June 2010 • Issue #396 AmstatNews The Membership Magazine of the American Statistical Association • http://magazine.amstat.org
Board Approves Accreditation Guidelines ALSO: ASA Releases Statement on Risk-Limiting Publications Agreement No. 41544521 Audits ASA Endorses Recommendations to Improve Forensic Methods
JUNE 2010 • Issue #396
Executive Director Ron Wasserstein: [email protected]
Associate Executive Director and Director of Operations FEATURE S Stephen Porzio: [email protected]
Director of Education 3 President’s Corner Martha Aliaga: [email protected] 5 Highlights of the April 2010 ASA Board of Director of Science Policy Directors Meeting Steve Pierson: [email protected] 7 ASA Releases Statement on Risk-Limiting Audits Managing Editor 7 FREE Writing Workshop at JSM Megan Murphy: [email protected] 8 ASA Endorses Recommendations to Improve Production Coordinators/Graphic Designers Forensic Methods Melissa Muko: [email protected] Kathryn Wright: [email protected] 8 sPAIG to Provide List of Advisers Publications Coordinator 9 ASA President Appointed to NSF Directorship Val Nirala: [email protected]
Advertising Manager Claudine Donovan: [email protected]
Contributing Staff Members Amy Farris • Eric Sampson • Keith Crank Board Approves Amstat News welcomes news items and letters from readers on matters Accreditation Guidelines of interest to the association and the profession. Address correspondence p. 10 to Managing Editor, Amstat News, American Statistical Association, 732 North Washington Street, Alexandria VA 22314-1943 USA, or email [email protected]. Items must be received by the first day of the preced- ing month to ensure appearance in the next issue (for example, June 1 for the July issue). Material can be sent as a Microsoft Word document, PDF, or within an email. Articles will be edited for space. Accompanying art- work will be accepted in graphics file formats only (.jpg, etc.), minimum 300 dpi. No material in WordPerfect will be accepted. Amstat News (ISSN 0163-9617) is published monthly by the American 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 North Washington Street, Alexandria VA 22314-1943 9 Nominations Sought for JSM USA. Send Canadian address changes to Station A, P.O. Box 54, Windsor Mentoring Program ON N9A 6J5; [email protected]. Annual subscriptions are $50 per year for nonmembers. Amstat News is the member publication of the 10 Board Approves Accreditation Guidelines ASA. For annual membership rates, see www.amstat.org/join or contact ASA Member Services at (888) 231-3473. 12 ASA Offers Free Article Downloads American Statistical Association 732 North Washington Street 13 statistics Without Borders to Meet at JSM 2010 Alexandria, VA 22314–1943 USA (703) 684–1221 • FAX: (703) 684-2036 13 Editor Sought for CHANCE Magazine
ASA GENERAL: [email protected] 15 Congress Views Results of NSF-Funded ADDRESS CHANGES: [email protected] Research Projects AMSTAT EDITORIAL: [email protected] 17 CoC Sponsors Sessions Featuring Scientific ADVERTISING: [email protected] Applications of Statistics Web Site: www.amstat.org Printed in USA © 2010 17 Members Elected to National Academy American Statistical Association of Sciences
VISION STATEMENT 18 Journal of Nonparametric Statistics Highlights To be a world leader in promoting statistical practice, applications, 19 Statistical Analysis and Data Mining Highlights and research; publishing statistical journals; improving statistical education; and advancing the statistics profession 20 Significance Highlights
MISSION STATEMENT 21 The American Statistician Highlights Support excellence in statistical practice, research, journals, and 22 UMBC Statistics Program Celebrates meetings. Work for the improvement of statistical education 25th Anniversary at all levels. Promote the proper application of statistics. Anticipate and meet the needs of our members. Use our discipline to enhance human welfare. Seek opportunities to advance the statistics profession. Column Contributors Funding Opportunities Who Wants to Be a Biostatistician (or Environmental Statistician, or Social Science Statistician, or …)? p. 31
This column highlights research activities that may be of interest to ASA members. These brief articles include information about new research solicitations and the federal budget for statistics. Comments or suggestions for future articles may be sent to ASA Research and Graduate Education Manager Keith Crank at [email protected]. DEPARTMENTS Contributing Editor 25 Statistician’s View Keith Crank has a BS in mathematics education and an MS in mathematics from Michigan State Feedback: Odds Are, It’s Wrong University and a PhD in statistics from Purdue statisticians Comment on Status of Climate University. Prior to joining the ASA as research and graduate education manager, he was a program Change Science officer at the National Science Foundation, primar- 37 Meetings ily in the probability program. 2010 FDA/Industry Statistics Workshop Registration Form Crank
Science Policy News Members Affect ASA Science Policy p. 33
This column is written to inform ASA members about what the ASA is doing to promote the inclusion of statistics in policymaking and the funding of statistics research. To suggest science policy topics for the ASA to address, contact ASA Director of Science Policy Steve Pierson at [email protected].
Contributing Editor Steve Pierson earned his PhD in physics from the University of Minnesota. He spent eight years in the physics department of Worcester Polytechnic Institute and later became head of government relations at the American Physical Society.
Find Your Fit p. 35 Pierson
MEMBER NEWS Master’s Notebook 39 People News Find Your Fit p. 35 43 Award Deadlines and Information This column is written for statisticians with master’s degrees and highlights areas 45 Section News of employment that will benefit statisticians at the master’s level. Comments 55 Calendar of Events and suggestions should be sent to ASA Research and Graduate Education “ Manager Keith Crank at [email protected]. 63 Professional Opportunities Contributing Editor Jack Nyberg is a senior manager within the statistics and pharmacokinetics department of Covance, a drug development services company. He holds a master’s degree in statistics and a bachelor’s degree in econom- ics from the University of Kentucky. cover design by Melissa Muko
Nyberg PRESIDENT’S CORNER
Half-Time Report
t is the time of year when proud parents and Members of the Membership Growth Working other family members congratulate their loved Group, chaired by Jeri Mulrow, have been working ones at graduation ceremonies across the coun- hard with various existing committees to propose try.I It is also the time when graduates receive good strategies for enhancing growth. Suggestions so far advice about how to use the core, computational, include the following: and communication skills they acquired in their programs to solve major challenges related to health, • Strongly encouraging department chairs/ energy, and the environment. As President Barack heads to get students in their departments to Obama said in one of his addresses, “The United participate in ASA activities States is still a land of infinite possibilities waiting to Sastry Pantula be seized, if you are willing to seize them.” • Supporting the creation of a mentoring Even in the current economic climate, it is good program to see that statisticians have options in all three • Providing opportunities for communicat- sectors—academia, industry, and government. For ing and networking among students and those who are looking for jobs, JSM 2010 will pro- recent graduates vide many opportunities, including a session titled “Recruitment for the Federal Sector.” There also will • Making use of peers and others to encourage be many mentoring opportunities for our younger lapsed members and ASA Fellows to renew statisticians. And I am excited about the President’s their memberships Invited Speaker, SAS CEO Jim Goodnight, who • Strongly encouraging higher-level industry will give a talk titled “The Forecast for Predictive statisticians to support ASA membership and Analytics: Hot and Getting Hotter.” Analytics is an involvement in ASA activities area in which statisticians have a future. All this talk about talks reminds me I have to • Continuing to actively engage the heads of get ready for mine! I can’t believe how the time is federal statistical agencies by regularly meeting flying by and that I am already halfway through my with them and encouraging them to support term as ASA president. Certainly, exciting things are and promote ASA membership and participa- happening within our association—our relationship tion in ASA activities with RSS through Significance, accreditation imple- mentation, social media, electronic publications, • Promoting diversity within the ASA and a new winter conference on statistical practice, • Attracting new members from quantitative just to name a few. Amstat News issues are full of areas such as mathematics, computer science, goodies; there is something of interest to all our and the physical and social sciences who have members and the public. Also, there has been good a strong interest in statistics progress made on the initiatives related to Growth, Impact, Visibility, and Education.
JUNE 2010 Amstat News 3 Ideas aimed at new statisticians include to inform decisionmaking in public policy and sci- the following: ence policy.” The Visibility and Impact in Science Policy Working Group, chaired by Past President • Creating a group for new statisticians Sally Morton, is focusing on standardizing the pro- in research and a group for new applied cess for identifying emerging issues and providing statisticians a timely response to public and science policy mat- ters in collaboration with ASA Director of Science • Promoting forums and publications aimed at Policy Steve Pierson and other statistical associations new statisticians (see Science Policy News on Page 33). The group • Dedicating a web site to new statisti- also is working to identify key issues—climate cians that includes information about change, STEM education, election audits—and to employment opportunities, awards, and issue position statements on them. Informational mentoring activities outreach to the public and stakeholders is another important topic being discussed by the group. • Providing more career assistance Jessica Utts is leading the Education Working Group, which is organizing a panel at JSM 2010 I am thrilled with the energy of this working titled “Statistics Degree Programs in a Data-Centric group and their follow-up plans. On a related topic, World: What Needs to Change?” In addition, there a couple of member-initiated proposals supported will be three P.M. roundtables on preparing statisti- by the board are titled “Connecting the ASA to cians for the needs of industry. Young Statisticians Through Outreach to High- A subgroup of this working group is developing School Statistics and Mathematics Teachers: A First a process to update the undergraduate statistics cur- Step” and “JSM Conference Mentoring Program.” riculum guidelines approved by the board in 2000 The Public Awareness and Impact Working (see www.amstat.org/education/curriculumguidelines. Group is chaired by ASA Executive Director Ron cfm). Other subgroups are looking at professional Wasserstein. In April, board members discussed master’s degree programs and a plan to gather infor- ideas for an ASA tagline and two- or three-sentence mation from potential employers regarding educa- “elevator pitch” that describes what the ASA does tional expectations for PhD, MS, and BS graduates. or stands for. As you may know, the American On a related topic, a member-initiated proposal, Mathematical Society’s tagline is “Maintaining titled “Methodology for Measuring the Quality Excellence in Mathematical Sciences Research” and of Graduate Programs: A Workshop Focusing on the National Institute of Statistical Science’s tagline Programs in the Statistical Sciences,” received sup- is “The Statistics Community Serving the Nation.” port by the board recently. ASA Public Relations Specialist Rosanne I am absolutely thrilled with the energy of our Desmone also has been working to increase pub- member volunteers. The ASA thrives on your lic awareness of the ASA by regularly sending out dedication to the profession and our association. alerts to various news outlets and helping gather Constructive suggestions are always welcome. Thank Statisticians in the News articles for the ASA’s you for all you do for the ASA. I look forward to the web page. remaining half of this year. Another way the ASA has raised public aware- ness of its activities is by putting Amstat News online and making it accessible to the general public. Meanwhile, members of the working group con- tinue to develop other innovative ideas. One of the initiatives of the ASA’s strategic plan is to “promote the need for sound statistical practice
4 Amstat News JUNE 2010 Highlights of the April 2010 ASA Board of Directors Meeting
SA President Sastry Pantula led the board through a full agenda of policy matters Board of Directors during its April 16–17 meeting at the ASA Sastry Pantula, president officeA in Alexandria, Virginia. Following are the highlights: Nancy Geller, president-elect
• The board endorsed an ASA statement Sally Morton, past president on risk-limiting post-election audits and Strengthening Forensic Science in the United Nat Schenker, third-year vice president States: A Path Forward, a report by the Christy Chuang-Stein, second-year vice president National Academies (see Page 7). Rod Little, first-year vice president • Iain Johnstone presented the report of the Accreditation Committee. The board David Banks, third-year Council of Sections representative approved the detailed plan for moving for- ward with the accreditation program (see Jeri Mulrow, second-year Council of Sections representative Page 10 and www.amstat.org/accreditation). Jessica Utts, first-year Council of Sections representative • The board heard reports from the four stra- tegic plan working groups currently at work. Susan Hilsenbeck, third-year Council of Chapters representative In addition, President-elect Nancy Geller David Marker, second-year Council of Chapters representative presented her ideas for strategic initiatives for next year. Tom Short, first-year Council of Chapters representative • The board held a creative session with the Geert Verbeke, international representative Public Awareness Workgroup to discuss development of a tagline for the ASA and a Karen Kafadar, publications representative brief description (an “elevator pitch”) for the association. The session was an important Keith Ord, treasurer part of the working group’s efforts to develop Ron Wasserstein, executive director a comprehensive public awareness plan for the association. • The board heard proposals for developing a statistics portal, a “one-stop shop” for a vari- ety of statistical content. Discussions about • The board approved a revised development the concept continue. policy for the ASA. The policy governs aspects of the ASA’s fundraising activities. • As always, the board heard a report from ASA Treasurer Keith Ord. He noted that invest- • Michelle Dunn, a National Cancer Institute ments are up, but still below their 2007 val- program director, discussed the importance ues; that membership and subscription reve- of the statistics community’s involvement nues held steady because of rate increases, but in the Biostatistical Methods and Research numbers were down; and that the 2009 audit Design (BMRD) study section of the was concluded successfully. He emphasized National Institutes for Health and ideas for the importance of paying close attention to increasing the number of grant applications revenue sources. to the section.
