September 2019 • Issue #507 AMSTATNEWS the Membership Magazine of the American Statistical Association •

Home , Do You Want It Right Now

September 2019 • Issue #507 AMSTATNEWS The Membership Magazine of the American Statistical Association • http://magazine.amstat.org

practice

mentorship communication

BEST PRACTICES: Practice and Mentorship

Also: ALSO: NegotiatingTeaser 1 Fair Compensation Teaser 2 The Dos and Don’ts/Ins and Outs of the Side Hustle SPECIAL OFFER FOR ALL ASA MEMBERS SAVE 30% on Book Purchases with discount code ASA18. Visit the ASA Membership page to unlock savings on the latest books, access exclusive content and review our latest journal articles!

Also Available... MEASURING MEASURING CRIME NEW NEW SOCIETY Behind the Statistics Chaitra H. Nagaraja Sharon L. Lohr Collecting and analyzing data Crime statistics are everywhere, on unemployment, inflation, but how do you know when and inequality help describe they’re valid? If a newspaper the complex world around us. report says “the rate of overall When published by the violent crime decreased by 0.9 government, they are called percent,” how can you tell where official statistics. They are that statistic came from, what it reported by the media, used by measures, and how accurate it politicians to lend weight to is? Is it worth repeating or their arguments, and by sharing? Measuring Crime: Behind economic commentators to the Statistics gives you the tools opine about the state of society. Despite such widescale use, to interpret and evaluate crime statistics’ quality and usefulness. 9781138707603 explanations about how these measures are constructed are seldom provided for a non-technical reader. March 2019: 158pp: Pb: 9781138489073: $29.95 $23.96 www.crcpress.com/9781138489073 This book is a short, accessible guide to six topics: jobs, house prices, inequality, prices for goods and services, poverty, and deprivation. Each relates to concepts we use on a personal level to form an understanding of the society in which we live: We need a job, a place to live, and food to eat.

July 2019: 176pp: Pb: 9781138035980: $29.95 $23.96 www.crcpress.com/9781138035980

NEW STATISTICS AND NEW IMPROVING YOUR HEALTH CARE NCAA® BRACKET FRAUD WITH STATISTICS How to Save Billions Tom Adams 9781498782753 Tahir Ekin Twenty-four million people This book helps the public to wager nearly $3 billion on become more informed citizens college basketball pools each through discussions of real year, but few are aware that world health care examples and winning strategies have been fraud assessment applications. developed by researchers at The author presents statistical Harvard, Yale, and other and analytical methods used in universities over the past two health care fraud audits without decades. This book is both an requiring any mathematical background. The public suffers easy-to-use tip sheet to improve from health care overpayments either directly as patients or your winning odds, and an intellectual history of how statistical indirectly as taxpayers, and fraud analytics provides ways to reasoning has been applied to the bracket pool using standard handle the large size and complexity of these claims. and innovative methods. It covers bracket improvement methods ranging from those that require only the information February 2019: 142pp: Pb: 9781138197428: $29.95 $23.96 in the seeded bracket to sophisticated estimation techniques www.crcpress.com/9781138197428 available via online simulations.

December 2018: 183pp: Pb: 9781138597747: $29.95 $23.96 9781498795784 www.crcpress.com/9781138597747

Visit the ASA Membership hub to access this special offer and more! http://bit.ly/CRCASA2018 AMSTATNEWS SEPTEMBER 2019 • ISSUE #507

Executive Director Ron Wasserstein: [email protected]

Associate Executive Director and Director of Operations Stephen Porzio: [email protected] influential consultants

Director of Science Policy James Steve Pierson: [email protected] Rosenberger Director of Strategic Initiatives and Outreach Donna LaLonde: [email protected] Page 5

Director of Education [email protected] Rebecca Nichols: Doug Managing Editor Megan Murphy: [email protected] Zahn Editor and Content Strategist Page 7 Val Nirala: [email protected]

Production Coordinators/Graphic Designers Olivia Brown: [email protected] Megan Ruyle: [email protected] Linda Young Advertising Manager Claudine Donovan: [email protected] Page 9

Amstat News welcomes news items and letters from readers on matters of interest to the association and the profession. Address correspondence to Managing Editor, Amstat News, American Statistical Association, 732 North Washington Street, Alexandria VA 22314-1943 USA, or email amstat@ amstat.org. Items must be received by the first day of the preceding month to ensure appearance in the next issue (for example, June 1 for the July issue). Material can be sent as a Microsoft Word document, PDF, or within an email. Articles will be edited for space. Accompanying artwork will be accepted consulting career in graphics file formats only (.jpg, etc.), minimum 300 dpi. No material in WordPerfect will be accepted. Amstat News (ISSN 0163-9617) is published monthly by the American Communicating Statistics to Nonstatisticians 11 Statistical Association, 732 North Washington Street, Alexandria VA 22314- 1943 USA. Periodicals postage paid at Alexandria, Virginia, and additional mailing offices. POSTMASTER: Send address changes to Amstat News, 732 Negotiating Fair Compensation 14 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 The Dos and Don’ts/Ins and Outs of theSept embeSider 2019 • IsHustlesue #507 16 News is the member publication of the ASA. For annual membership rates, see www.amstat.org/join or contact ASA Member Services at (888) 231-3473. AMSTATNEWS Build YourThe MSuccessfulembership Magazine of th Consultinge American Statistical Associ Practice:ation • http://magazin e .amstat.org American Statistical Association 732 North Washington Street Advice from a Blog Series 18 Alexandria, VA 22314–1943 USA practice mentorship (703) 684–1221 communication ASA GENERAL: [email protected] ADDRESS CHANGES: [email protected] AMSTAT EDITORIAL: [email protected] ADVERTISING: [email protected] WEBSITE: http://magazine.amstat.org Page 20 Printed in USA © 2019 American Statistical Association

BEST PRACTICES: Practice and Mentorship ® The American Statistical Association is the world’s largest community of statisticians. The ASA supports excellence in the development, application, and dissemination of statistical Statistical Consulting Clients: How to Get Them and science through meetings, publications, membership services, education, accreditation, and advocacy. Our members serve in How to Keep Them 22 industry, government, and academia in more than 90 countries, advancing research and promoting sound statistical practice to inform public policy and improve human welfare. columns

24 STATtr@k Why Be a Statistical Consultant? STATtr@k is a column in Amstat News and a website geared toward Sally Morton people who are in a statistics program, recently graduated from a statistics program, or recently entered the job world. To read more articles like this one, visit the website at http://stattrak.amstat.org. If you have suggestions for future articles, or would like to submit an article, please email Megan Murphy, Amstat News managing editor, at [email protected].

28 STATS4GOOD Back to School With Data for Good 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].

College of Science Dean Sally C. Morton mingles with students on the Virginia Tech campus. Photo courtesy of Virginia Tech. consulting community What Does the ASA Section on Statistical Throughout the past 20 years, the Consulting Mean to You? 30 International Conference on Health Policy Statistics (ICHPS) has played a Science and Technology Policy Fellows vital role in the dissemination of statisti- Ready to Effect Change 35 cal methods in health policy and health services research. In preparation for the next conference, which will take place in January 2020, we’re running a series of in- terviews with previous Health Policy Sta- tistics Section award winners. This month, we feature a conversation between Sally C. Morton and Bonnie Ghosh-Dastidar. Morton was awarded the ASA Health Policy Andrew Althouse 37 Charles Kincaid 41 Statistics Section’s (HPSS) Long-Term Robyn Ball 38 LeAnna Stork 42 Excellence Award in January 2018 at the 12th International Conference on Health Marlene Egger 39 Todd Coffey 43 Policy Statistics, held in Charleston, South Carolina. Read the interview at https:// Gregory Csikos 40 magazine.amstat.org. member news 3 President's Corner 29 Survey of Publications Generated by ClinicalStudyDataRequest Consortium 44 Professional Opportunities

2 amstat news september 2019 president's corner

Statistics and Science Policy

hortly after being appointed the ASA’s In the case against the Department of Commerce Executive Director in 2007, Ron Wasserstein brought by the State of New York, the words “stat- said, “Statisticians complain that no one ever istician” and “statistics” appear 11 times in the oral listensS to us. We need to be saying something!” argument to the Supreme Court on April 23, 2019. As one of his first tasks at the ASA, Ron followed Although the court seemingly made its decision (to through on the 2006 board motion that arose from disallow the question) on other grounds (adminis- 2006 ASA President Sallie Keller’s initiative on tration’s case was not compelling), the ASA’s silence science policy: to hire a director of science policy. on this issue might have been interpreted as sup- Shortly after, Steve Pierson joined the ASA and porting inadequate research on a topic as important found multiple areas in which we all know statistics as the decennial census. Karen Kafadar should be involved. But he discovered our legisla- tors, policy experts, and decision-makers often do USDA Economic Research Service not know statistics should be involved. So, he set The Economic Research Service (ERS) is one of the about identifying those key individuals and has con- 13 principal federal statistical agencies and, with the tinued to do so, despite changing administrations. National Agricultural Statistics Service, is a critical At no time since then has his role been more critical research arm of the US Department of Agriculture. than it is today. Its mission is to provide objective and reliable anal- At each meeting of the ASA Board of Directors, yses on issues regarding the farm, rural economy, Steve and his science policy fellow, Daniel Elchert food safety and nutrition, natural resources, trade (see http://bit.ly/2YT5m7K), provide a report of their and international agriculture, and environment. Its activities, and the list of their efforts is extensive. I work depends heavily on accessibility to data from mention here a few times they have kept statistics in the other agencies in Washington, DC. Yet, for the science limelight. somewhat mysterious reasons, the current admin- istration decided to relocate the ERS 1,000 miles Census 2020 away—without congressional approval—and pro- Editor’s Note: The Washington Post published an opinion by Ron posed to cut its funding by almost half. Again, the Daniel Elchert (March 13, 2019) strongly objecting to the decision ASA leadership thought its silence would be viewed and Steve Pierson by the secretary of commerce to add a question to as support for the decision and joined 170 agricul- contributed to the decennial census asking whether the respondent tural, food, and science organizations in opposing this column. is a US citizen. the move. Ron also prepared an opinion piece for Although this question is on the American The Kansas City Star, where ERS will be relocated. Community Survey (sent to a sample of about 300,000 households each month), its inclusion on National Center for Education Statistics the federally mandated decennial census without The ASA also has advocated on behalf of the adequate field testing could grossly affect either the National Center for Education Statistics (NCES), response rate or the accuracy of answers on all ques- another of the 13 principal federal statistical agen- tions, or possibly both. In addition, the ASA released cies whose mission is to collect and analyze data an official statement (http://bit.ly/33xXiI6) strongly related to education (at all levels) in the US. Its staff advising against the inclusion of the question on size has been so severely limited that it relies increas- the grounds of inadequate testing, and future ASA ingly on contractors. The ASA joined the American President (2021) Rob Santos was quoted in Science Educational Research Association and Council of (July 11, 2019) (http://bit.ly/2N2A9Yv) speaking Professional Associations on Federal Statistics in out against adding the question. Several sessions at supporting increased resources for NCES, as the last month’s JSM also focused on the effects of low data sets it makes public on a timely basis provide response rate and/or mistrust in the confidentiality valuable information for assessing educational prog- of census responses (see http://bit.ly/2OTXmyt). ress and trends in the US.

september 2019 amstat news 3 To Comment or Not to Comment? You may be wondering what criteria the ASA uses to determine whether it should publicly comment on an issue. Several years ago, the board identified the following five criteria that would inspire the ASA to comment: Forensic Science The ASA issued two statements on the importance 1. Does the issue impact the statistics profession or of statistical research in strengthening forensic sci- the ASA itself? ence: an endorsement of the National Academies’ 2. Does this issue impact the quality or integrity of 2009 report, Strengthening Forensic Science in the science or the ability to inform public policy? United States: A Path Forward (April 17, 2010), and recommendations for the use of statistical 3. Does the issue provide an opportunity to educate statements used in expressing the strength (or lack the public about statistics or the statistical aspects thereof) of forensic evidence (January 2, 2019). involved in this issue? Thanks to Steve’s efforts to keep this item front and 4. In our judgment, would our members expect us center on Congress’ radar, federal funds were made available to the National Institute for Standards to be involved in the issue? and Technology to fund the Center of Excellence 5. Does the issue relate to possible infringement on in Forensic Science, which led to the largest con- scientific freedom or human/civil rights? centration of statisticians conducting research in forensic science (CSAFE: forensicstats.org) and the The ASA Board of Directors considers these questions ASA Ad-Hoc Advisory Committee on Statistics in in deciding whether to speak on any policy matter. Forensic Science. Sometimes the answers are clear, but sometimes much less so. Are there issues on which, in your opinion, the Letters of Support ASA should comment? The ASA also supported the Protecting American Votes and Elections Act (PAVE) to enhance federal election security and mandate risk-limiting audits and a bill to restore the Department of Justice Science Advisory Board. We Still Have Work to Do Sallie reminded us (http://bit.ly/2YPccem) that “statistical sciences through the ASA leadership has become part of the voice of science policy.” As a result, statisticians today are now “more visible at the National Institutes of Health, the National Science Foundation, and the White House Office Count on Stats of Science and Technology Policy, making the case Count on Stats (countonstats.org) is a public rela- that statisticians improve scientific research.” tions campaign to build public confidence and trust We have seen the respect our work has attracted in government statistics. In recent months, it has from people such as DJ Patil, 2015 chief data sci- focused on ensuring no attack on federal statistical entist (http://bit.ly/2Mi6pam), and the University of agencies goes unanswered. With ASA staff, it has Virginia’s Teresa Sullivan. By raising the visibility of advocated for a rigorous 2020 Census and the value statisticians’ contributions to questions of scientific and necessity of the ERS, distributed weekly social research and policy, we create respect for our own media features, and provided general agency sup- profession. Please let us know in what areas you port though congressional outreach. This advocacy think the ASA’s advocacy can help you—either for contributed to outcomes such as heightened media your own work or for the betterment of society—at attention on statistical agencies, including publica- [email protected]. Steve, Daniel, and I look for- tions in The Washington Post, Des Moines Register, ward to hearing from you! and The Kansas City Star.

