how their contributions relate to public should be accompanied by enough informa- health concerns, up to three letters of support, tion to allow the judges to effectively use and and the candidate’s CV to the award committee evaluate the software (including its design [email protected] considerations). This information can be chair, Amy Herring, at . provided in a variety of ways, including a user Nominations are due by April 1, 2015. manual (paper or electronic), a paper, a URL, Deming, Noether Lecturers and online help to the system. Consider nominating a colleague for the All materials must be in English. Electronic text prestigious Deming and Noether lec- submitted in Postscript or PDF is preferred. Entries tures that honor the life and work of W. will be judged on a variety of dimensions, including Edwards Deming and Gottfried E. Noether. the importance and relevance for statistical practice Awardees will give the lectures at the Joint of the tasks performed by the software, ease of use, Statistical Meetings and receive a cash prize clarity of description, elegance, and availability for and engraved award. For details, visit www. use by the statistical community. Preference will be amstat.org/awards/awardsscholarships.cfm. given to those entries that are grounded in software design, rather than calculation. The decision of the 2015 Fellows, Awards award committee is final. We believe it is important to recognize All application materials must be received by individuals who have made significant con- 5:00 p.m. EST, February 17, 2015. The winner tributions in the field of statistics. Please will be announced in May, and the award will be consider nominating one of your colleagues given at the 2015 Joint Statistical Meetings. for an ASA award or Fellow. For more Application materials for either the student information, visit www.amstat.org/awards/ paper award or John Chambers Award can be sent awardsscholarships.cfm. to Aarti Munjal at [email protected]. n

november 2014 amstat news 33 professional opportunities

Professional Opportunity listings may not exceed 65 words, plus equal oppor- tunity information. The deadline for their receipt is the 20th of the month two Arizona months prior to when the ad is to be published (e.g., May 20 for the July issue). n Arizona State University. The School Ads will be published in the next available issue following receipt. of Mathematical and Statistical Sciences Listings are shown alphabetically by state, followed by international listings. at Arizona State University invites appli- Vacancy listings may include the institutional name and address or be identi- cations for two positions in statistics fied by number, as desired. beginning in fall 2015. One is tenure- Professional Opportunities vacancies also will be published on the ASA’s track (assistant professor) and the other website (www.amstat.org). Vacancy listings will appear on the website for the is tenured (associate or full). These posi- entire calendar month. Ads may not be placed for publication in the magazine tions are part of a substantial expansion only; all ads will be published both electronically and in print. of the statistics program at Arizona State. Rates: $320 for nonprofit organizations (with proof of nonprofit status), Information and application available $475 for all others. Member discounts are not given. For display and online through Mathjobs at www.mathjobs.org/ advertising rates, go to www.amstat.org/ads. jobs/ASU. EO/AA Employer. Listings will be invoiced following publication. All payments should be made to the American Statistical Association. All material should be sent California to Amstat News, 732 North Washington Street, Alexandria, VA 22314- n San Diego State University's MIS 1943; fax (703) 684-2036; email [email protected]. Department seeks applicants for one full- Employers are expected to acknowledge all responses resulting from pub- time, tenure-track business analytics/sta- lication of their ads. Personnel advertising is accepted with the understand- tistics position at the assistant-professor ing that the advertiser does not discriminate among applicants on the basis of level starting August 2015. Applicants race, sex, religion, age, color, national origin, handicap, or sexual orientation. should possess a PhD in statistics or relat- Also, look for job ads on the ASA website at www.amstat.org/jobweb. ed discipline by August 2015 and dem- onstrate potential for excellence in both teaching and research. Apply at http:// apply.interfolio.com/26335. Applications will be reviewed until the position is filled. SDSU is an equal opportunity employer and does not discriminate against persons based on race, religion, national origin, sexual orientation, e Department of Biostatistics and the Lineberger Comprehensive Cancer Center (LCCC) at the gender, gender identity and expression, University of North Carolina at Chapel Hill are seeking a non-tenure track faculty member with exper- tise in basic cancer research including but not limited to clinical trials, outcomes research, and protocol marital status, age, disability, pregnancy, development to collaborate with cancer researchers in these areas. is position is open at the Research or medical condition, or covered veteran Clinical track and at the Assistant or Associate Professor level. e faculty member is expected to engage status. The person holding this position is in independent methodological research. e LCCC is one of 27 NCI-designated comprehensive cancer considered a “mandated reporter” under centers. Applicants should hold a PhD in biostatistics or statistics, and possess good communication the California Child Abuse and Neglect skills. is position will remain open until  lled. Reporting Act and is required to comply To apply, use the electronic submission website at https://unc.peopleadmin.com/postings/50034 with the requirements set forth in CSU and upload PDF versions of your CV, cover letter, and research and teaching statements. Candidates Executive Order 1083 as a condition of must also arrange for three letters of recommendation to be emailed to [email protected] addressed to: employment.

Faculty Search Committee n c/o Vera Bennett The department of mathematics at Department of Biostatistics California State University, Fullerton, CB #7420, McGavran-Greenberg Hall invites applications for a tenure-track University of North Carolina at Chapel Hill faculty position in statistics at the assis- Chapel Hill, NC 27599-7420 tant professor level beginning August At the UNC Gillings School of Global Public Health, diversity, inclusiveness and civility are core values 2015. Applicants must have a solid as well as characteristics of the School. We strongly encourage applications from diverse individuals, theoretical background in statistics. including but not limited to diversity in such characteristics as race/ethnicity, color, national origin, age, A complete description of the posi- gender, socioeconomic background, religion, creed, veteran’s status, gender identity, gender expression, tion is available at www.mathjobs.org. sexual orientation and disability. e University of North Carolina at Chapel Hill is an equal opportu- Applicants must submit their materials nity employer that welcomes all to apply, including protected veterans and individuals with disabilities. through www.mathjobs.org. Application deadline is December 1, 2014. EOE.

