WSS
NEWS

News about activities of the Washington Statistical Society
A Chapter of the American Statistical Association

WSS Home | Newsletter | WSS Info | Seminars | Courses | Employment | Feedback | Join!

September 2001

Contents:


Congratulations

The balloting results are in and the following people were elected to the Board of Directors of the Washington Statistical Society:

President-elect
Brenda Cox

Secretary
Thomas Mule

WSS Board Representatives at Large
Renee Miller
Jeri Mulrow

Methodology Program Chair
Mary Batcher

A round of applause goes to the following individuals who were recently elected Fellows of the American Statistical Society (ASA). This honor goes to:

Andrew White, National Academy of Sciences

Return to top

New Scholarship Fund

WSS is contributing to the Wray Smith Scholarship Fund and encourages its members to also contribute.

The Government Statistics Section is please to announce a new scholarship fund in memory of Wray Jackson Smith, a founding member of the Section, for his contributions to the Section and to the practice of statistics within the Federal government. The scholarship fund that will bear his name is intended to reward promising young statisticians for their diligence and thereby encourage them to consider a future in government statistics.

Dr. Wray Smith has had a long and distinguished career, both in the private and public sectors. He began his career in 1948, when he received his bachelor's degree in mathematics and physics from the George Washington University; in 1980, Smith received a doctorate of science in operations research and statistics from The George Washington University. His Federal career spanned four decades and included the Office of Economic Opportunity, where he was responsible for the operational start-up of the Job Corps and evaluation of manpower development programs; the Office of the Assistant Secretary for Planning and Evaluation in what was then the Department of Health, Education, and Welfare, where he oversaw the policy research data center and directed the Income Survey Development Program (the predecessor to the Survey of Income and Program Participation); and the Energy Information Administration, U.S. Department of Energy, where he directed energy modeling and data systems development and provided oversight of the Department's surveys of energy consumption in multiple sectors.

After retiring from the Federal government in 1983, Wray Smith continued to play a role in Federal statistics from the private sector, first as a visiting senior fellow at Mathematica Policy Research, then as research consultant, research director, and advanced studies director at Synectics for Management Decisions. With Carl Harris he founded Harris-Smith Research, an operations research/management science consulting group, which he chaired from 1985 to 1996. In 1997 Smith established the independent, nonprofit Harris-Smith Institutes, to provide statistical research and a range of other services to other nonprofit organizations and governmental entities. He served as president of Harris-Smith Institutes until his death in May 2000.

An active member of ASA, Wray Smith served on the Board of Directors and as a committee chair and Section officer. As Chair of the ASA Committee on Subnational Statistics, Smith appeared before the ASA Board in December 1987 to speak on behalf of a proposal to create a new Section on Government Statistics that would replace his committee and the ASA Committee on Government Statistics. The Board approved the proposal, pending the outcome of a membership vote, and the Government Statistics Section was established on January 1, 1989.

Beyond this list of accomplishments, Wray Smith is remembered fondly for his contributions to the career development of a number of young statisticians. The scholarship fund that will bear his name is intended to reward promising young statisticians for their diligence and thereby encourage them to consider a future in government statistics.

The Government Statistics Section is accepting contributions to help fund this new effort. For more information, please contact Pat Doyle at patricia.j.doyle@census.gov or by phone at (301) 457-3822. Or, please stop by the Wray Smith Scholarship Fund booth at the Joint Statistical Meetings in Atlanta. Individual contributions to the Smith Scholarship Fund are welcome. Checks should be made payable to "American Statistical Association." Please include your name, address, and phone number on your check and write "Wray Smith Scholarship Fund" in the comment line on your check. Contributions should be mailed to:

American Statistical Association
Attn: Stephen Porzio
1429 Duke Street
Alexandria VA 22314-3415

Return to top

Graduate Study in Statistics at GMU

Statistical Science is an information technology subject since it is focused on extraction of information from raw data. The Department of Applied and Engineering Statistics (AES) at George Mason University is one of six departments in the School of Information Technology and Engineering (IT&E).

Department faculty also head two interdisciplinary Centers. The Center for Computational Statistics is a research center focusing on the interaction of computing and statistics. It operates a virtual reality laboratory, a high performance computing laboratory and a technical library. The Statistical Consulting Center is a service center that works with faculty and graduate students across the disciplines on research projects and grant proposals. It operates a laboratory with state of the art software.

Two major factors have shaped the statistics curriculum at GMU:

  • The computing revolution
  • Proximity to Federal Agencies

Advances in computing power over the last two decades have spawned the fields of computational statistics, data mining and data visualization, in which GMU has an international reputation for innovative research. As a result of faculty interaction with federal agencies as consultants or through research grants and contracts, other areas of research at GMU include robust statistics, statistical signal processing, statistical computing software, and survey sampling.