4 amstat news september 2019 influential consultants James Rosenberger: A Collaborator Across Disciplines Coleman Harris JAMES ROSENBERGER ON CONSULTING Would you consider yourself a statistical consultant? If not, what would you call yourself? Well, I have used the term statistical consultant most of my career. I would, however, recommend for most academics involved in statistical consulting to actually think of themselves as a collaborator. It’s a term that implies a broader role, a more equal status with the project leader. I think over the years, that’s become quite clear to me, and it resolves a lot of the stressful issues statistical consultants face getting adequate credit for our contributions.

nspired by a high-school What have been some of the most rewarding collaborations teacher who taught the beau- throughout your career? ty of mathematics, James L. It’s more of a bundle of applications I found rewarding, and that RosenbergerI chose to study had to do with my work with agricultural scientists. The reward mathematics at Eastern for me was understanding their reasons, challenging appropriately, Mennonite University. In the and then applying the best methodology or experimental designs final year of his undergraduate that fit their situation [so they could] do their work more efficient- degree, he was introduced to sta- ly. Another area I worked in for quite a while was transportation tistics. He recalls he particularly research. And again, in that setting, I was often the statistician who enjoyed this course in statistics understood the research and, by learning enough of a particular and FORTRAN programming discipline, you become a really vital member of the team. that, combined with his math Even before coming to Penn State, I worked in a medical center degree, landed him his first sta- where my role was to practice the art and science of statistics when tistics job as a data assistant for a things were not as rigorously designed as modern science dictates. cardiovascular research group at New York University. While learning to curate data What advice would you give someone who wants to have a and perform statistical analysis, career as a statistical consultant? Rosenberger earned his master’s One bit of advice would be to pick good scientists to work with— from the Polytechnic University those who bring the best of their side of the collaborative role to the of New York. Subsequently, process. I would also say pick an area you know enough about that his doctoral work at Cornell you can communicate with the folks you’re collaborating with. And University cemented his inter- last, pick areas you’re genuinely excited about, so it’s not a drudgery, est in applied statistics, as he had but you have a passion for really getting into the heart of the science the opportunity to collaborate you’re collaborating on. I think that goes a long way toward success- and consult with researchers in ful collaboration. both agriculture and medicine. What makes a good statistical consultant? Rosenberger took a tenure-track One important thing is the ability to communicate well and to be position in the statistics depart- interested in what the collaborator is doing. So, in general, the ability ment at Penn State, where one to communicate both orally and in writing, so you can express yourself of his major responsibilities and be understood—those communication skills are very valuable. was to lead the recently formed Statistical Consulting Center, fol- What in your training best prepared you to be a good, well- lowing his PhD. He then took a rounded statistical consultant? brief leave of absence to pursue a I understand a lot about broad scientific questions and how things postdoctoral position at Harvard work. And partly, that’s a matter of growing up in the small business University working on robust environment, where I understood the financial side, the transpor- statistics and scientific reporting tation, logistics—a whole broad range of things one learns if they with esteemed statistician Fred grow up on a farm or [around] a small business. So, I think that Mosteller—another opportunity was part of why I could communicate with a broad spectrum of to collaborate across disciplines.

september 2019 amstat news 5 influential consultants

scientists, just my ability to understand what they were trying to do and interact and focus my contribution on the statistical issues that needed to be solved. land-grant university, which focused on useful research for the What is the most important career advice you have received? local economy and agricultural Pursue the things that bring you joy. You can’t be successful doing sector, particularly for the benefit something that is drudgery to you. Ask yourself periodically what of the region. Returning to the part of your life or part of your job is most rewarding, and then United States, he spent 15 years pursue that aspect. That’s why when I stepped out as department as the department head at Penn head, I jumped into the political arena. I enjoyed the interaction State, seeking to build a broad with a wide range of people, and I think I was fairly good at resolv- department—from theoretically ing conflict rather than contributing to it. rigorous statisticians to strong applied statisticians. He then How did your work in academia prepare you for your govern- spent two years on sabbatical at ment roles, both with the NSF and NISS? the National Science Foundation I think my general understanding of science and how things work as a statistics program officer, provided a great background for being an effective statistical con- enjoying the exposure to various sultant. I was usually a quick study and able to understand the disciplines using statistical meth- needs of a scientist asking for advice or wanting to collaborate on a odology to improve their respec- research project. I think the ability to listen and understand the per- tive scientific fields. spective of your collaborator is very important. Then, bringing my When he returned to Penn understanding of statistics and statistical tools—including design of State, Rosenberger created a bio- experiments and general linear and nonlinear modeling—allowed informatics consulting center to me to be successful working with scientists in a variety of fields. embrace the rapidly changing field, one that includes constant How does a statistical consultant role differ for someone who is statistical challenges associated in academia compared to someone working in government? Was with large data dimensional- there anything new you had to learn to succeed in the govern- ity and reproducible results. ment roles? Stepping down as department During my time at the NSF, I was exposed to a broad range of head, he served as a local coun- research proposals and had to ensure the review process was fair to cilman for the Borough of State innovative and even unconventional ideas. What I enjoyed most College Pennsylvania Council. was working on cross-disciplinary initiatives that linked the math- He cites the experience as an ematical sciences—including statistics—with various disciplines enjoyable venue to flex his lead- that needed good statistical methodology to make valid and repro- ership and collaboration skills, ducible conclusions. teaching him much about the From a long-term perspective, my most challenging role was value statisticians can have in as chair of the Mathematical Sciences Working Group, which consulting relationships. was directed to develop an initiative to support NSF proposals to Since 2017, Rosenberger increase the budget. We built our case on the important role the has served as the director of the mathematical sciences could play in improving the quality of the National Institute of Statistical sciences broadly. The impact of this strategy was seen by a doubling Sciences (NISS)—a national of the division’s budget over the next five years. non-profit organization that brings together the academic, government, and industry sec- tors around statistical and data science issues. Among its diverse Rosenberger turned the theo- advice to researchers; improving interests, NISS focuses on deliv- retical norms of his statistics the research quality of clients using ering sound research related to department on its head and was the center; and finding stimulating public policy and statistical issues awarded tenure with a body of research opportunities requiring in the rapidly growing field of work concentrated on applied new statistical methodology. data science. Rosenberger sums statistics upon his return to Penn A researcher with a propensity up this vision best: “As statisti- State. He then focused on shaping for side trips, Rosenberger took cians, we need to recognize that the philosophy of the consulting two years to help develop a col- we are not the whole field of data center around three goals: training lege of agriculture in Zimbabwe science, but that statistics is a students to deliver sound statistical in the model of an American vital component of what needs to be done.” n 6 amstat news september 2019 influential consultants

A Conversation with Doug Zahn:SELF IMPROVEMENT The Power of In both business and personal life, we engage in hundreds of conversations each day. Some are successful. Some are not. In Stumbling Blocks to Stepping Stones, you will learn from author Doug Zahn how to improve your conversations,

thereby building more effective relationships, whether with TO STEPPING-STONES TO STUMBLING BLOCKS friends or those you perceive as adversaries. “… a user’s guide to disciplined caring.” This book offers a way to begin, continue, and end a successful conversation, helping assure participants are satisfied with the —Roger Kaufman, Professor Emeritus, Conversation in Consultation outcome. Zahn offers strategies to monitor and adjust your input during the interaction. To successfully use his approach, Florida State University you must commit to self-observation and change.

Successful relationships are the key to completing most jobs whether at home or at work. Through the analysis of difficult Doug Zahn—recipient of the 2016 ASA Mentoring Award—is professorconversations emeritus (those you often prefer to avoid) andof the use ofthe methodical building blocks, you will be able to have successful Florida State University Department of Statistics, where he taught appliedinteractions wherever statistics you go. and

Doug Zahn earned a PhD in statistics from Harvard University. He is a professor emeritus statistical consulting courses for 35 years. He has also coached consulting professionalsof the Florida State University Department of and Statistics, where he taught applied statistics and statistical consulting courses for thirty-five and students. years.For over He provided nine years consulting he coached services consulting to faculty taught a course on consultancy skills in the United Kingdom for nine years. So,professionals at the United Kingdom it Office was for National no Statistics while co-teaching a course on consultancy skills. Zahn is the co-author of and Quality ManagementThe Human Plus: The Side Continuous of Statistical Improvement Consulting ZAHN DOUG of Education. He and his wife, Andrea, live in Tallahassee, surprise to us when he shared his knowledge and wrote a book for consultantsFlorida. called Stum- bling Blocks to Stepping Stones: A Guide to Successful Meetings and Working Relationships. U.S. $XX.XX We chatted with Zahn recently to find out more about the book and how it can help consultants.

What or who inspired you to write this book? “OPEN” entails an intentional sequence of A book for The students from the consulting courses I taught thoughtful steps to ensure you have the informa- consultants over the years. After almost every course, one or tion required to produce an effective conversation: by Doug more students made an impassioned plea that I Zahn write a book that would pass on the material from • Reach an understanding with your client on the course to a wider audience. At every ASA meet- how much time each of you has for the con- ing over the past two decades, a group of them versation. Not having a time conversation is would reiterate this request. a guarantee for disaster unless one of you is After I retired from FSU, I finally chose to write a mind reader. the book. This was more daunting and time con- • Compile a list of your client’s conditions for suming than I had ever envisioned. Finally, in April satisfaction. Not having a wanted conversa- 2019, I finished the book and published it with tion results in guessing at what your client iUniverse. A long list of colleagues patiently helped wants from the consultation. me along the way. Many of them are listed in the • Identify what you are willing to do to acknowledgements section of the book. Without help achieve your client’s goals. Be aware them, I doubt I would have completed the book. of your boundaries, such as not being willing to work 24/7 for each of the next You base your book on three core processes— three weekends. Zahn POWER, RAPID, and LEARN. Can you tell us what • Identify what you are able to do to help they are and describe how they are important to achieve your client’s goals. Be realistic. having an effective conversation? • Verify that what your client wants from this The POWER process is a sequence of five steps— conversation matches what you are willing prepare, open, work, end, and reflect. This process and able to do. This will avoid a breakdown, is designed to help you consistently produce effec- which is more effective and efficient than tive consultations. “Power” in this context does not recovering from one. mean controlling, dominating, or manipulating your client; instead, POWER is an acronym for “WORK” is the step in which you respond steps to achieve desired results. with clear content that answers your client’s questions comprehensively, concisely, and accurately. “PREPARE” is an essential first step toward An essential part of all these steps is treating your an effective conversation. In this step, one pre- client with respect, by both your client’s standards pares mentally, physically, and emotionally for and yours. In addition, the goal is to produce a the upcoming conversation. Completing this step relationship that is cooperative, rather than adver- ensures you are ready to engage in conversation. sarial, and collaborative, rather than hierarchical.

september 2019 amstat news 7 influential consultants “END” is often overlooked because people think The LEARN process is also a sequence of five they are done when they complete the “work” steps—listen, evaluate, act, review, and next. It is a step. This is simply not true. Essential work that structured process designed to help you systemati- remains to be done is reviewing and agreeing on cally improve your interactions. any decisions made during the conversation and reviewing and agreeing on who will do what by “LISTEN” to your interactions. Identify one to when to what standards. If another meeting would improve. be helpful, agree on the purpose, time, and place “EVALUATE” this interaction. Identify a seg- of that meeting. ment you want to improve. “REFLECT” is the final step in the POWER “ACT” out a role-play of this interaction, taking an process. Now is the time to consider the effective- action to address the segment you want to improve. ness of your words and actions in the session just completed. You can do this by yourself or with a “REVIEW” the video of the role-play. What partner. If you have videoed this conversation, you worked? What did not work? What will you do dif- have a treasure trove of data you can use to improve ferently in the future? the quality of your conversations. Even if you are ambivalent about videoing yourself, I encourage you “NEXT,” implement what you learned in an actual to do so (provided your client agrees) because you consultation. Organize a different role-play relating can always erase the video, whereas you can never to another issue. Look at larger questions relating to video the conversation after it is over. your career. The RAPID process is also a sequence of five steps—recognize, address, pinpoint, identify, and Why is conversation so important to consulting, and do it. This process is designed to help you rapidly particularly statistical consulting? recover from breakdowns and the resulting emotional We must become facile in conversations, especially clouds that will inevitably occur in your practice. those that have a lot riding on them. This is often the case in statistical consulting. Yet, it continues to “RECOGNIZE” your emotional cloud is the surprise me how often we make assumptions about first step to recovering from your breakdown, be it important aspects of our consultations, rather than large or small. Inevitably, you will encounter sur- having a long enough conversation to identify and prises in your practice when things do not go as you address the issues at hand. Conversation eliminates think they will. These surprises are usually accom- the necessity of becoming a mind reader. panied by emotional clouds. Emotional clouds are similar to thunderstorms in that it is difficult to see What can statistical consultants look forward to tak- your way out once you are in one. In fact, if the ing away from this book? emotions are strong enough, it is even difficult to My hope is that statistical consultants learn and apply recognize that you are in an emotional cloud. these strategies for producing effective consultations, recovering from breakdowns, and learning from their “ADDRESS” is a natural next step to take breakdowns by using video. once you realize you are in an emotional cloud. A strategy that is often helpful here is to learn what Any tips for how the reader can use your book your early warning signs are when heading for successfully? an emotional cloud. Once you are aware of these Find a partner to journey through the book with as is signs, you can deal with the emotional clouds described in the introduction. Yes, you can read the while they are still small, before they become a book by yourself. However, to get maximum value raging thunderstorm. from the book, a partner is essential so you can role- play the situations that cause you trouble in your pro- “PINPOINT” the breakdown is an essential fessional practice. You would not think of going into step toward isolating the components of your an interview without checking how you look in the emotional cloud. mirror. Yet, many of us consistently make the same “IDENTIFY” is an easier task when you are mistakes in a consultation because we are not using looking at the individual components of your emo- the mirror that is available, namely video, to assess tional cloud, rather than the whole gnarly mess. what we do and what we do not do in consultations. I encourage you to address your barriers by using “DO IT” is more manageable when you take on video starting today. one individual component at a time. I wish you much success in systematically improving the quality of your consultations. n

8 amstat news september 2019 influential consultants

For Linda Young, It’s About Learning the Right Questions to Ask Coleman Harris

LINDA YOUNG ON CONSULTING Would you consider yourself a statistical consultant? If not, what would you call yourself? I think of myself more as an applied researcher. I have done consulting projects throughout my career, but, in the end, some solutions require innovation. And so, there’s usually a gap between the material that’s first published in journals and what can be done in practice— that interface between the science and the statistics is where I basically live. So, I’ve done consulting where I didn’t need to do anything new, but often it’s an applied problem that requires research and innovation to actually address the issues at hand.

inda J. Young, who grew How does a statistical consultant role change between academia up in the small town of and government? White Deer, Texas, The emphasis is different in the two places. At a university, the enjoyedL math from an early age. emphasis is on publication, and in the government, it’s on the prod- Before her senior year in high uct you put out. You can’t always implement the optimal solution in school, she applied for a National the government, because there’s a strict timeline you have to adhere Science Foundation program on to; if you have a Bayesian model that takes a week to run, it’s not probability and FORTRAN pro- suitable. At the university, the secondary goal is the project answer, gramming at the University of getting something that addresses the real-world problem; if they don’t Tennessee-Knoxville. After thor- quite solve the problem, they can still publish. So, you have those oughly enjoying her experience kinds of trade-offs. And it requires a real understanding of the gives with statistics and computer sci- and takes and what’s important—what you really mean with your set ence, she ended up at West Texas of assumptions. Then, you have to consider what method would be State University (now West Texas robust enough to implement, even if it’s not the optimal solution. A&M University) to pursue her bachelor’s and master’s degrees in mathematics. What advice would you give someone who wants to have a career Hailing from an academ- as a statistical consultant? ic family, Young planned to Think about what you’d like other than statistics, because you cannot earn a PhD in mathematics at be an effective statistical consultant without learning about the sub- Oklahoma State University but ject matter. I could not do what I’m doing at NASS if I had not been found herself inclined toward involved with farming. So, I won’t be making a mistake simply because the statistics department. This I don’t know winter wheat is planted in the fall and harvested the next introduced her to a formative year. You have to like not just the statistics, but you have to like what project in her career: Why were you’re applying it to and enjoy learning about that subject matter. all the golden marmosets at the Oklahoma City Zoo getting sick? Are there any experiences you’d recommend to students inter- After going behind the scenes ested in statistical consulting, like particular jobs, internships, and learning the right questions courses, etc.? to ask, she was hooked. She then The big thing is to figure out what it is you really like. I learned a lot traversed Oklahoma’s scorch- working in the entomology department for two or three summers, ing hot cotton fields to develop going out to the cotton fields and breeding the plants, and also the sequential sampling plans for experience I had with the Oklahoma City Zoo. And you can do cotton pests, culminating in her internships, because not everyone wants to do consulting in agricul- doctoral thesis work. ture, and you could work under a consultant. But the big question is, do you really have fun looking at those kinds of problems?

