AMSTATNEWS the Membership Magazine of the American Statistical Association •

April 2018 • Issue #490 AMSTATNEWS The Membership Magazine of the American Statistical Association • http://magazine.amstat.org

MATH & STATISTICS AWARENESS MONTH DATA Is My Four ASA members JOBshare career insights

ALSO: The Value of CSP Climate Science Day Participants See Change of Tone on Capitol Hill Thank you 2018 SPONSORS

PLATINUM

GOLD

SILVER

For sponsorship opportunities, visit ww2.amstat.org/meetings/jsm/2018/sponsors AMSTATNEWS APRIL 2018 • ISSUE #490

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 Recognizing the ASA’s Longtime Members Director of Strategic Initiatives and Outreach 13 Longtime Members Offer Wisdom Donna LaLonde: [email protected] 16 Q&A with Industry Leaders Director of Education Rebecca Nichols: [email protected] 22 STATCOM: Revitalization of Statistical Community Service at Universities Managing Editor Megan Murphy: [email protected] 24 Math and Statistics Awareness Month: Data Is My Job Editor and Content Strategist 28 Master’s and Doctoral Programs in Data Science and Analytics Val Nirala: [email protected] 33 Climate Science Day Participants See Change of Tone Production Coordinators/Graphic Designers Sara Davidson: [email protected] on Capitol Hill Megan Ruyle: [email protected] 35 Committee on Privacy and Confidentiality Offers Resources Advertising Manager Claudine Donovan: [email protected]

Contributing Staff Members Amanda Conageski • Amy Farris • Naomi Friedman

Amstat News welcomes news items and letters from readers on matters of interest to the association and the profession. Address correspondence to columns Managing Editor, Amstat News, American Statistical Association, 732 North Washington Street, Alexandria VA 22314-1943 USA, or email amstat@ amstat.org. Items must be received by the first day of the preceding month 36TAT S tr@k 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. On Biostatistics, Alan Alda, and Being Mission Driven Articles will be edited for space. Accompanying artwork will be accepted in graphics file formats only (.jpg, etc.), minimum 300 dpi. No material in STATtr@k is a column in Amstat News and a website geared toward people who WordPerfect will be accepted. are in a statistics program, recently graduated from a statistics program, or recently Amstat News (ISSN 0163-9617) is published monthly by the American entered the job world. To read more articles like this one, visit the website at Statistical Association, 732 North Washington Street, Alexandria VA 22314- http://stattrak.amstat.org. If you have suggestions for future articles, or would like 1943 USA. Periodicals postage paid at Alexandria, Virginia, and additional to submit an article, please email Megan Murphy, Amstat News managing editor, mailing offices. POSTMASTER: Send address changes to Amstat News, 732 at [email protected]. North Washington Street, Alexandria VA 22314-1943 USA. Send Canadian address changes to APC, PO Box 503, RPO West Beaver Creek, Rich Hill, ON L4B 4R6. Annual subscriptions are $50 per year for nonmembers. Amstat News is the member publication of the ASA. For annual membership rates, 38 PASTIMES OF STATISTICIANS see www.amstat.org/join or contact ASA Member Services at (888) 231-3473. What Does Claire Kelling Like to Do When She Is Not Being American Statistical Association a Statistician? 732 North Washington Street Alexandria, VA 22314–1943 USA This column focuses on what statisticians do when they are not being statisticians. (703) 684–1221 If you would like to share your pastime with readers, please email Megan Murphy, ASA GENERAL: [email protected] Amstat News managing editor, at [email protected]. ADDRESS CHANGES: [email protected] AMSTAT EDITORIAL: [email protected] ADVERTISING: [email protected] WEBSITE: http://magazine.amstat.org 39 STATS4GOOD Printed in USA © 2018 Data for Good Researchers Fight Human Trafficking American Statistical Association This column is written for those interested in learning about the world of Data for Good, where statistical analysis is dedicated to good causes that benefit our lives, our communities, and our world. If you would like to know more or have ideas for articles, contact David Corliss at [email protected].

® 40 MASTER’S NOTEBOOK The American Statistical Association is the world’s largest The Important Role of the Master’s Statistician community of statisticians. The ASA supports excellence in in Clinical Trials the development, application, and dissemination of statistical science through meetings, publications, membership services, This column is written for statisticians with master’s degrees and highlights areas of education, accreditation, and advocacy. Our members serve in employment that will benefit statisticians at the master’s level. Comments and suggestions industry, government, and academia in more than 90 countries, should be sent to Megan Murphy, Amstat News managing editor, at [email protected]. advancing research and promoting sound statistical practice to inform public policy and improve human welfare. In honor of National Poetry Month, Larry Lesser of The University of Texas at El Paso offers this poem (on the trending topic of mindfulness): Mindful Means departments © 2018 Lawrence M. Lesser 42 meetings Reprinted with permission The Value of CSP 2018

An explanatory variable has a response and Symposium on Data Science and Statistics The space Promises Solid Program, Networking Before response is deemed Freedom, Sought by degrees: More time to reflect If randomness is Uniform, if correlation is

Causal, chance, or complexity yet 2018-MSAM-poster-front-FINAL.pdf 1 3/19/18 4:19 PM Unnamed. MATH & STATISTICS AWARENESS MONTH ® In the space DATA IS MY JOB To celebrate Mathematics and Statistics Awareness Month, we highlight four ASA To scan members who have used their statistics skills and imaginations to snag sweet jobs. ROB SANTOS Chief Methodologist What’s there, what else Urban Institute WHEN WE Knowledge SAY “DATA,” YOU SAY: Connecting Could be: mediator or moderator JANET PresidentMcDOUGALL SUPERPOWER: research question to design Tailor your to analysis to insight McDougall ScienticRÉSUMÉ HACK: Tailor your résumé for each job to which Variable. résumé to the job require- RÉSUMÉ HACK: ments and the company. you apply; include a brief bio that weaves your career and SUPERPOWER: The ability Sending the same résumé to educational paths in a way that shows your passion for to synthesize the science everyone is not the way to be statistics/research/methods/whatever and how it ts behind the research question, noticed. Showing, based on exactly into the position you are applying to. ... Employers Not bearing burden of being notice. and the FDA regulations that your experience and education, will need to be satised, into a how you match the require- robust and efcient statistical ments sets you apart—you Certain, we seek habit design, then explain to the make the reviewer’s job easier researcher how the design and it shows you put effort addresses each of these into the response. YIHUI XIE essential components— To inhabit uncertainty – backed up by solid literature. Software Engineer WHEN WE SAY “DATA,” YOU SAY: Integrity RStudio, Inc. To describe, not deny – RÉSUMÉ HACK: Don’t create a Leading static PDF. Create a web page HAVE GOOD ADVICE? NANCY POTOK SUPERPOWER: instead (print it to PDF when Share your answers by tagging @AmstatNews and #mathaware. Chief Statistician people to “the road less necessary) and update it as And fail to reject. traveled” by creating frequently as you can. To be able of the United States simpler software. to update it frequently, you have to do things, which is the only Ofce of ManagementRÉSUMÉ HACKand: IncludingBudget WHEN WE SAY “DATA,” way to make yourself prolic. results I personally achieved, YOU SAY: data.frame() targeted to the résumé reader. SUPERPOWER: I never send the exact same Procient active listening résumé twice. Conducting a study of one WHEN WE SAY “DATA,” Treasure Being present in the moment, YOU SAY: Not lost in time series, Letting noise and outliers pass, Check out the Math and Statistics Awareness Month Centered in breath, poster in this month’s centerfold. Eyes shut to better see our mode Beyond expectation, Sitting in the lotus, No longer standing for Law Of The Unconscious Statistician. member news 45 Section • Chapter • Committee News Online Articles 46 Professional Opportunities The following articles in this issue can be found online at http://magazine.amstat.org.

STATISTICS TEACHER has a new editor! Jessica Cohen is new to the role, but not to the journal — she acted as associate editor for ST’s predecessor, Follow us on Twitter Statistics Teacher Network. Check out the publication www.twitter.com/AmstatNews online at www.statisticsteacher.org. Join the ASA Community http://community.amstat.org

IN MEMORIAM Sadly, Sharon Anderson, Lawrence Like us on Facebook www.facebook.com/AmstatNews (Larry) Brown, and Truc Truong Nguyen passed away recently. To read these members’ obituaries, visit Follow us on Instagram http://magazine.amstat.org. www.instagram.com/AmstatNews

2 amstat news april 2018 president's corner

Statisticians Leading with Justice for All

n the past two issues of Amstat News, I have product statements are misleading is the subject of focused on building the ASA Leadership FDA guidance documents, but these determina- Institute. This month, I want to highlight tions often end up in court, where a company’s anotherI 2018 presidential initiative—expert wit- first amendment rights in making claims about a ness training. The idea for such a program came product are weighed against the agency’s need to from our membership. ensure such claims do not mislead the public. Early in 2017, Executive Director Ron A recent example is the Federal Trade Photo by Jon Gardiner/ by Photo UNC-Chapel Hill Wasserstein heard from several members about Commission’s suit against Quincy Bioscience Lisa LaVange whether the ASA could help prepare consulting that questioned the use of secondary analysis to statisticians for service as expert witnesses in a support a claim about a treatment for trial or deposition. Around the same time, I had memory loss after the trial failed occasion to talk with a former University of North on its primary endpoints (https:// Carolina biostatistics student, Naomi Brownstein, goo.gl/64zjP8). The FTC lost the and she described her interest in being an expert claim, and no statisticians were witness. Now a statistics faculty member at Florida involved as expert witnesses due State University, Naomi had been approached to the nature of the suit, but sta- about serving in this capacity. So, Ron and I put tistical theory about inference our heads together to assess the need for this kind in general and multiplicity in of training and came up with a proposal. particular formed the basis There are areas of the law that involve quanti- of the FTC’s argument. tative expertise. Ensuring there are qualified sta- Given the interest in tistical professionals in the courtroom or other- training among ASA mem- wise involved with the legal process in those areas bers, requests for ASA’s would improve the quality of the legal process and assistance as expert wit- increase recognition of the important contribu- nesses from other parties, tions of statistics and statisticians. and my own interest in There are many leaders in our field who regu- this topic, Ron and larly step up to serve the courts on a variety of I presented our pro- important topics, and our sister fields—such as posal for an expert mathematics—are also stepping up to contrib- witness training pro- ute. Gerrymandering of legislative districts, for gram to the board, instance, is one topic that has drawn attention of and thus was born a late (see https://goo.gl/HW6csM). Gerrymandering presidential initiative. struck a chord with me, living in two states Our vision is to (Maryland and North Carolina) that have been develop a program that accused of extreme partisan gerrymandering— provides the general skills but in opposite political directions. Other topics and knowledge a statistician include investigations of Medicaid fraud by medi- should have to be an expert cal providers and using “risk-limiting audits” to witness, as well as prepare par- detect problems with elections. These topics often ticipants to speak as an expert end up in court, and statisticians should be pre- on at least one subject matter pared to chip in. area. To this end, we formed One aspect of the FDA’s mission to protect a working group made up of and promote the public health that I did not fully leading statisticians Joseph L. appreciate until working there was the need to Gastwirth, Mary W. ensure product claims made in labeling and adver- Gray, Nicholas P. tising were accurate and did not mislead the pub- Jewell, and Rochelle lic. Evaluating the evidence to determine whether E. Tractenberg and

april 2018 amstat news 3 The recent creation of the National Institute of asked Kathy Ensor of Rice University to be the Standards and Technology–supported Center chair. The assembled team includes a lawyer and for Statistics and Applications in Forensic statisticians with courtroom experience. I asked Kathy, as chair, to report on the delib- Evidence (CSAFE, forensicstats.org) recognizes erations of the working group to date, and here is the important role statisticians play in forensic what she had to say: science and hence the legal system. The expert witness working group also noted emerging Leaders in our field have often provided expertise areas that include statistical and machine to the courts and Congress, many learning learning algorithms potentially guiding by experience. The objective of this training decisions of the courts and the criminal justice program for statistical expert witnesses is to help system. Equally as broad as the societal issues our community, especially those new to our statisticians are asked to address are the areas of profession, expedite the learning curve on how to statistics in which statisticians serve as experts. best serve the courts as an expert statistician. And given the demand for our expertise, we as a committee were reminded that one important component to serving as an expert is knowing There were exciting suggestions for what should when to decline a request. and should not be included in a training workshop. Although our experiences varied, several common themes emerged, including the following: We are developing a training program with the following key goals in mind: • What it means to be an expert witness • Quality – Develop a program that provides • What it means to be an expert statistical wit- excellent training and meets, or even exceeds, ness the needs of our members • Voicing a clear unbiased statistical opinion at • Impact – Develop a program that can reach a all stages of the legal process substantial number of people over time • Ethical considerations as practicing • Sustainability – Develop a program that can be professional statisticians offered regularly and support itself financially • Common mistakes and pitfalls to avoid An RFP for training program development will The role our profession has played in the courts be announced, pending approval by the ASA throughout history is laudable. The role of the Board. The general goal is to begin the program expert statistician emerged strongly in the 1980s. either in the fall of 2018 or spring of 2019. Once As a young statistician, I recall reading with great developed, this program will become part of the enthusiasm the text Statistics and the Law by ASA Leadership Institute, offered at a frequency [Morris] DeGroot, [Stephen] Fienberg, and [Joseph] deemed helpful to our community. Kadane and then later Jay Kadane’s 2008 book, Statistics in the Law: A Practitioner’s Guide, Cases, and My thanks go to Kathy and her team for the Materials. I guess this speaks to my love of statistics work accomplished so far, and I look forward and its application, as I read the books with the to seeing this program roll out in the coming deep immersion great novels require. months. I believe the program will be a valuable resource to ASA members, and we have certainly The vast array of areas in which statisticians interact heard from several who are anxiously awaiting its with the courts simply boggles the mind—areas inception. This is yet another example of an area such as employment discrimination, DNA, medical in which strong statistical leadership can have an practice, environmental issues, patent challenges, impact that extends far beyond our membership. economic risk, and financial fraud. So, here’s to statisticians leading with justice for all!

4 amstat news april 2018 Recognizing the ASA’s Longtime Members

he American Statistical Association would Vancouver, British Columbia, Canada, please like to thank its longtime members by con- join us for a reception in your honor. If your tinuing its tradition of honoring those who name is not below and you believe it should be, joinedT the association 35 or more years ago. This contact Amy Farris at [email protected] to correct year, we recognize the members here for their dis- your record. tinguished and faithful membership. Following this list is a Q&A with a few of our If you are a longtime member and will be longtime members—find out why they have attending the Joint Statistical Meetings in remained members of the ASA for so long. 50+ Y ears Khazan C. Agrawal Thomas J. Boardman Thomas E. Doerfler Jane F. Gentleman Myles Hollander Dennis J. Aigner Donald J. Bogue Norman R. Draper Jean D. Gibbons J. Stuart Hunter Murray A. Aitkin Jack R. Borsting Satya D. Dubey Rudy A. Gideon Hiroshi Ikeda Jack Alanen Alan Bostrom George T. Duncan Dennis C. Gilliland Arthur G. Itkin Philip J. Ambrosini Kimiko O. Bowman Douglas M. Dunn Phil D. Gilliland Gudmund R. Iversen Carol H. Ammons David R. Brillinger Arthur M. Dutton Leon J. Gleser Thomas B. Jabine Sigmund J. Amster Lyle D. Broemeling Morris L. Eaton V. P. Godambe Laurence F. Jackson Theodore W. Anderson Donna J. Brogan Bradley Efron Charles H. Goldsmith Ronald L. Jacobson Gary M. Andrew Maurice C. Bryson Jonas H. Ellenberg Arnold F. Goodman Aridaman K. Jain Charles E. Antle Charles R. Buncher Robert C. Elston Leo A. Goodman Bruce Johnston Barry C. Arnold Norman Bush William B. Fairley Donald Guthrie Richard Hunn Jones Joseph R. Assenzo John M. Chambers William R. Fairweather Irwin Guttman Karl G. Joreskog Barbara A. Bailar Herman Chernoff Kenneth H. Falter Gerald J. Hahn Joseph B. Kadane R. Clifton Bailey Janet E. Cherry Paul I. Feder Martin A. Hamilton Graham Kalton John J. Bartko Robert P. Clickner Charles Federspiel J. Wayne Hamman Marvin J. Karson Noel S. Bartlett Arthur Cohen Ivan P. Fellegi Roy Dean Hardy Marvin A. Kastenbaum Rex J. Bates Ayala Cohen Arlin M. Feyerherm William L. Harkness Shriniwas K. Katti David L. Bayless Theodore Colton Stephen E. Fienberg David A. Harville Gordon M. Kaufman John J. Beauchamp William Jay Conover David John Finney Kenneth Harwood Kathleen M. Keenan Gerald J. Beck Richard G. Cornell Richard L. Forstall Victor Hasselblad William J. Kennedy Donald F. Behan John W. Cotton Alan B. Forsythe Ronald W. Helms Jon R. Kettenring Donald L. Bentley David R. Cox Martin R. Frankel William G. Henderson Benjamin F. King Rudolf J. Beran J. R. Crespo Ralph F. Frankowski Neil W. Henry Elizabeth S. King-Sloan Alan P. Berens Ralph B. D’Agostino Donald A. S. Fraser Jay Herson John J. Kinney Mark L. Berenson Martin H. David Edward L. Frome Milton C. Heuston Melville R. Klauber Robert H. Berk Herbert T. Davis Carol Holly E. Fuchs John E. Hewett Michael H. Klein Donald A. Berry Miles Davis Wayne A. Fuller William J. Hill Gary G. Koch Elmer S. Biles John J. Deely A. Ronald Gallant Bruce Hoadley Uwe Koehn William C. Blackwelder Arthur P. Dempster Joseph L. Gastwirth Vincent Hodgson Mel Kollander Saul Blumenthal Frank T. Denton Donald P. Gaver Paul W. Holland Stephen L. Kozarich Colin R. Blyth Dennis O. Dixon David W. Gaylor Lawrence L. Kupper

april 2018 amstat news 5 Thomas E. Kurtz Curtis Meinert Rusi K.N Patell Richard L. Scheaffer Lowell H. Tomlinson Michael H. Kutner Edward L. Melnick Ganapati P. Patil Robert R. Scheer James Tonascia Ronald E. Kutscher Peter F. Merenda Edward B. Perrin David Schenker J. Richard Trout Peter A. (Tony) Paul W. Mielke Roger C. Pfaffenberger J. Richard Schmid Bruce E. Trumbo Lachenbruch G. Arthur Mihram Walter Piesch Stanley Schor Chris P. Tsokos John C. Lambert Jerry L. Moreno S. R. S. Rao Poduri Stanley L. Sclove N. Scott Urquhart Eugene M. Laska Donald F. Morrison Ralph D. Pollard Donald T. Searls Constance van Eeden William J. Latzko John W. Morse Richard F. Potthoff Daniel G. Seigel Pearl A. Van Natta Malcolm R. Leadbetter Mervin E. Muller John W. Pratt Robert J. Serfling Willem R. Van Zwet Eun S. Lee Thomas D. Murphy S. James Press Norman C. Severo James R. Veale Ferdinand Lemus Janet M. Myhre Charles H. Proctor Babubhai V. Shah Harvey M. Wagner Fred C. Leone Patricia L. Nahas Dana Quade William F. Shaw Ray A. Waller David Levine Longtime members Longtime Charles B. Nam J. N. K. Rao Monroe G. Sirken Robert C. Walls Eugene Levine Joseph I. Naus Joan S. Reisch Betty J. Skipper James A. Walsh Robert A. Lew Wayne B. Nelson Gladys H. Reynolds Armand V. Smith William G. Warren Thomas M. Lewinson Marc Nerlove Robert H. Riffenburgh William Boyce Smith Steve Webb James M. Lucas John Neter Larry J. Ringer Ronald D. Snee Bruce S. Weir Edward MacNeal Mark J. Nicolich Naomi B. Robbins Daniel L. Solomon Herbert I. Weisberg Brian D. Macpherson W. Michael O’Fallon Bruce E. Rodda Edward J. Spar Raymond L. Wilder Albert Madansky Anthony R. Olsen Charles A. Rohde Douglas E. Splitstone John Williams Richard Maisel Richard A. Olshen Joan R. Rosenblatt Stephen M. Stigler William H. Williams Colin L. Mallows Anthony M. Orlando Paul F. Ross George P. H. Styan Othmar W. Winkler Charles R. Mann Bernard Ostle Richard S. Ross D. Derk Swain Robert L. Winkler Nancy R. Mann Vernon E. Palmour Robert A. Rutledge Paul Switzer John J. Wiorkowski Jack A. Marshall Louis A. Panek Harold B. Sackrowitz Douglas B. Tang John E. Witcher Harry F. Martz Takis Papaioannou David S. Salsburg Elliot A. Tanis Janet Wittes John I. McCool Robert P. Parker Charles B. Sampson Judith M. Tanur John Harmon Wolfe Robert L. McKnight Emanuel Parzen Innis G. Sande James R. Thompson Donald E. Young Robert A. McLean James L. Pate Eberhard G. Schaich Leo J. Tick Calvin Zippin