JUNE 2010 Amstat News 5 • Vice President Christy Chuang-Stein, chair • The LSC recommended a change to the of the Membership Council, and Holly bylaws (see Bylaw Change), requiring, with Shulman, council vice chair, presented the certain exceptions, that all members of ASA first report of the newly formed council. committees who are appointed by the ASA Chuang-Stein provided an impressive list of be full members of the association during the the major accomplishments of each commit- term of their appointments. tee in 2009, as well as an assessment of the progress of the committees to date. She also • The Advisory Committee on Teacher Enhancement (ACTE) was dissolved at its noted some issues for the board to consider own request. This request was reviewed by the for some of these committees. LSC before being considered by the board. • Geller, also chair of the Leadership Support The ACTE felt its functions were being ful- Council (LSC), presented the council’s first filled by the new Education Council. report. The LSC, the cornerstone of the • Karen Kafadar, publications representative, committee restructure approved by the ASA presented the annual report of the Committee Board of Directors last year, is fully under on Publications (COP). The COP has been way. The early results of this reorganization active and engaged under the leadership of are positive. The primary reasons for creat- its chair, David Scott, handling a variety of ing the new structure were to improve the matters and providing valuable advice and function of committees and the communica- guidance to the board. tion between committees and the board. The whole organization is more effective when its • One specific recommendation of the COP committees are operating in concert with the was that editors of the ASA’s wholly owned board to implement the strategic plan. The journals be required to be members of the time and effort of volunteers, the lifeblood of ASA during their terms of service. The the association, is better spent when we oper- COP believes these editors serve an impor- ate effectively and efficiently. tant role and occupy a visible place within the association. The board agreed that this requirement should apply to all future edito- rial appointments, those beginning in 2011 and beyond. Bylaw Change • The board reviewed proposals for the Member Initiative program and selected some of them The Leadership Support Council and for funding. ASA Board of Directors recommend the following change to the ASA bylaws, Article • David Judkins, JSM 2011 program chair, pre- IX (Committees), Part 2 (Membership): sented the results of the 2009 JSM Presenter Satisfaction Survey. The survey, sent to JSM From: “All members of Standing presenters, addressed the subject of satisfac- Committees, as well as the chairs of tion with the presentation experience, but Continuing and Ad hoc Committees, shall had tailored questions for the various session be full members of the Association.” types and presenter roles. This survey was in addition to the standard satisfaction survey To: “All members of ASA committees who sent out to registered JSM attendees. are appointed by the ASA shall be full members of the association during the term • The board, as it does at each meeting, heard of their appointment. Exceptions for certain about issues and concerns related to chap- ters and sections from the respective council committees or committee positions can be governing boards. granted by the Leadership Support Council when appropriate for the function of The board next meets on June 18 in Alexandria the committee.” for its annual budget meeting, then again in late July at JSM in Vancouver. n
6 Amstat News JUNE 2010 ASA Releases Statement on Risk-Limiting Audits
he ASA Board of batches (a batch is a group of bal- Directors recently released lots for which the machine total Previous Election a statement recommend- has been separately tallied). The Auditing Statements Ting that routine risk-limiting total number of ballots examined March 2008, www.amstat.org/about/ audits be conducted and reported to confirm a correct outcome gen- pressreleases/asacallsforaudits.pdf in all federal elections, as well as erally increases with 1) a smaller in statewide and other govern- margin of victory, 2) larger-sized August 2008, www.amstat.org/about/ mental election contests. batches of ballots, and 3) more pressreleases/asaendorsespostelectionaudits.pdf The statement also urges election errors in the machine count. officials to seek statistical advice Thus, states should seek to buy about how to efficiently attain electoral machinery that makes the desired accuracy. small-batch reporting easy and “Risk-limiting audits are minimizes errors in recording board may be viewed at designed to limit the risk of cer- voter intent. The audits, them- www.amstat.org/outreach/pdfs/ tifying incorrect winners,” said selves, can help identify the kinds Risk-Limiting_Endorsement.pdf. Sastry Pantula, ASA president. of machines and ballot designs Additional information about “A well-designed audit often that lead to the fewest errors. election audits can be found can confirm a correct electoral The complete two-page at www.amstat.org/outreach/ outcome after examining only statement approved by the electionauditingresources.cfm. n a small fraction of the ballots cast. If, however, an outcome is unclear or incorrect, a risk-limit- ing audit protects against prema- turely accepting it. We need laws that call for such audits—laws with teeth.” “Most current laws mandating FREE Writing Workshop at JSM post-election audits call for look- ing at a fixed percentage of bal- lots,” continued Pantula. “They free writing workshop for junior may select many more ballots researchers will take place dur- than are needed to do the job or, ing JSM on August 1 and 4. The workshopA will provide instruction in sometimes, not enough to pro- vide a satisfactory check on the how to write journal articles and outcome. Good statistical design grant proposals. can efficiently and economically As part of their application, participants guarantee a large chance of find- will be required to submit a recent sample of their writing, which will be reviewed by ing and correcting erroneous session on August 1 and lunch on August electoral outcomes. Statisticians a senior mentor. Applications are due by June 1, and 4. Funding is anticipated for partial can help design risk-limiting travel support. audits that make better use of successful applicants will be notified by June 30. Applications received after June Attendance will be limited and depend election officials’ time and tax- on the number of mentors available. payers’ money.” 1 will be considered if space is available. There is no fee for participation; however, Visit www.amstat.org/meetings/wwjr/ In a risk-limiting audit, bal- index.cfm?fuseaction=main for details. n lots are sampled for review in participants must agree to attend both the
JUNE 2010 Amstat News 7 ASA Endorses Recommendations to Improve Forensic Methods Rosanne Desmone, ASA Public Relations Specialist
he ASA Board of Directors recently voted to uncertainty, designing experiments for testing new endorse recommendations made in protocols or methodologies, and analyzing data Strengthening Forensic Science in the United from such experiments.” TStates: A Path Forward, a 2009 National Academies’ One of the recommendations of the report is to report. In its statement, the board recognized the establish an institute of forensic science, and the ASA pivotal role of forensic science in the U.S. judicial board noted that sound statistical practices are essen- system and cited the value of statistical methods tial for the proposed institute to achieve its mission. and research to improve forensic methods. Examples of such practices include the following: The National Academies’ report identified sev- eral deficiencies in the nation’s forensic science sys- • The assessment of current and newly devel- tem and called for major reforms and new research. oped forensic practices using properly designed It came after years of critiques of specific forensic experiments and data analytic methods science practices and calls for reform, but broke new ground by offering a comprehensive review of the • The use of statistical methods based on forensic science system. established principles and procedures for “Statisticians have played an important role in the analysis of data, including estimated the constructive criticism of current forensic sci- error rates ence practices,” said Sastry Pantula, ASA presi- • The review, in mainstream scientific journals, dent. “We can continue to play an important role of novel methods (beyond variants of estab- in the reforms urged by the National Academies. lished methods) developed for the analysis Statisticians can make vital contributions toward of data; reviewers should include statistically establishing measurement protocols, quantifying qualified experts • The employment of modern statistical qual- ity control and quality assurance procedures to ensure that measurements, procedures, and testimony are of high quality SPAIG to Provide • The application of proficiency tests that employ accepted statistical designs and, whenever possible, are double blind to avoid List of Advisers potential biases Statistics Partnerships among Academe, Industry, • The availability of all expert reports to inter- and Government (SPAIG) is updating its list of advi- ested parties and the provision of sufficient sory boards and review teams for academic statistics supporting data and information to permit programs, to be posted at www.svsu.edu/orgs/spaig. independent review The purpose of this initiative is to provide a list of statisticians working in government or industry who The complete statement approved by the ASA have an interest in serving as advisers to academic board, including background, may be viewed at statistics programs. To be included in this list, send www.amstat.org/outreach/pdfs/Forensic_Science_ your name and contact information to Morteza Endorsement.pdf. n Marzjarani at [email protected]. n
8 Amstat News JUNE 2010 ASA President Appointed to NSF Directorship
SA President Sastry and SAA-PAMS Department to research in time series analy- Pantula was recently Head Award in 2005 and 2008. sis; exemplary service in graduate appointed Division Pantula earned his doctoral education as a teacher, researcher, DirectorA for Mathematical degree in 1982 from Iowa State mentor, and recruiter of graduate Sciences in the Mathematical University, while both his bach- students and industrial partners; and Physical Sciences Directorate elor’s and master’s degrees are and contributions to the profes- at the National Science in statistics from the Indian sion. He also received the Young Foundation (NSF). Statistical Institute in Kolkata, Statistician Award from the Pantula will join NSF in India. His areas of research International Indian Statistical September from North Carolina include time series analysis and Association that year. “I look forward to working State University (NCSU), where linear and nonlinear models. Pantula he has been a professor and direc- In addition to a number of pub- with Dr. Pantula in this impor- tor of graduate programs in the lications in various journals, tant leadership role for mathe- department of statistics since he coauthored the textbook matical sciences,” said H. Edward 1994 and department head since Applied Regression Analysis: A Seidel, acting assistant director 2002. While at NCSU, Pantula Research Tool. for mathematical sciences in has worked with a number of Pantula was elected an ASA the Mathematical and Physical companies—including Becton Fellow in 2002 for contributions Sciences Directorate at NSF. n Dickinson, GlaxoSmithKline, Eli Lilly and Company, SAS Institute, and Merck—to increase the number of graduate industrial traineeships and develop graduate fellowships. During a scholarly leave in 1990–1991, he worked Nominations Sought for at SEMATECH, where he devel- oped and taught various courses JSM Mentoring Program in quality control and experi- mental design and collaborated Nominations are being accepted for the Cavell Brownie with engineers from semicon- Scholars JSM Mentoring Program, which will be offered dur- ductor industries in the United ing JSM 2010 in Vancouver, British Columbia. This pro- States and supplier companies gram brings faculty, minority graduate students, and postdoc- in Mexico. toral scholars together in a structured mentoring program. It Under Pantula’s leadership, the provides graduate students who are preparing for academic department of statistics at NCSU careers with information, mentoring, and a peer network; sta- received the Departmental tistics faculty members learn best practices for mentoring Teaching and Learning Excellence graduate students. Award in 2004–2005. Pantula For more information, a nomination form, and selection also became a member of the criteria, visit www.amstat-online.org/2010mentoringprogram/ NCSU Academy of Outstanding CavellBrownieScholarsProgram.php. Nominations are due by June 10. Teachers and received an out- Those received after June 10 will be considered if space is available. standing teacher award, the D.D. Contact Marcia Gumpertz at [email protected] with questions. Mason Faculty Award in 2001,