Graduate Programs

Statistical Science, M.S.

Students are trained in the theory and practice of statistical methodology with special attention to computational techniques. The M.S. program offers a choice of defined subject matter emphases, which currently include: Applied statistics, Computational statistics, Engineering statistics, Federal statistics and Statistical signal processing. A student will normally select one of these emphases or design a customized curriculum in conjunction with a faculty advisor.

The student also selects either the research or professional option. The research option is intended for students planning to continue with the Ph.D. degree or to begin careers in statistical methodology research. The professional option provides M.S. degree qualifications to those seeking an expanded knowledge base in modern statistical theory and practice, but not wishing to pursue a research career. Such students might plan to work in applied statistics at federal government agencies or private sector companies, go on to professional schools, or teach statistics at a secondary level.

Admission Requirements

In addition to satisfying the general admission requirements for graduate study, all applicants to the M.S. program in Statistical Science must hold a bachelor's degree from an accredited institution that demonstrates mastery in multivariate calculus, matrix algebra, probability, and basic computer skills (the equivalent of five undergraduate-level courses at GMU).

Financial Aid

A limited number of graduate teaching assistantships are available for full-time students. Students who demonstrate exceptional potential after one or two semesters may qualify for graduate research assistantships. Although GRE (or GMAT) scores are not required for admission, students seeking financial aid are encouraged to submit one.

M.S. Degree Requirements

All students must complete 30 graduate credits including four required core courses (12 credits) in Applied Probability, Applied Statistics, Statistical Inference, and Regression Analysis. These courses cover the basic elements of statistics at the graduate level. The student may choose one of the five defined emphases outlined above or design a customized curriculum, which may include courses from other departments with approval of the AES graduate coordinator. The professional option requires completion of 18 additional credits of coursework beyond the core. The research option requires 12 additional credits of coursework and 6 credits of independent research leading to a master's thesis.

Graduate Level Courses Offered in 2001-02

All courses are offered in late afternoon (4:30-7:10 pm) or evening (7:20-10:00 pm) for the convenience of part-time students. Courses carry 3 credits unless noted otherwise.

STAT 544: Applied Probability
      
 (fall and spring)
STAT 554: Applied Statistics
 (fall and spring)
STAT 652: Statistical Inference
 (spring)
STAT 501: SAS Language and Procedures
 (fall) (1 cr, meets for 5 weeks)
STAT 502: Introduction to SAS/GRAPH
 (fall) (1 cr, meets for 5 weeks)
STAT 503: SAS Macro Language
 (fall) (1 cr, meets for 5 weeks)
STAT 574: Survey Sampling I
 (fall)
STAT 655: Analysis of Variance
 (fall)
STAT 657: Nonparametric Statistics
 (summer)
STAT 658: Time Series Analysis & Forecasting
 (spring)
STAT 663: Statistical Graphics & Data Exploration
 (fall)
STAT 674: Survey Sampling II
 (spring)
STAT 682: Computational Methods in Engnr. & Statistics
 (fall)
STAT 779: Topics in Survey Design and Analysis
 (spring) (1-3 cr.)*
STAT 789: Advanced Topics in Statistics**
 (spring)
STAT 781: Data Mining & Knowledge Discovery
 (fall)
CSI 810: Scientific Databases
 (fall)
IT 875/CSI 803: Scientific & Statistical Visualization
 (spring)
IT/CSI 972-973: Mathematical Statistics I & II
 (fall-spring)
*  Tentative topics for STAT 779 include 'Confidentiality Issues in Federal Data' and 'Survey Quality'.
** Tentative topic for STAT 789 is 'Survival Analysis'.

Certificate in Federal Statistics

The graduate certificate in federal statistics is a professional program targeted at upgrading the skills of practitioners. It consists of 15 credits that are aimed at building the foundations of statistical analysis and survey methods. It is extremely flexible and can be tailored to individual student needs. Admission requirement is a bachelor's degree that includes two semesters of calculus and at least one course in probability or statistics. Students must complete 9 credits of coursework chosen from STAT 554 (or 510), 574, 634 (Case Studies in Data Analysis), 663, 665 (Categorical Data Analysis), 673 (Analysis of Longitudinal Data) and 674, plus 6 elective credits approved by the Certificate Coordinator.

Ph.D. Study in Statistics

The Ph.D. in Information Technology has an emphasis in statistical science with an engineering focus. The Ph.D. in Computational Sciences has an emphasis in computational statistics with a basic science focus. Both degrees are interdisciplinary and allow the student a broad range of course and research options.