september 2019 amstat news 9 influential consultants heard you might be interested in What makes a good statistical consultant? changing your snow shoes for sun- The more you can solve problems, the better you get, which is glasses.” She left for Florida, also where I was really fortunate. I worked for a professor in the ento- excited to trade agricultural prob- mology department, and he would ask me tough questions I had lems for new problems related to to think about. Then, after I came back with the answer, he would the effects of red tide and issues for have another tough question, and so on. And after a while, I could wildlife like manatees and sea tur- think about it while I sat there, but that didn’t come automatically tles. Subsequently, she formed new for me. It’s important to understand that when you start out, you collaborations with the Florida may not be able to answer the question immediately. There’s noth- Department of Health, looking at ing wrong with saying you need time to think about it; it takes a the impact of climate change and lot of practice. heat on the health of Floridians. Young began spending sum- What is the most important career advice you have received? mers in Washington, DC, after Find what you love, your passion, and pursue it. being recruited by the National Institute of Statistical Science Was there anything you had to change in the transition from (NISS) to work on a problem for academia to government? the USDA’s National Agricultural I’ve had to continue to update and learn more, and to continue Statistics Service (NASS). Her to grow as a statistician. But that’s what the PhD prepares you for, team comprised methodology to right? This continual evolution, so I have continued to evolve. And improve annual estimates of farm sometimes, I do miss teaching. If I wanted to, there are opportuni- counts as a supplement to the ties for me to teach in the area. But, in the government, vacations Census of Agriculture, measured are not restricted to the summer or between semester breaks, so it every five years. Pulling on her kind of spoils you after a while. But teaching and interacting with experiences in ecology, she hoped students is something I miss. to implement designs similar to capture-recapture methodology to If you could change one thing in your career thus far, what improve the severe underestimates would it be? NASS measured. I really can’t think of anything, because it took all of my right and Young eventually moved to wrong choices together to get me to the point I am now, and I’m NASS full time, taking on the happy with where I am. Some of my biggest failures taught me les- chief mathematical statistician and sons that were extremely painful, and I would have given anything director of research and develop- not to learn them at the moment. But it is through that growth, ment positions. There, she has through those experiences, that I really grew professionally. focused on making a national impact using statistics, namely Do you have any parting words for our readers? implementing the capture- I really encourage people to consider statistical consulting—research recapture methods into the Census at the interface of science and statistics—because there’s a great need. of Agriculture in 2012 and an People are becoming more statistically literate, but there is still quite enhanced methodology in 2017. a gap. And we can really help make data-based decisions far more The future, she says, is all about solid if we do a good job. incorporating other kinds of data to make the census estimates more accurate. At present, she is working on a method using web Young spent a few more a gap between what she thought scraping to supplement traditional years working with the faculty at researchers meant and what they estimates and aims to incorporate Oklahoma State after finishing her were actually saying. From zoo more administrative data into the PhD, collaborating with entomol- cages in Oklahoma City to sloped next decade. And even though she ogists and weed scientists. Sensing fields in Nebraska, this helped her moved to NASS to implement it was time to leave the univer- fully understand the challenges of capture-recapture, she is excited sity, she took a position working applied agricultural research. about the future including learn- at the University of Nebraska, col- Realizing again that she needed ing and implementing new types laborating again with those in the a change, Young recalls a memo- of data. It requires the learning agricultural sciences. She says a rable email from renowned stat- and growing process she says is key to her consulting success was istician George Casella, then a necessary to succeed as a statis- personally observing occasional faculty member at the University tician—the continual evolution experiments—there was always of Florida. He wrote, “Linda, I her PhD prepared her for nearly n 10 amstat news september 2019 three decades ago. consulting career

Communicating Statistics to Nonstatisticians

ne of the biggest challenges faced by any Focus on the client and project at hand, collaborative statistician is communicating rather than general statistical concepts. statistical information to those with less During my time at the University of Georgia Focus on knowledgeO of statistics. Many of us with a formal Statistical Consulting Center, I supervised many education in statistics receive extensive training in students who were just learning to become collab- Client theory, methods, and application; however, even orative statisticians. Across the board, when these with a PhD in statistics, it is not uncommon to students were initially challenged to explain a statis- have taken one or no courses that focus on com- tical method to a client, they provided equations full municating this knowledge to those who can ben- of Greek letters and other mathematical notation. efit from it. In other words, many of us leave school While each client’s needs should be evaluated indi- with little understanding of how to put our skills vidually, for many clients, this tends to add to their into effective practice. confusion about a method rather than mitigate it. What do our nonstatistician colleagues need from Here is my personal hierarchy, from greatest to us to get the most out of our interactions? As pos- least chance of success, of techniques to explain sta- ited by Janice Derr in her 1999 textbook Statistical tistical methods to most clients: Consulting: A Guide to Effective Communication, there are five dimensions of quality that nonstatisti- Plain English Kim Love is the cians evaluate when collaborating with statisticians: owner and lead Example: “Across your group of students, for collaborator at K. R. every additional point a student scores on the Love Quantitative Availability of support entrance exam, the final achievement score Consulting and Responsiveness of support increases by an average of about half a point.” Collaboration. She earned her PhD Timeliness of support Equations with your client’s variables written out in words from Virginia Tech Completeness of support in 2007 and, prior Example: “Estimated Average Achievement Score to starting her Pleasantness of support = 0.32 + (0.54 x Entrance Score)” practice, served as the associate director Note that none of these dimensions directly incor- Equations with mathematical notation of the University of porates correctness, technical savvy, or methodological Georgia Statistical Examples: “yˆ = 0.32 + 0.54 x1”; “yˆ = β0+ β1X1+ε” awareness. It’s not that those are unimportant; they Consulting Center. Visit her blog about are extremely important, and a collaborative statisti- Well-labeled figures are always helpful when it starting her private cian will not last long without solid abilities in those comes to understanding and should be used in tan- practice at www. areas. It’s simply that most nonstatisticians are unable dem with these techniques when possible. Don’t label krloveqcc.com/blog. to evaluate those aspects of a collaboration and have figures using the cryptic, abbreviated variable names to make the assumption that their statistical collabora- we often use as programming shortcuts; this is a bar- tor possesses those skills. rier to a client who does not think like a programmer The question then becomes, how can we improve and who would need to continuously remind him or our communication skills when working with non- herself of the meaning of those labels. statisticians so they will understand and appreciate our expertise? The following recommendations are Some addenda: Some clients require mathe- based on my personal experience and the advice of matical notation for their eventual research output, other statisticians in the consulting and collaboration including those who are publishing in quantita- community. I refer to nonstatisticians as “clients” in tive academic journals in their fields. Many of my these recommendations, but that term is not limited clients require a combination of these methods to to what one might view as a traditional consulting cli- both understand the concepts and be able to pro- ent; it could be a boss, a coworker, or even a friend vide a final product that meets the requirements of who asks for help with a quantitative problem. their stakeholders.

september 2019 amstat news 11 consulting career

Some clients do have an interest in learning more • Did you follow a proper structure for the inter- general information about statistics. When a client action (see the next recommendation)? Make asks a general question (e.g., What is power?), it is a checklist. still helpful to explain it in a way that is specific to that client and his or her research. For example, for • Most importantly, leave your review session with a client in agriculture, “Power is the probability that one or two goals for future client interactions. you will choose a sample of lettuce plants for your study that will result in a statistically significant dif- Statisticians also can improve communication ference between your two lettuce strains, assuming outside of live client sessions. New collaborators can a difference exists.” The wrong explanation would practice communication skills using role play scenar- be “Power is the probability of rejecting the null ios, in which one person acts as a client and the other hypothesis when the alternative hypothesis is true.” acts as a statistical collaborator (and these can be video reviewed as well). Workshops and continuing Actively improve your communication; education opportunities to help improve communi- Improve communication skills can be learned. cation throughout a statistician’s career also are avail- Why do so many new statisticians respond to cli- able at ASA conferences (such as the Conference on Skills ents’ difficulties in understanding with Greek let- Statistical Practice or the Joint Statistical Meetings) ters and nonproject-specific explanations? The short and through other professional organizations. answer is this is how most of us learned statistics. This notation and general conceptual discussion is Structure your useful in a classroom—it’s a shorthand language we interactions and their Structure have in common so we can learn advanced concepts outcomes. Interactions quickly. However, its usefulness diminishes greatly One of the keys to practicing when working with someone who does not share communication effectively is our background in quantitative sciences. The good to have a well-thought-out news is, contrary to what many of us believe (at least plan for your interactions. Several prominent col- in practice), communication abilities can be prac- laborative statisticians have presented structures ticed and learned. for interactions, including Derr (again in her text- One of the best ways to improve communica- book, Statistical Consulting: A Guide to Effective tion in client interactions is through the use of Communication) and Doug Zahn, who developed video review. Think of it as collecting data on client the POWER process for client interactions (see interactions (with the client’s approval, of course). http://magazine.amstat.org/blog/2009/09/01/hearto- Set the recording device in a location where both fastatcareersept09). While it is probably more impor- the client and statistical collaborator are visible on tant to have a structure in the first place than to the video, record the interaction, and then choose adopt a specific meeting structure, these structures some or all of the video to watch—preferably with do have a number of elements in common. There is a trusted colleague. A video by Eric Vance from the an initial period to prepare for the meeting; terms Laboratory for Interdisciplinary Statistical Analysis for the interaction and a mutual agenda are agreed at www.youtube.com/watch?v=uZq8GWp3Ims is a upon between the statistical collaborator and the helpful resource for implementing video coaching client; there is a work session in which information and feedback sessions. is exchanged productively, with opportunities to question and enhance understanding on both sides Focus on the following during video review: of the table; and time is allowed at the end to review the interaction and agree on steps going forward. • Was your communication effective? Did you pro- While it will take some time and experience to vide explanations that were appropriate for your implement a meeting structure smoothly and with client, and did he/she seem to respond to them? appropriate flexibility, the benefits to the partici- Did you ask the necessary questions when you pants are well worth it. When there is a structure in didn’t understand something? Listen to the ver- place and the statistician no longer needs to concen- bal exchange, but also examine body language. trate on the meeting logistics, it becomes much easier to focus on communication and gauge effective- • What was the relationship between you and the ness. Note that this idea can be extended to written client like? Was it collaborative or hierarchical? reports and other outcomes outside of immediate Was it constructive or combative? interactions—having a general, flexible structure in place to organize statistical information is immensely helpful to making that information understandable.

12 amstat news september 2019 consulting career

Never be afraid to ask a client if he Learn About Clients or she understands something.

Gauge your clients’ knowledge and com- • Directness: Some clients are very good about ask- munication needs on an individual basis. ing questions and guiding the direction of the Not all nonstatisticians who need statistical exper- meeting. Others are quieter and prefer to be invit- tise are the same. That may sound like common ed to contribute. Never be afraid to ask a client if sense, it is easy to begin treating all clients as if they he or she understands something or is happy with are the same over time. While I have seen many the pacing and direction of the meeting. attempts to categorize clients—some serious and some humorous—they fall short of describing the Practice respect. variety and nuances of nonstatisticians who seek Respect takes many forms during cli- statistical collaboration. ent interactions. General politeness With respect to background knowledge, the sim- (such as showing up on time, remain- plest approach is to ask clients what kind of experi- ing focused on your interaction, and Respect! ence they have had with statistics. It’s important to not interrupting your client) is one do this in a respectful manner (see my next recom- way to demonstrate respect. Some of mendation), as immediately firing off a barrage of what I have already discussed in my other questions related to specific statistical techniques recommendations also go toward demon- and courses can be intimidating. Instead, clarify strating respect and are more specific to collabo- that you are asking because you want to make sure ration: acknowledge your client as an individual; you use appropriate vocabulary and provide proper structure your interactions so they are efficient; and explanations. Also, invite them to ask questions any be patient with your clients’ hang-ups when needed. COMMENT time you are not being clear. There is also an issue of professional respect, which cannot be understated—clients collaborate ONLINE Please visit the Communication needs are a bit subtler, and with us because they acknowledge we have statis- web version of I’ve emphasized just a few points in the tical expertise they do not. It is important to rec- this article at following bullets: ognize in return that our clients have expertise in http://magazine. their areas that we do not have. Just like clients do amstat.org to • Atmosphere: Some clients prefer an atmosphere not become experts in statistics after an hour-long add your own that includes friendly conversation, while others meeting, we do not become experts in their fields comments and prefer a more polished, down-to-business envi- thoughts about or areas of expertise during that meeting. Allow and communication. ronment. I usually try to gauge this based on my encourage clients to contribute to the interaction. client’s demeanor after the introduction—is the This not only results in open lines of communica- discussion moving to the parking situation out- tion and positive relationships, but even improves side? Or is the client already placing material on the technical aspects of statistical work, as it pro- the table and starting to tell me about some of the vides a more complete view of clients’ research issues involved in his or her project? problems and objectives. • Attitudes: Unfortunately, some clients have pre- conceived negative perceptions about statistics; Mentor others. they have had little exposure and believe they One of the greatest contributions experienced stat- have poor abilities. They may be nervous or even isticians can make to the field is to share their expe- fearful about meeting with a statistician. Make an riences. How do you explain complicated statisti- effort to be particularly patient when you recog- cal procedures to nonstatisticians (think random nize a client has an emotional reaction to working effects, ordinal responses, and computationally with statistics, and that client will be more recep- intensive techniques)? How did you learn to expand tive to what is being communicated. your communication abilities? n

september 2019 amstat news 13 consulting career Negotiating Fair Compensation

I won’t go into the advantages think about being one minute and disadvantages of charging a late if you are charging a special flat fee versus charging by the hour. evening and weekend rate for A discussion of both approaches their tight deadline. appears in an earlier Amstat News And please don’t raise your article (http://bit.ly/31EFLMo). But consulting fee because the cli- here are some tips for how to get ent is asking for something you the compensation you deserve. know is unethical. You don’t take First, if you are charging by a $10,000 bonus when you find the hour, post your hourly rate a way to push that p-value below on your website. It will save you 0.05. There are some things that time by scaring off anyone who are not worth any amount of is trying to get statistics help on extra money. the cheap and give you some- Get some of your money thing to point to if someone tries up front. In legal circles, this f I had to find fault with my to negotiate a lower rate. Sorry, is known as a retainer fee. Tell colleagues in the statistics you can say, but my rate is what I someone what your minimum profession and myself, it have listed on my website. Non- cost is going to be and ask for wouldI be because we sell our- negotiable rates are the norm in half of this amount before you selves short. I can’t blame the the legal and accounting fields, so start your work. I had a cli- researchers we work with. If they you are not being unreasonable. ent who cut me off after seeing offer us support at 5% effort on a Charge a premium when the some of the preliminary results. Stephen Simon is research grant where everyone client asks for something unusual. I didn’t ask for a minimum, so a part-time else is getting 20% effort or more Your client needs you to run SAS I couldn’t complain. But it hurt independent and we accept, they are the smart programs and you are much more me much more than the loss of statistical consultant ones and we are the dumb ones. comfortable with R? Fine, but that expected income. I take pride and part-time It is not just in the research will be $75 more per hour. Your in my work and knew I had not faculty member in arena that we sell ourselves short. client wants to own your code and provided a quality data analysis for the department of biomedical and In our outside consulting work, not let you re-use it for any future this client. health informatics we don’t charge nearly enough. client? Same thing. Will you have Raise your consulting rates over at the University Much too low is the $130 aver- to work odd hours to meet a ridic- time. If you were working for a of Missouri-Kansas age hourly rate listed in a 2006 ulous deadline? Let them know big company, you’d see extra pay City. He writes about newsletter of the Statistical your evenings and weekends are because you’re more experienced statistics, evidence- based medicine, and Consulting Section (http://com- valuable to you. You’ll sacrifice and produce better quality work research ethics. munity.amstat.org/cnsl/documents/ them for an important deadline, using less time. The same must be Read more of his newsletters/archive/2000-2009). but it will cost 25% extra. true for your independent consult- work at http://blog. You already know how much Every time you ask for more ing work. So, think about asking pmean.com. a typical lawyer or accountant money, you do need to “up your for an extra $25 dollars per hour makes because you needed their game” to the same extent. Your at the beginning of a new year. If it help getting your business started. SAS code will have to be every bit helps, you can offer the earlier rate Your training and experience are as good as the R code you normal- for longtime customers. Hey, any- at least as good as theirs, so con- ly write. The programs your client one who sticks with you year after sider what you pay them a mini- owns will be meticulously docu- year probably deserves a better mum for what you should charge. mented. And you won’t even consulting rate than a new client.