JUNE 2010 Amstat News 9 Board Approves Accreditation Guidelines Committee moves carefully to test procedures
Iain Johnstone, ASA Accreditation Committee Chair, and Ron Wasserstein, ASA Executive Director
10 Amstat News JUNE 2010 he ASA Board of Directors recently approved a set of guidelines for accreditation devel- What Others Are Saying oped by the ASA Accreditation Committee. Colleagues in other countries have warmly TThe committee was formed last year, after the board welcomed ASA accreditation, expressing the endorsed a recommendation to begin a program of view that ASA participation adds value to voluntary individual accreditation of statisticians. accreditation and enhances the perception What Is Accreditation? of the statistics profession worldwide. First and foremost, accreditation is a service offered only to members of the ASA. Not all members will want to seek accreditation; however, the experi- Those who meet these guidelines will be awarded the ences of colleagues in Australia, Canada, and the designation “accredited professional statistician.” United Kingdom regarding accreditation have been encouraging. Additionally, the results of a mem- What’s Next? ber survey (see http://magazine.amstat.org/2009/09/ Because the concept of accreditation has caused accreditationsep09) indicate many members expect concern for some, the ASA will move carefully to find value in a credential that provides peer rec- and deliberately to positively develop the program. ognition for all of the following: Members of the ASA Board and ASA Accreditation Committee invite all members to read the guidelines • Having advanced statistical training for accreditation at www.amstat.org/accreditation and and knowledge provide comments, questions, and suggestions. • Having experience in applying statistical Those who are considering applying for accredi- expertise competently tation should fill out the intent-to-apply form at www.amstat.org/accreditation. This does not create • Maintaining appropriate professional any obligation, but will help the committee better development estimate the level of interest and manage the initial wave of applications. Also, committee members will • Agreeing to abide by ethical standards communicate regularly with those on the intent-to- of practice apply list to keep them abreast of developments in • Being able to communicate effectively the program. Between now and JSM, accreditation committee Accreditation is a portfolio-based—not examina- members and ASA staff will be alpha testing basic tion-based—credential that is renewable every five procedures and reviewing member comments. After years. Accreditation is also voluntary; applicants JSM, small groups from the intent-to-apply list will seek accreditation because they believe the creden- be invited to participate in beta testing systems and tial is worthwhile to them, but it is not a require- processes. Over time, as committee members learn ment for practice. and improve, they will move from testing into full- Accreditation applicants will submit materials to fledged operation. be reviewed by members of the ASA Accreditation For more information, visit www.amstat.org/ Committee, peers who will evaluate submissions accreditation. n based on the ASA’s Guidelines for Accreditation.
JUNE 2010 Amstat News 11 ASA Offers Free Article Downloads
“Who Is Teaching Introductory Statistics?” Frank P. Soler “The Business of Desire and Fear,” Rick Cleary and Sam Woolford “Response to ‘Desired and Feared—What Do We Do Now and Over the Next 50 Years’ by Xiao-Li Meng,” Elart von Collani “Rejoinder: Better Training, Deeper Thinking, and More Policing,” Xiao-Li Meng Available at http://pubs.amstat.org/toc/tas/64/1
Journal of Business & Economic Statistics or a limited time, several ASA journals will “Another Look at the Identification of Dynamic have featured articles available to download Discrete Decision Processes,” Victor Aguirregabiria for free. The following articles were chosen by theF journal editors as featured articles: Available at http://pubs.amstat.org/toc/jbes/28/2
Journal of the American Technometrics Statistical Association “Statistical Methods for Fighting Financial “Statistics: From Evidence to Policy,” ASA Past Crimes,” Agus Sudjianto et al. President Sally Morton’s address to the 2009 Joint “Fraud Detection in Telecommunications: History Statistical Meetings and Lessons Learned,” Richard A. Becker, Chris “A Moving Average Approach for Spatial Statistical Volinsky, and Allan R. Wilks Models of Stream Networks” (with discussion), Jay “Fraud Detection in Telecommunications and M. Ver Hoef and Erin E. Peterson Banking: Discussion of Becker, Volinsky, and “A Statistical Framework for Differential Privacy,” Wilks (2010) and Sudjianto et al. (2010),” David Larry Wasserman and Shuheng Zhou J. Hand Available at http://pubs.amstat.org/toc/jasa/105/489 “Statistical Challenges Facing Early Outbreak Detection in Biosurveillance,” Galit Shmueli and The American Statistician Howard Burkom “Passion-Driven Statistics,” Robert Easterling Available at http://pubs.amstat.org/toc/tech/52/1 “Desired and Feared—Quo vadis or Quid agis?” David R. Fox Journal of Computational and Graphical Statistics “Moving the Statistics Profession Forward to the “Rainbow Plots, Bagplots, and Boxplots for Next Level,” Roger W. Hoerl and Ronald D. Snee Functional Data,” Rob J. Hyndman and “Thoughts on the Importance of the Han Lin Shang Undergraduate Statistics Experience to the Available at http://pubs.amstat.org/toc/jcgs/19/1 n Discipline’s (and Society’s) Future,” Brian C. Kotz
12 Amstat News JUNE 2010 Statistics Without Borders to Meet at JSM 2010 James J. Cochran
tatistics Without Borders (SWB), an organi- SWB projects at various stages of completion. All zation committed to providing not-for-profit interested individuals are welcome to attend. organizations with pro bono statistical con- Also of interest to SWB volunteers is an invited Ssulting to assist in the resolution of international panel session, titled “Global Statistical Capacity health issues (broadly defined), will meet August 1 and the Role of Statistical Societies,” scheduled on from 5:00 p.m. – 6:00 p.m. during the 2010 Joint August 1 from 2:00 p.m. – 3:50 p.m. (See Page 48). Statistical Meetings in Vancouver, British Columbia. Additionally, a meeting of the ASA Volunteerism SWB officers and volunteers will learn more about Special Interest Group will follow the SWB meeting. related activities, discuss the progress SWB has For more information about SWB, or to volun- made in establishing working relationships with teer, contact James J. Cochran at [email protected]. other not-for-profit organizations, and talk about edu or Gary Shapiro at [email protected]. n
Editor Sought for CHANCE Magazine
ominations and applica- tions are being sought for the next editor of NCHANCE magazine. Working with the editorial board and the ASA’s magazine staff, the editor will provide direction and vision for the magazine, which has been published by the ASA for more than 20 years. The editor’s term will be from 2011 to 2013. To submit a nomination, include the name and contact information of the nominee and a brief description of the nomi- nee’s qualifications. niches for CHANCE as a To apply, submit a curricu- complement to Significance, lum vitae and the names of two a magazine published jointly references. Applicants also by the ASA and Royal should provide a vision state- Statistical Society. to Megan Murphy, ASA com- ment for CHANCE that Nominations and applications munications manager, at megan@ addresses opportunities and should be submitted by June 21 amstat.org. n
JUNE 2010 Amstat News 13 14 Amstat News JUNE 2010 Congress Views Results of NSF-Funded Research Projects Steve Pierson, ASA Director of Science Policy
ontserrat Fuentes, a professor in the sta- tistics department at North Carolina
State University, represented the ASA at PhotographyScavone of Scavone David by Photo theM 16th Annual Coalition for National Science Funding (CNSF) Capitol Hill Exhibition on April 14. The CNSF exhibit highlights to Congress research funded by the National Science Foundation (NSF). Fuentes also made Hill visits on behalf of
the ASA Climate Change Policy Advisory ( www.scavonephoto.com ) Committee (CCPAC). Fuentes’s poster summarized her NSF-funded research projecting the effects of climate change on tropospheric ozone in the eastern United States in the early 2040s. It also highlighted the contribu- tions statisticians bring to climate change and health impacts research, including the quantification of From left: Montserrat Fuentes, ASA President Sastry Pantula, Rep. Howard uncertainty, expression of projections in terms of Coble, Rep. Walter Jones, and Matt Peterson stand in front of the poster sum- probabilities, and the evaluation the climate models. marizing Fuentes’s NSF-funded research project on the effects of climate Because of the policymakers in attendance, Fuentes change on tropospheric ozone. emphasized the tools statisticians provide to facilitate policymaking and more efficient management of air quality and other environmental agents under lim- ited information and changing climatic conditions. A number of representatives and their staffs vis- PhotographyScavone of Scavone David by Photo ited with Fuentes at her poster. Overall, 250 people were estimated to have attended the event, includ- ing nine members of Congress. Prior to the evening exhibition, Fuentes partici- pated in seven meetings on Capitol Hill to advocate full funding for President Barack Obama’s fiscal year ( www.scavonephoto.com ) (FY) 2011 budget request for NSF and discussed the work of CCPAC. Accompanied by The Ohio State University professor Mark Berliner and ASA Director of Science Policy Steve Pierson, she met with Rep. David Price (NC-4), staff for North
Carolina senators Richard Burr and Kay Hagan, and several congressional committees. Steve Pierson, ASA director of science policy (right), discusses Montserrat One of the CCPAC documents Fuentes dis- Fuentes’s climate change poster with Rep. Brad Miller. cussed in her Hill meetings was a review of the research on the health impacts of climate research that she co-wrote with five other statisti- disease and pollution impacts. While finding no cians. (See www.amstat.org/committees/ccpac/pdfs/ “consensus among scientists regarding an increase ClimateChangeHealthImpacts.pdf.) Her group con- of infectious diseases under climate change,” they cluded that the research indicates a significant health did find that climate change is expected to result in impact of climate change is increased mortality from larger concentrations of tropospheric ozone, which global warming. The group also discussed infectious would lead to more ozone-related deaths. n
JUNE 2010 Amstat News 15 Have You Checked Out CHANCE Lately?