34 amstat news november 2014 n The department of mathematics and statistics at San Diego State University seeks applicants for a tenure-track assis- tant professor position in data science starting in fall 2015. Focused research areas include but are not limited to machine learning, large-scale statistical analysis, Bayesian methods, biostatistics, and/or business/financial analytics. PhD in statistics/biostatistics required. See www.math.sdsu.edu/department/jobops for a full posting. SDSU is a Title IX, equal opportunity employer and does not discriminate against persons on the basis of race, religion, national origin, sexual orientation, gender, gender iden- tity and expression, marital status, age, disability, pregnancy, medical condition, or covered veteran status. n Department of statistics & applied probability, University of California, invites applications for a tenure-track assistant professor position in statistics, starting July 1, 2015. Qualifications: research/teaching excellence; PhD in statistics, biostatistics or related fields. Candidates who can contribute to the diversity of excellence of the academic community through research, teaching, and service are particularly encour- aged to apply. An EO/AA employer. Additional information at www.pstat. ucsb.edu/employment.htm. n RAND Corporation is seeking PhD statisticians for exciting opportunities to collaborate on multidisciplinary pub- lic policy research projects. Openings exist for recent graduates and experi- enced statisticians. See our ad in the September Amstat News for details or go to www.rand.org/statistics/jobs. Applications received by December 1, 2014, will receive priority. Applications must be submitted online following the instructions at www.rand.org/statistics/ jobs (search "PhD Statistician"). Send questions to [email protected]. EO/AA Employer.

november 2014 amstat news 35 n Biostatistician. The Graduate School demonstrated commitment to depart- employer and strongly encourages women of Public Health at San Diego State ment service. Apply by November 14 at and members of underrepresented groups University invites applications for tenure- PeopleAdmin: http://indiana.peopleadmin. to apply. track assistant professor position starting com/postings/992 (full professor) or http:// fall 2015. Requires doctorate in biosta- indiana.peopleadmin.com/postings/990 n The department of statistics at Iowa tistics, or statistics with public health (assistant professor). Address questions to State University invites applications for applications. Complete information at Kelly Hanna [email protected]. EOE. an open-rank faculty position in the area http://publichealth.sdsu.edu/open-positions. of bioinformatics. We seek a candidate Applicants should apply via Interfolio Iowa who can develop statistical methodology at: http://apply.interfolio.com/25795. n The department of statistics at Iowa for complex biological data, collaborate AA/EOE. State University is seeking candidates with with biological scientists, and enhance strong research expertise in applied prob- bioinformatics education. Candidates n California State University, Fresno. ability and stochastic processes for an open- applying for a tenured position should Decision Sciences—Assistant/Associate rank tenure-track/tenured faculty position. have excellent records in research, exter- Professor Tenure-track. Requires an earned Candidates applying for tenured associate nal funding, teaching, and graduate doctorate (PhD or DBA) in business or full professorship should have excellent student supervision. For further informa- administration, decision sciences, statis- credentials and track records in research, tion, visit www.stat.iastate.edu/faculty- tics, management science, supply-chain external funding, teaching, and graduate open-position. Iowa State University is management, or related field from AACSB supervision. For further information about an affirmative action/equal opportu- or ABET accredited institution (or equiva- the position, visit www.stat.iastate.edu/ nity employer and strongly encourages lent); however, candidates nearing comple- faculty-open-position. Iowa State University women and members of underrepresent- tion of doctorate (ABD) may be consid- is an affirmative action/equal opportunity ed groups to apply. ered. For more information and to apply, visit http://apptrkr.com/517511 EOE. FACULTY POSITION AT THE Georgia n The H. Milton Stewart School of Ind. FRED HUTCHINSON CANCER RESEARCH CENTER & Systems Eng. Georgia Tech, Atlanta, invites applications for asst professor, e Clinical Research Division of the Fred Hutchinson Cancer Research Center assoc professor, and professor positions. (FHCRC) invites applications for a biostatistician at the Associate or Full Member level Consideration from research areas, includ- to lead the clinical statistics program and provide collaborative statistical support for the ing advanced mfg, analytics, informatics clinical research activities of the division and the a liated Cancer Consortium with the & control, economic decision analysis, University of Washington. health systems, supply chain & logistics, e Clinical Research Division of FHCRC is a world-renowned group of investigators quantitative & computational finance, who conduct cutting edge research both in the laboratory and in the clinic to elucidate optimization, simulation, statistics and the pathogenesis and treatment of cancer. In particular, the CRD is internationally stochastics. Application submission: www. recognized for its leadership in the  elds of hematopoietic cell transplantation and isye.gatech.edu/candidate. Interview pro- immunotherapy. e successful candidate for this position will be a member of the cess begins fall semester 2014. Deadline: CRD Research Executive Committee and will participate in determining the scienti c March 31, 2015. Georgia Tech is an and operational priorities of the CRD, will oversee the statistical and data management Affirmative Action/Equal Opportunity sta of the division, and maintain their own research program. Independent research Employer, A Unit of the U System of GA. in statistical methodology is strongly encouraged. A Ph.D. or equivalent in statistics or biostatistics and excellent managerial and communications skills are required. Prior Indiana experience in a biomedical research environment is highly desirable. A joint appointment n Two positions in data science, one with the Division of Public Health Sciences at the FHCRC is possible, as is an a liate appointment in a relevant department at the University of Washington. senior, at Indiana University Department of Statistics (www.stat.indiana.edu). Please send curriculum vitae, a letter describing research interests, and the names of four Research should complement existing references to Stephanie Lee, MD at http://bit.ly/FHCRC_ClinStats strengths in machine learning, spatial sta- tistics and Bayesian inference. Senior appli- e Fred Hutchinson Cancer Research Center is an equal opportunity employer, committed to workforce diversity. cants should be renowned scholars with