45–49 years Abdelmonem A. Afifi Orley Ashenfelter Brent A. Blumenstein Raymond J. Carroll Robert J. Costello Robert A. Agnew Corwin L. Atwood Lennart Bodin Walter H. Carter Bradford R. Crain Alan Agresti Abdolrahman Azari Gordon J. Brackstone Edwin H. Chen Giles L. Crane Per A. T. Akersten Charles K. Bayne Edwin L. Bradley Raj S. Chhikara John R. Crigler Arthur E. Albert Laurel A. Beckett Gary L. Brager Joseph J. Chmiel David S. Crosby Mir Masoom Ali Richard J. Beckman William M. Brelsford Lee-Jay Cho Larry H. Crow Mukhtar M. Ali Mary S. Beersman Robert L. Brennan Domenic V. Cicchetti Jonathan D. Cryer Stan Altan Kenneth N. Berk Dwight B. Brock William S. Cleveland Gary R. Cutter Alfred Jerry Anderson U. Narayan Bhat Mark Brown Jerry L. Coffey Gerard E. Dallal Robert L. Andrews Peter J. Bickel John A. Burkart Guy M. Cohen James M. Davenport W. Tad Archambault Lynne Billard Kenneth P. Burnham Stanley H. Cohen Robert L. Davis Jesse C. Arnold Christopher Bingham Patricia L. Busk Kimon J.E. Constas Enrique de Alba Ersen Arseven David S. Birkes William L. Carlson R. Dennis Cook Timothy A. DeRouen James N. Arvesen John A. Blessing Margaret D. Carroll Lewis Coopersmith Susan J. Devlin

6 amstat news april 2018 Longtime members Jay L. Devore Silas Halperin Lynn Roy LaMotte Glenn L. Nelson Josef Schmee Paula H. Diehr Chien-Pai Han Kenneth C. Land Anna B. Nevius James Schmeidler W. Erwin Diewert R. Choudary Hanumara James M. Landwehr S. Edward Nevius William R. Schucany Darryl J. Downing Lynne B. Hare Per Lange David S. Newman Eugene F. Schuster Dennis A. DuBose Larry D. Haugh Kinley Larntz Earl Nordbrock Neil C. Schwertman Francois A. Dupuis Robert M. Hauser William D. Lawing Julia A. Norton Stuart Scott Benjamin S. Duran Douglas M. Hawkins Jerald F. Lawless Marija J. Norusis Nell Sedransk Danny Dyer William F. Heiland Anthony James Robert L. Obenchain Subrata K. Sen Robert G. Easterling Karl W. Heiner Lawrance Peter C. O’Brien Jolayne W. Service Brenda Kay Edwards Robert W. Hertz Kenneth D. Lawrence Jerry L. Oglesby Jayaram Sethuraman Janet D. Elashoff Agnes M. Herzberg Russell V. Lenth Leonard Oppenheimer Juliet Popper Shaffer Milton C. Fan Klaus Hinkelmann Yves Lepage J. Keith Ord Nagambal D. Shah Robert E. Fay David C. Hoaglin Donald Lewin Albert C. Ovedovitz Paul Shaman Walter Feibes Theodore R. Holford David L. Libby Maurice E. B. Owens Gary M. Shapiro Alan H. Feiveson David W. Hosmer Robert G. Lovell William J. Padgett Ronald E. Shiffler Alan C. Fisher David C. Howell Stanley E. Lunde Darrel W. Parke Iris M. Shimizu Andrew J. Flatt Paul B. Huber Lars Lyberg Leonard J. Parsons Albert P. Shulte Jairus D. Flora William F. Hunt George W. Lynch Jon K. Peck Robert H. Shumway Sandra Forman Frank L. Hurley James R. Maar Arthur V. Peterson Jon J. Shuster James W. Frane Huynh Huynh Bruce E. Mackey Charles G. Pfeifer Moshe Sicron Daniel H. Freeman Dar-Shong Hwang Dennis R. Mar Eswar G. Phadia Jagbir Singh Mark C. Fulcomer Ronald L. Iman Helen Marcus-Roberts Louis A. Pingel Nozer D. Singpurwalla Mitchell H. Gail Peter B. Imrey Robert L. Mason Mike Pore Walter Sloboda Edward J. Gainer William G. Jackson Takashi Matsui Stephen L. Portnoy Dennis E. Smith Daniel J. Gans David Jacobson Clement J. Maurath Ross L. Prentice Mitchell Snyder Fernando L. Garagorry F. E. James George P. McCabe Philip J. Press F. Michael Speed John A. Gaudiosi Sreenivasa Rao James B. McDonald Bertram Price Randall K. Spoeri Douglas O. Gause Jammalamadaka Lyman L. McDonald Philip C. Prorok M. K. Srirama Stephen L. George J. D. Jobson John D. McKenzie Thomas W. Pullum Muni S. Srivastava Gauri L. Ghai Clifford L. Johnson Glen D. Meeden Madan L. Puri Allan Stewart-Oaten Prabhakar D. Ghangurde Dallas E. Johnson Jeff B. Meeker David A. Pyne Robert L. Stout Edward J. Gilroy Richard A. Johnson James I. Mellon J. G. Ramage Jerrell T. Stracener Howard Seth Gitlow Paul K. Jones Gayle T. Meltesen Calyampudi R. Rao William E. Strawderman John R. Gleason John D. Kalbfleisch William L. Mietlowski George F. Reed Nariaki Sugiura T. F. Glover John H. Kalbfleisch George A. Milliken Benjamin Reiser Moon W. Suh Prem K. Goel Balvant K. Kale Satish Chandra Misra Louise C. Remer Michael Sutherland Judith D. Goldberg William D. Kalsbeek Robert Mondschein Richard D. Rippe Aaron Tenenbein Robert N. Goldman Howard S. Kaplon Douglas C. Bernard Rosner Ronald A. Thisted Montgomery J. Douglas Gordon Joseph D. Kasile Donald C. Ross Carol B. Thompson Roderick Montgomery Louis Gordon Myron J. Katzoff Donald B. Rubin Steven F. Thomson Billy J. Moore Bernard S. Gorman Thomas Keefe Barbara J. Rutledge Ram C. Tripathi David S. Moore David M. Grether David L. Kimble Julia Sabella Bruce W. Turnbull John K. Moore William E. Griffiths Roger E. Kirk Susan T. Sacks Neil R. Ullman Effat A. Moussa Joseph A. Guarnieri David C. Korts Francisco J. Samaniego Gerald van Belle Govind S. Mudholkar Richard F. Gunst Neal Koss Douglas A. Samuelson Joseph G. Van Matre Robb J. Muirhead Shelby J. Haberman Helena C. Kraemer Patricia D. Saunders Lonnie C. Vance Henry D. Muse Hermann Habermann Richard J. Kryscio James J. Schlesselman Wayne F. Velicer Wayne L. Myers Robert E. Hale Arabinda Kundu Joyce A. Schlieter Paul F. Velleman Subhash C. Narula

april 2018 amstat news 7 Hrishikesh D. Vinod George H. Wang Lynn Weidman David G. Whitmore Morty Yalovsky R. Lakshmi James F. Ward Sanford Weisberg George W. Williams Ann Graham Zauber Vishnuvajjala James H. Ware K. Laurence Weldon Douglas A. Wolfe Eric R. Ziegel Kenneth W. Wachter Edward J. Wegman James P. Whipple Robert F. Woolson Stuart O. Zimmerman Sylvan Wallenstein Owen Whitby Gooloo S. Wunderlich 40–44 years Dennis Aaron Ernst R. Berndt John P. Chandler William D. Dupont Malay Ghosh Robert D. Abbott David J. Bernklau Judith-Anne W. L. Marlin Eby David E. Giles Sandra C. Abbott Bibhuti B. Chapman Marlene J. Egger Phyllis A. Gimotty

Longtime members Longtime John M. Abowd Bhattacharyya Yogendra P. Chaubey Kathleen Louise Emery Dennis R. Givens Bovas Abraham Wayne F. Bialas Gina G. Chen Wil B. Emmert Beth C. Gladen Judith Abrams William T. Bielby Michael R. Chernick Curtis S. Engelhard Marcia A. Glauberman Lee R. Abramson Paul P. Biemer Nanjamma Chinnappa Thomas W. Epps Joseph Glaz C. J. Adcock Robert H. Bigelow Joan Sander Chmiel Samuel M. Epstein Avni Goeksel Frances J. Adox Richard A. Bilonick Jai Won Choi William H. Epstein Richard F. Goldstein Robert W. Aldred Jeffrey B. Birch George W. Cobb Eugene P. Ericksen Joe Fred Gonzalez Francis B. Alt Herbert L. Bishop Timothy C. Coburn James W. Evans James H. Goodnight Dallas W. Anderson Richard M. Bittman Steven B. Cohen Michael J. Evans Robert D. Gordon Keaven M. Anderson Jan F. Bjornstad James J. Colaianne Ray E. Faith Stephanie J. Green Robert J. Anderson Mark M. Blanchard John R. Collins Thomas B. Farver Timothy A. Green Sharon Anderson Peter Bloomfield Salvatore V. Colucci Alan Fask William H. Greene Bengtung Ben Ang Harvey Blumberg Kennon R. Copeland John P. Fazio Joel B. Greenhouse Lawrence Annable Dan C. Boger Margaret D. Ronald S. Fecso John Vic Grice Copenhaver Steve Ascher Robert J. Boik Martin Feuerman Susan Groshen Charles D. Cowan Taka Ashikaga James A. Bolognese David F. Findley Marvin H. J. Gruber Brenda G. Cox Anthony C. Atkinson Dennis Boos Carl Thomas Finkbeiner Leslie S. Grunes Marie V. Bousfield Andrew Joseph Agustin F. Ayuso Cucchiara Nicholas I. Fisher Victor M. Guerrero John E. Boyer Stephen P. Baker William G. Cumberland Allen I. Fleishman Perry D. Haaland Norman M. Bradburn Saad T. Bakir L. Adrienne Cupples Nancy Flournoy Timothy O. Haifley Ellen F. Brewer Vincent P. Barabba Robert D. Curley Hans-Theo Forst David B. Hall J. Michael Brick William A. Barnett Andrew I. Dale Peter E. Fortini James L. Hall David R. Bristol John L. Barone Prithwis Dasgupta Mary A. Foulkes Nancy R. Hall Ron Brookmeyer Michael P. Battaglia Roberta W. Day Janet F. Fowler David C. Hamilton Dean S. Bross Eileen J. Beachell Michael L. Deaton John D. Fox Janet M. Hanley Rocco L. Brunelle Patricia C. Becker Pierre C. Delfiner Leroy A. Franklin Robert C. Hannum Edward C. Bryant Richard A. Becker David L. DeMets Martin D. Fraser David Hardison Richard K. Burdick Jay H. Beder Lorraine Denby David Frontz William V. Harper John M. Bushery Steven Belle Wayne S. Desarbo Barbara A. Gabianelli Frank E. Harrell Thomas J. Bzik Robert B. Bendel Thomas F. Devlin Stephen J. Ganocy Stephen P. Harris Lawrence S. Cahoon James O. Berger E. Jacquelin Dietz Turkan K. Gardenier Kenneth R. Hartmann Patrick J. Cantwell Roger L. Berger David P. Doane Edward E. Gbur Maurine A. Haver Grant D. Capps Timothy M. Bergquist Allan P. Donner Robin T. Geiger Ronald W. Hawkinson Arthur Carpenter James S. Bergum Joseph R. Donovan Alan E. Gelfand Richard M. Heiberger Daniel B. Carr Catherine S. Berkey Janice L. Dubien Fredric C. Genter Lance K. Heilbrun Frank C. Castronova Jose Miguel Bernardo Joseph W. Duncan Cynthia D. Gentillon Harold V. Henderson Amrut M. Champaneri

8 amstat news april 2018 Ellen Hertzmark Beat Kleiner Kathleen S. Madsen H. Joseph Newton Philip H. Ramsey Longtime members Thomas Herzog Stuart A. Klugman Jay Magidson El-Sayed E. Nour Gopa Ray James L. Hess Kenneth J. Koehler Linda C. Malone Thomas S. Rose M. Ray Eugene R. Heyman Edward L. Korn Charles F. Manski Nunnikhoven William J. Raynor James J. Higgins Kenneth J. Koury Kanti V. Mardia Barry D. Nussbaum Kenneth J. Resser Steven C. Hillmer Abba M. Krieger Mary A. Marion David Oakes Mark William Riggs Susan M. Hinkins S. David Kriska Ray L. Marr Kevin F. O’Brien Paula K. Roberson Jerry L. Hintze Pieter M. Kroonenberg David B. Marx Ralph G. O’Brien Rosemary A. Roberts Raymond G. Hoffmann Naoto Kunitomo Donald L. Marx Michael W. O’’Donnell Jeffrey A. Robinson Thomas P. Hogan Robert Kushler Victor M. Matthews Judith Rich O’Fallon Frank W. Rockhold Alan Hopkins Alan H. Kvanli LeRoy T. Mattson Walter W. Offen Robert N. Rodriguez Berne Martin Howard John M. Lachin Timothy A. Max Patrick D. O’Meara Russell H. Roegner Ina P. Howell James R. Lackritz Margaret W. Maxfield Bernard V. O’Neill John E. Rolph Elizabeth T. Huang Nan Laird Scott E. Maxwell Terence John O’Neill Paul R. Rosenbaum Marla L. Huddleston Mansum A. Lam Michael J. Mazu Joyce Orsini James L. Rosenberger Mark Hudes Carol J. Lancaster Janet Elizabeth Melvin L. Ott N. Phillip Ross Mohammad F. Huque J. Richard Landis McDougall Willis L. Owen Lawrence V. Rubinstein David N. Ikle Nicolaas F. Laubscher Stephen A. McGuire Mari Palta David Ruppert Duane M. Ilstrup Philip T. Lavin Joseph W. McKean William S. Pan Estelle Russek-Cohen John M. Irvine Sheila M. Lawrence Geoffrey J. McLachlan Swamy A.V.B. Paravastu Carl T. Russell Alan J. Izenman Johannes Ledolter Christine E. McLaren Won J. Park Thomas P. Ryan Allen E. Izu Kelvin K. Lee Don L. McLeish Rudolph S. Parrish Michael S. Saccucci Kirk A. Jackson Kerry L. Lee William Q. Meeker Van L. Parsons Thomas W. Sager William E. Jackson Martin L. Lee Cyrus R. Mehta Jeffrey S. Passel John P. Sall David Jaspen James D. Leeper Robert J. Meier Kevin Pate William M. Sallas Jean G. Jenkins Stanley A. Lemeshow Kathleen A. Mellars Charles L. Paule Allan R. Sampson Linda W. Jennings Heryee H. Leong Roy Mendelssohn Karl E. Peace Gilles F. M. Santini Robert W. Jernigan James M. Lepkowski Michael M. Meyer N. Shirlene Pearson Thomas J. Santner Bruce E. Johnson Trudy J. Lerer Terry G. Meyer Raymond C. Peck Robert L. Santos Paulette M. Johnson Martin L. Lesser Joel E. Michalek Peter H. Peskun Nathan E. Savin Gerald A. Joireman Marcia J. Levenstein Richard O. Michaud A. John Petkau Richard L. Sawyer Ian T. Jolliffe Bruce Levin Mary-Jane Mietlowski Daniel Pfeffermann William G. Saylor David C. Jordan Charles Lewis John A. Miller John G. Phillips Patricia A. Scanlan Harmon S. Jordan Shou-Hua Li John Francis Monahan Philip J. Pichotta Stephen Schacht Henry D. Kahn Walter S. Liggett Katherine L. Monti Linda Williams Pickle Nancy K. Schatz Lee D. Kaiser Lawrence I-Kuei Lin Thomas F. Moore Joseph G. Pigeon Kenneth Schechtman Theodore G. Karrison Carol L. Link George E. Morgan Dale J. Poirier Mildred E. Schmidt Daniel Kasprzyk Robert E. Little David R. Morganstein William E. Pollard David A. Schoenfeld Richard W. Katz George A. Livingston Max D. Morris Darwin H. Poritz Timothy L. Schofield Robert M. Katz Greta M. Ljung Barbara G. Frank J. Potter Friedrich W. Scholz Mroczkowski Sheryl F. Kelsey Roger Longbotham Randall W. Potter John H. Schuenemeyer Lawrence H. Muhlbaier James L. Kenkel Michael T. Longnecker Manfred Precht Donald J. Schuirmann Leigh W. Murray James L. Kepner Thomas A. Louis Dale L. Preston Alastair John Scott John C. Nash Andre I. Khuri Milton W. Loyer Kevin Price William L. Seaver Elliott Nebenzahl Byung-Soo Kim Jay H. Lubin Lloyd P. Provost Joseph Sedransk Reinhard Neck Ignatius A. Kinsella Donald M. Luery John N. Quiring Teddy I. Seidenfeld Gary L. Neidert Nancy J. Kirkendall James Lynch Tony K. S. Quon Thomas R. Sexton James W. Neill Syed N.U.A. Kirmani John MacIntyre Alfred W. Rademaker Glenn R. Shafer Margaret A. Nemeth Rudolf G. Kittlitz Michael E. Mack Volker W. Rahlfs Arvind K. Shah

april 2018 amstat news 9 Mohammed A. Shayib Gerald R. Stewart John H. Thompson Denton R. Vaughan Robert M. Wharton Shingo Shirahata John A. Stewart Mary E. Thompson Niels H. Veldhuijzen Andrew A. White Patrick E. Shrout Robert A. Stine Theodore J. Thompson Joseph S. Verducci Roy W. Whitmore Andrew F. Siegel Sandra S. Stinnett Richard B. Tiller Hajime Wago Howard L. Wiener Richard S. Sigman Michael A. Stoto Ronald R. Titus Grace Wahba Rand R. Wilcox Judith D. Singer Miron L. Straf Jerome D. Toporek Joel A. Waksman Leland Wilkinson Robert D. Small Donna F. Stroup Robert D. Tortora Joseph J. Walker Jean F. Williams Martyn R. Smith Walter W. Stroup David C. Trindade Katherine K. Wallman Stephen R. Williams Murray H. Smith Perla Subbaiah J.R. Roger Trudel Stephen D. Walter William J. Wilson William A. Sollecito Richard A. Sundheim L. Claire Tsao Chao Wang Michael A. Wincek Dan J. Sommers Robert Sutherland Kam-Wah Tsui Sophronia W. Ward Lawrence C. Wolfe Keith A. Soper David A. Swanson Alan R. Tupek Herbert W. Ware Kirk M. Wolter