Though CHANCE has been around since 1988, the magazine recently unveiled a new online version.
Now, free for student members, CHANCE features articles for anyone with an interest in the examination of data. CHANCE intends to inform and entertain with articles focused on current events and statistical practice.
Check out CHANCE today online at www.amstat.org/publications/chance or log in to ASA Members Only for full access.
Not currently a subscriber? Start your subscription today. ASA members pay only $30 a year.
16 Amstat News JUNE 2010 Have You Checked Out CoC Sponsors Sessions Featuring CHANCE Lately? Scientific Applications of Statistics
2010 he Council of Chapters is sponsoring two invited sessions at JSM 2010 that will fea- ture diverse scientific applications of statis- Ttics. One session will highlight statistical opportu- nities in large-scale international science endeavors. JSM The other will showcase the work of several CANADA Vancouver statisticians exploring the impact of cli- Vancouver, British Columbia mate change on British Columbia’s forests. Big Science: Opportunities for Statisticians in the World’s Most century unfolds, massive new threats are emerging. Massive Projects Mountain pine beetles have devastated vast tracts Organizer and chair: Kary Myers, Los Alamos of lodgepole pine, and wildfires have swept down National Laboratory dry hillsides to threaten valley-bottom communi- From atom smashes to gene sequences, huge ties. Warming temperatures are possible contribut- collaborative endeavors generate some of the most ing factors. Two of the three speakers in this session exciting scientific data and developments in the will assess scientific evidence for this hypothesized world today. Often called “big science,” these mas- influence. The third will report on investigations sive projects combine big machinery, big labora- on extracting greater benefit from the threatened tories, big computing power, and big budgets. In forest resources. many cases, big science also produces big data sets, Visit www.amstat.org/meetings/jsm/2010 to view offering many opportunities for statistical insights. the JSM 2010 online program for session times n In this session, scientists will introduce the techni- and locations. cal context and statistical challenges of three big sci- ence efforts: the hunt for the Higgs boson with the Large Hadron Collider (Isabel Trigger, TRIUMF), the mission of the Mars Science Laboratory Rover to assess whether Mars can support microbial life (Sam Clegg, LANL), and the discovery and characterization of Earth-approaching objects Members Elected to National such as asteroids and comets by way of the Pan- STARRS digital sky survey (Alex Szalay, The Johns Academy of Sciences Hopkins University). The National Academy of Sciences recently announced the election The BC Forest Resource in a of new members and foreign associates, of which three are ASA mem- Changing Climate bers. ASA members elected are Jerome H. Friedman of Stanford Organizer and chair: Rick Routledge, Simon University, Michael I. Jordan of the University of California, Berkeley, Fraser University and Donald B. Rubin of Harvard University. British Columbia forests are facing unprec- Election to the academy is considered one of the highest hon- edented change. As the last century wound down, ors that can be accorded a U.S. scientist or engineer. The entire the forests were the focus of high-profile protests list of new members and foreign associates is available at www8. over forest management practices. Yet as the current nationalacademies.org/onpinews/newsitem.aspx?RecordID=04272010. n
JUNE 2010 Amstat News 17 Journal of Nonparametric Statistics Highlights May Issue Features Discussion Paper Suojin Wang, Editor-in-Chief, Journal of Nonparametric Statistics
Special Rate for Members variables or independence between the two samples. In addition, the authors propose sieve quasi maximum Journal of Nonparametric Statistics is now a publication of the likelihood estimation for the parameter of interest American Statistical Association. ASA members are entitled to and investigate its properties when the two samples a personal online subscription for $15 per volume. To order, are independent and the nonlinear regression model visit www.tandf.co.uk/journals/offer/gnst-so1.asp. is parameterized. Five experts—Aurore Delaigle, Peter Hall, Han Hong, Marie-Luce Taupin, and Young he May 2010 issue features a discussion Truong—contribute interesting discussions from paper, “Identification and Estimation of both theoretical and practical viewpoints. They offer Nonlinear Models Using Two Samples with additional insight and ideas for further research and TNonclassical Measurement Errors,” by Raymond development. These discussions are followed by Carroll, Xiaohong Chen, and Yingyao Hu. As the a rejoinder. title indicates, this paper considers identification The discussion paper, comments, and rejoinder and estimation of a general nonlinear errors-in- will be freely available online until December 31. variables model using two samples. In addition to the discussion and rejoinder This statistical problem is especially appeal- papers, seven other papers appear in this issue. ing in current research in econometrics, biology, These papers cover a range of research interests in and medicine. The authors show that the general nonparametric statistics, from functional estima- latent nonlinear model is nonparametrically identi- tion to variable selection to goodness-of-fit test- fied using the two samples when both could have ing. To view the full table of contents, visit www. nonclassical errors, without either instrumental informaworld.com/gnst. n
Join Us for JSM 2010 in Vancouver, British Columbia
Connect with your colleagues and take advantage of hundreds of sessions about the state of the art of statistics this summer in Vancouver, JSM Registration Is Open! British Columbia.
Register early and SAVE! Early Bird deadline: June 29 Advance registration deadline: July 13
Visit the JSM website at www.amstat.org/meetings/jsm/2010/registration
18 Amstat News JUNE 2010 Statistical Analysis and Data Mining Highlights Three Interrelated Papers Spotlighted Joe Verducci, Editor, Statistical Analysis and Data Mining
olume 3, issue 3 features However, SVMs have some three interrelated papers. potential shortcomings. They can The first proposes a new be overly sensitive to outliers and automaticV criterion for selecting the number of support vectors the bandwidth to be used in needed to determine the classifi- Gaussian kernel support vector cation boundary grows linearly machines (SVMs). The second with sample size. This last prop- proposes a sequential version of erty is particularly troublesome SVMs, called twin prototype when large amounts of training SVMs (TVMs), which efficiently data are streaming and there is updates a fixed number of support only a fixed, budgeted amount of vectors when training data arrives storage. Sensitivity to outliers can sequentially and there is limited be fixed by replacing the SVMs’s storage capacity. The third paper hinge loss function with a ramp also covers processing data streams, loss that ignores all large deviations using only a minimal number of but summarization is in terms of from boundary, but this comes at distinct stocks. Novel techniques hidden factors that link multivari- a high computational expense. include regularizing the cross- ate inputs and responses. In “Online Training on a Budget covariance matrix M of X and Y In “A Stable Hyperparameter of Support Vector Machines Using to simplify partial least squares Selection for the Gaussian RBF Twin Prototypes,” Zhuang Wang (PLS) estimation to ordinary least Kernel for Discrimination,” and Slobodan Vucetic propose squares (OLS), which allows for Jeongyoun Ahn provides a geo- using a fixed number of proto- sparse estimation by penalizing metrical interpretation of the types in place of support vectors. the L1 norm of the coefficients. smoothing parameter in terms To accommodate a new example The incremental Sparse PLS of the feature mapping implied arriving near the current bound- (iS-PLS) algorithm is the first to by the radial basis function of a ary, either the prototype farthest combine tracking of latent fac- Gaussian kernel with bandwidth from the boundary is removed or tors with variable selection in an h. For small values of h, points two near prototypes are merged adaptive fashion for data streams. get mapped to near uniformity and the boundary is updated. In The iS-PLS procedure allows the on a hypersphere, whereas large addition to being computationally number of important latent fac- values preserve the original dis- efficient, this TVM attains com- tors and their weights to evolve tances between data points. Since parable accuracy to the uncon- over time; the important variables the SVM is essentially a linear strained SVM, as reported for 12 retained within each latent factor discriminator in the feature space, large benchmark data sets. also evolve over time, but their a natural (geometry-based) crite- An interesting extension of number does not. This method rion is GB (h) = difference of the learning from streaming data is validated on both simulated between-versus the within-sums occurs when the response is not and real data, including enhanced of squares in the feature space, and just a classification, but a real-val- index tracking in which individ- h is chosen to maximize GB (h). ued vector Y, and the objective is ual stocks are selected to outper- This choice applies to any linear to learn a linear regression linking form the indexes being tracked by discriminator in the feature space, to the input vector X. Giovanni a fixed percentage. is computationally very efficient, Montana and Brian McWilliams As a whole, these papers pro- has low variability under many tackle this problem in “Sparse vide a snapshot of current research underlying models, and tends to Partial Least Squares Regression in classification and regression, achieve better tuning than other for Online Variable Selection with making the procedures more self- methods in terms of minimizing Multivariate Data Streams.” adaptive and extending them to the misclassification rate. This A motivating problem is track- streaming data, either from a stable latter property is illustrated using ing multiple financial indexes, such distribution or one subject to local nine benchmark data sets. as the S&P 100 and the Nikkei, trends, such as a market factor. n
JUNE 2010 Amstat News 19 Significance Highlights First Joint Issue Offers Fair Bit to Enjoy Julian Champkin, Significance Editor
s the result of a partner- possible to predict the results of Also in this issue are an article ship between the ASA and such fundamental reform? Jasjeet about a better, statistically based Royal Statistical Society, Sekhon and Jonathan Gruber way of understanding the threats everyA member of the ASA debate the issue. to orangutans and contributing will receive Significance in addi- Meanwhile, Peter Brooker to their conservation; a statistical tion to Amstat News starting looks at the risk analysis behind analysis of carvings on thousand- in September. the decision to ground the year-old standing stones that The first joint issue will con- planes—and finds that the statis- shows they are not random, but tain a mix of the important, the tical tests that would have made form a written language; and an topical, the scientific, and, occa- the analysis meaningful have not interview with George Box, the sionally, the whimsical. been done. “Renaissance Grand Old Man” Two events that have domi- Nobel prizes will be announced of statistics. nated the news recently are the in October. Ahead of them, Add columns, letters, and health care reforms that passed Significance looks at just how even a statistical crossword into U.S. law in March and the objective they can really be. The and you get a magazine with Icelandic volcano that halted air statistics of women laureates over something to interest, some- traffic over Europe in April. The the years show a pattern that not thing to inform, and a fair bit health care changes are the most even an all-male conspiracy theo- to enjoy. n far-reaching in U.S. history. Is it ry could convincingly explain.