36 amstat news november 2014 Tenure Track Assistant Professor Position The Department of Statistical Science at Duke University invites applications for appointments at the level of Assistant Professor to begin in Fall 2015. Preference will be given to candidates n The department of statistics and Kentucky whose core statistical science research interests are the Center for Survey Statistics and n University of Kentucky. Assistant complemented by collaborative research interests Methodology at Iowa State University professor, tenure track, biostatistics (pub- in areas including economics, fi nance or other areas of the social and policy sciences, computer is seeking candidates for an open- lic health), beginning August of 2015. science, neurosciences, and environmental science. rank tenure-track/tenured faculty Potential for demonstrated excellence We are particularly interested in applicants with position. Excellence in research and in methodological and collaborative demonstrated experience in complex stochastic teaching expected, survey expertise research and in teaching public health modeling and computation or large-scale problems and data sets. desirable. Candidates applying for professional students and biostatistics tenured associate or full professorship students. Selection begins Dec. 1, 14 The Department is an internationally recognized should have excellent track records in and continues until filled. CV, three center of excellence for research and education in contemporary statistical methodology. With research, external funding, teaching, reference letters to Richard Kryscio, leading strengths in Bayesian analysis, interdisci- and graduate student supervision. For 230 Sanders Brown Center on Aging, plinary applications and computationally intensive further information, please visit www. 800 South Limestone St., University of methods, the Department offers outstanding stat.iastate.edu/faculty-open-position. Kentucky, Lexington, KY 40536-0236. computational facilities and opportunities for inter- disciplinary research. We currently have 18 regular Iowa State University is an affirmative [email protected]. The University of rank faculty, 16 visiting, adjunct, and postdoctoral action/equal opportunity employer Kentucky is an affirmative action, equal faculty, and over 75 graduate students. Beyond and strongly encourages women and opportunity employer. core research in statistical and computational members of underrepresented groups sciences, we have many collaborative interactions with multiple other Duke departments, institutes to apply. Massachusetts and centers, including the Information Initiative n Boston University mathematics at Duke (iiD), Duke Institute for Brain Sciences n The department of statistics at Iowa and statistics invites applications for (DIBS), and Social Science Research Initiative State University invites applications for tenure-track probability/stochastics/sta- (SSRI). Complementary interactions involve long-standing associations with the Statistical and an open-rank tenure-track/tenured faculty tistics assistant professor. PhD required. Applied Mathematical Sciences Institute (SAMSI) position. The position requires an ability Begins July 2015 pending final budget- and the National Institute of Statistical Science to engage in collaborative research with ary approval. Commitment to research (NISS), located nearby in the Research Triangle scientists in the College of Agriculture and teaching both undergraduate and Park, as well as many collaborators, institutes and companies around the US and worldwide. and Life Sciences. Candidates applying graduate level essential. Submit cover for tenured associate or full professorships letter, CV, research statement, teaching Our internationally recognized PhD program is should have excellent records in research, statement, and four recommendation complemented by our MS in Statistical Science, our MS in Statistical and Economic Modeling, and external funding, teaching, and graduate letters (one teaching): www.mathjobs. our Statistical Science undergraduate degree. More student supervision. For further infor- org. Alternatively, submit materials: information is available at the department website mation, please visit www.stat.iastate.edu/ Probability/Stochastics/Statistics Search, http://stat.duke.edu. faculty-open-position. Iowa State University Mathematics and Statistics, Boston To apply, submit a letter, curriculum vitae, is an affirmative action/equal opportunity University, 111 Cummington Mall, personal statement of research and teaching and employer and strongly encourages women Boston, MA 02215. Deadline Dec. names/letters from three references via https:// and members of underrepresented groups 15, 2014. We are an equal opportunity academicjobsonline.org/ajo. Enquiries can be to apply. employer and all qualified applicants will emailed to [email protected]. The application pool will remain open until the position is fi lled; receive consideration for employment screening will begin on December 1st 2014. Kansas without regard to race, color, religion, sex, n The department of statistics at national origin, disability status, protected Duke University, located in Durham, North Caro- Kansas State University has openings veteran status, or any other characteristic lina, is an Affi rmative Action/Equal Opportunity Employer committed to providing employment for up to three tenure-track assistant protected by law. We are a VEVRAA opportunity without regard to an individual’s age, professor positions to begin August federal contractor. color, disability, genetic information, gender, gen- 2015. Email a cover letter, a complete der identity, national origin, race, religion, sexual CV, transcripts of graduate-level course- n Boston University mathematics and orientation, or veteran status. Applications from women and minorities are strongly encouraged. work, and three recommendation letters statistics invites applications for a three-year Individuals in dual career couples are encouraged to [email protected]. Screening begins postdoctoral position in statistics/prob- to visit http://provost.duke.edu/faculty/partner/, the during fall 2014 and continues until ability, starting July 2015 pending final website on Duke’s Advantages for Faculty, for in- formation on opportunities for dual career couples positions are filled. KSU is EOE. For budgetary approval. Strong commitment to in the area and how the university can help. the full announcement, see www.ksu. research and teaching is essential. Submit edu/stats. EOE. cover letter, CV, research statement,

november 2014 amstat news 37 38 amstat news november 2014 teaching statement, and four recommenda- full classified and instructions to submit the power of a diverse community and tion letters (one teaching) to www.mathjobs. online, in October and November, at www. encourages applications from individuals org. Alternatively, submit all materials to: mathjobs.org. Applications should be com- with varied experiences, perspectives and Statistics/Probability Search, Mathematics plete by December 1, 2014. Massachusetts backgrounds. and Statistics, Boston University, 111 Institute of Technology is an Equal Cummington Mall, Boston, MA 02215. Opportunity Affirmative Action Employer. Michigan Deadline Dec. 15, 2014. We are an equal n Michigan State University invites opportunity employer and all qualified n The department of quantitative health applications for chairperson of the applicants will receive consideration for sciences at the University of Massachusetts department of statistics and probability. employment without regard to race, color, Medical School is recruiting a tenured or Required doctoral degree in statistics or religion, sex, national origin, disability sta- tenure-eligible faculty member at the asso- related field and internationally recog- tus, protected veteran status, or any other ciate professor or professor levels for the nized and externally funded research pro- characteristic protected by law. We are a division of biostatistics and health services gram. Must exhibit communication and VEVRAA federal contractor. research. Applicant should have an estab- interpersonal skills. Candidates should lished, nationally recognized research pro- be qualified to hold the rank of tenured n Mathematics Department MIT seeking gram. Responsibilities include sustained professor. Review of applications begins to fill combined teaching and research posi- excellence in independent research focused December 1, 2014. For detailed informa- tions as instructor, assistant professor, and on methodological innovation, doctoral- tion, go to https://jobs.msu.edu (position higher in statistics, beginning September level teaching and mentoring. Apply to: #0044) EOE. 2015. Appointments based mainly on https://academicjobsonline.org/ajo/jobs/4341. exceptional research qualifications. PhD As an equal opportunity and affirma- n Grand Valley State University invites required by employment start date. See tive action employer, UMMS recognizes applications for assistant professor of