14 amstat news september 2019 consulting career

On the flip side, you are not your exceptions are before you list, they’d lose a bit of time by anyone’s slave, and if a project enter into negotiations. I offer a having to go back to that list and isn’t worth it to you anymore substantial discount to graduate look for a different consultant. (usually because the aggravations students needing help on their But if you are the only consul- are not worth the money), just dissertations. I was a poor gradu- tant they know who has the right call it quits. Find a good stop- ate student once, so I can sym- skills and experience, then their ping place that does not leave pathize. I also offer volume dis- BATNA puts them in a poor your client hanging. That’s the counts. For one client, I cut my negotiating position. This is the big advantage of being an inde- normal consulting fee by 25% in case, more often than not. The cost pendent consultant. In a big exchange for a guarantee of 10 to a client of leaving a problem company, you might be able to hours of consulting a week for a unsolved is usually much greater get out of an aggravating project full year. But limit your exceptions than the cost to you of the lost that isn’t worth your time, but to cases where you see the need. business opportunity.

When you accept less than you truly are worth, you aren’t just hurting yourself. You are hurting every other statistician who is trying to get decent compensation for their work.

more often than not, your boss You must have a stiff spine Don’t use this edge to win will tell you to suck it up and be when you are negotiating pay concessions that are excessive a team player. with a new client. Any reasonable to the point of being coercive. Put everything in writing. person will understand if your You do want repeat business, I have never been burned by a fee is non-negotiable. They may so treat everyone fairly. When bad client, but enough consul- not like you for it, but they will you have the negotiating edge, tants have that they insist on a respect you. just use it to give yourself the contract. Put a payment sched- Roger Fisher and William Ury courage and determination to ule in the contract and include a have a great book about nego- keep your important condi- penalty if the company does not tiation, Getting to Yes. One of their tions non-negotiable. pay within an agreed-upon time top recommendations is to know When you accept less than frame (usually 30 or 60 days your BATNA (best alternative to you truly are worth, you aren’t from receipt of the invoice). If a negotiated agreement). Never just hurting yourself. You are you don’t like the idea of a pen- say yes to conditions in a con- hurting every other statistician alty, offer a discount for timely tract that are worse than what you who is trying to get decent com- payments and raise your base would have if you didn’t reach pensation for their work. So, go consulting rate to compensate. an agreement. out there and get every last penny If you have exceptions to your Also consider what your cli- you deserve. n stated consulting rate, that is ent’s BATNA would be. If they okay, but you should know what got your name from a consulting

september 2019 amstat news 15 consulting career The Dos and Don’ts/Ins and Outs of the Side Hustle Michael J. Anderson

aven’t we all desired to For example: better ourselves, espe- Do continue to work hard cially once we became Don’t run your side and be focused at your day job. Hadults and joined the work force? business during your day job. The goal should always be to We strive to better our position It is tempting to shoot a quick do what you can at your job to where we work, and many of us email to a client while at work, grow the company, not just go studied to earn our advanced but don’t do it! It’s not only through the motions to collect degree(s) while working full unethical, but with all the a paycheck. Jobs and Wozniak time. Since “work hard to get tracking software on the mar- accomplished much at their day ahead” is an American battle cry, ket, companies know what is jobs, even collaborating to create it’s no wonder the side hustle has being done on their computers. one of the most successful com- After a career in structural become so popular. Further, they have a legal right puter games, called Breakout. engineering, The side hustle (i.e., moon- to observe any email you send. Michael Anderson lighting) is nothing new, with If you must, take your lunch earned a business probably the most famous break outside the office and degree and followed and successful moonlight- send your emails from your lap- be open and honest if that with a master’s in Do finance. He advanced ers being Steve Jobs and Steve top or smart phone, but never asked about your side business. to the executive Wozniak, who founded Apple on the company’s time. … Honesty is always the best level, first as a vice Computers. They did so while policy, even if the short-term president and then working at Atari and Hewlett results are painful. as a CEO. His love for Packard, respectively. Similarly, statistics was born in Instagram’s founders created take on too business school, and Don’t he is now finishing a their company by working eve- much and show up at your day master’s program in nings and weekends. job late, distracted, stressed Do operate your business applied statistics. His As a corporate executive, I out, or over-tired. Be open and in a manner that you’ll be able focus and passion are exhort all of us to do our side up front with your side hustle to explain to your boss with in statistical analysis consulting ethically and legally. clients. Tell them this is a side your head held high. Treat your for improved business decisions. His email Our firm has experienced many business (for now) that you will employer the way you want your is MJAServiceData@ damaging actions by unethical be conducting after regular busi- future employees to treat you. gmail.com staff members, but those actions ness hours. Tell them to expect always come back to bite the cul- your emails when they have Having owned two success- prits—the world is much smaller likely gone home. ful corporations, I am one than it once was. Consequently, who always looks for business always discuss it with your firm if opportunities in whatever I you feel there could be a conflict pursue. Our side hustle should of interest. Our consulting work Don’t run a side busi- be approached as if starting a should never adversely affect ness that competes with your company, both structurally our performance for our current day job and/or do anything and operationally, so it will employer, and a good starting that goes against your contract flow smoothly as it grows. point is to use common sense with your employer. Conflicts and the golden rule. of interest are not only unethical, but can result in a lawsuit. Therefore, be cognizant of what projects you engage in.

16 amstat news september 2019 consulting career

Here are a few steps worth taking, whether you are just starting out or considering taking it to the next level. 4 Draft a contract/work agreement. Sadly, the days of the handshake Moonlighting for 1 are long gone. Consequently, Create a business plan— a solid agreement between statisticians isn’t like this should always be you and the client is a must. It job one! doesn’t have to be 20 pages long; A business plan is an invaluable binding contracts can be as short delivering pizzas after tool; it helps guide you and can as one page. However, you want answer difficult questions about to be sure you are protected, our day jobs. the venture. Tons of information and thus a lawyer should review exists on creating one in book the contract. In the past, I have stores and on the internet. drafted the contract in grave detail and then had a lawyer review it. If you’ve done good Moonlighting for statisti- research, the lawyer should only cians isn’t like delivering pizzas 2 need to make minor changes, after our day jobs. Our analysis Get organized. and thus the legal costs can be and subsequent recommenda- This is critical as your busi- substantially reduced. Be sure tions have the potential to ness grows, and if you don’t to discuss this with the attorney make or break a company. Not start early, it becomes over- in advance, because some don’t to mention, some firms can be whelming as the number of allow this type of arrangement. quite litigious. documents grows. Consider a There is such a diverse group structured and logical num- of analysts here that it would be bering system for all your files difficult to give ideas for obtain- (e.g., invoices, analytical files, ing clients; however, staying with- computer programs, etc.). 5 in your passion is always a good This allows for quick and easy Consider a business idea. Perhaps you desire to help access at all times. I often cre- structure. the legal, health care, or transportation fields. We become much ate database software to assist There are many business struc- more valuable as an expert in one with this, but use a spread- tures you should consider (e.g., or two areas. My advice would be sheet program at a minimum sole proprietorship, partnership to think creatively and outside to help you stay organized. (if you have a partner), LLC, the box. Sometimes, that’s what it S-Corp, C-Corp, etc.). There takes to get to the next level. are different tax advantages and The side hustle can be operational requirements with rewarding and exhausting— each that must be considered to physically, mentally, and even 3 determine which is right for you. emotionally—at the same time. Protect yourself with Be sure to consult with a CPA to Taking care of ourselves can insurance. help you weigh all your options. An errors and omissions liabil- keep us performing at our best ity policy is appropriate for (at both of our jobs). Make sure consulting work. Hopefully you to take time out for physical and will never need it, but one claim spiritual exercise, and regularly can make it well worth the cost, do absolutely nothing but relax n not to mention the peace of and recharge emotionally. mind it brings.

september 2019 amstat news 17 consulting career

Build Your Successful Consulting Practice: Advice from a Blog Series Anthony Carra, Business Improvement Group The Business Improvement Group has been busy the past few years! What started off as dabbling in the consulting space has developed into a thriving practice with several team members and contributors. One of our recent activities was to create a blog series to share some of the lessons we have learned over time. Here are some of the high- lights, with links to the full articles.

They say a journey of a thousand miles begins with the first step. How often do individuals come up with an idea for a product or service Dream BIG, to offer, only to be disheartened by the sup- posed hardships of starting a business? How many dreams have been dropped, only to be As principal consultant Start Small and a co-founder of the picked up by others willing to make a go of it? Business Improvement We often find ourselves encouraging others to Group, Anthony Carra at least start off with what they have and slow- brings more than ly build toward their BIG dreams. In “Dream 20 years of experience BIG, Start Small,” we share how we walked a in technical services, group of aspiring entrepreneurs through their process engineering, equipment engineering, limiting beliefs and demonstrated how they applications engineering, could start their businesses that very day with sales, and quality the resources they already had. http://bigbpi. engineering. Carra is com/dream-big-start-small an ASQ Certified Six Sigma Master Black Belt and holds a degree in electrical engineering from the New Jersey Institute of Technology The 3 Key Roles for a and a master’s in applied statistics from Penn State Most of us who offer consulting, collabora- Successful Business University. tion, and other professional services do so as a skilled practitioner in a particular field. This is the most common way to get started, leveraging the network you have and meeting an immediate need. However, there are other roles one must consider to ensure a sustainable business for years to come. While the practitioner is focused on completing the tactical work in the business, there are two other roles that transition from tactical to strategically working on the business for future growth and profit- ability. You will want to exercise all three as you build your practice. http://bigbpi.com/the- 3-key-roles-for-a-successful-business

18 amstat news september 2019 consulting career The Team Inside the Machine Just as there are three roles the professional service provider must take on to ensure business for tomorrow, there are four primary functions required. Most start off filling these functions themselves, but only on an as-needed basis (and most commonly in response to some sort of cri- sis). By proactively identifying these four functions and developing strategies, you can avoid crises and systematically build your practice. Unlike the three key roles that are more of a mindset (albeit, there are some skills involved), these functions can either be learned through training, given to employees, or contracted out to others providing these services. You’ll find the most successful consulting practices align the strengths of each team member with the functions they are most proficient with. http:// bigbpi.com/the-team-inside-the-machine

I often tell people there is a big difference Grow Your between consulting and owning a consulting business (as I have personally found). When Professional you are skilled in what you do, delivering your services by collaborating with your client is Services probably the most fun and rewarding part of the job. However, what happens when that Business project is over and you have to find another? Most professional service providers find themselves on a sort of roller coaster, reaping the benefits of project work and then roaming social media channels to announce their availability when that project is over. This approach is similar to holding the proverbial “will consult for food” cardboard placard and often results The 3 Key Roles for a in periods of downtime and no cash flow. A better way is to treat your practice like a gar- Successful Business den—plant seeds, nurture them, and then harvest a plentiful bounty according to your season. http://bigbpi.com/grow-your- professional-services-business

If you are thinking about starting your own practice or would like to learn more growth strategies, request a copy of The Small Business Owner’s Handbook at http://bigbpi.com/download. Also, follow our company page on LinkedIn for upcoming webinars and articles with tips to help you grow your successful consulting practice. n

september 2019 amstat news 19 September 2019 • Issue #507 AMSTATNEWS The Membership Magazine of the American Statistical Association • http://magazine.amstat.org

practice

mentorship communication

David Morganstein and Philip Rocco Scinto

BEST PRACTICES: Practice and Mentorship

any of us have been touting what you have your degree and do not wish to revisit a col- should do to be an effective consultant lege campus, then you will need to find the needed over the past several years, but we have training elsewhere. beenM neglecting the how. When we are talking about consulting, we are So, how do we successfully implement and exe- talking about interacting with humans. While some cute the necessary “nontechnical” habits, character- online training may be fine (humanities, philoso- istics, qualities, and practices? The answer is to use phy, and human behavior), steer away from com- the same approach as in mastering any subject. munication courses in which there is no active, live First, we assess our enthusiasm. We often start human participation. Active participation and role- down the path to statistics (biology, chemistry, engi- playing are necessary components of most learning neering, etc.) because we have an interest, desire, and training for communication and consulting. and/or passion for the subject. There is no way to We must admit that the ASA could do more to sugar coat this: If you do not have a passion, desire, offer quality courses, seminars, and tutorials in com- or inclination to dedicate yourself to helping and munication and consulting. The current offerings enabling people and organizations to collect, ana- are good, but they could be better. While several lyze, and learn from data, then statistical consulting attempts have been made, we need to recruit more is probably not for you. This is not a bad thing; experts in the actual communication and consulting statistical consulting just isn’t your thing. fields to teach these subjects. Could these instruc- Next, we identify the gaps in education. If you tors also be statisticians? Yes, but not at the expense wish to be a biologist, you study biology. If you of bringing in the most capable instructors. wish to be an engineer, you study engineering. If you wish to be a consultant … Well, what do you Practice, Practice, Practice do? Since there is no formal degree in consulting, You have the passion for consulting. Check. You are look for bachelor’s, master’s, and doctoral programs actively learning the skills involved in communication that incorporate consulting and consulting practice and consulting. Check. What’s next? Practice. in the curriculum. Make sure you include a good If you do not practice what you have learned, you dose of electives involving communication, writ- will not improve. Consulting is no different than golf, ing, speaking, humanities, philosophy, and human tennis, speaking a new language, or playing an instru- behavior. Anthropology and sociology (focusing ment. New skills cannot be mastered by reading a on culture) are valuable, as well. If you already book or attending lectures. They must be practiced.