20 Amstat News JUNE 2010 The American Statistician Highlights Statistics Curricula, March Madness Featured in May Issue John Stufken, Editor, The American Statistician
he featured article in the May issue is “Computing Book Reviews in the Statistics TCurricula,” by Deborah Nolan Applied Spatial Data Analysis with R Roger S. Bivand, Edzer J. Pebesma, and Virgilio Gómez-Rubio and Duncan Temple Lang. The authors argue that this is an The Art of Conjecturing, together with Letter to a Friend on Sets in opportune time to make signifi- Court Tennis cant changes in statistics curri- Jacob Bernoulli, translated by Edith Dudley Sylla cula and focus their attention on the integration of computational Bioconductor Case Studies topics with the teaching of mod- Florian Hahne, Wolfgang Huber, Robert Gentleman, and Seth ern statistical methods. Nolan Falcon and Lang advocate that comput- Business Statistics ing must be central to the statis- Norean R. Sharpe, Richard D. De Veaux, and Paul F. Velleman tics curricula at both the under- graduate and graduate levels and Dynamical Processes on Complex Networks that computational topics must Alain Barrat, Marc Barthélemy, and Alessandro Vespignani be integrated into traditional sta- Graph Theory: A Problem Oriented Approach tistics courses. The article pres- Daniel A. Marcus ents a broad set of computational topics with ideas on how to Introduction to Probability and Statistics for Science, Engineering, and teach them. Finance In the General section, Paul Walter A. Rosenkrantz Fearnhead and Benjamin M. The Monty Hall Problem: The Remarkable Story of Math’s Most Taylor extend the madness of Contentious Brain Teaser the NCAA basketball season Jason Rosenhouse in “Calculating Strength of Schedule and Choosing Teams New Drug Development: Design, Methodology, and Analysis for March Madness.” The authors J. Rick Turner propose a method to select teams for the NCAA tournament based SAS for Data Analysis: Intermediate Statistical Methods Mervyn G. Marasinghe and William J. Kennedy on their win-loss records after accounting for the different The Unfinished Game: Pascal, Fermat, and the Seventeenth-Century opponents they have played. Letter That Made the World Modern David H. Annis, Peter C. Keith Devlin Kiessler, Robert Lund, and Tara L. Steuber propose a sym- Wavelet Methods in Statistics with R metrized moment estimator for G. P. Nason estimation of the one-step-ahead Applied Survival Analysis: Regression Modeling of Time-to-Event Data transition probabilities in a revers- (2nd ed.) ible Markov chain on a count- David W. Hosmer, Stanley Lemeshow, and Susanne May able state space in “Estimation in Reversible Markov Chains.” The Little SAS Book: A Primer (4th ed.) Devan V. Mehrotra, Xiaomin Lora D. Delwiche and Susan J. Slaughter Lu, and Xiaoming Li propose Statistics: Informed Decisions Using Data (3rd ed.) and study two alternatives to Michael Sullivan III the van Elteren test, which uses
JUNE 2010 Amstat News 21 within-stratum ranks, for com- paring two treatments in a strati- fied experiment. UMBC Statistics Program In “Three Examples of Accurate Likelihood Inference,” Celebrates 25th Anniversary C. Lozada-Can and A. C. Elliot Hirshman, Neerchal Nagaraj, and Bimal Sinha Davison demonstrate through illuminating examples that, com- putationally, it can be relatively straightforward to use higher- order corrections for improving standard likelihood inference. In the final contribution to this section, Gunnar Taraldsen and Bo Henry Lindqvist present an elementary introduction to a the- oretical framework for statistics that includes improper priors, considering both Bayesian and non-Bayesian models. The Statistical Practice sec- tion opens with “A Note on Bayesian Inference After Multiple Imputation,” by Xiang Zhou and Jerome P. Reiter, who consider From left: Elliott Hirshman, UMBC provost; Nagaraj Neerchal, department Bayesian inference on multiply chair; and Bimal Sinha, founding faculty member imputed data sets in settings in which posterior distributions of the parameters of interest are he University of and poster presentation. Later, not approximately Gaussian. Maryland Baltimore Dean Rous of UMBC’s College Nicholas T. Longford follows County (UMBC) of Natural and Mathematical with “Bayesian Decision Making TDepartment of Mathematics and Sciences recognized every statis- About Small Binomial Rates with Statistics celebrated the 25th tics doctoral student to graduate Uncertainty About the Prior.” anniversary of its statistics gradu- in the past 25 years. In “Is the t Confidence ate program April 25 in conjunc- Rous also recognized the Interval: ± t (n − 1)s/ Optimal?”, tion with their Fourth Annual contributions of Bimal Sinha, Yijun Zuo ‹proposes a new pro- Probability and Statistics Day. the program’s senior most and cedure for constructing a confi- More than 100 participants founding faculty member, and dence interval for an unknown attended the silver jubilee, where commended the leadership of mean parameter that can outper- Ray Carroll, Bruce Lindsey, Ron Nagaraj Neerchal as department form not only the classical inter- Butler, and Nancy Geller were the chair and Anindya Roy as gradu- val in the title but also bootstrap keynote speakers. Many alumni ate program director. confidence intervals. also presented talks. The conference ended with In the History Corner section, The program was inaugu- a banquet and speech by ASA Michael Friendly, Pedro Valero- rated by UMBC’s provost, Elliot President Sastry Pantula, titled Mora, and Joaquín Ibáñez Hirshman. In addition to plenary “All I Want to Do When I Grow Ulargui take the reader on an and parallel sessions covering a Up … Be an ASA President.” interesting journey featuring variety of topics, the jubilee fea- For more about the celebra- Michael Florent van Langren and tured oral and poster presentations tion, visit www.retrieverweekly. the first (?) instance of data visu- by graduate students and the pre- com/?module=displaystory&story_ alization in “The First (Known) sentation of awards for best oral id=5602&format=html. n Statistical Graph: Michael Florent van Langren and the ‘Secret’ of Longitude.” n
22 Amstat News JUNE 2010 Titles from the ASA-SIAM SERIES on statistics and applied probability
Data Clustering: Theory, Algorithms, and Applications Guojun Gan, Chaoqun Ma, and Jianhong Wu 2007 · xviii + 466 pages · Soft · ISBN 978-0-898716-23-8 List $114.00 · ASA/SIAM Member $79.80 · Code SA20 The Structural Representation of Proximity Matrices with MATLAB Lawrence Hubert, Phipps Arabie, and Jacqueline Meulman ASA and SIAM MEMBERS 2006 · xvi + 214 pages · Soft · ISBN 978-0-898716-07-8 List $83.00 · ASA/SIAM Member $58.10 · Code SA19 GET 30% OFF LIST PRICES! The Analysis of Means: A Graphical Method for Comparing Means, Bayesian Nonparametrics Eliciting and Analyzing Expert Rates, and Proportions via Neural Networks* Judgment: A Practical Guide Peter R. Nelson, Peter S. Wludyka, Herbert K. H. Lee Mary A. Meyer and Jane M. Booker and Karen A. F. Copeland 2004 · x + 96 pages · Soft · ISBN 978-0-898715-63-7 2001 · xxxii + 459 pages · Hard · ISBN 978-0-898714-74-6 2005 · xii + 247 pages · Soft · ISBN 978-0-898715-92-7 List $49.50 · ASA/SIAM Member $34.65 · Code SA13 List $109.50 · ASA/SIAM Member $76.65 · Code SA07 List $89.00 · ASA/SIAM Member $62.30 · Code SA18 Applied Adaptive Statistical Introduction to Matrix Analytic Design and Analysis of Gauge R&R Methods: Tests of Significance Methods in Stochastic Modeling Studies: Making Decisions with and Confidence Intervals G. Latouche and V. Ramaswami Confidence Intervals in Random Thomas W. O’Gorman 1999 · xiv + 334 pages · Soft · ISBN 978-0-898714-25-8 and Mixed ANOVA Models 2003 · xiv + 174 pages · Soft · ISBN 978-0-898715-53-8 List $77.00 · ASA/SIAM Member $53.90 · Code SA05 Richard K. Burdick, Connie M. Borror, List $72.00 · ASA/SIAM Member $50.40 · Code SA12 Statistical Case Studies: and Douglas C. Montgomery Fuzzy Logic and Probability A Collaboration Between Academe 2005 · xviii + 201 pages · Soft · ISBN 978-0-898715-88-0 List $63.00 · ASA/SIAM Member $44.10 · Code SA17 Applications: Bridging the Gap and Industry Edited by Timothy J. Ross, Jane M. Edited by Roxy Peck, Larry D. Haugh, Anthology of Statistics in Sports Booker, and W. Jerry Parkinson and Arnold Goodman Edited by Jim Albert, Jay Bennett, 2002 · xxiii + 409 pages · Hard · ISBN 978-0-898715-25-5 Instructor Ed.: 1998 · xxxi + 282 pages · Soft and James J. Cochran List $137.50 · ASA/SIAM Member $96.25 · Code SA11 ISBN 978-0-898714-13-5 · List $70.50 2005 · x + 322 pages · Soft · ISBN 978-0-898715-87-3 ASA/SIAM Member $49.35 · Code SA03 List $68.00 · ASA/SIAM Member $47.60 · Code SA16 Recurrent Events Data Analysis Student Ed.: 1998 · xxxi + 181 pages · Soft for Product Repairs, Disease ISBN 978-0-898714-21-0 · List $43.00 Experimental Design Recurrences, and Other ASA/SIAM Member $30.10 · Code SA04 for Formulation* Applications Engineering Reliability Wendell F. Smith Wayne B. Nelson Richard E. Barlow 2005 · xx + 367 pages · Soft · ISBN 978-0-898715-80-4 2002 · xii + 151 pages · Hard · ISBN 978-0-898715-22-4 List $110.00 · ASA/SIAM Member $77.00 · Code SA15 1998 · xx + 199 pages · Soft · ISBN 978-0-898714-05-0 List $102.50 · ASA/SIAM Member $71.75 · Code SA10 List $75.00 · ASA/SIAM Member $52.50 · Code SA02 Mathematica Laboratories Multivariate Statistical Process Instructor’s Manual: ISBN 978-0-898714-26-5 SA06 free with book for Mathematical Statistics: Control with Industrial Applications Emphasizing Simulation and Robert L. Mason and John C. Young Statistical Case Studies for Computer Intensive Methods 2001 · xiii + 263 pages · Hard · ISBN 978-0-898714-96-8 Industrial Process Improvement Jenny Baglivo List $84.50 · ASA/SIAM Member $59.15 · Code SA09 Edited by Veronica Czitrom 2004 · xx + 260 pages · Soft · ISBN 978-0-898715-66-8 Includes CD-ROM and Patrick D. Spagon List $78.50 · ASA/SIAM Member $54.95 · Code SA14 A Primer for Sampling Solids, 1997 · xxviii + 514 pages · Soft · ISBN 978-0-898713-94-7 Includes Student CD-ROM · An Instructor’s CD-ROM List $79.50 · ASA/SIAM Member $55.65 · Code SA01 with complete solutions (ISBN 978-0-898715-70-5) is Liquids, and Gases: Based on the available upon adoption of text. Seven Sampling Errors of Pierre Gy *A portion of the royalties from the sale of this book Patricia L. Smith are contributed to the SIAM student travel fund. 2001 · xx + 96 pages · Hard · ISBN 978-0-898714-73-9 List $62.50 · ASA/SIAM Member $43.75 · Code SA08
Visit www.siam.org/catalog for more books and to order online. www.siam.org · www.amstat.org · 1-800-447-SIAM · +1-215-382-9800 6/10
JUNE 2010 Amstat News 23 24 Amstat News JUNE 2010 STATISTICIAN’S VIEW
Feedback: Odds Are, It’s Wrong
Below is a response to the article “Odds Are, It’s Wrong,” published in the March 27 issue of Science News. ASA members wrote the letter to the editor and cited the writer’s failure to clearly distinguish between the misuses of statistics and its methodological limitations. The letter, as edited by the magazine, was printed in the May 8 issue of Science News (Page 32) and is now online. To view the original article, visit www.sciencenews.org/view/feature/id/57091/ title/Odds_Are,_Its_Wrong. To view the edited response and other responses to the article, visit www.sciencenews.org/view/generic/id/58594/title/Feedback.