DEPARTMENT OF BIOSTATISTICS

The Department of Biostatistics at the Johns Hopkins Bloomberg School of Public Health is seeking an outstanding applicant to join our tenure track faculty. Rank of appointment will be commensurate with experience, and new PhDs and recent postdoctoral fellows are encouraged to apply. Candidates should have a PhD or equivalent in statistics, biostatistics, or a comparable data science field. In 2014-15 the Department particularly seeks new faculty members who are committed to education and to engagement in cutting edge research on statistical inference and methods and subject-area content for the population health, clinical and social health sciences. Women and under-represented minority candidates are particularly encouraged to apply.

Track Faculty Positions The Hopkins Department of Biostatistics, founded in 1918, was the first degree-granting department of statistical science in the US and has ranked among the best throughout its history. The Johns Hopkins Health Institutions (Schools of Public Health, Medicine, and Nursing, and the Johns Hopkins Hospital) are among the top worldwide and provide a research environment in which energetic faculty can promulgate scientific excellence. Today, the Department comprises 21 tenure track faculty members, 18 research track faculty members, 13 postdoctoral fellows, 49 PhD students, 12 full-time master degree students, and 10 students pursuing joint master degrees together with doctorates in other departments.

TO APPLY Email cover letter, CV, contact information for three references, a statement of research interests and goals, a teaching statement, and two manuscripts or articles representing your most important work to: Faculty Search Committee at [email protected].

The Johns Hopkins University is an affirmative action/equal opportunity employer.

november 2014 amstat news 39 Faculty Positions in Biostatistics statistics. Qualifications include a PhD in statistics or related field, evidence e Department of Biostatistics at e University of North Carolina at Chapel Hill is seeking applications for two tenure-track Assistant Professors. e appointments will start in the summer or fall of 2015. A doc- of commitment to teaching excellence toral degree in Biostatistics, Statistics or equivalent is required. is search is not restricted to any speci c and active scholarship, and an inter- area of biostatistics. Applicants should have broad research and teaching interests, the potential to direct est in using technology in teaching at PhD-level research, and the ability to engage in collaborative research with other faculty members at the all levels. For information regarding University. Review of applications will start in December, 2014. e positions will remain open until  lled. application materials, see our complete www.gvsu.edu/ To apply, use the electronic submission website at http://unc.peopleadmin.com/postings/50978 and upload position description at PDF versions of your CV, cover letter, and research and teaching statements. Candidates must also arrange stat. Applications only accepted online; for four letters of recommendation to arrive via email at [email protected] addressed to: review begins on November 21, 2014. AA/EOE. Faculty Search Committee c/o Vera Bennett Minnesota Department of Biostatistics n CB #7420, McGavran-Greenberg Hall The biostatistics division, school of University of North Carolina at Chapel Hill public health, University of Minnesota, Chapel Hill, NC 27599-7420 seeks applicants for open-rank ten- ured/tenure-track faculty position. At the UNC Gillings School of Global Public Health, diversity, inclusiveness and civility are core values as well as characteristics of the School. We strongly encourage applications from diverse individuals, including Especially interested in individuals with but not limited to diversity in such characteristics as race/ethnicity, color, national origin, age, gender, socio- academic and research records in the economic background, religion, creed, veteran’s status, gender identity, gender expression, sexual orientation development of innovative approaches, and disability. e University of North Carolina at Chapel Hill is an equal opportunity employer that methods and software for the manipu- welcomes all to apply, including protected veterans and individuals with disabilities. lation and analysis of "big data" in the biomedical sciences, especially using machine learning techniques. Longer ad: www.sph.umn.edu/biosta- tistics. Applicants should apply online: http://employment.umn.edu/applicants/ FACULTY POSITIONS AT PENN STATE Central?quickFind=124351. EOE. The Department of Stati sti cs at The Pennsylvania State University seeks to fi ll multi ple faculty positi ons beginning in the 2015–16 academic year: Missouri • Open-Rank Tenure-Track positi on(s): We welcome applicants in any area of stati sti cs, with parti cular interest n in those areas that strengthen existi ng interdisciplinary ti es, which include but are not limited to astrostati sti cs, University of Missouri Department bioinformati cs and “omics” sciences, and computati onally oriented cross-disciplinary research (see htt p://ics.psu. of Statistics is seeking one tenure-track edu/what-we-do/hire/). assistant professor starting fall 2015. • Director of Undergraduate Program: The successful candidate will lead a robust and growing program that Exceptional candidates considered at rank currently boasts about 150 students majoring in stati sti cs and an additi onal 150 minoring in stati sti cs. Possibiliti es also exist for leadership in our sizable undergraduate service teaching mission, both in-residence and online, of associate professor. PhD in statistics or depending upon interest. This non-tenure-track faculty positi on includes the same teaching load as our tenure- related field required by August 15, 2015. track faculty.