Longtime members Longtime Terence P. Speed Gerald R. Swope David L. Turner Stanley Wasserman Wayne A. Woodward Bruce D. Spencer Ajit C. Tamhane Gregory W. Ulferts William L. Weber Farroll T. Wright Clifford H. Spiegelman Robert M. Tardiff Jessica M. Utts William E. Wecker Tommy Wright Nancy L. Spruill Erica S. Taucher Pamela M. Vacek Thomas E. Wehrly Marvin Yablon Donald M. Stablein Marcia A. Testa Richard L. Valliant William W. S. Wei Michael G. Yochmowitz William M. Stanish A. Cole Thies Richard Craig Van Daniel L. Weiner Sarah T. Young Richard M. Stanley Hanspeter Thoeni Nostrand Jon August Wellner Daniel Zelterman Robert R. Starbuck John M. Thomas Kerstin Vannman Roy E. Welsch James R. zumBrunnen David W. Stewart Stephen B. Vardeman Fredrick S. Whaley

35–39 years Michael A. Adena Jenny A. Baglivo David K. Blough Lynda T. Carlson Peter D. Christenson Joseph John Bailer Carol Joyce Blumberg B. Thomas Carr Tin Chiu Chua Adwere-Boamah Steven P. Bailey Douglas G. Bonett John F. Carter Christy Chuang-Stein Dennis P. Affholter James A. Baldwin David E. Booth Nancy J. Carter Constance F. Citro Mohammad David L. Banks Richard C. Borden L. Douglas Case B. Christine Clark Ahsanullah Thomas A. Bass Victor Marek Borun Aki N. Caszatt Cynthia Z.F. Clark Adelin I. Albert Karin M. Bauer H. Christine Bourquin Deborah A. Cernauskas Murray K. Clayton James H. Albert Andrew Lewis Robert D. Bowser Subhabrata Daren B. H. Cline Jeanne M. Aldred Baughman Michael N. Boyd Chakraborti Avital Cnaan Melvin T. Alexander Moraye B. Bear Nancy J. Boynton Raymond L. Chambers Paul E. Coffman Rich Allen Mark P. Becker Rollin F. Brant Charles W. Champ Mark E. Cohen Paul D. Allison Paula J. Beitler Mary-Lynn Brecht Promod K. Chandhok Michael P. Cohen Wendy L. Alvey Alexander E. Belinfante James E. Breneman Richard A. Chechile Michael L. Cohen W. Gregory Alvord Michael E. Bellow Thomas W. Broene Chen-Hsin Chen Stephen H. Cohen Yasuo Amemiya Julia L. Benson Roger L. Brown James J. Chen Richard Daniel Cohn John Angle Peter M. Bentler Judith A. Buchino William W.S. Chen Hans Colonius Clifford W. Angstman Nancy Berman Shelley B. Bull Ching-Shui Cheng Charles F. Contant Thomas Arbutiski Debra H. Bernstein Christine M. Bunck Richard P. Chiacchierini Richard S. Conway Susanne Aref Charles C. Berry Lawrence F. Burant Yu-Kun Chiang Bruce K. Cooil Vincent C. Arena James Calvin Berry Thomas E. Burk Vernon M. Chinchilli Nancy R. Cook Stephan Arndt Jonas V. Bilenas Harry F. Bushar Paul C. Chiou Peyton J. Cook Arlene S. Ash Thomas E. Billings Richard J. Caplan Youn-Min Chou Patricia S. Costello Tim Baer Thomas R. Birkett David A. Carlin Ronald Christensen Lawrence H. Cox

10 amstat news april 2018 John R. Crammer Dean H. Fearn Michael S. Hamada David R. Judkins Steven A. Lewis Longtime members Keith N. Crank Michael B. Feil Herbert Hamilton Karen Kafadar Wai K. Li James A. Creiman Luisa T. Fernholz John B. Hannon Leslie A. Kalish Lillian S. Lin Noel A. Cressie G. Donald Ferree J. Michael Hardin Tzu-Cheg Kao Wayne S. Lindsay Douglas E. Critchlow Christopher A. Field Jeffrey D. Hart Roxanne Kapikian Barbara A. Lingg Leonard A. Cupingood Dianne M. Finkelstein Rachel M. Harter Bruce A. Kaplan Charles L. Liss Estella Bee Dagum Patrick E. Flanagan Nancy C. Hassett John M. Karon Jen-Pei Liu Robin A. Darton T. A. Foster Trevor J. Hastie Charles R. Katholi Wei-Yin Loh Marie Davidian Anne E. Freeny Gary D. Hatfield Barry P. Katz Stephen W. Looney Bruce M. Davis Larry D. Freese William D. Heavlin Darryl Katz James T. Love Charles S. Davis Stephen A. Freitas Charles E. Heckler Joseph L. Katz Joseph F. Lucke Richard A. Davis Frederick L. Freme Wolf-Dieter Heller Jerome P. Keating Helmut Luetkepohl Roger B. Davis Arthur Fries David H. Henry Sallie Keller Michael J. Luvalle Thomas M. Davis Shayne C. Gad Michael Hesney Elizabeth J. Kelly Michael F. Macaluso Willis L. Davis Paul Gallo Richard P. Heydorn Joan Kempthorne- Stephen P. Mack Rawson Thomas C. Dawe Michael A. Gates Susan G. Hilsenbeck David Peter Mackinnon Arthur J. Kendall Virginia A. de Wolf Constantine Gatsonis Chihiro Hirotsu Donald Macnaughton Harry J. Khamis Angela M. Dean Jeffrey J. Gaynor Joseph G. Hirschberg Laurentius Marais KyungMann Kim Roger L. Deaton Philip M. Gbur Edward C. Hirschland Mervyn G. Marasinghe John E. Kimmel Michael R. Delozier Joseph C. Gfroerer Douglas A. Hlavacek James C March Robin Laurence Kirby Richard A. Derrig Subir Ghosh Myron Hlynka Sue M. Marcus Genshiro Kitagawa Jeanne A. Devin John A. Gillespie Lorrie L. Hoffman David A. Marker John C. Klensin Marie Diener-West Michael E. Ginevan Howard R. Hogan Paul J. Marovich George J. Knafl Ralph Digaetano William J. Glynn Larry R. Holden James Martin Henryka K. Komanska Thomas W. Dobbins A. Blanton Godfrey Paul S. Horn Adam T. Martinsek David P. Kopcso John E. Donmyer Alfred D. Godfrey Carol C. House Michael P. Massagli Samuel Koslowsky Gerald A. Dorfman Susan C. Goebel Wei-Min Huang Joe Matsuoka Kallappa M. Koti Gaylen W. Drape Miguel A. Esther Sid Hudes Carl A. Mauro Ken G. Kowalski Kevin Ward Drummey Gomez-Villegas Arthur L. Hughes Fred M. Mayes Lawrence Krasnoff Bonnie P. Dumas Nancy M. Gordon Edward Hughes Charles Maynard Jeffrey P. Krischer Barry I. Graubard Paul R. McAllister Charles L. Dunn Allen C. Humbolt Alok Krishen J. Brian Gray Harold E. Dyck Luis H. Hurtado Katherine B. Krystinik Donna K. McClish Janis G. Grechko Jean L. Dyer Deborah D. Ingram Richard A. Kulka Joseph P. McCloskey Edwin J. Green Kirk A. Easley Henry F. Inman Lynn Kuo Kenneth F. McCue John W. Green Jeffrey H. Ebersole Patricia A. Jacobs Edward Lakatos Peter McCullagh Daniel A. Greer Robert G. Edson Paul D. James Kuang-Kuo Gordon Lan Daniel L. McGee James A. Grimes Don Edwards Denis George Janky Jurate M. Landwehr Philip G. McGuire David J. Groggel Thomas Barry Edwards Guillermina Jasso Jerry Langley Raymond E. McIntyre Shulamith T. Gross Bruce P. Ekholm Christopher Jennison Stephen S. Langley Kenneth B. McRae Berton H. Gunter Paul J. Elson Herbert Y. Jew Linda B. Lannom Shailendra S. Menjoge Ramesh C. Gupta John D. Emerson Karl-Heinz Jockel Lisa M. LaVange Michael Meredith Yesvy Gustasp Patricia A. English B. Alan Johnson Barbara A. Leczynski Samuel Merrill Sam Gutterman Eugene A. Enneking Gary R. Johnson Hyunshik J. Lee Marianne E. Messina Josue Guzman Neil R. Ericsson LuAnn K. Johnson Kwan R. Lee R. Daniel Meyer Michael Haber Kent M. Eskridge Robert E. Johnson John J. Lefante H. Andrew Michener Michael D. Hale Mark A. Espeland Albyn C. Jones Robert George Lehr Rosemarie Mick Marc Hallin Sylvia R. Esterby Karen C. Jones Greg M. Lepak William E. Mihalo David Fairley William A. Halteman Michael P. Jones Donald K. Lewis Eva R. Miller Frederick W. Faltin Katherine T. Halvorsen Shelton M. Jones Richard A. Lewis Michael F. Miller

april 2018 amstat news 11 Renee H. Miller J. Lynn Palmer Anthony M. Roman Mack C. Shelley Brian J. Thelen David H. Moen Deborah L. Panebianco Elvezio Ronchetti John T. Shelton David M. Thissen Leyla K. Mohadjer Sastry G. Pantula Robin L. Rose Mark R. Shenkman Neal Thomas Donna L. Mohr Corette Breeden Parker Gary L. Rosner Malcolm J. Sherman Peter James Thomson Robert J. Mokken Mary R. Parker Peter J. Rousseeuw Weichung J. Shih Anthony D. Thrall Leslie M. Moore Robert A. Parker Andrew L. Rukhin Lucy Shneyer Luke-Jon Tierney Walter T. Morgan Jeffrey R. Parno George C. Runger Gary L. Shoop Naitee Ting Elizabeth A. Robert E. Parson Keith F. Rust Stanley A. Shulman Ruey-Shiong Tsay Morgenthien Lee Parsons Roland T. Rust Arthur R. Silverberg Siu-Keung Tse June Morita Sharon M. Passe Steven W. Rust Jeffrey S. Simonoff Clyde Tucker Christopher H. Morrell Robert J. Pavur Jim Rutherford Terry L. Sincich Thomas P. Turiel David T. Morse Roxy L. Peck Pedro J. Saavedra Douglas R. Sizemore Thomas J. Uryniak Michael J. Morton William H. Sachs Christopher John Esa Ilkka Uusipaikka Longtime members Longtime Jane F. Pendergast Linda L. C. Moss Elgin S. Perry Jerome Sacks Skinner Mark J. VanRaden Tetsuro Motoyama Kimberly T. Perry Mehmet Sahinoglu Joan H. Skurnick Steven J. Verhulst Mohamed Ahmed John D. Pesek Ulderico Santarelli Charles Eugene Smith Steve P. Verrill Amin Moussa David W. Peterson Michael J. Santulli Elizabeth C. Smith Joseph G. Voelkel Ronald P. Mowers John J. Peterson Adriano L. Sarmiento Richard J. Smith Robert L. Vogel Daniel H. Mowrey Joseph D. Petruccelli Miles M. Sato Richard L. Smith Joachim Vollmar Nitis Mukhopadhyay Gerald L. Phillips John W. Sawyer Robert A. Smith Edward F. Vonesh Keith E. Muller Walter W. Piegorsch Stephen M. Scariano Stephen J. Smith Howard Wainer Alvaro Munoz Gregory F. Piepel David J. Schaeffer Tom A.B. Snijders Paul G. Wakim Jay Munson Chester H. Ponikowski Daniel W. Schafer Karen L. Snowdon-Way Lars Walloe Bengt Muthen Dudley L. Poston Nathaniel Schenker Ying C. So Chai-Ho C. Wang Haikady N. Nagaraja Paul N. Powell Mark J. Schervish Jose Francisco Soares Ann E. Watkins Jayalakshmi Natarajan J. Michael Price Mark F. Schilling Francisco P. Soler Carol Weideman William Navidi Louis H. Primavera Brian R. Schlain Eric R. Sowey David L. Weimer Larry Alan Nelson Howard M. Proskin David C. Schlotzhauer Floyd W. Spencer Clarice R. Weinberg Dean V. Neubauer Jamie K. Pugh Mark D. Schluchter John J. Spinelli William J. Welch Tie-Hua Ng William M. Pugh Paul R. Schneeman Gene D. Sprechini James G. Wendelberger Truc Truong Nguyen Clifford R. Qualls Daniel James Schnell Kadaba P. Srinath Joanne R. Joyce C. Niland James O. Ramsay John R. Schoenfelder Edward J. Stanek Wendelberger David Butcher Nolle Dabeeru C. Rao John D. Schoolfield Joel H. Steckel Glenn D. White Michael A. Nolte Richard F. Raubertas Donald E. Schreiner Leonard A. Stefanski David C. Whitford Phillip N. Norton Howard L. Rauch Charles B. Schriver Seth M. Steinberg David A. Whitney Robert M. Norton Domenic J. Reda Linda Kay Schultz Lorraine C. Steiner Dexter C. Whittinghill Douglas W. Nychka Nancy Reid Phyllis A. Schumacher Barbara Stevens Priya J. Wickramaratne Thomas W. O’Gorman William K. Rice Steven J. Schwager David S. Stoffer Christopher John Wild Francis G. Ogrinc Wasima N. Rida Lawrence A. Schwartz Maura E. Stokes William E. Wilkinson William P. O’Hare William J. Riley Sidney H. Schwartz S. Lynne Stokes Thomas R. Willemain Akinori Ohashi James S. Roberts Michael Schwarzschild Mark C. Strong Jeffrey R. Wilson Noboru Ohsumi Edwin L. Robison James R. Schwenke Mark Lionel Suda William E. Winkler Thomas H. Oliphant David M. Rocke David W. Scott Stephen R. Sulpor Luke G. Wolfe Frank Olken Richard A. Rode Joanne B. Severe James J. Swain Glenn S. Wolfgang John A. Ondrasik Mark H. Rodeffer Joseph Severs Yoshio Takane F. Lennie Wong George Ostrouchov Jack Rodgers Bahman Shafii Roy Noriki Tamura John Charles Wurst Soo Peter Ouyang Ward Rodriguez Ramalingam Deborah L. Tasky Emmanuel Yashchin William J. Owen Nestor Rohowsky Shanmugam Robert L. Taylor Linda J. Young Albert Palachek Javier Rojo Steven J. Shapiro George R. Terrell Elizabeth R. Zell Alberto Palloni

12 amstat news april 2018 Longtime Members Offer Wisdom We interviewed a few longtime members to find out why they remain members of the ASA and how they have benefitted from being a member. Here is what they had to say:

Janet Myhre Betty Skipper Member for more than 50 years Member since 1963 Being a mathematical applied statistician has allowed I majored in mathematics at Oberlin College, and I me to have a career in which I am continuously was the first in my family to graduate from college. learning, always involved in problem solving, and During my senior year in college, I saw an never bored. My first interest in statistics occurred as announcement on a bulletin board about a bio- a graduate student at the University of Washington. statistics graduate program at Western Reserve Janet Myhre Z. W. Birnbaum was instrumental in founding my University (WRU). A professor from that program interest in probability and mathematical statistics. came to campus that afternoon to talk to students. One of my careers has been as a professor I had never heard of biostatistics. It was explained of mathematics at the Claremont Colleges and as an opportunity to combine mathematics and University. My teaching involved instructing cours- science. The coursework consisted of two years of es in probability, theoretical and applied statistics, statistics at Iowa State University (ISU) and the problem solving, data analysis, and statistical think- first year of medical school at WRU. I applied and ing for engineers. Other activities included found- received a pre-doctoral fellowship. As I was finishing my dissertation, a professor ing the Reed Institute for Applied Statistics. This Betty Skipper institute funded summer research for undergradu- at WRU told me he was moving to a new medical ates and facilitated the process of obtaining fund- school at the University of New Mexico (UNM) ing for courses in applied statistics. In these courses, and would be hiring biostatisticians. I applied, problems were elicited from government and busi- intending to move back to the Midwest in two ness, which involved data and could be solved using years. Forty-eight years later, I retired as a tenured applied statistics. A course in applied statistics was professor from UNM and now work part time. I designed to analyze one of these problems during a met my late husband in Albuquerque and have two semester course. Students received course credit and grown children and four grandchildren. provided the client with written and oral reports. I have spent my career as an applied biostatistician, My second career, concurrent with my first collaborating with students and health professional career and still active, has been as a statistical colleagues, as well as teaching biostatistics courses. consultant to the US Navy. This career has been I am particularly interested in mentoring students extremely rewarding. The problems to be solved and junior faculty and teaching statistical concepts are often complex and require additional theory. to health professionals who are not statisticians. My The solutions are made possible by teamwork combined training in statistics and medical sciences with exceptionally well-informed and dedicated has been an important asset. Over the years, I have naval officers, engineers, and scientists. seen major changes in statistical practice—from How has your professional and/or personal life mechanical calculators and computer punch cards to modern computers and statistical software. been affected by being a member of the ASA? Although I didn’t start with specific plans for The ASA has helped my professional life by pro- this career, it has been a rewarding career. I have viding statistical problem solving information both been privileged to work with many students and online and in professional meetings. My years as faculty colleagues at the University of New Mexico. an associate editor of Technometrics and later as chair of the Committee for Careers in Statistics How has your professional and/or personal life were both informative and rewarding. been affected by being a member of the ASA? Is there anything you wish someone had told you Continuing education through local chapter meet- when you embarked upon your career that you ings, publications, and short courses. would like to tell others now? What or who inspired you to become a statistician? One wants to find a career where the work is so There was really no one who inspired me to become enjoyable you never want to stop, where being a statistician. As I mentioned in the career sum- involved in your work is one of the most reward- mary, I saw an announcement on a college bulletin ing things you do. board and decided to apply without really knowing much about it.