20 amstat news september 2019 The opportunities to practice are much more We make these suggestions available than you might think. Toastmasters is for many reasons, but the pri- an effective and convenient way to practice public mary one is we have followed these speaking. Writing in a journal and/or on a blog to suggestions in our own careers and communicate your thoughts and/or convey infor- found them helpful. mation is a great way to practice writing. If you We both chose statistics and our do not wish to share your writing with the public, sub-areas of survey statistics (David) write for family and friends. Develop your ability to and industrial design (Phil) because of convey simple ideas and concepts to more compli- our keen interest in what we could accom- cated ones. These can be topics related to statistics or plish and because the field offered a discipline we not, but if you can write about collinearity to your valued: Enthusiasm! grandparents and have them understand it, then you We’ve filled in the gaps in our educations by will probably do well with your clients. Speaking of taking or offering to teach short courses in relevant family and friends, there is no reason you cannot and related topics. There is nothing like having to practice empathy, patience, and listening with them. teach a topic to deepen your own understanding! Reading your presentation for a meeting aloud We often look to short courses at professional meet- and practicing it in front of a mirror is effective. ings such as JSM or the Conference on Statistical You get direct feedback on your timing, expressions, Practice for opportunities to learn new methods approach, and style in addition to mastering what and improve our understanding. you want to say and how to say it. Other opportuni- Improving our writing and speaking skills have ties to practice communication skills include volun- been fundamental to both of us for professional teering for local clubs or professional organizations. growth. Here we are, preparing an article for Amstat Take a leadership position such as chair, group lead- News. We seek out opportunities to present ideas in er, secretary, and/or treasurer for your local running various forums, via both papers and presentations. MORE ONLINE When we have talks to give, our motto is “practice, Nearly seven years ago, practice, practice.” At work, conducting a dry run David Morganstein of any talk is a helpful way to improve and sharpen wrote a summary of the Develop your ability to convey it (not to mention testing to make sure it fits into characteristics of a good statistical consultant. the allotted time). To view that article, visit simple ideas and concepts to Mentors? Here we have some disagreement. Phil https://magazine.amstat. believes he had the best mentors in the world, and org/blog/2012/09/01/ David argues he did. What mattered to both of consulting-best- more complicated ones. us? We were willing to accept the help offered. It practices. is a natural evolution for many to reach a point in their careers when they realize how important it is club, an ASA section, or ASA chapter. Volunteer as a to help bring along the next generation. We were spokesperson for a local charity and/or coach a team. both fortunate to have superb senior statisticians who taught us so much. We realize the value of Reach Out to a Mentor what we received and felt the calling to do the same. What to learn, where to learn it, and how to prac- Throughout our careers, we have both worked to tice and improve are just as important for consult- create and grow mentoring opportunities at our ing as they are for any effort, whether it’s statistics workplaces, at conferences, in ASA sections, and in or tennis. Awareness is a great first step, but action ASA chapters. What is more, we helped acknowl- is required. But what action, how often, and how edge the outstanding mentors in our profession by much? How do you know if the action is effective initiating an annual recognition bestowed to those and leading to improvement? In addition to action, who have contributed so much to our profession. we all need guidance, direction, and feedback. The best way to plan your path may be through a men- In Summary tor. You may not realize it, but you’ve been men- We would like to congratulate you on taking this tored before. Ever play baseball? You probably had a step to improve as a consultant. Step? Absolutely. coach. Play the piano? You probably had a teacher. You have made clear your interest, desire, and And most of us have had the fortune of a parent enthusiasm by reading this article. Next, deter- or guardian to help guide us through childhood. mine and reduce the gaps in education and training Why not a mentor for consulting? A good mentor you may have in communication and consulting. should have the experience, wisdom, and patience Practice what you learn, and practice the identified to work with you to explore your passion; identify effective techniques. Finally, a mentor is of tremen- the skills needed; develop a plan to learn, practice, dous benefit in providing guidance, direction, and and strengthen those skills; and provide feedback. feedback in your quest for excellence. n september 2019 amstat news 21 consulting career

STATISTICAL CONSULTING CLIENTS: How to Get Them and How to Keep Them Acquiring and nurturing your statistical consulting client base

If you’re a statistical consultant or you want to be a statistical consultant, how will you build and maintain a client base?

The Answer Melissa Kovacs, PStat®, is the founder of Word-of-mouth. While some consul- Network. Don’t let this data analytics, tants advertise, specifically in niche word lead to a pit in the stom- program areas to niche audiences, most of us ach of your introverted statisti- evaluation, and cal self. This isn’t as bad as it litigation support get clients because someone recom- firm FirstEval, sounds. Networking is as simple LLC. She is also mended us. as telling everyone you know a part-time that you are a statistical con- associate director sultant.1 Speak it into existence. at Arizona State Announce it on LinkedIn; tell University's How to Make It Happen your friends; tell your family; Morrison Institute for I have six thoughts about your ob- and tell your professional col- Public Policy. leagues you are looking for proj- vious next question: How can I be ects. You’d be surprised whose personally recommended for statisti- colleague’s roommate’s girl- friend’s colleague is looking for a cal consulting work? My thoughts statistician for a project. If you’re here span from the origination of bold, join a networking group your business cycle to your work with in a particular area—maybe one for women, tech in your city, or clients. Here we go: data visualization lovers.

22 amstat news september 2019 consulting career

Consider your brand. As a statistical4 consultant, you do have a brand. What is your brand? Give this Market yourself. some serious thought. What do Maintain your Think of marketing as being you do that signals your brand integrity. After careful relentlessly helpful2 and working to your client? Is it high-quality listening, you may realize you to actively solve people’s prob- statistical consulting? (Well, are not the right consultant for lems. You don’t have to think of that should be a given). Is it a client. Their project6 may be it as placing ads on Google, but the professional data visualiza- too much of a stretch for you, do educate yourself about what tions and graphics you deliver you may not have time to deliv- it is. As a statistical consultant, with your end product? Is it er a high-quality product, or the marketing will encompass talk- your witty repartee on Twitter? topic isn’t in a space for which ing to others about what you do Is it your go-to blog posts? Is it you feel comfortable providing (see #1) and providing helpful your friendliness and ease when advice. It is alright to decline a ideas to solve business problems doing business? Is it your help- project for these reasons, or in using statistics. ful monthly newsletters? Is it any situation you don’t feel you your nerdiness with a touch of can deliver high-quality work. humor? Is it your blunt honesty Be honest and be yourself. Even when delivering results? Is it if you lose income in the short your community ties? Is it your term, you’ll maintain your trustworthiness or remarkable integrity and gain respect in the attention to detail? All these long term. elements are a brand. Are they Treat your your brand? Consider what you consulting like want to be known for and how a business, you’ll impart your brand to your because3 that’s potential clients. what it is. All those biostatistics, applied econometrics, Listen. This is the most and statistical theory courses will important item on this list. only help you with your actual Once a client engagement client work, not with building originates, a statistical consultant and maintaining a client base. should be in listening mode until You are officially in the busi- the finish. From that first phone ness world now, so immerse call and initial client meeting,5 the yourself in it and embrace it. I consultant should be listening recommend not assuming you carefully to understand and rec- All these factors, combined can just “figure it out” with- ognize the unique situation and with your rock-solid knowl- out some help and education. story of each client. Remember edge of applied statistics, sum Attend a business boot camp. that this is part of marketing— to an incredible statistical con- Take classes for entrepreneurs listening to actively find and solve sultant—one who will be asked at your local library. Read some people’s problems and the prob- to return for more work and be Seth Godin books. Steer your lems in their organizations (using recommended by past clients to podcasts toward advice for busi- statistics, of course). Recognizing new clients. n ness owners and entrepreneurs. people’s problems and their orga- Join a business founders’ meetup nization’s needs requires empathy, group. You don’t need to get which requires listening. Your an MBA; the information and clients may not know what help is out there for free, in your they need, but if you listen long community and online. enough, you’ll know. Or, they may think they know what they need, but after careful listening, you may be able to offer a better solution to their problem. september 2019 amstat news 23 columns

STATtr@k

s relatively new members of the statistics profession, you might be considering a WHY BE A career as a statistical consultant. You may alsoA be wondering what it is exactly that statistical consultants do. (As a more established statistician, I hear that question a lot. I also want to reply that we STATISTICAL don’t do anything “exactly”; we allow for error, CONSULTANT? unlike mathematicians.) Mary J. Kwasny

24 amstat news september 2019 columns

Well, as statisticians, we know how to deal with appropriate type of statistical analysis, and some- data—collect it, clean it, analyze it, and interpret times be the “honest broker” when interpreting the findings. We are not, however, experts in the the results to avoid bias. subject matter in which the data was collected. But, Apart from knowing statistics and learning about the subject matter experts, or clients, who collect the research field you work in, you will also need data are—guess what!—not experts in statistics. communication skills to be able to talk to clients, Working in a particular industry or research field, understand their language, and communicate sta- or on a specific project, the statistical consultant tistics effectively. You will need business skills, busi- Mary Kwasny is an needs to not only understand, but be able to trans- ness acumen, time-management skills, networking associate professor late, the subject matter into a statistical problem talents, and a host of other skills. Putting yourself in the department of preventive and then translate findings back to the client. out there as a new professional is hard, isn’t it? medicine and an Anyone who has read a translation of a transla- active member of tion, or played Telestrations or Telephone, knows Consulting Comes in Many Colors the Biostatistics this is not always easy. As statistical consultants, There are different types of consulting and different Collaboration Center we have to be able to understand how the data are types of consulting jobs available. On one end of the at Northwestern University Feinberg generated and what measures are being used. We spectrum, there are short-term consultations, which School of Medicine. have to translate research questions into problems may just be one or two meetings. These typically She has been that can be analyzed with data, determine the are for quantitatively skilled individuals who are enjoying the art of comfortable running their own analyses, but statistical consulting appreciate an expert looking over their and collaboration for work. On the other end of the spec- more than 20 years in academic medical trum, there are long-term consul- centers and external tations; these are occasionally nonprofits. called collaborations in an academic setting. These are projects in which you are an integral part of the research team from start to finish, and they may span many years. And there are projects that could be anywhere in between. Consulting may include being an expert witness in a trial or helping someone with a master’s thesis in physical therapy. As for settings, you could work on your own as an independent contractor or in a small group. You could also work in a large group of consultants, in a consulting firm, or be the “in-house” statistician in

september 2019 amstat news 25 columns

a large company. I personally enjoy working in an Short consults can be difficult this way, but academic environment; I am a member of a collabo- not everyone needs a full-time statistician. Other ration center in which part of my job is consulting. projects provide great opportunities to learn new Consulting may not be part of your job description topic domains, as well as to delve into potentially at all, but you may want to do some consulting “on different statistical methods or skills. For these the side” if your current employer allows it. projects, I use the analogy that statistics are the scaffolding that helps construct or renovate the Consulting Takes You to the Unknown building. Statistical consulting provides a means Almost every consulting project is a giant step into to keep statistical and communication skills sharp. the unknown. You meet with a client, discuss their Additionally, like teaching an introductory statis- research questions who knows how many times, tics course, it provides the opportunity to discover learn more about their problem, and then you do different and potentially better ways to explain the statistics part—possibly including, but not lim- things. All told, statistical consulting is educational, ited to, providing assistance for study design, data exciting, and challenging. collection, proposing an analysis plan, conducting When I was a graduate student, I did a summer an analysis, and writing up the results. But it doesn’t internship and had my first “outside the classroom” stop here. You need to estimate the time and cost statistical consult. A pediatric surgical fellow wanted for your work and later invoice for your work—and to look at the importance of staging laparotomy in all this is for the straightforward projects. pediatric Hodgkin’s disease. As an intern, I had the It’s definitely not like coursework, in which good fortune to work on this consult under the you can expect to fit longitudinal models for your supervision of a faculty member, who was there to Longitudinal Analysis class or survival models for facilitate the consult. It was the first time I used cat- your Survival Analysis class. Any consulting project egorization and regression trees (in my defense, I has the potential to be … anything! don’t think random forests had been developed at My colleague and friend Masha Kocherginsky that time). I learned something about medical pro- said that in more than 15 years of consulting, she cedures and treatments. The investigator was really has never encountered a real-world consulting proj- nice and brought doughnuts to our meetings. And ect that’s straight out of a textbook, and it’s never I got my first collaborative publication out of it. It the same analysis twice. As a result, it’s never boring! was amazing. I was hooked. I never worked with that investigator again, but recently looked him up So Why Would You Want to Do This? only to discover it was his first publication as well. For me, it was precisely because of all the above. He is now an endowed chair in surgical research at John Tukey once said, “The best thing about being a prominent children’s hospital. Apparently, he was a statistician is that you get to play in everyone’s hooked, too. backyard.” I love learning and problem solving. I love seeing what other people are excited about. Consulting Perks and Struggles I love helping them learn more about their data, Not all projects have gone as smoothly or been which, in turn, helps them learn more about their as productive—and I quickly realized doughnuts subject matter. Topics change, people are different, were not the norm. There have been problems with and the one thing you know will happen is you will investigators wanting to run inappropriate meth- learn something and help someone. This alone can ods, times when I thought I understood a data set be addictive. and was—in fact—wrong, and times when I sent There are times I call consulting a mix between an invoice only to have the recipient email back, statistical triage and statistical improvisation. In “FOR WHAT!!!” I would hazard a guess that these triage, the key is to quickly assess, diagnose, and problems are common and that every statistical con- assist a patient. Any actor will tell you their craft is sultant, anywhere in the world, has dealt with one honed when they are forced to improvise. Not that or more of the same struggles. But, over the years, I you are “making things up” as you go, but rather have learned to be a better collaborator and to make you are attempting to give unscripted advice to an it a priority to align expectations so those problems unknown question. don’t occur quite as much. Granted, bad consulting experiences will happen, but, fortunately, there are

26 amstat news september 2019 columns

many more positive experiences and enough enjoy- more statistically efficient or better methods that ment in the “random consult” to continue including could be developed over time to address these collaboration in my job description. adaptations. The bonus with these motivational Another perk of consulting is that, if you are inter- examples is that, provided you have the investiga- ested in methodologic research, it is a hotbed for moti- tor’s consent, you automatically have data to use as vating examples and finding omissions and gaps in an example! knowledge. Clients may seek statistical help after data Apart from the statistical, educational, and com- collection, and it may be clear that traditional statisti- munication opportunities consulting provides, it is cal methods are not appropriate. Other times, when also a convenient way to expand your professional investigators are collecting data, they change protocol network. Being newer to the workforce, this may not out of necessity. These changes may not have gotten a seem like an important reason to become a consultant, statistical stamp of approval beforehand. For example, but thanks to different collaborators over the years, I I once worked with an investigator who failed to tell have developed a network of experts in various fields. me they changed urine collection from a 24-hour col- These individuals have referred other investigators in lection to a 12-hour collection until the end of the need of statistical assistance to me and provide me trial. While pragmatic and understandable, had they with feedback if I’m looking for second opinions or consulted me on the change, I would have requested clarifications when working on another project—or they had a period of time where both were collected since I work in a medical school, when looking for a so we could properly adjust numbers taken under the trusted medical specialist. Additionally, you will learn different conditions. that not every statistician gets along with every poten- In many of these situations, we can make tial collaborator. When that does happen, it is nice to assumptions, appropriately and clearly state those know there are investigators out there who appreciate assumptions and potential limitations, and still your unique talents and style. provide a timely analysis. However, there may be Consulting for New Statisticians As new statisticians, I would strongly recommend some consulting, be it on the job or as pro bono work. It is a great way to keep learning and honing your statistical craft. I would also suggest you don’t go it alone; find a mentor, colleague, or boss who is available for advice. This is especially helpful when you have one of those difficult clients. Independent consulting can be a fabulous career choice, but first it is important to have a network of potential clients and a portfolio of work. As you develop as a consultant, you will be able to better estimate how long a project might take, more easily figure out the real question the investigator is asking, and develop an expanding toolbox of statistical methods to be used for other projects and new problems. While you learn and hone these skills, it is vital to have backup and advice. The Statistical Consulting Section is active in the online ASA Community and caters to all types of consulting—within universities, pro-bono, on the side, sole proprietorships, or wherever and however consulting is done. This is a tremendous resource for those who are thinking of, or actively consulting in, any of those environments. So, go! Consult! Have fun! n

september 2019 amstat news 27 columns STATS4GOOD

With Data for Good

ith the start of the aca- that motivates the group to work would be only too happy to see demic year, students, together and go above any course modern methods of text analytics, professors, and staff are requirements. econometrics, and other statisti- lookingW forward to the opportuni- Of course, student papers cal research techniques applied to ties the new year brings. One order in Data for Good aren’t just for literature, history, philosophy, and of business many students face— STEM courses. Classes outside the the arts. and sometimes delay as long as sciences can benefit from scientific possible—is choosing a subject for research. Here, the context is often Outside the Classroom term papers, class projects, and the reverse of science classes: The Beyond classes, there is an abun- other research. To turn this poten- subject is narrowly defined by the dance of educational opportunities tially worrisome task into a unique class but the research methods vary in Data for Good. Hackathons are opportunity, students might con- widely. For example, document a great way to expand your skills sider how Data for Good subjects vectorization and feature extrac- and horizons at the same time. can meet the educational specification of Abraham Lincoln’s speech- The ASA’s DataFest (ww2.amstat. With a PhD in statistical tions given by their professors. es could be plotted on a time org/education/datafest/index.cfm) is astrophysics, David This can create an opportunity series—like a search on Google one part hackathon and one part Corliss leads a for an interesting paper or project Trends—to trace the development Kaggle competition, with events data science team that addresses a question of need of his thought and language from at a number of universities over a at Fiat Chrysler. in the community and develops opposing the expansion of slavery, weekend. Multiple small teams at He serves on the steering committee D4G experience while complet- to “send them back” to Africa or each site explore the same data set for the Conference ing class requirements. Central America, to emancipa- over several weekends late in the on Statistical While the goal of the class is tion, to reconciliation in the sec- school year, with applications to Practice and is the learning new analytic methods, ond inaugural address … a gradual host an event likely due in January. founder of Peace- there can often be a fair amount development reflected in our wider Another ASA activity is the Work, a volunteer of latitude in how these methods society today. annual Data Challenge. The 2019 cooperative of statisticians and data are applied: The techniques are Good statistical science and a Data Challenge looked at data scientists providing determined by the class, but the little creativity can go a long way from the New York City Housing analytic support for specific cases may vary widely. toward developing both your ana- and Vacancy Survey. Next year’s charitable groups Programming courses can find lytic skills and D4G practice while JSM has a D4G theme, so the and applying practical applications in any sub- delivering a powerful, creative, Data Challenge should be a good statistical methods in issue-driven ject to make a difference in an area and unique essay to professors opportunity to get involved. More advocacy. students care deeply about. A proj- tired of reading term papers that information about the 2020 event ect for a big data or machine learn- merely regurgitate lecture points. will be available around the end of ing class can mine social media Of course, it’s always important to the year. Keep watching this col- and extract features. Professors can read the syllabus carefully and dis- umn to find out more. inspire students with projects that cuss ideas for papers and class proj- Many organizations have make the grade and make a differ- ects with your professors to be sure resources to support projects and ence. D4G projects can facilitate they meet the requirements for the events in which students can team projects by choosing a subject course. I think most professors participate. The AAAS Science