Apply online http://hrs.missouri.edu/find-a- Requirements for these positi ons include a PhD in stati sti cs or a related fi eld, excellent teaching, and research/ job/acamic. Cover letter, CV, and transcripts undergraduate educati on credenti als as appropriate to the positi on. Screening begins in mid-November of 2014. required. Email three letters of reference to The Pennsylvania State University is located in the center of the state, in a valley surrounded by the Appalachian Mountains and state forestland. The adjoining town of State College is part of a metropolitan area with ample health [email protected]. Deadline care, indoor and outdoor recreati on, and excellent quality of life. All the ameniti es of a metropolitan area—fi rst-rate December 1, 2014. The University public transportati on, world-class theater and concert events, advanced technology and research faciliti es—are available of Missouri is an Equal Opportunity/ without the att ending stress. Applicants must apply online and complete the Penn State applicati on at htt p://psu.jobs/ (search job #53770 for the Affirmative Action/ADA Employer. tenure-track positi on; search #53748 for the undergraduate program positi on) and must apply online and submit an applicati on and materials, including cover lett er, CV, and reference lett ers, through mathjobs.org (htt ps://www.mathjobs. n Assistant/Associate Professor org/jobs). Additi onal informati on about the department can be found at htt p://www.stat.psu.edu/. Statistics with Emphasis in Biological CAMPUS SECURITY CRIME STATISTICS: For more about safety at Penn State, and to review the Annual Security Report Science. Statistics & biological sciences which contains informati on about crime stati sti cs and other safety and security matt ers, please go to htt p://www.police. departments, University of Missouri, psu.edu/clery/, which will also provide you with details on how to request a hard copy of the Annual Security Report. announce a tenure joint position with Penn State is an equal opportunity, affi rmati ve acti on employer, and is committ ed to providing employment emphasis on current research trends in opportuniti es to minoriti es, women, veterans, disabled individuals, and other protected groups. biology and a research program that

40 amstat news november 2014 complements each department's and experience in a clinical trial environ- n Merck, a global health care leader campus strengths. Required: Statistics ment. Apply online at www.merck.com/ with a diversified portfolio of prescrip- or biostatistics PhD. Apply http://hrs. careers. Req# STA000521 EOE. tion medicines, vaccines and consumer missouri.edu/find-a-job/academic. Cover health products, as well as animal health letter, vita, research & teaching state- n Assistant Professor, Biomathematics/ products seeks a: senior scientist, biosta- ments, three references. Deadline: Statistical Computing, Rowan tistics, Rahway, NJ. Req# BIO004013. December 1, 2014. The University University. Full-Time Tenure-Track. — Required: Master's degree in statistics of Missouri is an Equal Opportunity/ PhD in bomathematics, statistics or or related field. — Minimum of 3 years Affirmative Action/ADA Employer. related disciplines. Responsibilities experience in statistical planning and data include teaching mathematics and data analysis. Apply online at www.merck.com/ New Jersey analytics. Must demonstrate ability to careers. Req# BIO004013. EOE/AA. n Merck, a global health care leader conduct research in computational biol- with a diversified portfolio of pre- ogy/statistical computing, with applica- n Merck, a global health care leader with scription medicines, vaccines and tions to biological or biomedical sci- a diversified portfolio of prescription consumer health products, as well as ences. Must show potential for obtaining medicines, vaccines and consumer health animal health products seeks an associ- external grants, and help develop courses products, as well as animal health prod- ate principle programmer, Rahway, NJ. in a master's program in biomathemat- ucts seeks a: senior programmer, Rahway, Req# STA000521- Required: bachelor's ics. Application deadline November NJ. Req# STA000522. — Required: degree in computer science, statistics, 15th. Details and procedure: http:// bachelor's degree in computer science, sta- biological sciences, or engineering- rowanuniversity.hodesiq.com/job tistics, biological sciences, or engineering. Minimum of 9 years SAS programming -details.aspx?jobid=4620400. EOE. Minimum of 5 years SAS programming

BIOSTATISTICS FACULTY POSITION e School of Public Health is seeking quali ed candidates for a tenure track faculty of biostatistics in the Department of Epidemiology and Biostatistics. is position is 12-month, open-rank, and state-funded with a possible start date of January 2015. Applicants are expected to hold a Ph.D., Sc.D., or equivalent degree in biostatistics or statistics. Applicants with interests in all areas of health-related statistical research are highly encouraged to apply. Particular interest is in those candidates with consulting and/or online teaching experience. Pending budgetary approval, a second position that focuses on collaborative research in cancer will be available. e successful candidates will be expected to contribute to the research, teaching and service mission of the Department and the School.

Current departmental statistical expertise is in biological processes, survival and longitudinal data analysis, semiparametric inference, bioinformatics, measurement error models, radiation epidemiology, statistical proteomics, and statistical computing. Faculty members actively collaborate with colleagues in the Department of Statistics, College of Medicine, Veterinary Medicine, and other departments of the school.

e School of Public Health is located in College Station, and is part of the Texas A&M Health Science Center. e Health Science Center o ers a supportive environment for professional development as a unit of Texas A&M University, a research-intensive  agship university with more than 50,000 students — including 10,000 graduate students — studying in more than 120 undergraduate and 240 graduate degree programs in ten colleges. e A&M System is one of the largest systems of higher educa- tion in the nation, with a budget of $3.8 B.

Interested individuals should submit a letter describing their interest in the position, research and teaching experiences deemed most relevant to this position, current curriculum vitae, and three references (letters will be requested at a later date) to

Search Committee for Biostatistics Faculty Position c/o Samantha Payton Department of Epidemiology and Biostatistics Texas A&M Health Science Center, School of Public Health 1266 TAMU College Station, TX 77843-1266 or email: [email protected].