april 2018 amstat news 13 Dennis Boos preventive medicine. In addition to teaching and Member since 1974 research, I served as the campus’ only consulting How has your professional and/or personal life biostatistician. This resulted in involvement in a been affected by being a member of the ASA? great variety of fascinating research areas. I continued to be active with the San Francisco Bay Area The ASA has been the central professional orga- Chapter of the ASA. nization in my career. I have attended most JSMs As head of the UCSF cancer registry and with over the last 40 years. In fact, we often made the Dennis Boos ties to state and national cancer data systems, my JSM a family vacation. (My children loved the work became focused on the biometry and epi- hotels when they were young!) I have been involved demiology of cancer. Long-term continuing sup- in a number of committees and have appreciated port from the National Cancer Institute fueled my the organizational professionalism of the ASA. Of research and led to travel throughout the world and course, JASA and TAS have been important jour- service on national and international committees. nals for me, and I still get hard copies. Some of the ways in which I was further reward- What or who inspired you to become a statistician? ed was election to fellowship in the ASA, service as president of the Western North American Region of In the summer of 1973, I visited the department the Biometric Society, membership in COPSS, and of statistics at Florida State looking for a career receipt of a lifetime achievement and leadership award change. My undergraduate degree was in phys- from the National Cancer Institute. I am currently ics, but I had become disillusioned with the heavy professor emeritus of epidemiology in the department dependence of physics on the defense industry for of epidemiology and biostatistics at UCSF. funding. I first talked to a math professor friend about applied math, but he suggested walking How has your professional and/or personal life down the hall and talking to someone in statistics. been affected by being a member of the ASA? (He saw the future!) So, the associate head, Fred Attending ASA meetings in the early days brought Leysieffer, told me about statistics—I had no clue me in contact with persons with similar and diverse about the field—and I applied to graduate school interests within statistics, and subject matter talks soon after. helped me try to keep current on developments Is there anything you wish someone had told you within the field. when you embarked upon your career that you Membership in the ASA provided much of the basic grounding for my career. Rather than cite an would like to tell others now? individual experience, I will list several that stand I guess it would have been nice for me to under- out in my memory: stand the entrepreneurial nature of an academic career. As a junior, I wasn’t proactive enough in • Four trips to the Soviet Union during the making connections with scientists and other statis- Cold War for an international cancer congress ticians. Fortunately, North Carolina State is a warm and as part of a US-USSR collaborative project on breast cancer and nurturing environment for young faculty. • Participation in an international World Health Calvin Zippin Organization meeting on the importance of Member since 1947 teaching statistics to medical students held in the unusual location of Karachi, Pakistan In 1947, upon graduation from SUNY-Albany with majors in biology and mathematics and a course • Meeting with staff of the Atom Bomb in statistics, I was hired by the Sterling-Winthrop Casualty Commission in Hiroshima, Japan, Research Institute—the research arm of a large phar- while doing research on late effects of radiation Calvin Zippin maceutical firm—to do a variety of chores, includ- • At the request of the American Association ing statistical analysis of laboratory and clinical data. for the Advancement of Science, I interviewed That same year, I joined the Albany Chapter of the in 1978 in Buenos Aires mothers of missing ASA. Three years later, with a full tuition scholar- abductees (some scientists) during the so- ship, I began graduate work in biostatistics at Johns called “dirty war” in Argentina. Hopkins under the revered William G. Cochran. With my ScD degree in hand, I accepted a fac- What or who inspired you to become a statistician? ulty position in the school of public health at the I credit Lloyd C. Miller, my boss at the Sterling- University of California, Berkeley in 1953 and Winthrop Research Institute, with inspiring me to transferred in 1955 to the school of medicine in go into biostatistics. Although he was primarily a San Francisco (UCSF) with appointments in the pharmacologist, he had worked with Chester Bliss Cancer Research Institute and the department of of Yale on bioassay and became aware of the critical

14 amstat news april 2018 importance of statistical methodology in research. and career days for statistic graduate students and He went on to become director of revision of the high-school students. United States Pharmacopoeia for 20 years. His I retired in May of 2016 and still keep active with encouragement led to my going on to graduate the ASA New Jersey Chapter and review papers for work in biostatistics. two veterinary journals and one sports in statistics Is there anything you wish someone had told you journal. In addition, I do horse show announcing and when you embarked upon your career that you keep busy with my sports memorabilia collection. would like to tell others now? How has your professional and/or personal life For anyone interested in concentrating on an been affected by being a member of the ASA? applied area of statistics, I would emphasize the Being a member has allowed me to keep up with importance of learning as much about the subject the latest advances in statistics through subscrib- matter of that field as well as that of statistics. ing to journals, belonging to various chapters and I was interested in biology and the health sci- sections, and attending events for knowledge and ences. At the time of my graduate work at Johns networking. I have recommended membership to Hopkins, we were required to take courses com- students and colleagues throughout my career. prising most of the first two years of the medical curriculum in addition to work in biostatistics. Will you share an experience that stands out to For me, this was a blessing. you regarding your ASA membership? Also, I would say a career in statistics can What stands out for me is getting more involved in open itself to the most exciting and unexpected the ASA New Jersey Chapter. I have been president avenues of fulfillment. Keep your sights high. (four years), vice president (two years), and am Expect the unexpected! currently secretary. I have gotten to work with my wonderful and dedicated officers and have always Steve Ascher been very appreciative that all of the speakers we Member since 1974 have had at various events do this for no financial gain. The volunteer spirit and giving back to the I earned my PhD in 1978 from SUNY at Buffalo. Statistics VC-Community is alive and thriving! My first job was as an assistant professor at Temple Through the New Jersey Chapter, I have also got- University (1978–1983). At Temple, I revived the ten involved in career days for both statistics gradu- undergraduate major in statistics, which had been ate students and high-school students. The future dormant for many years. of our profession lies with them. Steve Ascher In 1983, I decided to go into industry and joined McNeil Pharmaceutical as a statistician. It What or who inspired you to become a statistician? was there that I learned about the complexities of I was good at math growing up and, as a baseball fan, drug development and how incredibly complex it wondered how baseball statistics were computed. is to get a new drug on the market. From 1993– Whereas many of my friends wanted to be doctors, 1999, I worked at IBRD and Covance (two con- lawyers, etc., I wanted to be the statistician for the tract research organizations (CROs)). I received New York Mets. When I started my undergraduate my first managerial experience at both organiza- studies at SUNY at Buffalo in 1970, I was a math tions as I headed two small statistics groups. major. A friend thought I might like to take a statis- Working on both the client side and the con- tics course as a way to apply math. I did and earned tractor side gave me a better understanding of how my BA in math/stat in 1974, my MS in statistics in to build beneficial relationships between clients and 1976, and my PhD in statistics in 1978—all from CROs. I returned to Johnson & Johnson (Janssen) SUNY at Buffalo. in 2000 to build a phase 4 statistics/programming/ data management group. Is there anything you wish someone had told you My CRO experience was helpful, as our busi- when you embarked upon your career that you ness model at Janssen was to use CROs. While would like to tell others now? at Janssen, my group supported numerous world- The importance of written and oral communica- wide neuroscience trials. We also assisted in writ- tions. I stress this when I give presentations at career ing posters, abstracts, and manuscripts. In addi- days. Today, communications all too often are in tion, I co-founded a mentoring program for J&J tweets, emails, Instagram, etc., where people “talk” in statisticians and programmers. During my J&J shorthand (e.g., LOL, UR, etc.). The art of face-to- tenure, I became involved with the ASA New face communication seems to be fading. As I learned Jersey Chapter and served as president for four during my career, one needs to be able to present years and vice president for two years. I am cur- findings to people in a clear and concise manner. rently secretary. We sponsor workshops, webinars, Statisticians do not just compute p-values! n

april 2018 amstat news 15 Q&A with Industry Leaders Statistician and data scientist continue to be ranked among the top jobs, most recently by U.S. News & World Report in 2017, which listed statistician as #1 on both its lists of Best Business Jobs and Best STEM Jobs, and Glassdoor in 2018, which had six analytics and data science jobs included in its 50 best jobs in America for 2018. We asked leaders in industry to answer the following set of questions to help students and statistics departments better prepare for jobs in the technology industry. We also hope the Q&A’s will help companies attract the statistical talent they desire.

research staff member, my responsibilities involved TALKSPACE working directly with other IBM divisions to apply Bonnie Ray currently leads data and/or develop new statistical methods to solve science efforts at Talkspace. Talkspace immediate business problems, while also filing patent is a NYC-based startup that enables disclosures and publishing applied research papers. behavioral health care for all through Over time, I took on responsibility for managing a small team of researchers and ultimately serving as providing a secure, affordable platform director for an organization within research focused for messaging-based psychotherapy. on development and application of AI algorithms to Number of employees: Talkspace currently business challenges facing IBM and IBM’s clients. has about 60 full-time employees, five of whom make up the data Moving into a management role required me to refine science team—part of the larger technology team that also includes my business communication skills, as I was expected software engineers and product managers. to present the work of my team to both internal and external clients on a regular basis, grow my strategic Distribution of highest degrees: The backgrounds of the data thinking skills to develop a solid research and fund- science team members vary, with some transitioning into data ing roadmap for my team, and develop my people science after having worked for several years in another technical management skills to effectively motivate and provide discipline and others having just completed master’s or PhD degrees feedback to my team. I also had to learn how to think that included extensive data science training. more broadly, without necessarily understanding all the details of a research area. Growth (or lack thereof) in number of data scientists over last several More recently, as I’ve transitioned to the tech years? The data science team was only created in 2017, but may grow by start-up world where I build and lead small teams another one or two individuals over the course of 2018. of data scientists, I’ve needed to become comfortable with engineering concepts and agile development skills, while also continuing to hone my ability to I completed a PhD in statistics from Columbia translate business problems into defensible statisti- University and began my career as an academ- cal approaches and skill at communicating statistical ic, first as a postdoctoral fellow at the Naval findings across different parts of the business. Postgraduate School in Monterey, California, and then as an assistant/associate professor at the New What do you like about working in the start-up Jersey Institute of Technology. During my time space? What are the challenges? as a faculty member, I was consistently drawn to Working for an early-stage start-up allows me the interdisciplinary work that had clear and tangible opportunity to learn about all parts of the business, impact. I also came to realize standing in front of a working directly with product managers to determine class of undergraduate or master’s students to teach how data science can contribute to new product fea- introductory statistics classes was somewhat nerve- tures, with the marketing team to understand the racking for me, rather than fulfilling. impact of advertising on customer acquisition and I decided to leave academia to take a position that retention, with the clinical team to develop quantita- allowed me to continue doing research while also tive approaches to measure the quality of the service providing opportunities for direct business impact. provided by therapists on the Talkspace platform, I took a position as a research staff member in the and with the commercial team to understand analysis statistical forecasting group at IBM Research. As a and reporting needs of our commercial clients.

16 amstat news april 2018 What advice do you have for students interested One challenge for me, personally, is the need in working in the technology industry? Any advice to have a deeper understanding of the existing and for students in general? potential computing infrastructure available to the data science team to ensure our data and computing I would advise students to make sure their coding needs are met. I’ve had to come up to speed on the skills are more than competent, as working with any various tool sets and platforms available for doing of the data sets typically collected will require effec- data science (e.g., Hadoop, Spark, Jupyter, GPUs, tive and efficient data wrangling skills—and some- etc.) to successfully lobby for the needs of the team. times knowledge of multiprocessor, distributed, or Additionally, the team is sometimes asked to rap- GPU-based computing techniques. idly pivot from one priority to another, which can I would also advise students to become comfort- sometimes affect team morale—something I need to able with delivering the 80% solution (i.e., a solu- make sure I effectively motivate and manage. tion for which an initial version can be developed quickly, but that may not address the complexity of How is the demand for statisticians in the the problem in its entirety). technology industry? What are the main Additionally, joining local data science–oriented degrees you consider when looking for meet-ups is really useful for establishing a network candidates with statistical expertise? of colleagues from which one can learn and grow, as The demand for statisticians/data scientists in the well as finding potential new career opportunities. start-up industry is currently huge, with newly Many individuals in the tech industry also maintain minted graduates often starting at salaries 1.2x–1.5x personal blogs or social media accounts focused on that of new assistant professors. data science topics, which gives them exposure in While I do not look specifically for candidates the tech community. with a degree in statistics when filling a data science role, I do look for evidence of basic statistical think- What advice do you have for statistics and ing and an understanding of fundamental statistical biostatistics departments (e.g., coursework, modeling approaches, such as linear modeling, sur- nontraditional training suggestions, research vival analysis, and Bayesian techniques. Most of the experience)? candidates I look for at the bachelor’s and master’s I would encourage statistics and biostatistics depart- levels have degrees in either statistics/data science or ments to incorporate coursework on text analysis If you would like a computer science (with a focus on machine learning techniques, given the huge amount of information representative of and/or natural language processing). I also am open gathered in free text format today. I would also your company— to candidates with degrees in electrical engineer- encourage exposure to additional computer science whatever the ing, physics, computational biology, econometrics, concepts, particularly those having to do with algo- sector—to respond and the quantitative social sciences—assuming they rithmic complexity, as well as an overview of state- to these questions for a future issue, show a solid understanding of statistics fundamen- of-the art optimization approaches. Also, intern- please email Steve tals and appropriate coding skills. ships or work in the university statistical consulting Pierson at pierson@ center should be a required part of the curriculum amstat.org or Megan What do you see as the most important statistical to allow students to practice their problem defini- Murphy at megan@ skills in the tech industry? What are the other tion, data wrangling, and communication skills. amstat.org. important skills necessary for a successful career in this sector? What opportunities for advancement and It is important that an individual have not only a firm professional growth exist for data scientists grasp of core statistical modeling approaches such as and statisticians in industry, and what advice covered in a strong master’s program, but also be for young professionals would you have to take competent enough to identify and understand appro- advantage of those opportunities? priate statistical techniques s/he may not initially be Advancement in the tech industry can be quite familiar with to appropriately address the problem rapid, with individuals moving from individual con- at hand. Solid programming skills are also necessary, tributors to team leads to chief data scientist roles with the specific language or tool set and the depth in the course of only a few years, depending on the of technical knowledge dependent on the particular size of the company. I would suggest an individual role. The most important skill needed for a successful think deeply about where he wants to take his career career, however, is the ability to communicate effec- before taking on a management role, as often a tively with colleagues across the business, both to move to management takes one away from further understand the business needs for which data science development of one’s core technical skills. However, expertise is needed and to communicate the results working in the tech industry, particularly at a smaller of analyses back to the business with minimal jargon. company, often enables an individual to move into

april 2018 amstat news 17 a role he may not have previously considered, such are responsible for all machine learning and statisti- as a product management or business strategy role. cal modeling at LinkedIn. The ability to use data and For career advancement in general, I would statistics to help connect talent with opportunity at advise a young professional to be proactive in pre- scale is inspiring. senting his results to the business to gain exposure for his work, take advantage of opportunities to give What do you like about working in the talks at local data science community meet-ups, technology industry? What are the challenges? and/or participate in hackathons. These experiences I really like the broad impact one can have on the provide broader exposure and networking opportu- society and the agility with which one can accom- nities that can lead to career advancement. plish things in the technology industry. The cycle from ideation to deployment is weeks, not months LINKEDIN or years. Improving products through data-based algorithms is an integral part of the work. Statistical Deepak Agarwal is a vice president thinking and mindset is critical to success in this of engineering at LinkedIn, where he is area. The volume and heterogeneity of data avail- responsible for all machine learning and able to solve problems provide unparalleled oppor- statistical modeling efforts across the tunities to do novel methodological research. For company. LinkedIn is the largest profes- instance, we changed the job recommendation sional networking site available today. algorithm last year to use random effects models. This improved the job application rates on the The site provides a way to connect with site by roughly 30%. The innovation was not so other professionals and helps you stay in much the statistical model, but more scaling the contact with millions of users. estimation to billions of parameters and running Number of employees: LinkedIn has more than 11,500 full-time such a large model to recommend jobs on the site. employees, including more than 400 data scientists and more than 30 In addition to the technical innovation, that such statisticians. work can create job opportunities for so many professionals on the planet is fulfilling. Distribution of highest degrees: ~350 PhDs, with many more While statisticians can have significant impact, employees who hold other graduate/postgraduate degrees the nature of the work is highly interdisciplinary. How many hires per year? More than 1,000 new hires per year One has to collaborate closely with product management, engineering, security and privacy, legal, Growth (or lack thereof) in number of data scientists over last and others. Formulating the problem is often more several years? 25% growth in data scientists last year of an issue and very challenging. Even when things get formulated, it can change quickly if there is a I earned a PhD in statistics from the University of change in strategy or some new evidence emerges. Connecticut in 2001, with Alan Gelfand as my the- Being adaptive in such a fast-paced environment sis adviser. My thesis involved fitting spatial models can sometimes be challenging. to large data obtained from disparate sources like satellite imagery, GIS, and census. How is the demand for statisticians in the I got very interested in doing statistics for large technology industry? What are the main degrees data and joined the statistics department at AT&T you consider when looking for candidates with Research Labs. After five fruitful years at AT&T, statistical expertise? I decided to move to Yahoo! Research, where I Statisticians play a key role in many areas, and the became the chief statistician for the company. This demand for statistical expertise is growing at a rapid was the best part of my technical career. I had the pace. Given the dearth of statisticians currently opportunity to create new statistical methodolo- working in this industry, many in computer science gies for large-scale problems that arise in consumer and machine learning filled the gaps. internet space. My work in this area had a signifi- Success in areas like experimental design, causal cant impact on the business and resulted in sev- analysis, and fraud prevention that is essential for eral publications (including a book by Cambridge almost every technology company today requires University Press). deep statistical expertise. In addition, large-scale sta- After spending six years at Yahoo!, I decided to tistical modeling is the core of all search and recom- join LinkedIn in a management role. The last six mendation systems. We are interested in statisticians years at LinkedIn have been the most fulfilling so with both master’s and PhD degrees in statistics and a far in terms of impact and job satisfaction. I lead a genuine interest in learning computational techniques team of roughly 300 engineers and scientists who that can help them apply statistics to large data.