28 amstat news september 2019 columns and Human Rights Commission know if I can help refine an idea, girls with multiple barriers to offers a Campus Event Toolkit suggest possible speakers to com- career development. I can write (www.aaas.org/resources/campus- plete a proposal, or facilitate in about this vitally important work event-toolkit) that includes guid- other ways. by Soroptimists and others but ance and materials for events can never do what is needed and ways to apply science to Get Involved most, which is to be the role human rights research. Next Throughout the school year, model these young women need. year is a census year, and the US many organizations host events Through programs such as this, Census Bureau’s Statistics in to encourage girls and young professional women in statistics Schools (www.census.gov/school women to develop their careers, and data science make a huge s/?eml=20190725mssiss1ccpupr often in STEM fields. One impact on the future of Data for s&utm_medium=email&utm_ such program is Dream It, Be It Good by making a difference in source=govdelivery) program pro- (www.soroptimist.org/our-work/ the lives of girls at a critical point vides resources for K–12 activities dream-it-be-it/index.html) from in their development. n that could be implemented by col- Soroptimist International. This lege students in statistics education. particular program focuses on Teach Data Science maintains a list of Data for Good groups (https:// Survey of Publications Generated by teachdatascience.com/data4good), each full of opportunities for stu- ClinicalStudyDataRequest Consortium dent and faculty research. (If you The worldwide research community has embraced the efforts of the are part of a D4G organization, ClinicalStudyDataRequest.com consortium (CSDR) to share patient-level data from look into getting connected here!) clinical studies. Internships and research partnerships with industry are Here are some numbers to back up that claim: another way students can become • More than 500 research proposals have been submitted to CSDR since its inception in 2014. involved in Data for Good projects while at school. Whether • 40+ papers have been published using clinical research data that has been shared by scouting places to work after CSDR member companies. graduation or just trying new • These papers feature data collected in more than 193 studies. experiences, Data for Good can be on your list of considerations. • A closer look at the resulting publications reveals a diverse profile: Students, professors, and • Two-thirds of the researchers were based at universities, and the others were from research partners can all find medical institutions. ways the important work of Data • Half of the researchers were based in the EU, a third in the US, and the rest from other for Good intersects with course parts of the globe. requirements and other educational activities. Finding your The list of journals associated with these papers is also diverse. Examples include the following: project for next year will bring • The BMJ new opportunities for learning and service. It also might be the • Cancer Prevention Research beginning of a lifetime of activity • Journal of Urology in Data for Good. The theme for next year’s The success of CSDR demonstrates the model can and does work. Sponsors post a list of avail- JSM is “Everyone Counts: Data able studies, researchers prepare a formal proposal, an independent review panel reviews the for the Public Good”! Watch this proposal with respect to scientific merit, and the clinical data associated with the proposal is space for more information as shared (fully de-identified, of course). statistics’ biggest event in North This work has enabled sponsors to meet the growing transparency demands of world-wide America focuses on Data for Good. Now is the time to orga- regulatory agencies (such as the European Medicines Agency) and world-wide scientific as- nize topic-contributed session sociations (such as the International Committee of Medical Journal Editors). proposals. I’m putting together In addition to handling day-to-day research proposals, CSDR is also working to let the global one on human rights analytics research community know of its existence and the plethora of opportunities available. (i.e., genocide forecasting) and doing what I can to help develop Note that full data on the published work is shared! It can be found at other D4G proposals. Let me https://clinicalstudydatarequest.com/Metrics/Published-Proposals.aspx.

september 2019 amstat news 29 What DoesWHY? the ASA Section on Statistical Consulting Mean to You? LeAnna Stork, chair of the ASA Section on Statistical Consulting, asked members on the ASA Community to unlock the “secret sauce” and answer the following questions about what the section means to them. Following are some of the responses she received.

Allison Grove What else would you like to see the Owner/Statistician and Scientific Writer, AG section provide as a member service? Research LLC Networking and/or mentoring opportunities and marketing assistance. Anthony Carra’s blog posts Why did you join this section? on building a successful consulting business have I’m a new member of ASA and the Statistical been great, but would I find a marketing resource Consulting Section. It was recommended to me by that understands this kind of consulting work? My a member of ASA. I have an MS in animal science desire to focus my business more toward statistics and worked in this research area for 20+ years. Ten is somewhat of a career shift, and I could use some years ago, I started a business that provides research guidance along the way. I’ve recently begun taking support on a freelance basis, primarily statistical online stats courses toward obtaining a graduate analysis and scientific writing (agresearchllc.com). certificate and/or MS in statistics. I often help a researcher take a thesis or summary report and turn it into a manuscript for submis- sion to a peer-reviewed scientific journal. This has involved redoing the statistics and learning new Michael Schell techniques. I discovered the part I enjoyed most in Senior Member, Biostatistics and Bioinformatics, this process was analyzing and interpreting the data. Moffitt Cancer Center I’d like to shift my business more toward statistical analysis and thought this section would be a good Why did you join this section? way to network and learn. Keith Muller advised me to join. When we were What value do you get from being both biostatistics faculty at UNC, he said this was a a member? great section for people who cared about statistical I just joined in the last month, but have already practice, which I do. I had thought that this section found the discussion threads and blog posts very was really for statisticians in consulting or business helpful. My next task is to create a listing in the only. Thank you, Keith! ASA Directory of Statistical Consultants. 30 amstat news september 2019 JOIN? consulting community

What value do you get from being What value do you get from being a member? a member? While I don’t post very often to this online mes- I have enjoyed and benefited from the discus- sage group, I enjoy seeing the various issues, both sions—whether they were about business aspects statistical problems and consulting business issues. of consulting, consulting relationships, or specific The JSM Statistical Consulting Section meetings issues related to statistical practice. I have learned a are also good for meeting others who care about great deal from other members of this section. I par- similar issues. ticularly appreciate the helpful and respectful tone of the discourse. What else would you like to see the section provide as a member service? What else would you like to see the Translation of good ideas from theory to prac- section provide as a member service? tice remains a major problem for our profession. Besides more of the same great discussions, I would Whatever this section can do to help facilitate that agree with Michael Schell’s wonderful comments WHY?translation will be beneficial. about the translation from theory to practice. Deborah Dawson Ronald Mowers Professor of Biostatistics, Iowa Institute for Oral Senior Statistician Health Research, University of Iowa Why did you join this section? Why did you join this section? I wish to do some consultancy work. I did stat con- I joined quite a while ago, basically out of an inter- sulting work for a private ag startup the past two est in the consultatory aspect of being a statistician. years (my only client), after a 30+-year career with a This was despite the initial impression that I also major seed company (and six years of teaching math shared that this section was primarily intended for and statistics at major universities). Would like to do those in consulting or business only. I am an aca- more part-time statistical consultant work for seed demic and collaborate with many groups and indi- or ag product companies, or for cognitive decline viduals, but do little other consulting. But it has research (AD, Bredesen-type intervention research). turned out to be one of the sections most helpful to The ag economy has been down this year, and I have me and one that always keeps my interest. not had additional customers; this is an example of the ups and downs of a private consulting business.

JOIN? september 2019 amstat news 31 What value do you get from being a member? COMMUNITYI get some good pointers from the blog, especially starting my own training and consulting business. interactions of experienced consultants with “new- At the time, I tried to get some support and guid- bies” who are asking questions. I am still wanting to ance on how to do so via the Statistical Society be active in statistics, even after a long and (dare I of Canada, but didn’t get very far, so I somehow say) productive career. stumbled upon the ASA Section on Statistical Consulting and was impressed with the level of engagement and support I found within this sec- What else would you like to see the tion. ASA has provided a welcoming professional section provide as a member service? home for me and given me a sense of belonging. I Networking in specific topic areas. For example, am grateful in particular to two local consultants, would like to connect with some individuals or a Dr. Carl Schwarz and Dr. Jonathan Berkowitz, group who are currently working in either the world for offering guidance and support when I first food production area or in the area of Alzheimer’s started. And to my former professor, Dr. John disease/cognitive-decline research. Thanks for the Petkau, who is an outstanding teacher and men- opportunity for giving feedback. tor of aspiring consultants.

What value do you get from being Isabella R. Ghement a member? Ghement Statistical Consulting Company Ltd. I’ve met many wonderful colleagues by being a member of this section. I also tried to give back to the section by working with some of these colleagues Why did you join this section? to organize JSM sessions such as Communicating When I took a consulting course in grad school, I Statistical Findings to Consulting Clients Operating was the least likely candidate to become a statistical in a Decision-Making Climate: Best and Worst consultant. I had a passion for teaching, but con- Practices (whose materials are available at www. sulting seemed terrifying and I remember crying for ghement.ca/jsm2014.html) and Secrets to Effective a whole weekend because I just didn’t know how to Communication for Statistical Consultants (www. write a consulting report. But then I graduated and ghement.ca/resources.html). needed to find a job option that allowed me to stay in Vancouver, which is how I ended up considering

32 amstat news september 2019 What else would you like to see the What else would you like to see the section provide as a member service? section provide as a member service? COMMUNITYI would like to see that the section does more to sup- I think we have a fairly diverse set of interests and port consultants such as myself who are profession- a range of experience. It might be neat to set up ally isolated. There have been some attempts to do mentoring relationships so that more junior con- this in the past, but they didn’t pan out long-term. sultants have some guidance as we venture out into the consulting world. Someone else also mentioned special interest groups, which likely could accom- plish the same goal in a group setting. Jackie Szymonifka It also might be nice to set up an online “library” PhD Candidate, NYU School of Medicine, of sorts that has consulting-related documents for Department of Public Health, Division of others to refer to. Again, given our range of experi- MORE ONLINE Log in to view the Biostatistics (with 11 years of experience in ence, items like consulting agreements, invoices, entire thread at http:// academia/consulting) proposed budgets, etc. all could be really useful to bit.ly/30ppWIY. have sample documents to refer to, as well as general tips and guidelines for consulting relationships. Why did you join this section? A previous thread was posted by someone looking I’ve been consulting for about a decade now, but for advice on how to approach initial consultations, only recently rejoined the ASA, so when I was look- questions to ask, etc., and I think having a summa- ing at the various section options, this one seemed ry of that type of information in one place would very appropriate. be extremely valuable. There also have been a few threads about What value do you get from being administrative/logistical details (obtaining insur- a member? ance, what type of company to form, etc.) that I think would be useful to have in a single document As others have said, it’s been very educational to to refer back to. n read responses to technical questions. It’s also helpful, and at times cathartic, to read about others’ experiences, both positive and negative.

september 2019 amstat news 33 Electronic UndergraduateFREE Statistics Research Conference

The Electronic Undergraduate Statistics Research Conference (eUSR) will take place November 1.

The conference will feature talks from the Under- KEYNOTE ADDRESS graduate Statistics Research Project Competition Jennifer Thompson winners and sessions on graduate school and Biostatistician at Devoted Health careers in statistics and data science.

Undergraduate students who want to share their work and register for the conference should submit an abstract by October 23, 2019. Prizes will be given for the Best Video Presentations!

For details visit https://www.causeweb.org/usproc/eusrc/2019

Co-sponsored by the American Statistical Association and Consortium for the Advancement of Undergraduate Statistics Education

34 amstat news september 2019 consulting community

Science and Technology Policy Fellows Ready to Effect Change

illiam Adler, Kyle digitally map countries such as Statistics (IMS), Mathematical Novak, and Jiayang Côte d’Ivoire. He will serve as a Association of America (MAA), Sun were selected as congressional fellow. and Society for Industrial and Wfellows for the inaugural ASA/ Sun earned her PhD in statis- Applied Mathematics (SIAM) ACM/AMS/IMS/MAA/SIAM tics from Stanford University in to recruit applicants with exper- 2020–2021 Science and 1989 and is a fellow of the ASA. tise in statistics and data science. Technology Policy Fellowship Until recently, she was a pro- These partners used their net- and will begin their appoint- fessor at Case Western Reserve works to promote the program ments this month. University in the department through organizations devoted to Adler earned his PhD in of population and quantitative increasing diversity in the math- computational neuroscience health sciences. She will begin ematical sciences. from New York University in a new position as professor and The fellows will help real- 2018. He has been working chair of the department of statis- ize a future envisioned by the with the Princeton University tics at George Mason University Commission on Evidence-Based Gerrymandering Project in the at the conclusion of her fellow- Policy Making: A future in which role of computational research ship. Her interdisciplinary work rigorous evidence is created effi- specialist to develop tools to help includes cancer epidemiology, ciently as a routine part of gov- citizens understand district maps environmental science, imaging, ernment operations and used to and fairness in redistricting. His neuroscience, surgery, wound construct effective public policy. community service includes care, astronomy, computer sci- The fellows are supported by working as a data scientist with ence, energy, and law. Her ser- a grant from the Alfred P. Sloan DataKind on a project with an vice includes work with the Foundation and will help estab- education startup to track col- Caucus for Women in Statistics lish the “learning agenda” for the lege persistence. Adler will serve (CWS), for which she was the federal government envisioned as a congressional fellow. 2016 president. As president, she by the commission’s 2017 report Novak earned his PhD in led an effort to modernize the by bringing expertise to such applied mathematics from CWS web presence. Her fellow- areas as data science, machine the University of Wisconsin - ship placement is with the US learning, data visualization, and Madison in 2006. Since 2017, he Department of Agriculture. causal inference. has worked with the US Agency When asked what he hopes The selection committee was for International Development as to achieve, Adler responded, “I chaired by ASA Past President a AAAS Science and Technology hope that I will be able to use my Lisa LaVange, ASA President Policy Fellow researching the technical skills to help whichever Karen Kafadar, and ASA use and impact of emerging and office I join make meaningful President-elect Wendy Martinez. digital technologies like machine changes to policy. I also hope to Also serving on the committee learning, blockchain, big data, gain a stronger understanding of were representatives from the open data, and real-time data in how statistical findings and ways ASA Committee on Minorities low-income countries. He also of thinking can have an impact in Statistics, Government has worked with DataKind, col- in Congress.” Statistics Section, and Section laborating with other data scien- As the fellowship name indi- on Statistical Learning and Data tists to analyze air quality data cates, the ASA collaborated with Science, along with represen- and assess impacts of humanitar- the Association for Computing tatives from the AAAS, AMS, ian aid. Additionally, Novak con- Machinery (ACM), American ACM, IMS, MAA, and SIAM. n tributed to the OpenStreetMap’s Mathematical Society (AMS), Missing Maps project to help Institute for Mathematical

september 2019 amstat news 35 We interviewed a handful of statistical consultants to find out who inspired them, why they became consultants, and what skills they think consultants need to succeed.