Review of applications will begin immediately and continue until the position is  lled. e TAMHSC SRPH is an A rmative Action/Equal Opportunity employer; women and minority individuals as well as persons with disabilities are actively encouraged to apply.

november 2014 amstat news 41 experience in a clinical trial environment. Apply online at www.merck.com/careers Req# STA000522. AA/EOE.

n Merck, a global health care leader with a diversified portfolio of prescrip- tion medicines, vaccines and con- sumer health products, as well as animal health products seeks an associate prin- ciple programmer, Upper Gwynedd, PA. Req# STA000520. — Required: Bachelor's degree in computer science, statistics, biological sciences, or engi- neering. — Minimum of 7 years experi- ence using SAS in clinical trials setting. Apply online at www.merck.com/careers Req# STA000520. AA/EOE.

n Assistant Professor, Biostatistics, Rowan University. Full-time, tenure- track. PhD in biostatistics or related discipline. Responsibilities include teaching undergraduate and graduate statistics courses. Must demonstrate ability to conduct research in biostatis- tics or statistical theory and/or meth- odology applicable to the health sci- ences and medical research. Must show potential for obtaining external grants and contributing to new graduate data analytics program. Application deadline November 30. Details and procedure: http://rowanuniversity.hodesiq.com/job- details.aspx?jobid=4772480. EOE.

n The department of statistics and bio- statistics and the Center for Integrative Proteomics Research at Rutgers University seek to hire an outstanding tenure-track assistant professor with strong research and teaching interests in statistical analyses of biological and biomedical information. Applicants must have a PhD in statistics or related field by Sept. 1, 2015. Apply online through the Rutgers Interfolio website link (http://apply.interfolio.com/25963). Rutgers is an Equal Opportunity/ Affirmative Action Employer.

42 amstat news november 2014 New York Provides statistical expertise for study 43403-0267 postmarked by Dec. 12, n NYU Stern School of Business design, statistical analysis and data analy- 2014. Visit www.bgsu.edu/business/asor. Statistics Group, tenure-track assistant sis. MS + expertise with SAS or R. Mail html. BGSU is an AA/EO Employer and professor appointment in statistics. resumes to: Jill Ortman, Lerner Research encourages applications from women, Candidates should have evidence of Institute, Cleveland Clinic, 9500 Euclid minorities, veterans, and individuals with boundary-spanning interests across Ave., ND46, Cleveland, OH 44195. No disabilities. fields that reflect significant interfaces calls (principals only). EOE. of statistics with areas of relevance in a Oregon business school. Expected that candi- n Tenure-track assistant professor posi- n The Fariborz Maseeh Department of date will be productive researcher and tion beginning August 2015. PhD in Mathematics and Statistics of Portland effective teacher at both undergraduate statistics/OR/management science/related State University invites applications for and graduate levels. See http://apply. quantitative area, demonstrated potential a tenure-track, associate-level Maseeh interfolio.com/25737 for full details, for excellence in teaching and research. Professorship in Mathematical Sciences including information on application Research and experience in data mining/ to begin fall 2015 (Position # D93383). procedure. New York University is business analytics highly desirable. Send For full job description and applica- an Affirmative Action/Equal cover letter, CV, three recommendation tion instructions, see Portland State Opportunity Institution. letters, official transcripts to cmathis@ University's employment website: bgsu.edu or The Search Committee, https://jobs.hrc.pdx.edu/postings/13663. Ohio Department of Applied Statistics and Consideration of applications will begin n Biostatisticians (statisticians) (four Operations Research, Bowling Green Nov. 15, 2014, and continue until the openings) needed in Cleveland, Ohio. State University, Bowling Green, OH position is filled. AA/EOE.

Academia Sinica Institute of Statistical Science Research Positions

The Institute of Statistical Science, Academia Sinica, is seeking candidates for tenure-track or tenured research positions at the level of DEPARTMENT HEAD AND PROFESSOR OF STATISTICS assistant or associate research fellow available in 2015. Candidates in all areas of Statistics The College of Science at Virginia Tech is seeking an accomplished, visionary head for its Department will be considered. Candidates should have a of Statistics. The successful candidate will be appointed as a tenured full professor in the department PhD degree in statistics or areas related to data at our Blacksburg campus and will be expected to teach graduate and/or undergraduate courses. sciences. Application materials must include (1) a curriculum vitae, (2) three letters of recom- mendation, and (3) representative publications All applicants must: 1) show evidence of visionary leadership in supporting and promoting and/or technical reports. Additional support- excellence in teaching, research, and outreach, 2) have an outstanding academic record includ- ing materials such as transcripts for new PhD ing an internationally recognized level of professional achievement and commensurate funding degree recipients may also be included. Except appropriate to appointment at the rank of full professor with tenure at Virginia Tech, 3) have the for the letters of recommendation, electronic ability to eff ectively mentor and manage human and fi nancial resources, 4) have an earned doctor- submissions are encouraged. Applications ate in statistics, biostatistics, or closely related area, 5) have demonstrated the ability to advise and should be submitted to teach students and to work eff ectively with a diverse faculty, and 6) travel on occasion to attend professional meetings and conferences. The hiring process requires the candidate to successfully Dr. Ching-Kang Ing pass a criminal background check. Chair of the Search Committee Institute of Statistical Science, Academia Sinica 128 Sec. 2 Academia Road, Applicants must complete the faculty application online at https://listings.jobs.vt.edu/postings/51540. Taipei 11529, Taiwan, R.O.C. To be included are a cover letter, a curriculum vitae, the names of three references and a statement Fax: +886-2-27831523 summarizing the applicant’s administrative philosophy and vision for the position and the depart- E-mail: [email protected] ment. Review of applications will begin on October 15, 2014 and will continue until the position is fi lled. Direct inquiries to Bill Woodall, Chair of Search Committee, [email protected], (540) 231-7792. Applications should be received by December 31, 2014 for consideration. Virginia Tech is an EO/AA university. Individuals with disabilities desiring accommodation in the ap- plication process should notify Ms. Betty Higginbotham, Department of Statistics, (540) 231-5657, or For more information, please visit http://www. call TTY 1-800-828-1120 stat.sinica.edu.tw/statnewsite/?locale=en_US