18 amstat news april 2018 What do you see as the most important statistical scenarios (small data, large data) and a balanced skills in the technology industry? What are the portfolio of statistical inference techniques without other important skills necessary for a successful over-emphasizing one particular area (e.g., only fre- career in this sector? quentist, only Bayesian) would be good. It is more The most important statistical skill is, of course, a important for students to develop strong statisti- deep knowledge of statistical methods and the ability cal intuition and understand the pros and cons of to solve practical problems. In addition, it is impor- different approaches. This would help them enor- tant to have a strong background in statistical com- mously in their day-to-day job when working in puting and a genuine interest in learning new com- the technology sector. putational techniques such as distributed computing to apply statistical techniques to massive data. It is also important to not just like, but enjoy, WELLIO working in an interdisciplinary environment. Often, Sivan Aldor-Noiman the most successful folks in the technology sector are Wellio is a start-up of about a dozen those who gain a deep understanding of the entire people whose mission is to make it more end-to-end process over the years. Given how many opportunities exist to innovate using data-based algo- convenient for people to eat healthier and rithms, such individuals often end up becoming suc- better food, which they accomplish through cessful leaders and entrepreneurs. the personalization of variety, healthiness, The advent of cloud computing has made man- and cost, taking into account factors such as cooking aging and computing with large data more of a ability, taste preference, health needs, and shopping preferences. commodity; the next big quest is to “commoditize” Emphasizing the importance of statistics and modeling, four members the extraction of intelligence (aka statistical inference) from data. There was never a better time for of the Wellio team are data scientists, with two having PhDs in statistics statisticians to join the technology sector than now. (including Aldor-Noiman), one a PhD in applied math, and another with a bachelor’s and master’s in computer science. There are also four What advice do you have for students interested software engineers. in working in the technology industry? Any advice for students in general? Those interested in learning about Wellio’s internship opportunities should contact Aldor-Noiman at [email protected]. I have seen students with classical statistics training being a bit skeptical about joining the technology sector. They are not sure about the impact relative I was drawn to statistics as an undergrad at The to others with background in computer science and Technion – Israel Institute of Technology, where machine learning. I would advise them to seriously statistics is housed in the industrial engineering consider a career in the technology sector. There is department. After earning a master’s in statistics significant opportunity to make an impact and a there, I did my PhD in statistics at the University high demand for their skills. Things that they may of Pennsylvania Wharton School with Larry Brown want consider are a) do they like technology in gen- and Bob Stine. I loved teaching, but knew I was eral; b) do they like doing applied work and enjoy not destined for academics and applied for jobs working in an interdisciplinary environment; and with technology companies. c) do they enjoy learning new computational tech- I received several offers—including from well- niques to work with massive data. If the answer to known companies—but opted for the Climate all three is yes, they must consider a statistics career Corporation, which was then a smaller, but grow- in the technology sector seriously. There is no better ing, start-up working in the ag-tech sector. I chose time to do so than now. them because of the wide-ranging statistical challenges I would have as one of their first statisticians What advice do you have for statistics and doing modeling, including with spatial statistics biostatistics departments? and spatio-temporal dynamics. I also was really I would encourage statistics departments across impressed with the impact and importance data sci- the country to emphasize areas such as experience had in the product itself; it was clear this com- mental design a lot more from both a theoretical pany wanted to use data science to its full extent. and practical perspective. A bit more emphasis on It was also quite clear to me that if I were to go to modern computational paradigms like distributed one of the more established companies with many computing with Hadoop/Spark would be useful. other statisticians, the statistical challenges would When teaching classical statistics, emphasizing have been narrower and professional advancement what methodologies work well in different data would have been much slower.

april 2018 amstat news 19 I spent five years with the Climate Corporation, The other great satisfaction of working in indus- an exciting period when it grew from 120 employees try is that I get to see the impact of my work with to more than 500 and was purchased by Monsanto. the consumer using the product. Getting feedback Early in my career there, I led a group doing weath- and improving the models and products brings er modeling and risk insurance, which addressed endless technical challenges, which I really like. many challenging and interesting problems statistically speaking (e.g., how weather impacts yields). I How is the demand for people with statistics later started a remote sensing team (e.g., satellites, degrees in the technology industry? What are the drones), which helped develop products that moni- main degrees you consider when looking for tor crop development throughout the growing sea- candidates with statistical expertise? son. Both of these groups were extremely diverse There is a big demand for statisticians across degree and included statisticians, mathematicians, physi- levels (bachelor’s, master’s, doctoral). The Bay Area cists, and remote sensing experts with mostly mas- in particular is a crazy bubble for statisticians. There ter’s and doctoral degrees. is such huge demand here that statisticians are being I subsequently started two more teams. The pinged almost weekly by recruiters. Most of the data first focused on fields experiments and observa- scientist positions are filled by people with degrees in tional studies, which unlike the small experiments statistics, applied math, computer science, econom- Fisher conducted, were massive experiments across ics, physics, or disciplines that use more applied statis- many soil and weather environments, agriculture tics like remote sensing. Analyst positions are gener- practices, and crops. The second team focused on ally geared toward those with undergraduate degrees, developing best practices for data science across though. As I discuss below, one has to be careful the company. The challenges in this team quickly about job titles; they are often quite misleading. became more focused around culture development for helping data science grow. What do you see as the most important statistical Today, I lead the team of data scientists at skills in the technology industry? What are the Wellio. I really like both the mission we set for other important skills necessary for a successful ourselves and the team. We are building food rec- career in this sector? ommendation systems using advanced statistical What differentiates those with statistics degrees from and deep learning models to solve NLP and com- others is the ability to reason through an argument, puter vision problems. This team pushes me to the data, and the model and then communicate it. develop and improve my statistical, engineering, All these are fundamental skills of statisticians. We and leadership skills. learn it in our first EDA class. You just can’t explain a model without these skills. When I haven’t hired What do you like about working in the technology someone, not having these skills is the main reason. industry? What are the challenges? So many people don’t know where to start with a I like the many hats I wear working in the tech- problem. When I am interviewing candidates, I ask nology industry, including statistician, data engi- them to look at the data and tell me what they see. neer, leader, and communicator of our findings. I also assess their ability to build a model and assess The latter presents a challenge in that one must its limitations, implicit in which are the questions: also present the strengths, weaknesses, and limits What is a model? What are the assumptions? of the results to clients and customers who want I also see an increased desire for candidates with certainty. It’s hard for people to accept uncertain- causal inference expertise, with the technology ty; it’s a difficult concept. A data scientist must industry starting to recognize anew that correlation also understand what’s at stake for the customers. is not causation. I know it sounds like old news for For example, when I was in the agriculture indus- statisticians, but you would be surprised how people try, a farmer’s livelihood was directly affected by really look at results based on correlation and tell our products. themselves the right story to fit their goals instead of The expectation of providing certainty has only developing the hypothesis in advance … like we are increased with the recent emergence of artificial being taught in Stat101. intelligence, which is perceived as almost having an air of magic. There is a great responsibility for data What advice do you have for students interested scientists to understand the limits of their models in working in the technology industry? Any advice and yet produce and communicate something use- for students in general? ful. One has to strike a balance between use and I always provide the same advice: Go to a company limitations, being careful that communication of looking for people to learn from, instead of a com- the limits does not scare off the clients. pany where you would be the first data scientist

20 amstat news april 2018 and would have to teach statistics to everyone. In such a situation, you won’t learn as much about models and the many skills—technical and non- technical—that will help advance your career over the long run. It’s also important to ask what you are going MASTER’S IN to do and what the job looks like on a day-to-day basis. It might be they hire statisticians to produce BIOSTATISTICS dashboards and summary statistics, which may not be very challenging and therefore not helpful in the long term. 1 YEAR OPTION! What advice do you have for statistics and biostatistics departments (e.g., coursework, nontraditional training 4 TRACKS: suggestions, research experience)? My beef is that some of them haven’t acknowl- Biostatistics edged that applied statistics is very important, just as is the need to understand theory. One must have balance between the two. Someone trained Genomics and for industry needs to understand the theory and Bioinformatics must be able to analyze data. Students must be exposed to real problems. Taking classes in computer science and pro- Health Care Analytics gramming is also a very good thing. I had to learn programming on the side, and it’s harder for sure to Social and Behavioral do it this way. Furthermore, industry requirements for modeling and programming are very different Science than for statistical academia, which emphasizes statistical accuracy. Accuracy may not be sufficient in industry. Industry conditions can be much harsher. Sophisticated models may not be the first ones to try. A model that performs better from an engineering standpoint is likely the preferred one. What opportunities for advancement and professional growth exist for data scientists Internship and statisticians in industry, and what advice for young professionals would you have to take #2 US Hospital advantage of those opportunities? Industry is still trying to figure out what data sci- Affiliated with School ence is and what its career path is. Data analyst is often a junior position, but there is lots of vari- ance. So, a senior analyst could be a PhD statis- MONTHLY WEBINARS tician. Don’t be tempted by titles, which can be ON PROGRAM almost meaningless. As I advised above, ask lots of questions about what the position will entail. http://epbiwww.case.edu/ Those questions should also be about career ms-biostatistics path, but one shouldn’t be surprised if they struggle to answer the questions about career path. Two possible career paths are modeler-to-manager and statistician-to-principle data scientist. For the former, there is a big shortage generally of technical leaders MORE INFO? who are good managers of people. More specifically [email protected] for this audience, there is a big shortage of statisticians in the technology industry, but a much bigger shortage of statisticians who can manage. n

april 2018 amstat news 21 STATCOM: Revitalization of Statistical Community Service at Universities Evan Reynolds and Timothy NeCamp All project tasks are completed Finding pro-bono statisti- niversities provide valu- by student volunteers, including cal projects with local nonprofit able resources for provid- project obtainment, client com- and government organizations ing pro-bono statistical munication, consultation, and can be difficult. Though statisti- Uservices, including community analysis. Faculty members serve cians know and believe in the connections and many statistically as project advisers. Project tasks benefits of data, community part- inclined students eager to apply include developing surveys and ners may be less aware of data’s their skills. STATCOM at the data-collection practices, defining power. Starting a project with a University of Michigan is leverag- questions that can be answered potential partner requires raising ing these resources to increase its with available data, providing awareness of how STATCOM benefit to both the community numerical and graphical summa- can help, establishing trust in the and the university. While several ries of the organization’s data, and benefits of statistical consulting, universities with graduate statis- building statistical models and per- and ensuring that working with tics and biostatistics programs forming analyses to help answer STATCOM will not be a burden. founded STATCOM chapters in the organization’s key questions. By This is a not a trivial task. the 2000s, much of their activity participating in projects, students Two years ago, STATCOM has declined since then. In con- learn consulting skills that are diffi- started a collaboration with the trast, STATCOM is larger and cult to teach in a classroom setting. Community Technical Assistance more active than ever at the In the past few years, the Collaborative (CTAC, https:// University of Michigan. University of Michigan’s goo.gl/NDw2EN), a group at the Statistics in the Community STATCOM chapter has seen University of Michigan consist- (STATCOM) is a university- tremendous growth. In 2014, it ing of mostly social work students based, student-run organization had 45 student members with directed through the Edward that provides pro-bono statistical five projects. In 2017, it grew to Ginsberg Center. As the primary consulting services to nonprofit 147 student members with 10 office for community engagement and government organizations. projects. The chapter has devel- at the University of Michigan for STATCOM began in 2001 in the oped new collaborations and the past 20 years, the Ginsberg department of statistics at Purdue organizational features that have Center has a large established net- University. Following grant sup- helped the chapter both expand work of community organizations. port from the American Statistical and cope with the expansion. By collaborating with the Ginsberg Association, STATCOM pro- STATCOM is made up of a Center, STATCOM was able to grams were initiated in universi- multidisciplinary group of gradu- break down this initial barrier to ties across the country. ate students. Though most mem- community entry and acquire The STATCOM chapter at the bers are majoring in biostatistics many more community partners. University of Michigan (https:// and statistics, STATCOM also The collaboration also provided goo.gl/2VpfoC) was founded in has students majoring in survey benefits such as workshops teach- December 2006 under the super- methodology, social work, epide- ing STATCOM members prac- vision of Michael Elliott. It is a miology, information, and indus- tices for thoughtfully engaging community outreach program led trial and operations engineering. with communities and the ability by graduate students that provides The enrollment in many of these to work on projects that required free statistical consulting to orga- programs, especially biostatistics both social work and evaluation nizations in Southeast Michigan. and statistics, has increased sub- skills (i.e., CTAC) and statistical Currently, the group is led by stantially in the past few years. As expertise (i.e., STATCOM). co-presidents Evan Reynolds these programs grow, more stu- The rise of data science initia- and Timothy NeCamp and dents are looking for ways to prac- tives at the University of Michigan supervised by Elliott and Cathie tice statistical consulting in real- has fostered other beneficial part- Spino. STATCOM is project world settings, especially in ways nerships. Through collabora- based; as new projects arise, that can lead to positive impact. tion with the Michigan Institute teams of 3–5 are assembled to STATCOM has helped provide for Data Science (MIDAS), help the requesting organization. those experiences. STATCOM has worked with

22 amstat news april 2018 students from other data science disciplines to broaden its impact. Also, STATCOM is teaming up this spring with CTAC and the Ann Arbor Data Dive for an inaugural showcase of all the pro- bono quantitative work going on across the university. A recent project provides an excellent example of STATCOM’s work and the benefits of these new collaborations. The Ann Arbor Area Community Foundation (AAACF) received $13 million The University of Michigan’s STATCOM had 45 student members with five to improve services for the senior projects in 2014. In 2017, it grew to 147 student members with 10 projects. population of Washtenaw County and sought to assess which sub- STATCOM has partnered with the biostatistics, and data science grad- populations of the county were Detroit College Access Network to uate programs is well documented. most in need. Working together, evaluate how their counseling pro- The influx of students wishing to STATCOM and CTAC devel- gram has affected college applica- get real-world analytical practice is oped a survey to assess quality of tion and enrollment for Detroit at an all-time high. Additionally, life of seniors within Washtenaw high-school students. working on projects within the County. Quality of life was mea- STATCOM recently expand- local community guarantees a low sured through the Older People’s ed its student leadership team to overhead cost and easy access to Quality of Life questionnaire distribute some of the increasing projects. These types of collabo- developed by Ann Bowling. amount of administrative work rations can be replicated at other CTAC members implemented required with the influx of new universities where, with high prob- the survey and received more than projects. New leadership positions ability, established university-based 600 survey responses. Once the include a communications chair community engagement organiza- data was compiled, STATCOM who helps communicate with col- tions already exist. provided analysis with methods laborators and community part- Readers hoping to establish ranging from graphical summaries ners, a president-elect who will similar organizations at their to regression trees. The analyses become president in the upcom- universities should not be afraid indicated geographic location— ing year and who receives training to get started. The collection of specifically within two ZIP codes to ensure a smooth transition, and student volunteers and potential of Washtenaw County—and a project obtainment chair who projects can be slow, especially financial vulnerability were most actively seeks projects with new at first. Keep in mind, there associated with lower quality of and old community partners. are always organizations doing life. Surprisingly, living alone— As exciting as STATCOM’s impactful work for the commu- which was previously hypothesized growth and accomplishments nity, many of which can enhance to negatively affect quality of life— have been over the past several the scope of their effect with the was not a significant predictor. years, STATCOM is still nowhere help of statistically inclined stu- This year, STATCOM is con- close to its project capacity. dents. Also, while STATCOM at tinuing to work on a spectrum of Whenever a call for volunteers is the University of Michigan has projects with a variety of commu- sent out, there are always more established wonderful university- nity partners. STATCOM is col- interested volunteers than need- wide collaborations, these are by laborating with the Catholic Social ed. With the help of new student no means a requirement for start- Services of Washtenaw County leaders, STATCOM is hoping to ing a new chapter. All it takes to analyze therapy appointment increase its capacity and impact in is a few students interested in data to optimize scheduling. A 2018 and beyond. In 2018, there applying their statistical skills to STATCOM team has assisted the is a goal of 12 projects. improve their community. In the Ozone House of Ann Arbor to Many factors leading to end, the work will be an invalu- build a predictive time-series model STATCOM’s success at the able experience for students and to schedule staff volunteers to best University of Michigan are pres- provide a major benefit to the handle crisis calls from youth in ent at other universities. The rise community, especially in an the surrounding area. Additionally, of student enrollment in statistics, increasingly data-driven world. n

april 2018 amstat news 23 MATH & STATISTICS AWARENESS MONTH DATA Is MyJOB Four ASA members share career insights

Career opportunities are limitless with a degree in statistics or data science. To celebrate Mathematics and Statistics Awareness Month (#mathaware), we highlight four ASA members who have used their statistics skills and imaginations to snag sweet jobs.

What do you during a typical day at work? • Meetings. Both internal—project manage- McDougall: Like most professionals/managers, ment, product development, finance, HR, there is not really a typical day, but there are tasks marketing—and external—going to meet you regularly perform. I balance managing the with clients offsite or having teleconferences company with keeping up as a statistician and with them working with clients. • Research. New statistical methods, therapeutic • Email. Updates on ongoing projects (some- areas, regulations—usually as part of a project times resolving issues); answering clients’ or a work-up for bidding on the project questions; reaching out to potential clients • Training. Finding, organizing, and attend- or collaborators; setting up meetings; keeping webinars in statistics, data management, ing up on industry changes, regulations, etc.; also having statistical discussions with and statistics by being on mailing lists and staff about design and analysis issues reviewing the content The one big omission is programming. I don’t

24 amstat news april 2018 Who Are These People?

Janet McDougall is the Rob Santos is the chief Yihui Xie is a data Nancy Potok is the chief founder and president methodologist and scientist and software statistician of the United of McDougall Scientific director of the Statistical engineer at RStudio, Inc. States. consulting firm. Methods Group at the Urban Institute.

meet our standards for a programmer—because … [P]ursue your passions, keep your of all the other distractions—and I do miss that part of the job. options open, have fun, and always Santos: On any given day, I will be overseeing policy research projects in diverse topics like housing challenge yourself beyond JOB discrimination, refugee resettlement, urban community attachment, food insecurity, driving travel your self-perceived limitations. behavior, client feedback loops, and firefighter safety and risk assessment, as well as attending to internal administrative projects like IRB reviews, Life’s all about the journey, not advancing diversity/inclusion and community engagement research methods, and chairing insti- the destination. tutional awards committee deliberations (this is but a portion of my current portfolio).