Dear Editor,
dds are it’s wrong, but the chances that sta- fundamental, factors involved, such as the pressure tistics is to blame are slim and fat. Tom to publish and obtain funding (positive results sell); Siegfried accurately portrays the impor- the public’s, and hence the media’s, appetite for pal- tanceO of statistics in the conduct of science. However, atable findings (chocolate is good for your health); his failure to clearly distinguish between the misuses and data hoarding (no one can see how thoroughly of statistics and its methodological limitations leads I searched my data to find a p = .049). to misleading conclusions about the role of statistics A chisel in a skillful artist’s hand can produce a in the proliferation of erroneous scientific results. beautiful sculpture, and a scalpel in an experienced Furthermore, his characterization of statistics as a surgeon’s hand can save a person’s life. Similarly, sta- mutant form of mathematics rooted in the same tistical techniques used properly by an honest and principles that guarantee profits for Las Vegas casi- knowledgeable scientist can be equally impressive nos is unscientific at best. It only hinders the real- at illuminating complex phenomena, thus promot- ization that statistics is the discipline that is best ing scientific understanding and shortening the positioned to contribute to the solution of the prob- time between scientific discovery and its impact on lems that he so entertainingly describes. societal problems. If misused, they can produce the Statisticians have long recognized the challenges counterproductive results that Siegfried describes. presented by multiple testing, the interpretation of Such erroneous results, however, should not be observational data, and, more recently, the analy- viewed as a failing of statistics. sis of high-dimensional data. Siegfried rightfully acknowledges the many statisticians and biostatis- Sastry G. Pantula ticians who have persistently and repeatedly writ- President, American Statistical Association ten eloquently on these issues. He also notes that appropriate methods, such as those for false discov- Jef Teugels ery control, are available to ameliorate the problems. President, International Statistical Institute Yet, he curiously persists with the theme that statis- Len Stefanski tics is defective when it is the misuse of statistical Editor, Journal of the American Statistical methods that is the main culprit in the situations Association (Theory and Methods) he describes. Siegfried has fired a shot across the bow of sci- ence that, although not perfectly on target, serves as a call for further discussion among statistical scien- tists and researchers in the relevant disciplines, such Correction as the medical, social, and behavioral sciences. There is a need to educate statistical practitioners at all lev- In the May issue of Amstat News, the “Treasures from the els, as gross misuse of statistical methods borders on ASA Archives” mistakenly referred to Frances Perkins as a he. scientific misconduct. However, it is also important Perkins, the first woman appointed to a presidential cabinet to realize that while statistics usually plays the role post, served as U.S. Secretary of Labor from 1933 to 1945. of the fall guy in these matters, there are other, more
JUNE 2010 Amstat News 25 Statisticians Comment on Status of Climate Change Science
The following are letters from ASA members in response to the March Amstat News article “Statisticians Comment on Status of Climate Change Science,” by Richard L. Smith, L. Mark Berliner, and Peter Guttorp. The authors answered additional questions online on March 31. The transcript of the discussion can be viewed at http://magazine.amstat.org/2010/03/ climatemar10.
Dear Editor,
read the recent article by Smith, Berliner, and My understanding is that temperature recon- Guttorp explaining why the ASA Climate structions are based on combining inverse regres- Change Policy Advisory Committee recom- sions of various surrogates (such as tree rings). The mendedI that Sally Morton sign the letter of consen- opportunities for unknown biases and uncertainties sus of climate change science. entering such a calculation seem endless. It defies I appreciate the hard work the committee statistical intuition that we could accurately recon- obviously put into this and I recognize their sin- struct the average temperature of the Earth hun- cerity and deep concern about the environment. dreds or thousands of years ago from inverse regres- However, I don’t think the article addressed the sions based on nonrandomly sampled surrogates. correct questions. Imagine the numerous omitted variables, not to No one doubts that the climate is changing; it mention lack of random sampling. And remember, is always changing. No one doubts that there has we need to estimate the derivative, which is difficult been a rise in average temperature. No one doubts even from the best data. that carbon dioxide plays a role in average tempera- Question 2: Carbon dioxide is a small part of the ture. And no one doubts that humans release carbon greenhouse gas in the atmosphere, water vapor being dioxide into the atmosphere. Below, I list what I see the largest component. Humans account for only as the real questions. I would be interested in know- part of the carbon dioxide that is in the atmosphere. ing the committee’s view on these questions. What fraction of warming is due to human-released Question 1: The main concern is not that the carbon dioxide compared to the other greenhouse temperature is increasing. The temperature of the gases? How accurately can this number be deter- Earth is never constant. The concern is that the rate mined? How large are the biases and variance? of increase is possibly much larger than during past Question 3: The climate is always changing. increases. In other words, the concern is about the There is nothing optimal about our current climate. derivative of the current warming period compared A warmer climate will have both positive and nega- to the derivative of past warmings. How certain can tive consequences. (For example, far more people we be that the derivative of the current warming die from cold than from heat.) How strong is the is exceptional? statistical evidence that the negative consequences To answer this question, we need to compare the outweigh the positive consequences? current climate to past climate. And to do this, we Let me emphasize that I appreciate the commit- need to use climate reconstructions. You acknowl- tee’s efforts. And I am not suggesting we should be edge that some reconstructions have been the object unconcerned about the environment. But, if anyone of great skepticism. As you noted, Ed Wegman should hold climate scientists to the highest statisti- affirmed, in his testimony before Congress, that cal standards, it should be statisticians. I’d like some the criticisms due to McIntyre and McKitrick are assurance that the ASA is doing so. legitimate. Yet, you dismiss the criticisms by merely saying “… the research community has responded Larry Wasserman successfully to these challenges.” I don’t feel that you Professor, Department of Statistics have really addressed this issue. Carnegie Mellon University
26 Amstat News JUNE 2010 Dear Editor,
am embarrassed by the “defense” of the anthro- about those projections.” So, accepting solar influ- pogenic global warming (AGW) theory pub- ence makes no sense because the projections are lished in the March 2010 issue of Amstat News. so uncertain already that a little more uncertainty IAs an agnostic on the issue, I was excited about the wouldn’t be noticed? article. I expected an objective, dispassionate analy- That’s statistics? sis—and possibly conclusion—by professional stat- I have a simple question: Does anyone have a isticians trained in judging theories based on evi- model, based on GHG readings of a specific location dence supported by data. Instead, the authors either or locations, that successfully predicts any future, tried to cast doubt on any contradictory evidence, observable, measurable temperatures—whether it or, worse, determined the state of knowledge of is a monthly average of satellite readings, a set of alternative theories was not sufficient to overcome specific ground temperature stations for a specific the presumption of validity of AGW. month, or even the thermometer on my deck—for Listen to the tortured denial of solar activity as five or 10 years in the future? Given the years of data a cause of global warming: “There is no credible and analysis thereof we supposedly have, that seems physical theory that would deny the GHG [green like a pretty low bar. Is anyone so certain of AGW house gas] influence.” I believe that there is also that they would publish such projections? no credible physical theory that would deny the I don’t believe the ASA should have signed onto solar influence. the letter to lawmakers based on the article I read. Or consider this gem: Solar activity theory could make future temperature projections more uncer- Terry G. Meyer tain, but “there is already plenty of uncertainty President, TECH Consulting
Dear Editor,
he article “Statisticians Comment on Status include important factors leads to precarious con- of Climate Change Science” in the March clusions, at best. It seems to me that because there is 2010 issue of Amstat News left me wonder- evidence that CO2 levels were twice as high as they ingT whether the dues I pay to the American are now millions of years ago, long before human- Statistical Association are paying for any part of the kind walked the Earth, there are surely factors work by the ASA’s Climate Change Policy Advisory missing from current models, factors that would Committee. Please know that I do not want my have caused prior warming periods and periods of dues funding this committee, and I do not want the elevated CO2. ASA making position statements on the status of Furthermore, plotting temperature data collected “climate change science” on its behalf. Additionally, from small patches of the Earth’s surface over the if the ASA makes statements on behalf of its mem- past 140 years would have no discernible (hockey bers, which includes me, about the notion of stick) trend (i.e., be part of the noise) if plotted over anthropogenic global warming, I will terminate my 400,000 years or 4,000,000 years or 4,000,0000,000 membership in the association. years. Don’t we teach our statistics students to plot I remain a skeptic of the “overwhelming” sci- data scaled in context to avoid making false claims? ence of anthropogenic global warming because, as an industrial statistician, model bias from failing to Linda Trocine
JUNE 2010 Amstat News 27 Dear Editor,
n the March 2010 issue of Amstat News, Richard • What are the projected effects of global Smith, Mark Berliner, and Peter Guttorp dis- warming, and what is the strength of the evi- cuss various aspects of global warming and con- dence that they are happening and caused by cludeI that (1) the climate is warming, (2) humans global warming? are likely responsible, and (3) mitigation measures are needed. They discuss (1) and (2) at length, but • How much harm (or benefit) will those con- mention (3) only in the last sentence, as if it is self- sequences bring? evident. The whole climate debate in general, in • What are the feasibility and cost of fact, largely ignores this part of the problem. A “sec- potential remedies? ond front” is needed in the discussion, one that takes as a given that warming is happening and Peter Bruce tackles questions like the following: President, statistics.com
Authors’ Response