36 amstat news september 2019 Andrew D. Althouse

Assistant Professor of What or who inspired you Medicine, University of to be a statistician/data Pittsburgh scientist? I liked the class! Seriously, it was my favorite class as a freshman in college, and I just went with What is the most excit- What do you enjoy doing it. Over time, I began to develop ing part of consulting? in your spare time? more specific interests (statistical Seeing the results come to frui- Lifting weights! See applications in medicine), but it tion! You put so much work into https://magazine.amstat.org/ all started with college freshman- these projects. The rewarding blog/2018/08/01/ year statistics. part is seeing the finished prod- pastimesalthouse. uct in a journal and, hopefully, Why did you become Name one or two favor- a consultant? seeing that it has a meaningful impact on clinical practice. ite blogs or books you Because I like to collaborate! For have read and would rec- me, the most fun part of this What is the best piece of ommend to others. job is working with a physician advice you received from Four Days to Glory, about two or medical professional who has a mentor regarding sta- high-school wrestlers in pursuit a question. We can help them tistics or consulting? of their fourth state champion- define and refine it, figure out As much as possible, try to work ship. An inspiring story about the right data source and study with good collaborators. The the importance of hard work and design to answer their question, work experience of many statis- determination. n create an analytic plan, and exe- ticians is dictated by the environ- cute that plan! ment and people you work with, Name a few specific so you’ll be best suited to find a skills you need to be a place where you have some say in consultant. the people you work with. Communication is critical. You What advice would you can have all the technical and give to statisticians just programming knowledge in the beginning their careers? world, but if you cannot commu- Learning the field and practice nicate (both listening and speak- of statistics is an ongoing pro- ing), it is very difficult to be an cess. Don’t expect to know it all effective consultant. on your first day as a staff mem-

ber or faculty. There are always more models, tests, programming tricks to learn. Keep an open mind, and don’t be afraid to engage or ask questions when you aren’t sure of something.

september 2019 amstat news 37 What is the best piece of Robyn L. Ball advice you received from a mentor regarding sta- Statistical Consultant, Stanford tistics or consulting? University Center for Artificial I’m not sure who told me this, Intelligence in Medicine & Imaging (AIMI) but it has stood the test of time: “You may have the most brilliant idea, but if you can’t express it or if it’s so complicated that no one understands it, no one will want to use it. Learn to communicate and What or who inspired Name a few specific choose the simplest, most straight- you to be a statistician/ skills you need to be forward approach first.” data scientist? a consultant. When working on my BS and I think it’s more qualities, rather What advice would you MS in mathematics, I thought than skills, though skills will give to statisticians just I wanted to be a theoretical grow over time. In my opinion, beginning their careers? mathematician and philoso- the personal qualities needed for It’s the same advice I give every- pher. While a master’s student, I success are the following: one. Spend some time with yourself. Write down what is most took a statistics class from Blair • A natural curiosity to learn Sterba-Boatwright, who elegant- and a willingness to ask important to you in every area of ly explained the fundamentals questions your life: where you live; relation- of statistical concepts. Then, for ships; community; your work; etc. my master’s thesis, I developed • Perseverance and integrity Don’t consider any barriers. Just a machine learning algorithm (relationships—and trust— take the time to write down what with Philippe Tissot and Beate take time and persistence). you want because you are either Zimmer used for many years Be true to your word. moving toward your goal or you by the Texas Parks & Wildlife • Enjoyment of the creative are moving away from it; there is Department to preserve wildlife. process of finding the right no standing still. Then, every deci- I was so excited that what I built statistical framework for the sion becomes clearer. could actually be used—and problem at hand Above all, trust yourself used for good—that I decided to and be trustworthy. Trusting pursue statistics. What is the most excit- relationships with friends and Plus, options. When I was in ing part of consulting? colleagues will always see you high school, someone gave me For me, I get excited when the through. Don’t be afraid of a chal- the best advice: “Make the choice investigator/collaborator lights lenge or a change; walk through that gives you the most options— up and says, “Yes, that’s it!” the open doors. Importantly you can always turn something Sometimes, they don’t know the though, listen to yourself when down if you decide you don’t question they truly want to ask, it’s time to walk away. want to do it.” Statistics gives me and getting there is half the fun. What do you enjoy doing options; I can work in just about Also, finding a way to get the data in your spare time? any field. back into the hands of the expert I love hiking, camping, snow- Why did you become is powerful. They know what shoeing, staring at mountains, a consultant? to look for, and so providing a and watching waves crash. Also, I means for them to interrogate enjoy spending time with family Mainly, it’s fun! I love getting to their data—perhaps in an inter- learn all the time, and I enjoy the and friends, whether just talking active way—has led to insightful or going on adventures together. creative process of understanding discoveries. As Mary Ann Handel what the investigator/collaborator put it, “I don’t have to send you Name one or two favor- wants to know, why they want to an email at 2 a.m. when I have an ite blogs or books you know it, how they plan to use it, idea and a question about a gene’s and then translating that into a have read and would expression; I can just log in and recommend to others. statistical framework that provides look for myself!” a pathway to a trustworthy result. These aren’t statistics books, I know, but ... On the Road - Jack Kerouac

Together Is Better: A Little Book of Inspiration - Simon Sinek n 38 amstat news september 2019 What or who inspired you Name a few specific to be a statistician/data skills you need to be Marlene J. Egger scientist? a consultant. In 1970, I went to Knox College Keeping up with developments Professor (tenured), Division in Illinois. My favorite professors in statistics, statistical software, of Public Health, Department were Rothwell Stephens and Bob and advances in computing is of Family and Preventive Oberg (another ASA member). essential, as that is the capital a Medicine, University of Utah Rothwell Stephens suggested statistical consultant brings to the School of Medicine biostatistics, as I was a double table. The ASA has fine continu- major in math and biology and ing education programs in this had done well in his intermedi- regard, though their value was not ate theory of statistics course. He always as well understood as it is also wrote a letter of recommen- now. Proactiveness and time man- Disease Control and Prevention dation for me to Stanford. I was agement are essential. Diplomatic statistics branch. Gladys helped further inspired by Ingram Olkin, skills with grant writers are essen- found the ASA Section on Brad Efron, Persi Diaconis, Ted tial. Contract writing is not as Statistics in Epidemiology. I have Anderson, Alice Whittemore, and formal for an internal consultant known her to stop for a panhan- others. When it was time to do as an external consultant, but dler and ask, “What do you really my dissertation, I did it with Bill mutual, experience-based trust is want?” In one case, she gave a (Byron William) Brown in biosta- essential. For example, it is gen- young woman a bus ticket to tistics, and he assisted me in get- erally a norm that the statistician get off the street and go home. ting a position at the University who wrote the grant proposal will Gladys told me, “It is always time of Utah, where I still work today. be the statistician doing the work to right a wrong.” When I see the of the grant. But occasionally, I Me Too and Black Lives Matter Why did you become have been shocked by research- movements, I think of what my a consultant? ers one might characterize as generation encountered, and the It just happened. At the nouveau-riche, who were more generations before me, and what University of Utah, there was/ absorbed in their own power than we accomplished or failed to is no statistics department nor a in any sense of collegial norms. solve. And I am heartened by the department of biostatistics. So, courage of a new generation. It is departments knew they needed What is the most exciting always time to right a wrong. statistical scientists, though they part of consulting? didn’t call it statistical science back My favorite part of statistical con- What advice would you then. In my position in 1979, sulting is the ability to go into a give to statisticians just there were teaching, research, and new substantive area equipped beginning their careers? administrative responsibilities, with the skills to make strong It is such a different world now but “research” really meant sta- analytical contributions. In about than in 1979, when there were tistical collaboration or consulta- 2006, I met another of my favor- no cell phones and it took five tion on biomedical research, as it ite colleagues, Ingrid E. Nygaard, minutes to run a logistic regres- paid our salaries even in the ten- who was very interested in a sion on a good computer! I think ure track. I had the good fortune previously untouchable subject: the best advice I can give is to to work with the Cooperative women’s pelvic floor disorders. continue to learn. Trans-oceanic Systematic Studies of Rheumatic Earlier in my lifetime, this was collaborations have become rou- Diseases with John R. Ward, and far too private to be discussed, tine. Bayesian statistics and mod- Bill Brown was on our advisory and yet 50% of women will have ern causal methods will be very committee. John Ward turned a pelvic floor disorder by age 75. important in the lifetimes of new out to be one of my best men- We are still collaborating. statisticians. Big Data and super- tors. Thus began a collaborative computing will continue to draw research effort that lasted through What is the best piece of on our expertise. Look broadly the mid-1990s, with side trips advice you received from at what influences statistics and in cancer biostatistics, then into a mentor regarding statis- stay current with internet and health services research, mater- tics or consulting? computer advances, program- nal and child public health, and Well, this is about neither, but ming languages, databases, and now urogynecology statistics. So, rather advice for life. I received related software. I am an internal consultant, with it from Gladys Reynolds, the all the benefits and constraints of statistician who did pioneering internal consulting. research on STDs and was the first woman to run a Centers for

september 2019 amstat news 39 What do you enjoy doing in your spare time? Gregory Csikos I have had many recreational activities over the years: kayaking; Sole Practitioner (Gregory cooking Indian food; poetry; oil Csikos, CPA, CFE, GStat) painting; and gardening, to name a few. In this modern world, it is important to have your own life, and the analytic mind requires What or who inspired What is the most exciting counterbalances. Whitewater you to be a statistician/ part of consulting? kayaking was a blessing to me, as data scientist? I don’t think my work is “excit- it was physical, nonmathematical, I can’t identify anything or any- ing,” but it is enjoyable. I get a and focused me in the moment. one specific who inspired me lot of satisfaction from making As I begin to look toward retire- to become a statistician. I was order out of chaos and provid- ment, I have a new hobby—ama- working as an accountant at a ing an answer to a puzzle faced teur astronomy—and I am look- CPA firm and realized the grow- by a client. ing forward to the time when I ing importance of data analysis, will be able to allocate my statis- including descriptive and infer- What is the best piece of tical expertise to questions that ential statistics, to answering advice you received from attract me in this area. business questions. At that point, a mentor regarding statis- I decided to enter the Master of tics or consulting? Name one or two favorite Applied Statistics program at Not everyone should be your blogs or books you have Penn State. client. It is normal and healthy read and would recom- to not always match with every mend to others. Why did you become a potential client, whether that’s Statistical or nonstatistical? I think consultant? regarding the client’s professional Peter Block’s Flawless Consulting I started my solo practice in needs, financial resources, or is a classic, as is James Boen and 2016, after eight years of work- working style. Douglas Zahn’s The Human Side ing at public accounting firms. I of Statistical Consulting, though provide tax, financial accounting, What advice would you I very much like the content and forensic accounting services. give to statisticians just of Javier Cabrera and Andrew My statistical experience is most beginning their careers? McDougall’s Statistical Consulting. relevant to the forensic account- Know what you’re worth. I’ve I read the Women’s Review of Books ing aspect of my practice. noticed statisticians undervalue attentively, as it tells me what I decided to start my own their skills and seem ignorant of young women are thinking. n practice because self-employ- what other highly trained profes- ment gives me autonomy and sionals are paid. long-term security. What do you enjoy doing Name a few specific in your spare time? skills you need to be a Exploring my adopted city, consultant. Philadelphia, and listening to Being able to translate what a Mary Beard lectures on YouTube. client wants into services you Name one or two favorite can provide that will result in blogs or books you have the appropriate answers, knowing how to find and retain cli- read and would recom- ents, ensuring you get paid, and mend to others. dealing with the administrative Of what I’ve read most recent- aspects of being self-employed. ly, I’d recommend 12 Rules for Life: An Antidote to Chaos by Jordan Peterson. n

40 amstat news september 2019 Charles Kincaid

Product Owner / Principal Data Scientist, Experis Solutions

What or who inspired you to be a statistician/data What is the most exciting What advice would you scientist? part of consulting? give to statisticians just The undergraduate statistics The feeling of accomplishment beginning their careers? courses taught at Kansas State and excitement from helping If you want to be a statistician, University were standard courses, people solve problems is defi- do statistics. Find opportunities but they excited me with what nitely exciting. We make people’s to do, or help with, a wide variety could be done. When I started tak- jobs easier, help them create bet- of consulting projects. Learn the ing graduate classes, the instructors ter products, and even make the skills that cut across disciplines. If made it so alive with possibilities world a better place. you want to specialize, the broad that I never looked back. skills and experience will help What is the best piece of you more than specializing from Why did you become a advice you received from the beginning. consultant? a mentor regarding statis- I’ve always been a consultant. tics or consulting? What do you enjoy doing Kansas State had a consulting “Clients always lie to you.” This in your spare time? center in which students helped advice from a professor of com- Over the years, I’ve had a lot of faculty and grad students from puter science at Kansas State has hobbies and interests. In this sea- other departments. It fits well been helpful my entire career. son of my life, I enjoy playing into my desire to help others. They don’t do it intentionally. It’s board games, reading theology just that people come to us with books, doing tough mudders, Name a few specific problems and questions they’ve playing softball, dancing, watch- skills you need to be a filtered through their under- ing movies, and playing with my consultant. standing and experience, which grand kids. n The ability to listen to others and often gives an inadequate repre- dig into the questions behind sentation of what they need. their questions is very important, as well as the ability to build up trust with your client so they become more of a collaborator. The ability to learn other disciplines and grasp the technical aspects quickly is an immense help for doing more than just order taking. And, of course, the many other skills necessary. :-)

september 2019 amstat news 41 LeAnna Stork

Soy Crop Modeling Lead, The Climate Corporation - Bayer Crop Science

What or who inspired you What is the most exciting What do you enjoy doing to be a statistician/data part of consulting? in your spare time? scientist? Meeting new people and learn- Chasing after my kids, traveling My dad. I was good at math in ing new spaces. Thinking outside to new places, and tasting wine! high school but was unsure at the the box has really accelerated my time of what types of careers might career development. Name one or two favorite be possible. He had a friend who blogs or books you have was an actuary (I had never heard What is the best piece of read and would recom- of an actuary at the time!) and sug- advice you received from mend to others. gested I take some statistics classes a mentor regarding statis- The Oz Principle: Getting in college. I loved the stats classes tics or consulting? Results Through Individual and so much that I ended up major- Put yourself in the researcher’s Organizational Accountability by ing in statistics and pursuing this shoes; they may have spent years Roger Connors, Tom Smith, and as my career path. planning and conducting the Craig Hickman. The concept in study you are consulting on. this book is simple: Individuals Why did you become a Think creatively of the “best” sta- are accountable for results. The consultant? tistical solution that is practical authors highlight the “See It. Because I like learning from people and achievable for the researcher. Own It. Solve It. Do It.” concept. about diverse business problems. Sometimes the best solution is I believe most people are good at For me, the most rewarding part the simple one. the first three steps: See It. Own of being a consultant is listening It. Solve It. But to cross the fin- to customers’ problems and asking What advice would you ish line and truly deliver impact- them thought-provoking ques- give to statisticians just ful results, those who master “Do tions about those problems they beginning their careers? It” will achieve greater success. may not have even considered. Walk the hallways and meet with Simply seeing the problem, iden- your collaborators; don’t hide out tifying flaws, and blaming oth- Name a few specific in your cubicle! Visit their labora- ers for failures will not deliver skills you need to be a tories, fields, factories, etc. to see results. People who take account- consultant. their processes and the data they ability for actions, identify solu- The ability to ask your client are collecting. Ask questions about tions, and implement them really good questions—not just the process so you can truly under- will achieve great results. I hold about their data, but also about stand the variability associated myself and my team accountable the “why.” What is the purpose with the data you are analyzing. to this standard. n and objective of their study? What questions will the results of the study answer? How will the results of the analysis be used? Who/what will be affected by the results? What are important assumptions in the collected data?