november 2014 amstat news 43 Open Rank Tenure-Eligible Faculty Positions Pennsylvania Department of Biostatistics n The Wharton Statistics Department, Positions #F38130 and #F38140 University of Pennsylvania, seeks appli- cants for a full-time, tenure-track assistant professor position, appointment begin- The Department of Biostatistics at Virginia Commonwealth University (VCU) is seeking to fi ll two tenured/tenure-eligible faculty positions at the level of assistant, associate, or full profes- ning July 2015. Candidates should show sor. For position #F38130, we are seeking applicants with training and research interest in the outstanding capacity and achievement design and statistical analysis of high-throughput genomic data (e.g., next generation sequenc- in research, along with excellent teach- ing, microarray, proteomic technologies), bioinformatics, computational biology, or closely ing skills. Applicants must have a PhD related area. For position #F38140, the candidate should have interdisciplinary research (expected completion by June 30, 2016, interests with the potential for external funding and scholarship that complements existing is acceptable) from an accredited institu- expertise in the department. Research areas of particular interest for position #F38140 include (but are not limited to) statistical learning and data mining, non-linear modeling, longitudi- tion. Please visit our website to apply: nal and hierarchical modeling, survival analysis, stochastic processes, Bayesian statistics, https://statistics.wharton.upenn.edu/ statistical computing, and risk analysis. Additionally, applicants for both positions should have recruiting/facultypositions. Questions can be collaborative research experience. Faculty are expected to maintain extramural grant support, sent to [email protected]. teach and advise graduate students, and provide departmental and university service. edu. The University of Pennsylvania is an EOE. Minorities/women/individuals The Department of Biostatistics has a 40+ year history in the VCU School of Medicine and is committed to excellence in both biostatistical research and graduate education. The depart- with disabilities/protected veterans are ment offers both M.S. and Ph.D. programs in Biostatistics, including a concentration in Ge- encouraged to apply. nomic Biostatistics and a M.S. in Clinical Research in Biostatistics. Our biostatistics faculty, students, and staff collaborate with clinical investigators on the Medical College of Virginia Utah Campus (which includes the Schools of Medicine, Dentistry, Pharmacy, Nursing, and Allied n The department of mathematics at Health) in a wide variety of biomedical research projects. Located in Richmond, Virginia, the University of Utah invites applica- VCU has established relationships with the Virginia Department of Health as well as local and regional health departments. In addition to other computational resources at VCU, our depart- tions for the following faculty posi- ment supports its own high-performance computing cluster. tions: Full-time tenure-track or tenured appointments at the level of assistant, Qualifi cations: For all levels, candidates should have a Ph.D. in biostatistics, statistics or a re- associate or full professor in all areas of lated fi eld. Applicants must demonstrate research experience in one or more of the focus areas, mathematics and statistics. Applications have a record of publication/presentation in the area of expertise, have teaching experience must be completed through the website preferably at the graduate level, and excellent oral and written communication skills. www.mathjobs.org/jobs/Utah, and will be By Level of Appointment: accepted until the position(s) have been Full Professor: Applicants should have an established track record publishing in peer-re- filled. For more information see www. viewed journals, have national or international prominence in their area of expertise, and have math.utah.edu/positions. The University demonstrated experience obtaining extramural research support. of Utah is an AA/EOE and is committed to diversity in its work force. In compli- Associate Professor: Applicants should have an established track record publishing in peer- reviewed journals and have demonstrated experience obtaining extramural research support. ance with applicable federal and state laws, university policy of equal employ- Assistant Professor: Applicants should have at least two years of experience beyond ment opportunity prohibits discrimina- completion of their degree program and must demonstrate excellent oral and written com- tion on the basis of race or ethnicity, munication skills. religion, color, national origin, sex, age, sexual orientation, gender identity/ All candidates should have demonstrated experience working in and fostering a diverse faculty, staff, and student environment or commitment to do so as a faculty member at VCU. expression, veteran’s status, status as a Potential candidates can submit applications noting the position number, including a statement qualified person with a disability, or of research, teaching philosophy, curriculum vitae and contact information for three profes- genetic information. Individuals from sional references, via mail – to Mrs. Yvonne Hargrove, Department of Biostatistics, Virginia historically underrepresented groups, Commonwealth University, P.O. Box 980032, Richmond, VA 23298-0032 – or by e-mail to such as minorities, women, qualified per- [email protected]. sons with disabilities, and protected vet- Virginia Commonwealth University is an equal opportunity/affi rmative action employer. erans are strongly encouraged to apply. Women, minorities and persons with disabilities are encouraged to apply. Veterans’ preference is extended to quali- fied applicants, upon request and consis- tent with university policy and state law.

44 amstat news november 2014 Assistant Professor of the Practice The Department of Statistical Science at Duke University invites applications for faculty ap- n Faculty, Assistant Professor — or related field, and research focus in data pointment at the level of Assistant Professor of the Practice to begin in Fall 2015. This position Biostatistics. The department of math- analytics, machine learning, stochastic is a regular rank faculty with a term renew- ematics at Utah Valley University seeks modeling, or related computationally able appointment. Preference will be given to applicants at the rank of assistant profes- intensive statistical methods. Further candidates demonstrating outstanding teaching sor (tenure track) of biostatistics start- information is available at the links and strong interests in developing and growing ing August 2015. Applicants having above. Applications must be submitted our undergraduate major in Statistical Science and our MS in Statistical Science. We are also specialization in the area of biostatistics online at http://listings.jobs.vt.edu (posting interested in applicants with complementary or similar will be preferred. Major duties #TR0140103). Virginia Tech is an EOAA interests in Bayesian statistical science research comprise teaching statistics, serving as university, and offers a wide range of net- and collaborative applications. a consulting statistician, and pursuing working and development opportunities scholarly activities. To apply, visit www. to women and minorities in science and The Department is an internationally recognized center of excellence for research and educa- uvu.jobs. Utah Valley University is an engineering. tion in contemporary statistical methodology. Equal Opportunity Employer. With leading strengths in Bayesian analysis, n Department of statistics, University interdisciplinary applications and computation- Virginia of Virginia, invites applications for two ally intensive methods, the Department offers n The Virginia Tech Department tenure-track positions (assistant/associ- outstanding teaching support, computational facilities, and opportunities for interdisciplinary of Statistics (www.stat.vt.edu) has a ate/full professor) beginning fall 2015. statistics teaching and collaboration. We cur- tenure-track opening to support its Applicants must present evidence of out- rently have 18 regular rank faculty, 16 visiting, Computational Modeling and Data standing accomplishments and promise adjunct, and postdoctoral faculty, and over 75 Analytics program (www.science.vt.edu/ in research and teaching. Candidates are graduate students. ais/cmda). Required are a PhD in statistics expected to teach graduate and under- The educational program (graduate and under- graduate) as well as the Department’s research agenda benefi t from strong connections with many other groups at Duke, with the Statistical and Applied Mathematical Sciences Institute (SAMSI) and the National Institute of Statistical Sciences (NISS) located nearby in the Research Triangle, and with other collaborators, centers, companies and organizations nationwide. Visit http://www.stat.duke.edu for more information.