s a preface to our response to letters regard- were the warmest decade of the last millennium, ing our article, the following points are ger- though this statement is often made. Rather, we mane to most of the letters. Please recall that believe the main functions of paleoclimatology in weA wrote, “We comment on some of the most com- the context of projecting future climate change are mon arguments that climate change is not happen- to learn about feedback mechanisms and gain some ing, or humans are not responsible.” Our article was indication of the scale of natural variability that has not intended to be a summary of the basis of any occurred in the past and may be expected to occur recommendation regarding ASA endorsement of in the future. any letter. Any such recommendation would involve Recent work on paleoclimatic reconstruction much more than brief comments on selected points from a statistical point of view includes the forth- of controversy. In particular, we did not review coming paper by Li, Nychka, and Ammann in JASA the case for anthropogenic climate change. Such and a forthcoming paper by Tingley and Huybers a review would involve combinations of in Journal of Climate. Finally, an in-depth discus- statistical and scientific arguments leading to a sion of aspects of paleoclimatology can be found in weight of evidence. the U.S. Congress–commissioned 2006 National Responding specifically, now, to Academy of Sciences report Surface Temperature each of the four writers, start- Reconstructions for the Last 2,000 Years (available ing with Wasserman’s three free as a download from National Academy Press). questions: This report includes various studies confirming the Question 1: We do “hockey stick” shape. not agree that the main Question 2: This point relates to the general area of purpose of paleocli- attribution. Climate models are run with both natural matology is to prove and anthropogenic forcing. These results are a major the derivative dur- contributor to the confidence of the Intergovernmental ing the last 25 years Panel on Climate Change (IPCC) in their Fourth is greater than in the Assessment Report (AR4) that anthropogenic influ- past 1,000 years (or ences explain the observed temperatures. the past 100,000) We each have concerns regarding the quality and in particular, since uncertainties associated with the inputs and forcings the climate response used. Further, interactions and climate feedbacks to CO2 increase is far regarding forcings are subject to uncertainty and from instantaneous, nor substantial continuing research. Indeed, character- do we agree that it really izing the response of the hydrological cycle (water matters whether the 1990s vapor, cloud properties, etc.) has been a holy grail in
28 Amstat News JUNE 2010 climate science. We note that Solomon et al. (2010) a prediction of the indicates that the post-1998 period of stable tem- date of death due peratures can be explained in terms of changing pat- to lung cancer of a terns of water vapor. The paper also indicates that specified cigarette different greenhouse gases can have different effects, smoker. so that changes in CO2 do not necessarily indicate Responding to the same changes in other greenhouse gases. Trocine, the evi- Question 3: There are uncertainties associated dence for anthro- with climate-scale projections, with the resulting sea pogenic contri- level rise and regional and local weather impacts and bution to global with the resulting effects on human outcomes (e.g., warming is not lim- health) and human endeavors (e.g., agriculture). ited to CO2. Current We do not believe all these uncertainties have been climate models include quantified, nor taken into account. Fortunately, a variety of greenhouse researchers from many disciplines, including ours, gases, as well as other forcings are engaged in research in these directions. We also of the climate. There is interaction note that statisticians support decisionmaking in the and feedback in the relationship between tempera- presence of uncertainty. ture and CO2. While it is correct that the last 140 Finally, we have not seen any clear evidence that years are a blip compared to previous changes, what “far more people die from cold than from heat.” The one really needs to look at is a kind of residual plot, death rates in Europe and North America are higher namely the difference between what you get with- during the winter months, but that does not imply out anthropogenic forcing and what you get with that temperatures in winter are the cause of death. all observed forcings. That is the right scale of com- Generally speaking, the evidence points to long parison for the kind of scientific judgment we are stretches of extreme temperatures as being most talking about here. dangerous. For example, stretches of unusually cold Finally, we fully agree that the issues Bruce raised weather in Spain (where unusually cold may mean are critical and merit substantial development and just freezing) is much more harmful than stretch- participation by statisticians. As mentioned above, es of unusually cold weather in Finland (where it we had a limited scope in this article. n means temperatures below -40o F). In any case, this query is symptomatic of the dif- ficulty and potential misunderstanding of “climate change” versus “global warming.” Regarding the latter, scientists are not concerned that the future climate would be identical to the current one, save STATISTICIAN’S VIEW a uniform, simple shift of 1o F or 2o F over this century. Rather, climate change represented in sea Statistician’s View is a place for members to express viewpoints level rise and changes in the distribution of weather about ASA issues and important topics in statistics. If you have trends and extremes (perhaps both hot and cold) are an opinion you would like to present or you want to respond to the keys. something you have read in this or other issues of Amstat News, Regarding Meyer’s letter, there is credible theory please send your letter to: as to the solar influence on climate (which is sub- stantial from a historical perspective and explains Managing Editor—Amstat News much of the observed past climate change). However, there is no theory that both explains past American Statistical Association climate change and current observations based on 732 North Washington Street solar forcing alone. Alexandria, VA 22314-1943 The second-to-last paragraph betrays the writer’s or misunderstanding of climate and, perhaps, statistics. [email protected] “Climate” is the distribution of weather, and hence predicting the temperature at a particular location at Please indicate that you would like your letter to appear in a particular time would have similar uncertainty to Statistician’s View. Amstat News reserves the right to use, refuse, and predicting a single draw from a probability distribu- edit any submission. We will not accept anonymous material. tion. The writer’s challenge is a bit like asking for
JUNE 2010 Amstat News 29 30 Amstat News JUNE 2010 Funding Opportunities Who Wants to Be a Biostatistician (or Environmental Statistician, or Social Science Statistician, or …)? Keith Crank, ASA Research and Graduate Education Manager
n April, I visited the National for Mental Health. These were generally provided through the Institutes of Health (NIH) primarily people who fund neu- Ruth Kirschstein programs, which with Sastry Pantula, ASA roscience research and recognize include institutional training Ipresident; Sally Morton, a past the need to get more biostatisti- grants for students and individ- president of the ASA; and Steve cians involved. They also realize ual grants for postdocs. These are Pierson, ASA director of science biostatisticians are in demand and available through all the institutes policy. We met with Jeremy Berg, getting them to change their bio- at NIH. They are restricted to director of the National Institute medical area of research is a losing U.S. citizens, noncitizen nation- of General Medical Sciences battle. So, they are very interested als, and permanent residents. As (NIGMS), to talk about funding in training new researchers. mentioned previously, NIGMS for training biostatisticians. Opportunities for stu- would like to receive more We are pushing NIGMS to dents and postdocs are proposals for training PhD find ways to fund some of our master’s students. There is inter- est, but the current guidelines make it difficult. (Funding is definitely possible to retrain people with an MD, but it isn’t clear whether this can be extended to people with a PhD in another discipline.) We also are pushing for more PhD training grants in biostatis- tics. NIGMS would like to see more submissions of proposals for biostatistics training grants. They recognize the demand in this area (by the ease of find- ing jobs without going through multiple postdocs), but they cannot fund proposals they do not receive. Although our focus was on training, we also raised concerns about the visibility of statistics as a collaborative discipline (not just a consulting or data analysis dis- cipline) and about inappropriate reviews of statistics proposals by reviewers without an appropriate statistical background. Before our meeting at NIGMS, we met with program directors at the National Institute
JUNE 2010 Amstat News 31 who want to learn an area of Funding Resources application. The Interdisciplinary Interagency Modeling and Analysis Group Grants in the Mathematical www.nibib.nih.gov/research/multiscalemodeling/imag Sciences program provides up to Includes program directors from many federal agencies (including NIH $100,000 for training in another and NSF) scientific discipline. These awards are for just one year, but they are Biomedical Information Science and Technology Initiative not restricted to biomedical areas www.bisti.nih.gov/funding/index.asp and do not have the citizenship Provides links to funding announcements dealing with bioinformatics restrictions that the NIH awards and computation have (see www.nsf.gov/funding/ pgm_summ.jsp?pims_=5299&org StatFund =DMS&from=home). http://statfund.cancer.gov There are lots of opportuni- Identifies NIH funding opportunities for statistical methodology and provides a ties available for research fund- list of grants that have been made in statistical methodology ing from NIH and NSF. To learn more about the opportunities at NIH, I invite you to attend a ses- sion I organized for JSM 2010 students in biostatistics (See www. researchers to spend 3–5 years in Vancouver, British Columbia. nigms.nih.gov/Training/InstPredoc/ studying and doing research in The session will meet at 8:30 PredocTrainingDescription.htm). a biomedical area and provide a.m. on August 4. The panel- For those whose research funding for both research activi- ists are Michelle Dunn, Shawn careers are established, but who ties and salary. They are restrict- Drew, and Denise Wiesch from may be looking to do something ed to U.S. citizens, noncitizen NIH and Jeremy Taylor from different, NIH provides oppor- nationals, and permanent resi- the University of Michigan tunities for quantitative scientists dents and require a mentor who Biostatistics Department. I hope (including statisticians) to redirect is willing to assist in the devel- to see you there. their research toward biomedical opment and execution of the To contact me, send an email issues. This is done through their research plan. to [email protected]. Questions Mentored Quantitative Research The National Science or comments about this article, Development Award (K25) (See Foundation’s (NSF) Division as well as suggestions for future http://grants.nih.gov/grants/guide/ of Mathematical Sciences has a articles, are always welcome. n pa-files/PA-10-062.html). These similar, but less generous, pro- awards offer the opportunity for gram for mathematical scientists Volunteers Wanted Have you ever thought about serving on an ASA committee of interest to you? The success of the ASA depends on ASA committees and the volunteers who serve on them.
Each year, vacancies are filled in nearly all of the ASA’s 50+ committees, usually for a three-year term. A list and description of all the committees can be found at www.amstat.org/committees.
You can volunteer or recommend another by going to www.amstat.org/committees and clicking on “Volunteer or make committee member recommendations.”