Xie: I answer software questions from various chan- priority work areas. On other days, I am mostly in my nels (StackOverflow, GitHub, emails, etc.) and write office meeting with my small staff and guiding their code, documentation, and books. I have been trying activities. Some days, I brief the policy officials at the to publish one book a year since 2015, and—this Office of Management and Budget on high-priority year—I’m working on my fourth book. decisions that need to be made or documents that need to be cleared. If I am really lucky, I get to catch Potok: There is no typical day—every day is different. up on things I should be reading once in a while. A lot of my job involves external relationships with stakeholders, data users, and the US statistical agen- How did you end up in your current position? cies. As a result, I may be out of my office for the McDougall: By chance. I started freelance work in greater part of the day meeting with people, speaking the pharmaceutical industry after working at one at conferences or other events, attending workshops, company for about four years. Through contacts in or strategizing with the statistical agency heads on our the industry, the workload grew and I brought others

april 2018 amstat news 25 onboard, then had to move the business out of my University of Chicago. I finally realized I wanted to home and into a leased office. Taking on leases, it be closer to the action of putting research results to seemed appropriate to incorporate—so I did and, as use for the betterment of society and landed at the the owner, became the de facto president. Urban Institute, a public policy research think tank. For the next couple of years, I kept my title as I then was lured back to my home state of Texas statistician, as that was what I was proud to be. (Austin) to co-own a social science research firm, When a future client asked if this was my father’s but we sold the firm after six years of amazing business, I recognized I had to take the business— growth and interesting projects. and my image—more seriously and adopted the And then I returned to Urban Institute to resume title of president. I kept the title and have grown my passion for conducting research for the public into the position, learning to be more of a strate- good. And this is where I have been the past 11 years. gic thinker and planner, making important and I enjoy most the ability to play in everyone’s tough decisions, mentoring, delegating—even my tent, be it justice policy, immigration, housing, beloved programming—to the growing staff, deal- hunger issues, health, education, infrastructure, ing with all the financial administrative tasks. program evaluation, you name it. The past decade I was lucky to find good business coaches along has been the most rewarding and fulfilling period the way who not only guided me, but also two career-wise in my life. other staff members (also statisticians) to form a management team. My role as senior statistician, Xie: I wrote an R package named “knitr” in late where I keep abreast of the developing statistical 2011, which caught the attention of our current trends and advise on design and analysis, is still CEO. I met him in 2012 for the first time and what gives me the greatest pleasure. collaborated on a few talks, including a keynote at the useR! conference in 2012. We had a brief Santos: Through a journey-pursuing opportunity. phone call in 2013, when I was about to look for I spent much of my career in leadership positions a job. He basically said “Yes,” and that’s it. trying to promote the most rigorous research in I’m a statistician by training, but I love program- university-based survey research centers at Temple ming more than statistics. Sorry! I probably should University, University of Michigan, and The not have said that here. Anyway, I write software

PRACTICAL TRAINING IN UNIVERSITY OF MICHIGAN ISR INSTITUTE FOR SOCIAL RESEARCH RESPONSIVE ARE YOU COLLECTING SURVEY DATA SURVEY AS PART OF YOUR RESEARCH? WOULD YOU LIKE TO SAVE COSTS AND DESIGN INCREASE YOUR OPERATIONAL The Summer Institute in Survey EFFICIENCY? Research Techniques at the University of Michigan Institute for Social Research IF SO, YOUR WORK MAY BENEFIT FROM is excited to announce that after the LEARNING MORE ABOUT RESPONSIVE success of last year’s inaugural RSD SURVEY DESIGN (RSD). Program, we have added eight NEW hands-on workshops on different FOR MORE INFORMATION: aspects of RSD. http://rsdprogram.si.isr.umich.edu Email: [email protected] www.facebook.com/rsdprogram www.twitter.com/rsd_program

26 amstat news april 2018 SHARE WITH US! We want to see you sharing #Celebrate your cool data jobs with STATISTICS Celebrate future statisticians and data # ® DATA scientists. Snap a selfie with MATH & STATISTICS AWARENESS MONTH SCIENCE our #CelebrateStatistics or

MATH & STATISTICS AWARENESS MONTH #CelebrateDataScience printable ® posters and tag @AmstatNews.

primarily for statisticians and data analysts. ian. I loved reading, doing research, and uncovering new and interesting information. One of my favor- Potok: The short, technically correct answer is that ite activities at that age was to pick up a Funk & I applied for it on USAJobs.com and was selected Wagnalls encyclopedia, open it to a random page, through the federal government’s merit selection and just start reading. Of course, at age 13, I discov- process. Of course, that was after a long and var- ered boys and my career ambition shifted quickly to ied career that spanned two tours at the US Census wanting to be a go-go dancer for a rock band. It’s Bureau (first as the principal associate director and been an interesting ride since then. CFO and the second as the deputy director and COO); working in the private sector doing social What career advice would you give your science research and consulting; and—at various 20-year-old self? times—holding other federal positions, including McDougall: Keep learning—and not just in your deputy undersecretary for economic affairs at the narrow discipline. I still over-prepare for meetings Department of Commerce. I discovered I had a real and discussions about protocol development passion for strategically managing data and the peo- and analyses—therapeutic areas, publications ple who create and disseminate data to inform major for the disease or the design—also just general policy questions and provide high-quality informa- trends in our society. By reading widely, the doors tion that could change peoples’ lives for the better. to serendipity open more frequently and you get wonderful insights into a problem or an oppor- What did you want to be when you were 12? tunity that you might miss otherwise. You ‘see’ McDougall: A scientist. I got my first chemistry things other people miss, because your mind has set when I was around eight or nine years old been opened to different possibilities. and loved the sense of wonder and discovery. Santos: I’d say pursue your passions, keep your Santos: I totally wanted to be a math professor. I options open, have fun, and always challenge your- loved everything there was about mathematics and self beyond your self-perceived limitations. Life’s majored in math for my BA. But when it came time all about the journey, not the destination. A great to think about graduate school, my counseling pro- career and a good, fulfilling life can be had by heed- fessor insisted I consider a more applied area—sta- ing those few words. tistics—so I could always have a great job. Time has shown he certainly steered me well. Xie: Try more often to do things you don’t like but are important at the same time. It is easy to do things Xie: I just wanted to study super hard and obtain you like, but in the real world, there is no guarantee the highest possible educational degree, which was a you will like all tasks assigned to you. Don’t be afraid “postdoc” according to what people in my village told of the pain from challenges. If you don’t feel the pain me when I was a boy. I probably also wanted to be a in tackling a challenge, it basically means you are los- scientist. Now some people call me “data scientist,” ing the chance to learn more things and grow up. which is not what I intended to be when I was 12; I had no clue about statistics until I went to college. Potok: Conquer your fear of taking big risks to fol- To a child, a scientist who plays with colorful chemi- low your dreams—at 20 years old, there is plenty of cals looks like more fun. Playing with data is also fun, time to learn from your mistakes and discover both although it is a little more abstract. what you are great at and what makes you happy. n Potok: My career goal at age 12 was to be a librar-

april 2018 amstat news 27 Master’s and Doctoral Programs in Data Science and Analytics Steve Pierson, ASA Director of Science Policy

More and more universities are starting master’s and doctoral programs in data science and analytics—of which statistics is foundational—due to the increasing interest from students and employers. Amstat News reached out to those in the statistical community who are involved in such programs to find out more about them. Given their interdisciplinary nature, we identified programs involving faculty with expertise in different disciplines to jointly reply to our questions. We have profiled many universities in our April, June, and December 2017 issues and January 2018 issue; here are several more.

What was your primary motivation(s) for University of Central Florida developing a master’s (or doctoral) data science/ Liqiang Ni is associate professor of statistics in the analytics program? What’s been the reaction from department of statistics at the University of Central students so far? Florida. His research interests include dimension In the late 1990s, the statistics community began to reduction, multivariate analysis, actuarial science, and realize the great potential in data mining and data business intelligence. He has served as the graduate science. UCF created one of the earliest data mining coordinator since August 2017. programs, in part inspired by SAS Company and Shunpu Zhang is a professor of statistics and chair of with support from Disney, Florida Hospital, Blue the department of statistics at the University of Central Cross Blue Shield of Florida, Universal Studios, and Florida. His research interests include bioinformatics, many local business partners. functional estimation, health informatics, large-scale The students responded with a good deal of hypothesis testing, and big data analytics. enthusiasm. We have seen a growing need for an educated and talented workforce at the MS level and beyond that can contribute to industry, gov- MS in Statistical Computing—Data ernment, and academia through innovative applica- Mining Track tions of data analysis methodologies. Website: https://goo.gl/ndCqwg How do you view the relationship between Year in which first students graduated/are statistics and data science/analytics? expected to graduate: 2002 We believe statistical science is an integral part of data science/analytics. A good data scientist/data Number of students currently enrolled: 60 analyst must have adequate training in statistics. Program format: In-person, 36 credits, comprehensive exam required, either thesis or project-based, full What types of jobs are you preparing your time and part time, graduate assistantship offer on graduates for? competitive basis We are preparing MS graduates largely for indus- There are largely two components for required tries. Every year, a few graduates continue their courses in the curriculum. The first tilts to tradi- studies in PhD programs. tional statistics, including a two-semester course What advice do you have for students considering for theoretical statistics and one-semester course a data science/analytics degree? for regression analysis and logistic regression/GLM, respectively. The second component tilts to applica- We suggest students have a solid foundation in com- tions: two-semester course for data processing and puter programming, mathematics, and statistics to preparation, including coding, and two-semester be a good data analyst. They also should have a keen course for data mining. There is a variety of elective interest in the new developments in data science. courses students can choose from.

28 amstat news april 2018 Describe the employer demand for your Partnering departments: Biostatistics, Electrical graduates/students. Engineering and Computer Science, School of Information, Statistics (administrative unit) Demand for our graduates has always exceeded the supply, especially in recent years. Program format: Full time, on campus; requires at least 25 credit hours in core areas including databases, data Do you have any advice for institutions and web applications, regression, and statistical learning considering the establishment of such a degree? The program requires the students to have demon- We believe a data analytics/data science graduate strated competence in a basic computing sequence program resides best in a statistics department with and a basic statistics sequence. By taking graduate- concentrations in computer programming and soft- level courses, the students need to demonstrate ware development. Open-mindedness is the key to expertise in data management and manipulations, as a successful interdisciplinary program. well as statistical techniques relevant to data science. The students need to take at least one advanced elec- University of Michigan tive from each of the following buckets: principle of data science; data analysis; and data science compu- Michael Elliott is a professor of biostatistics and research professor tation. The students will also have an integrative cap- of survey methodology. His research stone experience through an approved project. interests include survey methods, Students with an undergraduate degree in data causal inference, missing data, and science would already have obtained a reasonable longitudinal data analysis with level of training toward the core skills and may fin- applications to social epidemiology, ish the master’s degree in one year. Students with cancer trials, women’s health, pediatrics, and injury. an undergraduate degree in mathematics or physics, statistics or biostatistics, computer science, and other H. V. Jagadish is Bernard A. Galler quantitative disciplines should be able to complete Collegiate Professor of Electrical all requirements within two years. Engineering and Computer Science. His research has spanned many What was your primary motivation(s) for developing a aspects of big data, including data master’s (or doctoral) data science/analytics program? usability when they come from What’s been the reaction from students so far? multiple heterogeneous sources, and has undergone many manipulations. The data science explosion is fueled organically by new data generated from diverse sources, devices, XuanLong Nguyen is associate web services, mobile communication, scientific professor and director of master’s studies, and social media. Data scientists require a programs in statistics. His research interests include Bayesian versatile and unique set of skills to manage, process, nonparametrics, hierarchical models, and extract data from these complex information and machine learning. streams, and then interrogate, analyze, visualize, and interpret the information. Nationally, there is Elizabeth Yakel is associate dean for a pressing need for data scientists, and, in fact, for academic affairs and professor in the people with every level of data science training. The school of information. Her research focuses on data reuse, teaching successful launch of our data science major, which with primary sources, and the has attracted almost 200 students across campus in development of standardized metrics its first two years, made it clear that our students to enhance repository processes and want to be part of data science. The collaborative the user experience. approach we take across departments and colleges enables us to pool resources and offer the best our Ji Zhu is a professor and director of university has for a truly cross-cutting program. the data science master’s program in statistics. His research interests How do you view the relationship between statistics include statistical learning; network analysis; and statistical modeling in and data science/analytics? finance, marketing, and biosciences. Statistics is undoubtedly a major part of data science. The advancement of statistics has always been Data Science Master’s Program driven by new data that arise in science or society, whether they are from agriculture measurements Website: https://goo.gl/NxAbnN or the industrial revolution or the internet. While Year in which first students graduated/are data science requires tools from multiple disciplines expected to graduate: 2019–2020 (e.g., mathematics, computer science) and must

april 2018 amstat news 29 The Johns Hopkins work with specific domains of applications (e.g., University business or health care analytics), statistics and data James Spall has four appointments science are inseparable. From design of experiments at The Johns Hopkins University: to probabilistic modeling, from data exploration to principal professional staff at JHU/ confirmatory testing, and from estimation to pre- APL; chair of the applied and diction, statistics has been the core to data analysis. computational math program; co- Statistics without data science will not thrive, and chair of the data science program; and data science without statistics is certainly unsound. research professor in the department of applied math and statistics. Spall What types of jobs are you preparing your has published extensively in the fields graduates for? of control systems and statistics. This is a new program, but we provide the training Master of Science in Data Science and the students need to work as data scientists in a wide range of industries, from financial services to health Post-Master’s Certificate (PMC) in Data care, from marketing to social networking. We invite Science companies to our career fair for the students, and we Website: https://goo.gl/RGZfVN encourage students to take internships to help them Year in which first students graduated/are understand what they need to prepare for in school. expected to graduate: Late 2018 Number of students currently enrolled: More than What advice do you have for students considering 120 fully matriculated students in the MS degree and a data science/analytics degree? 0 students in the PMC program. There are additional We offer two master’s degrees, one in applied statis- students who have been given a provisional admission status (additional evaluation and/or coursework tics and the other in data science. At the present time, required) for both the MS and PMC. the applied statistics degree focuses more on model- Partnering departments: Applied and computational ing and inference, and the data science degree focuses mathematics and computer science more on data handling and data mining. Some of the Student type: Nontraditional/part time/continuing students from the applied statistics degree pursue a education, although there are a few students pursuing doctoral degree in statistics, biostatistics, economics, the degree full time and other quantitative fields. We expect the data sci- Program format: Online/in-person/combination; 30 ence students to be versed in data management and credit hours required in five years for the MS; 18 credit programming. However, there is an increasing overlap hours required in three years for the PMC. between the two programs, as we offer more computing courses to applied statistics students and more sta- The program is a combination of selected offerings tistics courses to data science students. in two existing rigorous graduate degree programs in applied and computational mathematics (ACM) and Describe the employer demand for your computer science (CS). On the ACM side, students graduates/students. will take a foundational course in statistical methods We do not have data on our graduates from the and data analysis, followed by required courses in data science program, but the vast majority of our optimization, statistical models and regression, and graduates from the applied statistics program was computational statistics. On the CS side, students employed or went to PhD programs within six will take a foundational course in algorithms, fol- months of their graduation. lowed by required courses in databases, visualization, and data science. All students are also required to Do you have any advice for institutions take one upper-level ACM elective (e.g., data min- considering the establishment of such a degree? ing, queuing theory, or stochastic optimization) and Data science programs by nature cross tradition- one upper-level CS elective (e.g., machine learning al boundaries, but the department of statistics or big data processing using Hadoop). Qualified is a natural and ideal home for such programs. students will need to have taken three semesters of To make such programs successful, the statistics calculus (through multivariate), discrete mathemat- departments must be willing to modernize their ics, Java, and data structures. existing curriculum to embrace data science and reach out to work with the faculty from other What was your primary motivation(s) for developing a programs. At Michigan, different programs offer master’s (or doctoral) data science/analytics program? complementary courses in data science and, What’s been the reaction from students so far? together, we believe we can attract and accommo- The motivation for starting the program is clear date students from diverse backgrounds. to anybody even slightly paying attention to broad

30 amstat news april 2018 trends in society toward greater quantitative analysis in decision-making and the need for processing and interpreting massive data sets in many diverse The range of jobs associated with data fields. JHU had a well-received non-credit sequence in data science through Coursera and the school of science, broadly defined, is almost public health for several years, and the need for a graduate credit program was fairly clear. In response, limitless. It seems many large and small the JHU Whiting School of Engineering, through its engineering for professionals division, took on the challenge of creating a rigorous, credit data sci- employers have people doing data ence program based in both applied math and computer science. Relative to the number of applicants, science in some capacity, but without the data science program has had an overwhelming response since the program was rolled out in fall having that label in the job title. 2016. The cumulative number of applications grew from 0 to more than 2,000 in less than two years. How do you view the relationship between statistics and data science/analytics? the ground running.” Also, for the key demographic While there is a wide variety of data science proof students who are working full time or near full grams, all seem to have a substantial basis in statistics. time, it is recommended that students initially take That connection is not surprising when you consider only one course at a time. This allows a person to statistics is defined as the field devoted to “the prac- re-acclimate to academic life. tice or science of collecting and analyzing data”! While we will not proclaim to know “the” rela- What types of jobs are you preparing your graduates tionship between statistics and data science, the for? Describe the employer demand for your JHU program in data science is deeply connected to graduates/students. advanced methods in mathematical statistics, model- The range of jobs associated with data science, ing, and computational statistics. As such, the pre- broadly defined, is almost limitless. It seems many requisites for the data science program involve math- large and small employers have people doing ematics through multivariate calculus (Calculus III), data science in some capacity, but without hav- as well as a course in discrete mathematics and expo- ing that label in the job title. Given that most of sure to linear algebra and matrix theory. our students are part time and are partially or fully employer funded, the students are expected to con- What advice do you have for students considering a tinue with their current employer. For the minority data science/analytics degree? of students not employer funded, we currently have A prospective student needs to be strong in math little data regarding employer demand because the and adept at programming. Someone considering program is a new offering. That being said, given the program who has not taken mathematics or pro- the strong demand for the program, there is little gramming courses in several years prior to starting doubt that those students in the job market will be the program might consider taking a refresher to “hit able to find relevant positions.

BASS XXV Scheduled for Fall The 25th Biopharmaceutical Applied Statistics parallel, two-day short courses will be pre- Symposium (BASS XXV) will be held October sented October 17–19. BASS will also offer a 15–19 at the Hotel Indigo Savannah Historic poster session. District in Savannah, Georgia. One-hour tuto- BASS is a nonprofit entity established to rials on diverse topics pertinent to the research, support graduate studies in biostatistics. For clinical development, and regulation of phar- further information, contact the BASS regis- maceuticals will be presented by speakers from trar at [email protected] or BASS chair, academia, the pharmaceutical industry, and Tony Segreti, at [email protected]. Visit the US Food and Drug Administration. Two the BASS website at www.bassconference.org.

april 2018 amstat news 31 University of Vermont Students must have prior coursework or be able to establish competency in the following: James P. Bagrow is an assistant professor in mathematics and statistics at the University of Vermont • Calculus and a member of the Vermont Complex Systems • Coding (Python/R ideal, but not necessary) Center. He has degrees in liberal arts (AS) and physics (BS, MS, and PhD). • Data structures • Linear algebra Jeffrey S. Buzas is professor and chair of mathematics and statistics and director of the statistics program. He has • Probability and statistics degrees in mathematics (BS) and statistics (MS and PhD). What was your primary motivation(s) for developing a master’s (or doctoral) data science/analytics program? Margaret J. Eppstein is professor and chair of computer science at the University of Vermont and the What’s been the reaction from students so far? founding director of the Vermont Complex Systems The basic motivation was that we live in a renaissance Center. She has a BS in zoology, MS in computer time with so many fields moving from data-scarce to science, and PhD in environmental engineering. data-rich. Students need a suite of skills to be able to contend with the kinds of broad problem solving they Peter Sheridan Dodds is a professor in mathematics and statistics at the University of Vermont, where he will face in the real world, very likely as parts of teams. is also the director of the Vermont Complex Systems These students should not be cogs with narrow train- Center and co-director of the Computational Story Lab. ing. Student response has been extremely positive. How do you view the relationship between statistics PhD in Complex Systems and and data science/analytics? Data Science Our PhD and master’s incorporate training in Website: http://vermontcomplexsystems.org/education/phd computer science, statistics, mathematics, physics (mechanisms), and complex systems. Year in which first students graduated/are expected to graduate: 2021 What types of jobs are you preparing your grads for? (If Partnering departments: Vermont Complex Systems you have had graduates, please summarize the types Center (lead), Mathematics and Statistics, Computer of jobs they took and in what sector.) Science Program format: In-person (online being developed), Data science positions at corporations and in govern- thesis/project or coursework, 30 credit hours, traditional/ ments positions. Students with training that will be non-traditional/full-time/part-time/continuing education formally framed by our PhD have gone on to work for companies, as well as into careers in education. We provide students with broad training in computational and theoretical techniques for describ- What advice do you have for students considering a ing and understanding complex natural and socio- data science/analytics degree? technical systems, enabling them to then—as Students should look for data science programs that possible—predict, control, manage, and create are truly interdisciplinary. They should be able to such systems. develop skills that enable them to explain patterns, Our PhD is a natural addition to our educational and not just reproduce them or generate novel ones. platform, which already consists of an MS in com- While explanation is fundamental to science, it is plex systems and data science and a five-course grad- also crucial in real-world venues to be able to under- uate certificate in complex systems. UVM also now stand and defend, for example, decisions proffered has an undergraduate major in data science. by algorithms for maintenance of ethical, legal, and The major skill sets we aim to train include assurance standards. the following: • Data wrangling: Methods of data acquisi- Describe the employer demand for your grads/students. tion, storage, manipulation, and curation Very strong. We have increasingly received interest • Visualization techniques, with potential for in PhD students with a deeper training. building high-quality web-based applications Do you have any advice for institutions considering • Uncovering complex patterns and cor- the establishment of such a degree? relations in systems through data-fueled Just do it. The world has changed, and it is our machine learning and genetic programming responsibility to adapt. We have to frame educa- • Powerful ways of identifying and extracting tion so students will have a clear path to becoming explanatory, mechanistic stories underlying data scientists. Many essential courses will already complex systems—not just how to use black exist, but the development of hybrid core courses box techniques on data science will likely also be necessary. n