To apply, submit a letter, curriculum vitae, personal statement of teaching and research, and three reference letters via https://academicjob- sonline.org/ajo. For inquiries and e-mail cor- respondence please write to [email protected]. edu. The application pool will remain open until the position is fi lled; screening will begin on 1 December 2014.

Duke University, located in Durham, North Carolina, is an Affi rmative Action/Equal Op- portunity Employer committed to providing employment opportunity without regard to an individual’s age, color, disability, genetic infor- mation, gender, gender identity, national origin, race, religion, sexual orientation, or veteran status. Applications from women and minori- ties are strongly encouraged. Individuals in dual career couples are encouraged to visit http:// provost.duke.edu/faculty/partner/, the website on Duke’s Advantages for Faculty, for information on opportunities for dual career couples in the area and how the university can help.

november 2014 amstat news 45 graduate courses and provide service. We tact the search chair, Al Best, albest@ jobs/6473822/tenure-track-assistant seek candidates dedicated to UVA's mis- vcu.edu. Virginia Commonwealth -professor. Hong Kong University of sion of excellence in research and teach- University is an Equal Opportunity/ Science and Technology is an Equal ing. UVA is EO/AA Employer. See www. Affirmative Action University. Opportunity Employer. stat.virginia.edu. University of Virginia is an EO/AA Employer. International n Wang Yanan Institute for Studies in n Tenure-track assistant professor for Economics & School of Economics, n Biostatistician, Department of business statistics, Dept. of ISOM, Xiamen University, China. Full-time, Periodontics and Office of Research, HKUST. Demonstrated excellence in tenure-track/tenured professorship posi- Virginia Commonwealth University research and teaching and a doctoral tions in statistics. Preferred areas of spe- School of Dentistry. Position degree by July 1, 2015, are required. cializations are theoretical and applied F37990. Opening for a master's-level Prior business school experience or statistics. PhD degree in statistics or biostatistician in the school of dentistry. interests in business related statisti- probability theory must be completed Master's degree, 3–5 years of bio- cal research are especially welcome. by Aug. 1, 2015. Send applications, medical research experience required. Excellent computational skills in han- including cover letter, CV, samples Candidates must apply through the dling large and complex data sets is of research work, and three reference university's career website: www.vcujobs. a plus. Submit CV and three referees letters to [email protected] com. For further information, con- to: [email protected]. jobs.amstat.org/ before Nov. 30. EOE.

LearnSTAT OnDemand

Start enjoying additional benefits! • LearnSTAT OnDemand: FREE to all Accredited Professional Statisticians™ • JSM Meetings: 20% discount on regular professional developement offerings • Special JSM Recognition: PStat® designees enjoy a special ribbon designation at all future Joint Statistical Meetings, as well as a special PStat®-only reception at JSM • Conference on Statistical Practice: 20% discount for registration and short courses

Interested in becoming an Accredited Professional pstat Statistician™? Start the process today! ® TM Accredited Professional Statistician www.amstat.org/accreditation

46 amstat news november 2014 Visit the ASA JobWeb online Ontario n The department of statistics and actu- n The department of statistics and actu- arial science, University of Waterloo, arial science, University of Waterloo, TODAY! invites applications for one tenure-track invites applications for two tenure-track http://jobs.amstat.org or tenured position in actuarial science. or tenured positions in statistics or bio- PhD in area of actuarial, statistical, or statistics. PhD in any area of statistics mathematical sciences; research in actu- or biostatistics required. Apply through arial science or related disciplines. Apply www.mathjobs.org/jobs. Include cover let- through www.mathjobs.org/jobs. Include ter, CV, research/teaching statements, up cover letter, CV, research/teaching state- to three reprints/preprints, and three ref- JOBS ments, up to three reprints/preprints and erence letters. Full advertisement https:// three reference letters. Full advertisement math.uwaterloo.ca/statistics-and-actuarial https://math.uwaterloo.ca/statistics -science/available-positions. Closing: -and-actuarial-science/available-positions. Dec. 1, 2014. AA/EOE. n Closing: Dec. 1, 2014. EOE/AA.

november 2014 amstat news 47 NOVEMBER 2014 • ISSUE #449

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 SIAM...... centerfold

professional opportunities Academia Sinica Institute of Statistical Science...... p. 43 Duke University...... p. 37, 45 Fred Hutchinson Cancer Research Center...... p. 36 Johns Hopkins Bloomberg SPH...... p. 39 North Carolina State University...... p. 45 Penn State...... p. 40 SAMSI...... p. 38 Texas A&M University...... p. 41, 42 University of Iowa...... p. 46 The University of North Carolina at Chapel Hill...... p. 34, 40 U.S. Census Bureau...... p. 35 VCU Medical Center...... p. 44 Virginia Tech...... p. 43 Westat...... p. 42

software Cytel Inc...... p. 8 Dell Software...... cover 2 Minitab Inc...... centerfold NCSS...... p. 11 Salford Systems...... centerfold SAS Institute, Inc...... cover 3 SAS JMP software...... cover 4 Statistical Solutions...... p. 26

48 amstat news november 2014