32 Amstat News JUNE 2010 Science Policy News Members Affect ASA Science Policy Steve Pierson, ASA Director of Science Policy, [email protected]
ast spring, I wrote a col- email to develop a recommenda- parties of the ASA’s action, work- umn titled “State of ASA’s tion for the ASA Board. ing with the ASA’s public relations Science Policy” (www. When developing a recom- specialist when appropriate. Lamstat.org/outreach/pdfs/SP_ mendation, the panel considers AnApr09.pdf). I think it is impor- the optimal timing for ASA action ASA Board Statements tant to keep the ASA member- and the mode of input (e.g., state- The two statements endorsed ship apprised of such activities, ment, letter, meeting of ASA lead- by the ASA Board at its April but I think it is more important ership with decisionmaker, etc.). meeting (see Page 5) arose from to further engage members in The ASA’s executive director and ASA member activity. Knowing these efforts and the direction director of science policy provide that many states would be they take. As my examples below input and inform the ASA Board considering—or reconsider- indicate, member input is essen- or Executive Committee of the ing—election auditing legis- tial to ASA science policy success. panel’s activities, requesting input lation, certain ASA members So, rather than reporting a com- as necessary and soliciting outside recognized that the ASA could prehensive science policy update, input to make the ASA’s action as positively influence this legis- I report on activities sparked by constructive as possible. lation by endorsing the risk- members and strongly encourage Once a letter or statement limiting election audits (because members to contact me with sug- is finalized, I inform interested of their efficiency advantage over gestions or concerns. While I can- not guarantee that the ASA can act on all suggestions, it’s impor- tant for us to hear from you. Process Because I think ASA members should know the process for any ASA science policy action, let me explain it, emphasizing that I do not act alone when deciding how to proceed. After a topic has been sug- gested, the first step is to assess its importance and pertinence to the ASA and the value of an ASA action. The level of autho- rization needed for ASA action is also assessed, taking into account such considerations as accordance with previous ASA Board action, ASA bylaws, and the level of sup- port within the ASA. The next step is for me, the director of science policy, to identify a panel of ASA member experts by soliciting input from the ASA’s leadership, committees, sections, and chapters. This panel then communicates by phone and
JUNE 2010 Amstat News 33 and Confidentiality, suggested save the section, assuming the ASA Science Policy Actions the ASA send a letter to HHS numbers are sustained (see http:// ASA responds to an NIGMS request for input Secretary Kathleen Sebelius magazine.amstat.org/2010/04/ on training grants. explaining the tools statisticians fundingopps_apr10). It also led to bring to the table. The letter was broadening the scope of applica- ASA signs a letter supporting written and, as a result, Reiter was tions considered by BMRD. reauthorization of the America COMPETES invited to participate in a HIPAA The other driver was the Act this year. Privacy Rule deidentification response to a posting by ASA ASA signs a letter in support of the fiscal workshop this year. Research and Education Manager year 2011 budget for the Bureau of Seeing the plans of the Keith Crank to the electronic Michigan State Board of mailing list for the ASA Caucus Economic Analysis. Education to weaken the statistics of Academic Representatives ASA signs a letter to the President’s Council component in parts of its curricu- requesting questions to ask at the of Advisors on Science and Technology, lum, Detroit Chapter members— town hall meeting held by NIH urging its members to take the lead in led by David Fluharty—contacted Director Francis Collins last fall. affirming the behavioral and social sciences the ASA. After checking with the Questions covered meeting the as an important part of science, technology, ASA president, ASA staff mem- great demand for biostatisticians bers helped draft a letter, which and mentoring biostatisticians to engineering, and mathematics education. was revised and sent on December the review of biostatistics training 15, 2009. grant issues and involvement of After learning a bill before statisticians as collaborators. That so-called fixed-percentage audits the California legislature would input led to a letter from 2009 and ability to correct an outcome) contain a provision to pilot the ASA President Sally Morton to and urging that principles, rather use of risk-limiting audits, Philip Collins, an April visit to Jeremy than details, be legislated. ASA Stark—creator of the methods for M. Berg, director of the National experts devoted many hours to conducting such audits and an Institute of General Medical producing a statement to recom- advocate for their use—contacted Sciences by Morton and Pantula, mend to the ASA Board. They also the ASA. ASA President Sastry and the upcoming JSM session sought input from participants in Pantula then wrote a letter of sup- on NIH opportunities discussed an election auditing workshop held port for the provision. The bill in Crank’s column on Page 31. at the ASA last fall that included received the unanimous biparti- CCPAC is a result of a con- computer scientists, political scien- san support of the California State versation Amy Braverman had tists, and voting advocates. Assembly Committee on Elections with Executive Director Ron The forensic science statement and Redistricting and now awaits Wasserstein at a recent JSM, in was led by ASA members active further consideration. which she lamented the lack of in bringing statistical methods to engagement by the ASA in issues bear on forensic science, includ- NIH and Climate of the day. A follow-up conversa- ing a representative of the ASA Change Activities tion led to a group of statisticians Committee on Law and Justice Other ASA action instigated by discussing how ASA members Statistics and people who have members involves the National could inform Congress on climate served on National Academies Institutes of Health (NIH) and change issues. As the group became panels and/or written extensively the ASA’s Climate Change Policy more involved, it was made into an on this topic. Advisory Committee (CCPAC). ad hoc ASA committee. In NIH’s case, there were two pri- In short, ASA members deter- Letters from the mary drivers. One was an email mine ASA science policy activi- ASA President from member Marie Davidian ties. If you have a suggestion for The following three instances from to the electronic mailing list for ASA action on a policy matter, the last year illustrate ASA presi- the ASA Caucus of Academic please contact me at pierson@ dents sending letters as a result of Representatives about the possi- amstat.org. And if there is a sci- ASA member suggestions. ble elimination of the Biostatistics ence policy activity you are inter- Hearing that the Department Methods and Research Design ested in working on, let me know. of Health and Human Services (BMRD) study section or its Without member input, little (HHS) might soon reconsider its merger with another study sec- ASA science policy work would Health Insurance Portability and tion. The overwhelming response get done. Such input includes Accountability Act (HIPAA) pri- to Davidian’s call for grant sub- contacting members of Congress vacy rules, Jerry Reiter, chair of missions to BMRD led to enough and meeting with them or mem- the ASA Committee on Privacy of an increase in submissions to bers or their staffs. n
34 Amstat News JUNE 2010 Master’s Notebook Find Your Fit Jack Nyberg, Senior Manager, Covance
s summer approaches, my thoughts turn toward the coming warm weather. If you’reA a recent graduate, your thoughts may be turning toward summer fun, but also toward the practical matter of finding your first job as a statistician. I’m sure most of you have already thought about important factors that will influence your job search. Salary, location, and advancement opportunities are some of what we consider when ranking potential employers, but an easily overlooked consider- ation is “finding your fit.” Finding your fit means that when you evaluate a job opportu- nity, you must evaluate whether the work culture fits your person- ality in addition to the tangible factors of salary, location, title, etc. That is, beyond the techni- cal aspects of statistical work, What Kind of Personality or acquaintances who already every job is immersed in the cul- Do I Have? work there. An Internet search of ture of the organization, depart- Most of you probably have a good the organization may also help. ment, and project team. To find idea about your likes and dislikes, Finally, ask questions during the the right job for you, you must but have you ever really tried to interview process. Ask about the consider finding the right work identify your personality type? If organizational, departmental, and culture for you. not, there are a number of won- team cultures. Say you are lucky enough derful resources (e.g., Myers- Request to talk with potential to field two job offers. Both are Briggs Type Indicator Test, John coworkers and managers. At first, similar in almost every respect; Holland’s Theory of Vocational you may be reticent to ask many however, one is with a fast-grow- Choice, etc.) that can profile questions, but keep in mind that ing small organization and the your personality. Also, talk with the process is two-way. They’re other is with an established large your friends, family, or a trusted determining if you’d be a good organization. The smaller organi- adviser to determine how other fit there, and you’re determining zation embraces a fast-paced and people perceive you. These assess- if they’d be a good fit for you. If make-it-up-as-you-go-along cul- ments, combined with an honest both of you succeed, then both of ture and the other has a deliber- self-evaluation, will help you dis- you win. ate and process-oriented culture. cover your personality type. Which job would you take? Am I Willing to Work To help answer this question, What Is the Work Where My Personality you must evaluate your per- Culture? Doesn’t Fit? sonality, the potential employ- Once you know your personality After you’ve determined your per- er’s culture, and how the two type, identify the culture of the sonality profile and the culture of fit together. job under consideration. Good the jobs under consideration, you sources of information are friends might be in the enviable position
JUNE 2010 Amstat News 35 of evaluating taking the job. But, working in an organization. Finding your fit multiple job environment that goes against makes work easier because you’re offers. your personality is like swim- more likely to enjoy coming into Unlike our ming against the incoming work each day, collaborate effi- scenario above, tide. You may not notice it at ciently with your coworkers, and it is unlikely you’ll first, but it just makes it more respect your employer. be comparing jobs difficult to reach success at Although not easily quanti- that are essentially that organization. fiable, these factors intuitively the same. One job may combine to make work enjoy- pay substantially more than Will the Work Culture Fit able. When work is enjoyable, the others. Or, one job may be with My Personality? you are more productive, per- in your ideal location. Whatever This question is similar to the sonable, creative, and energetic. the mitigating circumstance, you previous one, but it powerfully These make you more valuable must determine if you’re willing shifts the focus. Instead of pres- to your employer, which leads to to take a job that may not be the suring you to adapt to an ill- career advancement opportuni- best fit for your personality. fitting culture, it requires the job ties. Although nothing is guar- It is important to note that to fit your needs. anteed, by taking a job that fits anyone can adapt to succeed any- Finding a job that allows you your personality, you are laying where. So if a work culture is not to honestly answer “yes” to this the groundwork for long-term a good personality fit, you should question makes it more likely you success even before you’ve worked not be totally dissuaded from will find long-term success at that one day. n
36 Amstat News JUNE 2010 MEETINGS REGISTRATIONFORM 2010 FDA/Industry Statistics Workshop Sponsored by the ASA Biopharmaceutical Section with cooperation from the FDA Statistical Association 3EPTEMBER n s 'RAND (YATT 7ASHINGTON7ASHINGTON $# INSTRUCTIONS 1. Print or type all information and retain a copy for your records. 2. Use a separate form for each registrant. 3. Mail form with payment to FDA/Industry Statistics Workshop Registration, 732 N. Washington Street Alexandria, VA 22314. Fax form (credit card only) to (703) 684-2037. www.amstat.org/ 4. Registration form must be received by August 26, 2010, to be processed at the reduced rate. meetings/fdaworkshop Forms Received Without Payment Will Not Be Processed.
ATTENDEE INFORMATION MEAL PREFERENCE Lunch on Tuesday, September 21, is included with your workshop registration. To assist in planning, please select the area of discussion that most interests you. Name See Page 2 of this form for the list of topics. 3ELECT ONE OF THE FOLLOWING MENU OPTIONS Tuesday, September 21 Preferred Name for Badge Topic:______
❑ Regular ❑ Vegetarian ❑ Not attending lunch ASA ID # (if known)
Organization REGISTRATION FEES Workshop Fee (required) By August 27 – Address August 26 September 13 ❑ Registrant $270 $295 $______❑ Academic (nonstudent) $230 $255 $______❑ Biopharm Section Member $230 $255 $______❑ Government $130 $155 $______❑ Student $130 $155 $______City State/Province ZIP/Postal Code Short Courses—-ONDAY, September 20 Add-ons to Workshop Fee: $100 each before Aug. 26; $105 each Aug. 27–Sept. 13 Country (non-U.S.) 8:30 a.m.–12:00 p.m. ❑ SC1: Emerging Challenges in Clinical Trial Phone Methodologies—James Hung & Sue Jane Wang $______❑ SC2: Beyond Survival Analysis: Recurrent Event Responses in Clinical Trials—Richard Cook & Jerry Lawless $______Email ❑ SC3: Interpreting Change and Responder Analysis for Patient-Reported Outcomes—Joseph Cappelleri, Lisa Kammerman, & Kathleen Wyrwich $______In case of emergency, list the name and phone number of the person we should contact (remains confidential). 1:30 p.m.–5:00 p.m. ❑ SC4: Hot Topics: Recent Development in Clinical Trial Emergency Contact’s Name ______Methodologies—Scott Evans & Jim Ware $______❑ SC5: Good Statistical Practice and Common Subtle Telephone Number ______Statistical Mistakes—Frank Harrell $______❑ SC6: Quantitative Pharmacovigilance: Statistical ❑ Check here if you would like your ASA customer contact Approaches to Medical Product Safety Surveillance— information updated with your meeting contact information. Jie Chen & Yi Tsong $______
This meeting is ADA accessible. TOTAL FEES: $______❑ Please check here if you need special services due to a disability and attach a statement regarding your needs. PAYMENT ❑ Check/money order payable to the American Statistical Association (in U.S. dollars on U.S. bank) CANCELLATION POLICY #ANCELLATIONS RECEIVED BY !UGUST WILL BE REFUNDED LESS A PRO- Credit Card ❑ VISA ❑ MasterCard ❑ American Express CESSING FEE AND LESS A PROCESSING FEE FOR EACH SHORT COURSE #ANCELLATIONS RECEIVED BY 3EPTEMBER WILL BE REFUNDED LESS A PROCESSING FEE AND LESS A PROCESSING FEE FOR EACH SHORT COURSE 2EQUESTS FOR REFUNDS RECEIVED Card Number AFTER 3EPTEMBER WILL NOT BE HONORED !LL CANCELLATIONS MUST BE MADE IN WRITING TO [email protected] VIA FAX TO OR MAILED TO &$!)NDUSTRY Exp. Date CVS# (3- digit number on back of card) 3TATISTICS 7ORKSHOP 2EGISTRATION . 7ASHINGTON 3TREET !LEXANDRIA 6! *0URCHASE ORDERS WILL NOT BE ACCEPTED .O EXCEPTIONS !3! &EDERAL