32 amstat news april 2018 Climate Science Day Participants See Change of Tone on Capitol Hill

Photo by Steve Pierson/ASA From left: The ASA’s 2018 Climate Science Day participants—Peter Bloomfield, Bo Li, Dorit Hammerling, and Leonard Smith—gather in front of the American Association for the Advancement of Science building.

he ASA Advisory Committee on Climate CSD’s purpose is to connect scientists with Change Policy and the Section on Statistics congressional lawmakers and their staffs to dis- and the Environment sent four statisticians cuss climate science, a purpose that is quite broad Tto the eighth annual Climate Science Day (CSD) considering the wide range of views and interests on Capitol Hill at the end of January. Peter across the 50 states and 435 congressional dis- Bloomfield, Dorit Hammerling, and Leonard tricts. The scientists attended a series of briefings Smith were return CSD participants, while Bo Li earlier in the day to prepare them for this range. participated for the first time. The briefings included a discussion with a panel

april 2018 amstat news 33 Photo courtesy of Carissa Bunge Bo Li with teammate Rachel Kirpes outside the Capitol

of Hill staffers, tips for Hill visits, strategies for communicating about climate science, a keynote speaker, and time for each team—two scientists of different disciplines and a science society staff person—to prepare specifically for their meetings. This year’s keynote speaker was Laura Helmuth, the health, science, and environment editor at The Washington Post. With a goal for productive, engaging discussion, the request made to congressional offices in the past was to consult CSD participants or the sponsoring organizations when they had questions regarding climate science. This year, however, CSD participants were encouraged to focus on connecting the mainstream scientific view with current and potential future impacts of climate change in their Photo courtesy of Julia Marsh districts for offices that have not yet acknowledged Dorit Hammerling and her teammate, Matthew Hurteau, the scientific community’s view of climate change outside the office of Sen. Tom Udall (D-NM)

34 amstat news april 2018 Photo by Steve Pierson/ASA Leonard Smith meets with his representative, John Rutherford (R-FL). as, for example, was stated in the 2016 letter signed once there was polite, but short and nonengaging by 31 scientific organizations, including the ASA: discussions, there was more open discussion about https://goo.gl/ycrG5d. the impacts of climate change in the district or state For offices whose views were more congruent and even the political challenges. with the mainstream scientific view, the discussions Hammerling, who participated in her second were about being a resource and the risks specific to CSD, also mentioned she enjoyed the interaction the district/state. with other CSD scientists and the preparatory The 19 2018 CSD participants—sponsored by briefings. Li commented on the importance—as 10 science associations—collectively had 70 meet- part of a scientist’s responsibility to society—of ings. As reported last year (https://goo.gl/BTcbCh), regular communication with policymakers and pro- participants again noted the changing tone of cli- viding updates from the research community about mate change discussions in many offices. Where new discoveries and developments. n Committee on Privacy and Confidentiality Offers Resources The ASA Committee on Privacy and Additionally, the committee’s website (https:// Confidentiality supports Data Privacy Day each goo.gl/1iGsVo) provides several resources for pol- year—held this year on January 28—to create icymakers, statisticians, and the public, includ- awareness of privacy issues and their importance ing an overview of methods to maintain data to the public. As part of this effort, a webinar confidentiality; training modules; and discus- titled “What Is a Privacy-Loss Budget, and How sions about various types of data, laws, and Is It Used to Design Privacy Protection for a regulations. The latest postings are a draft Confidential Database?” was presented February of guidelines for developing a privacy policy 1 by John Abowd, chief scientist and associate (https://goo.gl/HdHgWU) and a committee his- director for research and methodology at the tory piece written by Margo Anderson (https:// US Census Bureau and professor of econom- goo.gl/Qw5u7m). ics, statistics, and information science at Cornell Privacy and confidentiality problems and ques- University. Links to the recording and slides can tions can be sent to committee chair, Aleksandra be found at https://goo.gl/nAhcm9. Slavkovic, at [email protected].

april 2018 amstat news 35 columns

STATtr@k On Biostatistics, Alan Alda, and Being Mission Driven

EDITOR’S NOTE: ometimes, things have a I, too, was eager to head back to This originally ran way of coming into your New York City, but the winter on the Flatiron life at the moment you weather had other plans. I saw in blog in January mostS need them. the program there was an unusual 2018. (https://goo. Every year, our entire company gl/8s92Qo) workshop in the last slot of the day of almost 500 gathers at a venue hosted by the Alan Alda Center in New York City to hear updates for Communicating Science that from teams across the company, claimed to use improv theater get additional insight into our techniques to help scientists com- strategy from our co-founders, municate to a variety of audiences. and spend time with colleagues Stuck in Charleston with nothing based in other offices and cities. Alan Alda, actor, New York Times to do, I showed up—skeptical yet A biostatistician by training, bestselling author, and intrigued. This was not the kind I work on our quantitative sci- inspiration for the Alda Center of thing that usually happened at for Communicating Science ences team analyzing oncology a statistics conference. And what Sandy Griffith data collected in real-world set- did all of this have to do with is a principal tings. My job is to use statistics Neal, the research oncologist Alan Alda? methodologist at to learn from the experience of from Cleveland who recently It turns out, Alan Alda host- Flatiron Health, every patient with cancer—not joined the company after 30 ed a television show on PBS a health care just the small percentage enrolled years in traditional academic called “Scientific American technology company based in New in traditional clinical trials. medical center and university Frontiers” for many years, inter- York City. Prior to I was asked to present on environments. viewing scientists about their this role, she was my team’s research during this Ben, head of our provider research. Over time, he realized assistant professor “all-hands” meeting—an excit- sales team, who has spent years that his interview subjects tended of medicine in ing honor when I agreed to it understanding the software to stay within their safe, techni- the department of quantitative months earlier, but as the date challenges that community cal bubbles, not venturing out to health sciences at approached, my excitement start- oncology practices face. engage with the show’s diverse Cleveland Clinic ed to be eclipsed by nervousness. I Lise, a certified public accoun- audience. Scientists aren’t trained and Case Western give a lot of talks, but this one was tant who left a promising career in communications or presenta- Reserve University different. The all-hands was going in the financial services industry tion skills, but, through his cen- School of Medicine. to have a larger audience than I to join Flatiron when we were still ter, he hoped to change this. Griffith holds a PhD in biostatistics from had ever presented to, and it was a small start-up finding our way. At the beginning of the work- the University of going to be in front of my col- The week before our all- shop, the instructor handed out Pennsylvania and leagues—people who I see every hands meeting, with the pre- worksheets with Mad Libs–style can be found on day and whose opinions I care sentation still looming over me, exercises for us to ease into Twitter as @sgrifter. deeply about. Also, I was used I blocked it out of my mind explaining our research in lay- to presenting scientific results to because I had to focus on a more man’s terms. We filled in blanks: scientists. What I was not used to immediate task and talk at hand: • This is important was presenting scientific findings presenting statistical details of because people with cancer get to my colleagues and peers, most validating oncology endpoints at sicker or get better, but no one of whom are not scientists. the International Conference on learns from what worked or I thought of the people who Health Policy Statistics. Scientific didn’t work for most of them. would be in the audience. findings to a scientific audience. Rosie, the always-optimistic At the tail end of the last day • This is exciting because I manager of our office team who of the conference, the suitcases get to work with a big and helps make our New York City in the meeting rooms and the amazing source of information and, if I do my job well, office warm and welcoming limited attention spans signaled can improve and maybe save every single day. everyone’s eagerness to get home. peoples’ lives.

36 amstat news april 2018 columns

We paired up and acted out those patients will get treatments, That night at our company scenes, explaining our research to and might get sicker or better, party, I heard from Rosie, our our partner, who was instructed to but their experience likely won’t office manager. “Every day my react to our presentation as if they ever help others. For all of these team works hard to make sure were an 11-year-old who wasn’t people, different in so many ways, everyone has what they need, we familiar with technical scientific that wasn’t acceptable. That’s what are always so focused on that. One jargon. Equally as important as I’d speak to. I left Charleston with day a year we get to take a step the language I used to describe the a worksheet, a hastily scribbled back and hear about all the amaz- research, our instructor explained note, and newfound clarity. ing work Flatiron is doing, how that the findings had to mean The day of the presentation, it all comes together.” She told something to our partner in order I walked onto the stage filled me that the presentation brought to be understandable—we had to with nervous energy. I saw Ben tears to her eyes, and if this is convey the “so what” of the work. Reynolds, head of our provider what can come from her keeping So, instead of focusing on the nov- sales team, sitting in the front row, the lights on, she feels honored to elty and complexity of the statisti- but far removed from the day- get to be a part of our mission. cal methods of my research, I took to-day context of the work I was I spend my days surround- a different approach: “You might about to present. Then something ed by data and figures. As a have a family member who gets magical happened over the course biostatistician, of course I live sick one day, and you’ll want their of the presentation. The more data for this stuff. What I didn’t doctors to know about what hap- I showed, the more Ben’s face lit know was whether it meant as pened to other patients like them.” up. After I saw that, I began pre- much to everyone else. But then This meant something no matter senting directly to him, feeding a winter storm and Alan Alda who was on the listening side. off his engagement and reaction. showed up in my life at exactly Now I was ready to present to I couldn’t see the other 500 faces the right moment and helped a room of 11-year-olds, but I was in the darkened auditorium, but I me see the link between my lines still struggling with my all-hands knew that if it meant something to on a plot and the mission that talk because of the diverse back- Ben, it had meaning for everyone. connects us all. n grounds and experiences of my colleagues who would be in the audience. Some knew a lot about the research I was planning to present. Kaplan-Meier curves, a staple of oncology research, show the percentage of population alive over a period of time, but can be complex to interpret. Was oversimplifying the results going to be difficult for Neal, our seasoned oncologist, to swallow? Would Lise, part of our finance team, get lost in the clinical details of the findings? And then, it hit me. I scribbled a single note on the back of my worksheet “What do they have in common?” Underneath, I wrote two words: “mission driven.” Despite all their differences, I realized everyone in the audience would have something in common. They were all driven toward a mission of improving the lives of people with cancer—family and friends they’ve known, or those they will know. For some, it might even be their past or future selves. The idea that

april 2018 amstat news 37 columns

PASTIMES OF STATISTICIANS What Does Claire Kelling Like to Do When She Is Not Being a Statistician?

Claire Kelling’s largest project thus A tree of life Claire Kelling made for Claire Kelling completed her first far, a gift for her mother, contains her father. major beading project when she was 10,712 beads. in about the fourth grade and entered it into the 4-H fair. It won fourth place.

Who are you, and what is your statistics position? I am a dual PhD candidate in statistics and social data analytics at Penn State. Tell us about what you like to do for fun when you are not being a statistician. When I am not being a statistician, I like to weave! I mostly create beaded patterns for family members as Christmas and birthday gifts. My most recent patterns have used more than 10,000 beads per design! What drew you to this hobby, and what keeps you interested? I have been beading since I was quite young, start- Claire Kelling began beading when she was 10 years old. ing when I was about 10 years old, I think. I never enjoyed or was particularly good at the typical artis- hobby because, as a statistician perhaps, I am drawn to tic activities, like drawing or painting. I am a triplet, patterns, and beading involves patterns quite obvious- and my brother and sister were both quite talented ly in the designs, but also in the execution of the craft. artistically. Having hobbies was actively encouraged It is very methodical and careful, much like statistics. in my family, as we lived on 56 acres with little to I also enjoy creating something tangible through no TV, internet, or games. Therefore, I took up this my hobby. My beadworks are almost always gifts, hobby as a way to express myself artistically. It start- and I think a handmade gift is an excellent way to ed as weaving potholders and turned into weaving show your appreciation for someone. I think the pretty elaborate designs on a bead loom! fact that I have put 10,000+ beads individually onto My next big step for weaving will likely involve a needle to weave into a design shows a lot about a large fabric loom (about 5 x 5 x 5 ft). I enjoy this how much I care about someone! n

38 amstat news april 2018 columns

STATS4GOOD Data for Good Researchers Fight Human Trafficking

ata for Good researchers often tackle prob- identify key drivers of human trafficking. Applying lems reflecting the worst of society’s ills. One the model outcome to the statistics from smaller such area is human trafficking, which has locations, such as metropolitan areas, can indicate Dattracted considerable interest in recent years. The where relatively high trafficking levels are expected. Walk Free Foundation is dedicated to ending modern Differences between individual states in legislation, slavery and human trafficking and works to raise administration, and the anti-trafficking efforts and awareness; partners with governments, businesses, and organizations result in differences in reported levels With a PhD in NGOs; and performs research. Collecting and analyz- unrelated to the amount of human trafficking present. statistical astrophysics, ing data plays a critical role in this advocacy. The presence of these random factors in addition to David Corliss works in analytics Collecting data on modern slavery and human variables driving the outcomes makes meta-analysis a architecture at Ford trafficking faces many challenges. As these activities good choice for the modeling method, identifying fac- Motor Company are illegal in many areas, secrecy is required for them tors predicting a high level of human trafficking while while continuing to thrive. Victims can be reluctant to report their accounting for variations between states. The fixed astrophysics research plight, and perpetrators strive to prevent detection. effects found by the state-level meta-analysis were on the side. He is the founder of Peace- Walk Free worked with Gallup Inc. to develop then applied to data on large cities to yield potentially Work, a volunteer a world poll and has conducted 48 surveys since unidentified centers of human trafficking activity. cooperative of 2014. Tens of thousands of responses have been col- Peace-Work has begun to partner with local agen- statisticians and data lected. Senior researcher Davina Durgana used these cies to repeat meta-analysis of human trafficking at scientists providing to develop the Global Slavery Index (https://goo.gl/ the level of metropolitan areas. Variations between analytic support for zXTaet). Walk Free subsequently collaborated with states in the reported rate appear to be driven, in charitable groups and applying statistical the International Labour Organization (ILO) and part, by differences in state laws. As a result, work has methods to issue- International Organization for Migration (IOM) to begun to advocate for legislative changes to imple- driven advocacy in produce the first global estimate of modern slavery. ment best practices for finding traffickers and sup- poverty, education, This research found that more than 40 million people porting victims across the country. and social justice. around the world were held in slavery in 2016. At the SAS Institute, a recent project used text Another Data for Good organization active in analytics and machine learning to combat human MORE ONLINE You can learn more human trafficking research is Peace-Work. This all- trafficking. Starting with approximately 1,000 tex- about the Walk Free volunteer cooperative of statisticians, data scientists, tual reports from the US State Department, SAS Foundation and their and other researchers applies analytics to issues in applied supervised and unsupervised machine learn- fight against human poverty, education, and social justice. Peace-Work ing to identify patterns in the text between source and trafficking at www. projects are often in academic and policy research, destination countries, including types of trafficking walkfreefoundation. org. Peace-Work’s with volunteers as likely to be found working with and whether the countries involved were cooperating research, including government economic data to write a position paper to help solve the problem. As described by Principal the US human as working with a social justice organization. Projects Solutions Architect Tom Sabo, the next step was to trafficking study, can have included education performance metrics, root extract those patterns and present them in network be found at https:// cause analysis of homelessness, descriptive statistics data visualizations to share the insights gained with goo.gl/j92WVz. of privilege, and the impact of racial bias. law enforcement and advocacy groups. Text cluster- Peace-Work’s efforts focus on human trafficking ing was also used to identify groups of words in the in the United States. State-level summary data pub- text that could be associated with trafficking children. lished by Polaris was combined with demographic This study enables organizations fighting human and socio-economic data and other sources. A mixed trafficking to better leverage the intelligence to be model was developed to identify economic, demo- gained from the State Department reports, in addi- graphic, and other drivers of human trafficking in tion to finding key word clusters that can be used the United States. The number of victims reported to examine other documents. by state was divided by population to yield per capi- As much as the progress to date is encourag- ta victim rates as the model outcome. These relative ing, much more work is needed. The best efforts numbers are not able to estimate the number of peo- of statisticians and data scientists will continue to ple being trafficked; the goal of this project was to be needed in the fight against this terrible crime. n

april 2018 amstat news 39 columns

MASTER'S NOTEBOOK The Important Role of the Master’s Statistician in Clinical Trials

can’t live without my fellow master’s (MS) stat- often do not receive any formal training while in isticians! Their contributions to each project are school in clinical trials or NIH grant writing. invaluable. The Data Coordination Unit Once a trial is funded, the work scope turns (DCU)I serves as the data coordinating center for toward daily trial-related tasks, including devel- several NIH-funded multicenter clinical trials and opment of the statistical analysis plan, validation clinical trial networks. We currently have seven PhD of analysis and report programs, working closely statisticians and six MS statisticians on our team at with data managers on case report form develop- the DCU, in addition to several project and data ment and centralized data monitoring procedures, Valerie Durkalski- Mauldin is a managers and computer programmers. and report generation. Our MS statisticians are professor of Each of the clinical trials we coordinate involves savvy programmers (in SAS and/or R), which is biostatistics in a statistical team of at least one MS and one PhD essential for many aspects of their job. For central- the department statistician to work on the statistical aspects of the ized data monitoring, the MS statistician and data of public health trial and collaborate with the study team on the manager work together to define the missing data sciences at the Medical University coordination of the trial. Key tasks of the statisti- and logic checks to be programmed for each data of South Carolina cal team include study design and statistical analysis item. They also outline additional data checks and and serves as the plan development, data cleaning, generation of the define which will be performed by the data manag- director of the Data Data and Safety Monitoring Board (DSMB) reports, er and which will be performed by the statistician. Coordination Unit conduct of interim and final analyses (including pro- Examples of items within a trial that our statistical (DCU), a unit that gram validation), and creation of public use data sets team reviews include logic checks for sequences and specializes in the design, coordination, for submission to the NIH data repository. It is quite dates, comparison of data distributions within and and analysis of a workload for each project, and the contributions of across sites, and trends and patterns of data. multicenter clinical the MS statisticians are invaluable to the team. Our MS statisticians also have developed trials. Our PhD and MS statisticians work closely with graphics for centralized monitoring tasks to facili- one another and enjoy a mutually beneficial rela- tate the monitoring of site performance and data tionship. The PhD statistician often acts as an infor- quality. After the trial is completed, MS statisti- mal mentor, and the MS statistician has a trusted cians assist with primary and secondary analyses, source of advice and expertise when questions arise. manuscript preparation, and the creation of pub- The DCU MS statisticians are involved before lic use data sets. The list goes on for the contribu- trial initiation, during trial execution, and after trial tions our MS statisticians bring to each clinical completion. As several of our trials are NIH-funded, trial study team, and our PhD team members are the NIH grant submission process is an excellent grateful for their contributions. opportunity for the statistical team to work together In addition to the daily trial-related work tasks, with their clinical colleagues to develop the appro- our statistical group (both MS and PhD members) priate trial design to address the important clinical meets bimonthly to discuss statistical topics and questions of interest. challenges they have encountered with various tri- Our MS statisticians work on literature reviews to als. This provides the opportunity for relaxed dis- support the analysis plan and assist with sample size cussion within our group and is a great learning estimation and program simulation studies required opportunity for everyone. for more complex trial designs when we need to bet- Based on the questions, we identify topics of ter understand the trial’s operating characteristics. interest that are then presented to the group dur- This exposure to the trial-development and grant- ing a monthly “lunch n’ learn” session. Topics have writing process is the beginning of the team-building covered a wide range, including programming process. All our MS statisticians have enjoyed the graphics in R, review of FDA guidances, random- exposure to this aspect of trial development, as they ization techniques, handling of missing data, and