42 amstat news september 2019 Todd Coffey

Chair and Associate Professor, Department of Research and What or who inspired Biostatistics, Idaho College of you to be a statistician/ Osteopathic Medicine data scientist? I took a course in statistics as a sophomore in college. In one of the first lectures, the professor described how statistical meth- Name a few specific skills What is the best piece of ods could be used to design and you need to be an effec- advice you received from analyze a clinical trial to deter- tive consultant. a mentor regarding statis- mine if a novel therapy was bet- Effective consultants I have tics or consulting? ter than placebo. The recogni- worked with have a strong and On a recruiting trip to a gradu- tion that statistics held the key to broad methodological back- ate school I ultimately attended, understanding how to determine ground and a solid understand- I was taken to see a graduate of which experimental treatments ing of statistical theory. The best the program who was working were better, worse, or no different consultants add to that back- locally as a statistical consultant. fascinated me. I quickly became ground with active listening, After some small talk, he asked hooked on becoming a statisti- good questions, excellent writ- me what I thought was the most cian and never looked back. ten and verbal communication, important question a statistical creative thinking, appropriate consultant needed to ask. Before Why did you become a lack of fear toward trying some- I could answer, he responded that consultant? thing new, analytical thinking to consultants must understand com- Two reasons: I was fascinated understand advantages and dis- pletely the question the collabora- with the research being done advantages of various approach- tor is asking. It was an impression- in other fields and I loved the es, and social engagement with able moment, and I learned clearly intellectual challenges consult- the collaborator. that all the fancy statistics I would ing requires. These challenges learn in graduate school wouldn’t include understanding the What is the most exciting help the collaborator if I answered research question in the collabo- part of consulting? the wrong question. rator’s language, translating it to When a reluctant collaborator a statistical framework, develop- recognizes the power of statistical What advice would you ing a broad methodological back- thinking and transforms into a give to statisticians just ground, improving programming champion for implementing the beginning their careers? skills to implement solutions, methods proposed. This process Recognize your influence as a stat- retranslating the results into is the most satisfying when you istician will likely be proportional the language of the applicable can convince the collaborator to to the professional closeness of the field, and communicating how invest resources into an experi- relationships you develop with the research question has been mental design or study plan he your collaborators. Becoming a appropriately answered. The or she has never considered and full partner with those you work challenges vary across disciplines, then demonstrate results through with is critical. You may have an so a consultant can move from the analysis that could not have optimal technical solution to a one to another and continue to been achieved without statistical research question, but if the col- find opportunities. methodology. When this hap- laborator does not trust you, you’ll pens, the surprise and excitement never get the chance to show the on the face of the astonished col- power of statistics. n laborator is priceless.

september 2019 amstat news 43 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. Listings are shown alphabetically by state, followed by international listings. Vacancy listings may include the institutional name and address or be identified by number, Louisiana as desired. n Louisiana State University located Professional Opportunities vacancies also will be published on the ASA’s website in Baton Rouge, Louisiana, seeks (www.amstat.org). Vacancy listings will appear on the website for the entire calendar candidates for Department Head month. Ads may not be placed for publication in the magazine only; all ads will be of the Department of Experimental published both electronically and in print. Statistics (EXST). The successful These listings and additional information about the 65-word ads can be found at candidate will be a leader with a ww2.amstat.org/ads. shared vision to guide, support, and inspire a multidisciplinary faculty Employers are expected to acknowledge all responses resulting from publication of and staff in their tripartite mission their ads. Personnel advertising is accepted with the understanding that the adver- of scholarly teaching, research, and tiser does not discriminate among applicants on the basis of race, sex, religion, age, outreach. For more information visit color, national origin, handicap, or sexual orientation. the LSU Career site: https://lsu.wd1. Also, look for job ads on the ASA website at https://jobs.amstat.org/jobseekers. myworkdayjobs.com/LSU/job/LSU--- Baton-Rouge/Department-Head-Chair- --Tenured_R00036318. EOE.

New York n Applications are invited for a tenure- track position in statistics at Vassar College to begin fall 2020: see the full ad at MathJobs.org. The ideal candidate is committed to excellence in scholarship and undergraduate teaching, and working with the Department's two statisticians to further expand statistics and data science at Vassar. Application review will begin October 7, 2019, and continue until the position is filled. Vassar College is an affirmative action and equal opportunity employer with a strong commitment to increasing the diversity of the campus community and the curriculum, and promoting an environment of equality, inclusion, and HAVE YOU respect for difference. Candidates who can contribute to this goal through their teaching, research, advising, and other activities are encouraged to identify their strengths and experiences in this area. MOVED? Individuals from groups whose under- representation in the American profes- L S soriate has been severe and longstanding are particularly encouraged to apply. https://goo.g/h. Pennsylvania n The Biostatistics Core is seeking a full-time master’s-level biostatistician. The Biostatistics Core collaborates with researchers across Geisinger. The Core assists investigators throughout

44 amstat news september 2019

Head, Department of Statistics TEXAS A&M UNIVERSITY Texas A&M University, College Station, Texas Statistics A national search is underway to identify outstanding candidates for Head of the Department and holder of an endowed chair of Statistics at Texas A&M University, the fourth largest university in the nation with an enrollment of over 50,000 students. The position will be hired at the rank of Full Professor. The Department has a strong tradition of theoretical and interdisciplinary research. Faculty members have extensive research collaborations with faculty in all ten colleges on campus. The Department has 70 graduate students on campus with an additional 300 distance-learning M.S. graduate students, 150 undergraduate students, 26 research faculty members, and 13 academic professional track faculty. Further information about the Department, its faculty, and its facilities can be found at our website:www.stat.tamu.edu . The Department seeks an individual with a dynamic and internationally recognized research program, a sincere commitment to undergraduate and graduate teaching, and proven leadership skills to be the Head. The Head reports to the Dean of the College of Science and is responsible for the Department’s administrative, budgetary, and personnel matters. Candidates for Head must demonstrate a vision for supporting, directing, and enhancing the goals of the Department. The position requires a Ph.D. in Statistics or a related field and a record of scholarship and teaching consistent with the level of professor at a major research university. Any questions regarding this position may be directed to the Chair of the Search Committee, Dr. Bani Mallick at headsearch@ stat.tamu.edu. Interested applicants should apply at academicjobsonline.org/ajo/jobs/13998. Applications will continue to be accepted until the positions are filled. The Search Advisory Committee will review applications as they are received, and the review will continue until the position is filled. Texas A&M University is committed to enriching the learning and working environment for all visitors, students, faculty, and staff by promoting a culture that embraces inclusion, diversity, equity, and accountability. Diverse perspectives, talents, and identities are vital to accomplishing our mission and living our core values. The Texas A&M System is an Equal Opportunity/Affirmative Action/Veterans/Disability Employer committed to diversity. The College of Science and the Department of Statistics are dedicated to the goal of building an inclusive and culturally diverse faculty and staff who are committed to teaching and working in an environment of academic freedom and equality of opportunity.

Department of Statistics Tenure-Track Faculty Positions

Th e Department of Statistics at North Carolina State University seeks to hire multiple tenure-track faculty. All ranks will be considered. Th e start date is August 2020.

Applicants with interests and expertise in theoretical or methodological research in any area of statistics or biostatistics will be considered. Candidates with interests in data science, machine learning, and modern methods of data analysis more generally are encouraged to apply. Th e ability and desire to supervise graduate student research and to pursue excellence in teaching are essential.

To apply, please visit: https://jobs.ncsu.edu/postings/120549

Th e Department provides a dynamic environment for teaching, research and collaborations across disciplines. Inclusiveness and diversity are academic imperatives and are university goals: You will be expected to foster an environment that is supportive and welcoming of all groups. We are interested in candidates who have experience working with students from diverse backgrounds and have a demonstrated commitment to improving access to higher education for students from underrepresented groups.

Th e Department’s location in the Research Triangle provides rich opportunities for interactions with industry; other universities, including Duke Uni- versity and the University of North Carolina at Chapel Hill; and government agencies. Faculty enjoy collaborations with medical researchers at Duke, environmental scientists at the EPA research facility, pharmaceutical researchers at Glaxo-SmithKline, and software developers at SAS Institute, among many others. Th e Department is also a founding cooperator of the NSF-funded Statistical and Applied Mathematical Sciences Institute (SAMSI), located nearby in Research Triangle Park.

All applicants must have a Ph.D. in Statistics or Biostatistics or a related fi eld by the time of employment. Review of applications will begin soon, and continue until the positions are fi lled. Questions about the search may be directed to the Search Committee Chair ([email protected]).

NC State University is an equal opportunity and affi rmative action employer. Women and members of other underrepresented groups are encouraged to apply. In addition, NC State University welcomes all persons without regard to sexual orientation or genetic information.

september 2019 amstat news 45 all stages of research from proposal drafting to analysis and interpretation. Core members assist with research question and hypothesis specification, study design and variable selection, sample size and power estimation, analysis plan, data collection design, randomiza- tion, data management, and biostatisti- cal analysis. Apply: bjhicks@geisinger. edu. EOE.

n The Department of Epidemiology and Health Services Research and Geisinger's Biostatistics Core is recruiting a full-time PhD biostatistician. This position is eligible for a faculty appoint- ment depending on qualifications and career interests. The department currently consists of six faculty members and the core has nine staff. The department and the Biostatistics Core has strong interest to expand capabilities in multi-level modeling and spatial statistics. Contact us: [email protected]. EOE. n

Tenure-Track/Tenured Faculty Position

The Department of Statistics (www.stat.uci.edu) in the Donald Bren School of Information and Computer Sciences (ICS) at the University of California, Irvine (UCI) invites applications for a tenure-track or tenured faculty position at the assistant or open-rank level beginning July 1, 2020.

The Department of Statistics has a strong interdisciplinary flavor, focused on developing state-of-the-art methods for solving topical data-driven problems in science and engineering, and advancing the statistical theory that underlies those methods. We are searching for faculty with strong research potential, a commitment to excellence in teaching, and enthusiasm for helping our collegial department continue to grow. Applicants should hold a Ph.D. degree (or expected by fall 2020) in statistics, biostatistics or a related field. Candidates with research interests in all areas of statistics will be considered.

The University of California, Irvine is part of the premier public university system in the world. UCI is a member of the Association of American Universities (AAU), is ranked as a top ten public university by U.S. News and World Report, and was identified by the New York Times as No. 1 among U.S. universities that do the most for low-income students. UCI is located in Orange County, 4 miles from the Pacific Ocean, 45 miles south of Los Angeles, and 80 miles north of San Diego. Irvine is one of the safest communities in the U.S. and offers a very pleasant year-round climate, numerous recreational and cultural opportunities, and one of the highest- ranked public school systems in the nation. All positions would be eligible to participate in UCI’s faculty housing program.

Completed applications containing a cover letter, curriculum vita, graduate transcripts (for assistant professor candidates), statements on diversity, teaching, and research, three letters of recommendation, and sample research publications should be uploaded electronically. Please refer to the following web site for instructions https://recruit.ap.uci.edu/JPF05335. Applications received by November 1, 2019 will receive the fullest consideration.

The University of California, Irvine is an Equal Opportunity/Affirmative Action Employer advancing inclusive excellence. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, disability, age, protected veteran status, or other protected categories covered by the UC nondiscrimination policy. A recipient of an NSF ADVANCE award for gender equity, UCI is responsive to the needs of dual career couples, supports work-life balance through an array of family-friendly policies, and is dedicated to broadening participation in higher education.

46 amstat news september 2019 Possibilities and Probabilities AMSTATNEWS If working in an environment that values ADVERTISING DIRECTORY individuality and diversity and allows you to Listed below are our display advertisements innovate, engage in problem solving, and achieve only. If you are looking for job-placement ads, please see the professional opportunities sec- your professional goals appeals to you, then the tion. For more job listings or more information U.S. Census Bureau is the place for you. about advertising, please visit www.amstat.org.

misc. products and Your Work as a Mathematical Statistician services at the Census Bureau Taylor & Francis...... Cover 2

• Design sample surveys and analyze the data collected. professional opportunities North Carolina State University...... p. 45 • Design and analyze experiments to improve survey Texas A&M University...... p. 45 questionnaires and interview procedures. University of California, Irvine...... p. 46 US Census Bureau...... p. 47 • Improve statistical methods for modeling and∑ adjustment of seasonal time series. Westat...... p. 46

• Perform research on statistical methodology that will software improve the quality and value of the data collected. JMP Software from SAS...... Cover 4 StataCorp...... Cover 3 • Publish research papers and technical documentation of your work.

Requirements

• U.S. citizenship

• Bachelor’s, Master’s, or Ph.D with at least 24 semester hours in math and statistics (see Web site for more specifics on required coursework)

Apply at www.census.gov, click on Census Careers, Type of Position, Professional/Scientific/Technical, Math Statistician

The U.S. Census Bureau is an Equal Opportunity Employer.

september 2019 amstat news 47 SOCIAL CHATTER What is a question a good statistical consultant should FOLLOW US ask a potential client? community. amstat.org Milesius • @Milesius9 "Do you want it right or do you want it right now?" www.facebook. com/AmstatNews Jessica Lavery • @jessicalavs What's the sentence you want to be able @AmstatNews to say when the study is over? www.instagram • @randomjohn .com/AmstatNews JonJohnson What decision do you plan to make, or how will you make it? or Why didn't you involve me earlier? We could have avoided this!"

Read more of these responses on Elmen Damián D'Achiardi Barreto • @elmendachiardi Twitter at "If you had all the money in the world, what would you http://bit.ly/2yHO3H2 or by following us on want to know? Now, how much money do you have? Facebook. This phrase was said by my sampling professor, Leonardo Bautista.

Emily Griffith • @EmilyHGriffith1 I always start with, "Could you tell me a little about your proj- For next ect? How'd you get interested in this?" Let clients start by shar- month, we ing their own expertise. ask our

followers— Leslie McClure • @StatGirlLAM "What is your Yes! It's so important to establish rapport favorite pastime as a statistician?" and to establish trust! Make sure to tag Zach Weller • @wellerstats @AmstatNews One I almost always start with: "What is the question you are trying to answer?" in your response. Jane Coe Johnson • @jcjohnsonmo And I end up asking it over and over and ...

48 amstat news september 2019 Announcing

New in Stata 16 f Meta-analysis f Multiple-dose pharmacokinetic modeling f Lasso-based machine learning f Import data from SAS and SPSS f Python integration f Random-effects models with any combination of MNAR outcomes, f Bayesian predictions nonrandom treatment assignment, and unobserved confounding f Multiple chains in Bayesian estimation f Numerical integration f Truly reproducible reporting f Linear programming f Multiple-group IRT f Nonparametric series regression f Sample-size analysis for inference with confidence intervals stata.com/amstat-stata16 Stata is a registered trademark of StataCorp LLC, 4905 Lakeway Drive, College Station, TX 77845, USA. Jim Da iley Global Technical Key Account Manager, BASF Care Chemicals

Great software in the right hands can change the world.

At BASF, chemists created a breakthrough dishwashing formulation with Trilon® M chelate that is free of phosphates and delivers superior performance. Now clean is also green.

Read about BASF’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. 155242.0817US

155242-JMP-BASF-ad-AMSTAT.indd 1 8/4/17 10:37 AM