40 amstat news april 2018 columns

multiplicity. There is never a shortage of subjects, analysis plan is a key skill set and best learned by and it engages the group to stay in touch with the exposure to real projects. Practice using different current methodology research. techniques to achieve the same outcome through We also encourage our MS statisticians to publish programming and analyses so you are able to check methodology research. Having the combined statis- yourself for errors. Finally, realize you never stop tical team of PhD and MS members encourages this learning in both statistics and programming, so aspect, as several of our MS statisticians have signifi- don’t be afraid to take more courses, tutorials, and cant authorship in peer-reviewed journals and pres- webinars whenever possible. ent at national and clinical meetings. Jyoti Arora earned her master’s degree in statis- All these activities encourage professional devel- tics from Rashtrasant Tukadoji Maharaj University opment, which is important for any profession. in India with an undergraduate degree in statistics. Some MS statisticians may later want to pursue their She suggests you engage with real-world data as a PhD, while others are already doing what they love. student. Students should partici- Regardless of the goal or setting, I strongly believe in pate in data analysis competitions the collaboration between PhD and MS statisticians. such as those on DrivenData.org, So, what does it take to be a successful MS statis- Kaggle.com, and KDnuggets.com tician working in clinical trials? Let’s ask the experts. and ask their professors for help Here are a few suggestions from our team. if they have any questions. This Lydia Foster earned her Master of Science degree will offer them an opportunity in applied statistics from Kennesaw State University to work on the unstructured data with an undergraduate degree in mathematics. She and develop skills to identify pat- worked for the Centers for Disease terns, trends, and relationships Jyoti Arora Control and Prevention prior to within data. Data visualization is joining the DCU in 2011. Foster a tremendously useful skill to learn and maintain. believes strong statistical pro- Students earning a degree in the field of statistics gramming is key. In addition, should also focus on developing a broad perspec- being able to clean, manipulate, tive and understanding of the world by exposing and present data is helpful. This themselves to people, topics, and issues in disci- isn’t always the focus of a MS-level plines such as biology, psychology, public policy, program, so a student should con- social science, or computer science. This will help sider other resources such as SAS Lydia Foster in collaborating and communicating with people courses and certifications and R from other backgrounds. tutorials. Collaboration is another Angela Pauls earned her master’s degree in statis- important skill. Study teams often consist of several tics from Northern Illinois University with an under- people with different areas of expertise, and the proj- graduate degree in computer science. She highlights ect is most successful when each person can contrib- the importance of strong program- ute in their area of strength. ing skills, professional knowledge, Holly Tillman earned her master’s degree in and hands-on experience as sig- industrial statistics from the University of South nificant criteria for leading a suc- Carolina with an undergraduate degree in statis- cessful career. Continuous educa- tics. She worked for the Office of tion and good collaboration with Research and Statistics in South PhD-level statisticians can keep Carolina before joining the DCU you on the top of your game, as in 2011. Tillman stresses the well. Pay attention to the detail, as importance of having attention you will be the one who will work to detail and accuracy. “Once you with the data and provide analy- Angela Pauls graduate and find a position that sis results. If you don’t pay attention suits you, continue to be exposed to the detail and process carefully and patiently, it to new projects.” The ability to is easy to make a mistake, which not only reflects understand the study protocol/ poorly on you, but potentially those you work with Holly Tillman design and relate it to the data and, even worse, the trial. n

april 2018 amstat news 41 meetings

The Value of CSP 2018

Photo by Sara Davidson/ASA CSP 2018 attendees introduce themselves to one another at the ice-breaker during the keynote address on February 16. Conference focuses on people and interactions, not just talks and presentations

Jean V. Adams, CSP 2018 Steering Committee Chair

he 7th annual Conference on Statistical by inviting abstracts from potential speakers and Practice was held in Portland, Oregon, Feb instructors and selecting only the best for inclu- 15–17. There were talks, short courses, sion in the program. But, putting people in the Tposters, and blah, blah, blah. You’ve been to confer- same place at the same time is just the first hurdle. ences. You know what they’re about. Watching and What’s the second? listening to presentations. Right? Connecting them. Wrong. The value is in the interactions. To encourage Consider the TED talks. They are all available people to interact with each other, you need to cre- online for free. Yet people spend thousands, even ate a safe space and remove barriers to communica- tens of thousands, of dollars to attend the TED tion. The Conference on Statistical Practice focuses conference in person. Why? on this second hurdle to ensure attendees get the The people. most out of the conference. We limit attendance A conference is all about the people. The word to keep the conference small. We keep the meeting conference comes from the Latin word conferre, rooms close to the shared space to encourage mix- which means “to bring together.” To bring peo- ing. We include an ice-breaker at the keynote. We ple together, you need to have some enticement. encourage folks to gather together for meals. We The Conference on Statistical Practice does this connect mentors and mentees.

42 amstat news april 2018 meetings

Photo by Meg Ruyle/ASA Photo by Sara Davidson/ASA Theresa Henle, Brianna Heggeseth, and Christina Knudson get acquainted at CSP 2018 attendees chat during a break from a short the CSP 2018 Opening Mixer. course February 15.

I was surprised by how much people are using this meeting to connect with each other, even beyond the scheduled events. A lot of small professional networks (formal and informal) were taking advantage of the opportunity.

Photo by Sara Davidson/ASA ASA President Lisa LaVange gives the keynote address at CSP 2018 February 16.

Our focus on people and interactions was effec- “It was really interesting to meet people with tive. Just listen to what attendees had to say about such diverse backgrounds and at different stages in the 2018 Conference on Statistical Practice: their careers and studies.” “I was surprised by how much people are using Many thanks to ASA President Lisa LaVange, this meeting to connect with each other, even who kicked off the meeting with her keynote beyond the scheduled events. A lot of small profes- address, “Reflections on Career Opportunities and sional networks (formal and informal) were taking Leadership in Statistics.” LaVange also led a panel advantage of the opportunity.” session on her #LeadWithStatistics initiative to “I met a woman who started graduate school establish a leadership institute at the ASA. to get a degree in biostatistics. She was an opera Interact at next year’s Conference on Statistical singer before this! I think she wants to bring her Practice in New Orleans, February 14–16. To learn creativity to statistics.” more, visit ww2.amstat.org/meetings/csp/2019. n

april 2018 amstat news 43 meetings Symposium on Data Science and Statistics Promises Solid Program, Networking

he ASA Symposium on Data Science and Statistics (SDSS) is designed for data scientists, computer scientists, and statisti- Tcians who analyze and visualize complex data. The 2018 symposium returns to Reston, Virginia, the site of the original 1988 symposium held under the newly incorporated Interface Foundation of North America (IFNA). The SDSS SSS series, beginning this May, is a joint collaboration SYMPOSIUM ON of the ASA, assuming responsibility for administrative management, and the Interface DATA SCIENCE & STATISTICS Foundation, retaining responsibility for the direc- BEYOND BIG DATA: LEADING THE WAY tion and intellectual focus. RESTON, VIRGINIA • MAY 1619, 2018 The SDSS program offers short courses, con- Yasmin H. Said current sessions, electronic poster sessions, exhib- is the 2018 SDSS its, and many opportunities for networking. Program Chair. Emery N. Brown, a well-known scholar with a She holds a PhD in computational medical focus on anesthesiology and neurosci- statistics. Based ence, will give the keynote address: “Uncovering will present these short courses, as well as invited talks, on her research on the Mechanisms of General Anesthesia: Where including Cloudera, Databricks, Domino, H2O.ai, ecological alcohol Neuroscience Meets Statistics.” IBM, Microsoft, RStudio, and SAS. systems, she was The plenary talks will feature David Scott There will be many opportunities for network- awarded patent 7,800,616, Policy from Rice University, David Brillinger from ing and social interaction with ample breaks, con- Analysis and Action the University of California at Berkeley, Jerome tinental breakfasts on Thursday through Saturday, Decision Tool. Friedman from Stanford University, and Adalbert an opening mixer on Wednesday, and a sympo- She was a visiting Wilhelm from Jacobs University in Germany. sium banquet on Thursday. Barry Nussbaum, fellow at the Isaac The invited program includes session tracks on 2017 ASA president, will be the banquet speaker, Newton Institute data science, data visualization, machine learning, giving a light-hearted talk titled, “I Never Met a for Mathematical Sciences at the computational statistics, computing science, and Datum I Didn’t Like.” University of applications and features well-known scholars such For more information about SDSS, see ww2. Cambridge in as Leland Wilkinson, Roy Welch, Wayne Oldford, amstat.org/meetings/sdss/2018/index.cfm. England. She was a Edward George, William Cleveland, David The 2018 SDSS is being held in honor of founding co-editor- Banks, Michael Trosset, Menas Kafatos, Nozer Edward Wegman, the founder and a key person in-chief of WIRES: Computational Singpurwalla, Lynne Billard, Carey Priebe, Douglas in IFNA, serving as treasurer for some 30 years. Statistics, a Wiley Nychka, Kirk Borne, and Claudio Cioffi-Revilla. He was the founding chair of the statistics depart- journal. She is an In addition, there will be a number of contributed ment at George Mason University and devel- elected member and electronic poster sessions. In total, there will oped both the MS in statistical science and PhD of both the be approximately 300 presentations split nearly in computational science and informatics there. Research Society equally between invited and contributed talks, as He has been dissertation director for 44 doctoral on Alcoholism and International well as poster sessions spanning an array of topics. students, with seven additional students in can- Statistical Institute. A key feature of SDSS is a collection of short didacy. After a 32-year career at George Mason, courses. These short courses will focus on the latest Wegman will retire at the end of May as profes- software tools, technologies, and methodologies for sor emeritus. He earned his PhD in May 1968 data science—including the Hadoop, R, and Spark from the University of Iowa. Several sessions in ecosystems—and give participants hands-on experi- this first SDSS are dedicated to his contributions ence. A number of high-profile technology companies to the profession. n

44 amstat news april 2018 section • chapter • committee news sectionnews North Carolina Chapter Biometrics David Banks of Duke University moderated a discussion this past The Biometrics Section will sponsor seven con- December for the North Carolina Chapter of the ASA. The panel tinuing education (CE) courses at the 2018 Joint featured Jim Blum, Ashley Kesler, Kristen Foley, Kirsten Doehler, Joe Statistical Meetings in Vancouver. Here, we high- Ibrahim, and Bob Rodriguez. light four of them: Nearly 75 statisticians braved the snowy weather to discuss topics such as how to get the most out of your ASA and North Carolina Prediction in Event-Based Clinical Trials Chapter memberships, career directions, continued education, the Instructors: Daniel Heitjan and Gui-Shuang Ying mobility and flexibility of statisticians, and the future of statistics. Did you ever wish you could use the accumulat- Questions focused on how to decide between industry and academia, ing data from your event-based clinical trial to how to best use tools and skills on the job, and what the panelists’ favor- reliably predict its future course? Well, now you ite coding activities on a sunny afternoon are—to which the response can! Give these instructors a half day at JSM 2018 was, “Never let a data set best a beautiful Saturday afternoon!” and they will teach you how using their Bayesian Details about the panel can be found at https://goo.gl/gcMTpq. simulation methods coded in straightforward R. Participants will learn about flexible parametric and nonparametric prediction models for simulating future enrollment and event histories. The instructors relaxing linearity assumptions, interaction surfaces, will describe applications to real trials, showing how differences with machine learning, classification you can predict the timing of future interim analy- vs. prediction, quantifying predictive accuracy, ses, identify efficient enrollment strategies informed detailed case studies using R, and more. by current data, and give DSMBs the best possible Introduction to Bayesian Nonparametric information on the likelihood of trial success. Methods for Causal Inference Bring your own computer and data and give Instructors: Jason Roy and Michael Daniels their methods a try! Have you ever thought about trying more inno- Health Care Analytics in the Presence of Big Data vative approaches to causal inference, but you Instructor: Evan Carey didn’t know how to begin? Bayesian nonparamet- The phrase “big data” has become widespread, but ric methods (BNP) could be exactly what you are what does it mean for the practicing health care looking for! analyst? Come to this course to learn more! In this short course, expert instructors will In this course, participants will gain hands-on review BNP methods and illustrate their use for experience using cutting-edge software tools for the causal inference in the setting of point treatments, analysis of large administrative health care data sets, dynamic (longitudinal) treatments, and mediation. with a focus on Python and Apache Spark. Serial The BNP approach to causal inference has and parallel optimizations techniques using fre- several possible advantages over popular semi- quentist statistical frameworks and machine learn- parametric methods, including efficiency gains, ing frameworks will be demonstrated. the ease of causal inference on any functionals of This course will focus on methods and software, the distribution of potential outcomes, the use rather than the clinical context, but numerous real- of prior information, and capturing uncertainty world examples will be discussed that will offer a about causal assumption via informative prior broad perspective. Students will be provided with distributions. You’ll learn even more from their a copy of a functioning “virtual machine” with all wealth of examples, supported by detailed instruc- software and course materials pre-installed. tions for software implementation using R. For more information, visit the Biometrics Regression Modeling Strategies Section community site at https://goo.gl/bgUKmV. n Instructor: Frank Harrell When was the last time you had a “statistical mod- Physical and Engineering Sciences eling tune-up”? How do you keep up to date with Joanne Wendelberger, Joint Research Conference Chair methods for developing and validating predictive Make your plans now to head to Santa Fe, New models, dealing with common analytical challeng- Mexico, for the 2018 Joint Research Conference es, and graphically interpreting regression models? (JRC) on Statistics in Industry and Technology, This course is the answer! which will be hosted by the Los Alamos National Here is an enlightening and extremely popular Laboratory at the Drury Plaza Hotel June 11–14. course (that’s why we offer it nearly every year) that Further information about the conference is covers multivariable regression modeling strategies, available at jrc2018.lanl.gov n

april 2018 amstat news 45 professional opportunities

Professional Opportunity listings may not exceed 65 words, plus equal opportunity information. The deadline for their receipt is the 20th of the month two months prior to when the ad is to be published (e.g., May 20 for the July issue). Ads will be published in the next available issue following receipt. California Listings are shown alphabetically by state, followed by international list- n The Consortium for Data Analytics ings. Vacancy listings may include the institutional name and address or be in Risk (CDAR) invites applications for identified by number, as desired. a Swiss Re postdoctoral fellowship to Professional Opportunities vacancies also will be published on the ASA’s begin as soon as possible. Priority will website (www.amstat.org). Vacancy listings will appear on the website for be given to applicants with expertise the entire calendar month. Ads may not be placed for publication in the in data science, machine learning, and magazine only; all ads will be published both electronically and in print. financial economics. Apply online at Rates: $320 for nonprofit organizations (with proof of nonprofit status), http://jobs.amstat.org/jobs/10781779. The $475 for all others. Member discounts are not given. For display and online review process will continue until the n advertising rates, go to www.amstat.org/ads. position is filled. EOE. Listings will be invoiced following publication. All payments should be Indiana made to the American Statistical Association. All material should be sent n to Amstat News, 732 North Washington Street, Alexandria, VA 22314- The Department of Statistics at Indiana 1943; fax (703) 684-2036; email [email protected]. University, Bloomington, anticipates Employers are expected to acknowledge all responses resulting from two visiting assistant professor positions for 2018–2019. One year, with possible publication of their ads. Personnel advertising is accepted with the under- renewal for a second. Teaching load of standing that the advertiser does not discriminate among applicants on 2 courses per semester. See www.stat. the basis of race, sex, religion, age, color, national origin, handicap, or sex- indiana.edu for department information. ual orientation. Apply electronically at PeopleAdmin Also, look for job ads on the ASA website at www.amstat.org/jobweb. https://goo.gl/sPkacr. Indiana University is an equal employment and affirmative action employer and a provider of ADA services. n

TRAINING PROGRAM at Texas A&M University for New and Established Investigators in Bioinformatics and Biostatistics The Department of Statistics at Texas A&M University has openings for its NCI-funded T32 training program in Bioinformatics and Biostatistics with an emphasis on the Biology of Nutrition and Cancer (http://stat.tamu.edu/B3NC). Drs. Raymond Carroll in Statistics and OP Robert Chapkin in Nutrition direct the program. Program participants will receive training

via a structured format in biology, genetics, gene expression analyses, genomic signal pro- SH cessing, and the biological mechanisms of cancer that may be activated by nutrition-related factors. No teaching duties are required. Each participant will be mentored by a multidisciplinary team of experienced researchers from Statistics, Electrical Engineering, Nutrition and Biochemistry and will be provided with excellent computing support. Applicants should THEASASTORE have a Ph.D. in a quantitatively oriented discipline, such as statistics, electrical engineering and applied mathematics. Both recent and established investigators are invited to apply. Funding T-SHIRTS, RESOURCES, is restricted to U.S. citizens and permanent residents. The 2-year stipends are competitive ASA LOGO ITEMS, with initial tenure-track positions in statistics. Interested applicants should send a vita and AND MORE! three letters of reference (for new or recent Ph.D.s) by May 1, 2018 to: Dr. Raymond Carroll Visit the ASA Store at Department of Statistics www.amstat.org/asastore! Texas A&M University College Station TX 77843-3143 on your rst [email protected] SAVE purchase by entering 10% ASASTORE AA/EOE at checkout.

46 amstat news april 2018 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 Case Western Reserve University...... p. 21 University of Michigan...... p. 26 Westat...... p. 37 professional opportunities Texas A&M University...... p. 46 U.S. Census Bureau...... p. 47 software JMP Software from SAS...... cover 4 SAS Institute...... cover 3

april 2018 amstat news 47 SOCIAL CHATTER

community. amstat.org In honor of #MathStatmonth, we

www.facebook. com/AmstatNews asked our followers to tell us their job title—and what they actually do. @AmstatNews

www.instagram. IN State Data Center • @INsdc com/AmstatNews Data Librarian @INsdc We help people with statistical research and provide training on accessing public data to govt, academics, nonprofits, and the public!

ISS Association • @ISSA_MOSPI We are the official statisticians of the Government of India, the evidence providers, the backbone of information system...we are Karen Lamb • @drklamb the Indian Statistical Service #mathstatmonth Biostatistician. I help a range of researchers answer questions about James A. Joseph • @jmsjsp health, assisting with Proprietary systems trader — I write statistical specifying research programs to trade market price action during questions, designing uncertain times #mathstatmonth studies, collecting and analysing data and reporting findings. #MathStatMonth

NEXT MONTH: What do you like best about JSM? Sum it up in four or fewer words! #JSMin4words

48 amstat news april 2018

Tatsuo Ya mamoto Statistician and Statistical Consultant, W.LL. Gorore & Assocociaiatetes

Great software in the right hands can change the world.

At W.L. Gore, innovative experiments elevate the R&D process. Waterproof materials made from foolproof science.

Read about W.L. Gore’s success, and ﬁ nd out how JMP can help you change your world: www.great.jmp

SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. ® indicates USA registration. Other brand and product names are trademarks of their respective companies. Copyright © 2017, SAS Institute Inc. All rights reserved. 156200.0517US

JMP-Gore-ad-FINAL-AMSTAT.indd 1 8/9/17 5:27 PM