Information

Contact the graduate coordinator in Science and Technology II, Room 158, (703) 993-3645, statistics@gmu.edu, or visit the website http://www.galaxy.gmu.edu/~rbolstei/. The statistics section of the 2001-02 GMU catalog is on the web at http://www.gmu.edu/catalog/ite_aes.html.

Return to top

Graduate School, USDA

The Graduate School, USDA is a continuing education institution offering both career-related courses primarily to government employees as well as academic and personal enrichment courses open to all adults regardless of their place of employment or educational background. The School's mission is to improve the performance of government through education, training and related services and to provide opportunities for individual lifelong learning.

A Brief History

During the decade previous to the founding of the Graduate School, the Department of Agriculture had lost 400 scientists due largely to a policy that had discouraged university education for department employees. Elmer D. Ball, the School's first director, believed that the establishment of the Graduate School would help to meet the demand for formal education.

The Graduate School's first classes began on October 17, 1921. One hundred and seventy-six students took part in science classes, an economics class and a statistics class. The large majority of students and most of the faculty worked for the Department of Agriculture. Classes met twice a week and cost

During its first ten years of operation, most students came from the USDA. In the early 1930s, other agencies asked for and received permission to enroll their employees. As student enrollment increased, so did requests for more courses. By 1935, only 10 of 41 courses dealt with strictly scientific subjects, such as biology or chemistry. During the academic year 1942- 1943, annual enrollment increased to nearly 8,700, including 3,700 from outside the government.

By 1971 the Graduate School had grown tremendously from176 students and eight teachers to more than 20,000 registrants and a faculty of several hundred. The School had also shifted its focus away from graduate instruction to become a leading institution of continuing education. Today the Graduate School continues to serve adults who have "graduated" from full-time schooling, while it also serves adults who are preparing to resume their formal education.

Administration

Although affiliated with the Department of Agriculture, the Graduate School does not receive any government funding. It is a self-supporting government entity, also known as a "non- appropriated fund instrumentality" (NAFI). The secretary of agriculture appoints a General Administration Board to govern the Graduate School. Board members are drawn from senior positions in government, industry and academia. Advisory committee, whose members are also government, business and academic leaders, guide directors and managers in developing, delivering and evaluating academic and training programs.

Academic Status

The Graduate School, USDA is not a degree-granting institution. However, the School does offer certificates of accomplishment in a wide range of areas, including accounting, collegiate studies, computer programming, editorial practices, elder law, environmental law and policy, natural history field studies, and paralegal studies, to name a few. Most courses in these programs have been reviewed by the American Council on Education's College Credit Recommendation Service and recommended for a certain number of credits, most commonly two semester credits. Students have received credit for these courses from colleges and universities around the country, such as Strayer University, Trinity College, Johns Hopkins University, Columbia University and Harvard University.

Students can obtain transcripts of all courses taken at the Graduate School from the Office of the Registrar.

Faculty

The Graduate School, USDA contracts with more than 1,200 faculty to provide more than 1,500 different courses worldwide. Faculty members are leaders in the subject areas they teach, whether they work in government, academia or the private sector. Their work enables them to bring expertise and a practical approach to their classroom teaching.

Mathematics and Statistics Program

The mathematics and statistics program at the Graduate School is one of its oldest. During the 1930s, Dr. W. Edwards Deming, widely acknowledged as a founding father of the quality movement in industry and a pioneer in sample survey research, was a Graduate School instructor and chair of the mathematics and statistics advisory committee. The first course in sample survey theory offered in the Washington area (the second in the United States) was held at the Graduate School during his tenure. Through the mid-eighties the Graduate School, USDA offered a variety of basic, graduate level courses in statistics and remained especially active in survey theory and methods. Various federal agencies frequently sent their newly hired statisticians to the Graduate School to receive training in specialized areas of statistics relating to their positions.

Today most Graduate School students in mathematics or statistics are either pursuing an undergraduate degree that requires a basic understanding of statistics or they are preparing for graduate work, often in business, economics or international studies. While the most highly subscribed courses are Calculus I and II and Introductory Statistics I and II, higher level statistics courses with a special focus on sampling continue to be offered at the Graduate School.

Current Course Offerings

 
Fall 2001
      
  
Algebra I
 $239
Basic Technical Mathematics
 $239
Calculus I
 $239
Mathematics for Economists I
 $239
Introductory Statistics I
 $269
Introductory Statistics II
 $269
Introduction to Sample Surveys
 $289
Exploratory Data Analysis for Computers
 $289
Winter 2002
Algebra II
 $239
Basic Technical Mathematics
 $239
Calculus I
 $239
Calculus II
 $239
Mathematics for Economists II
 $239
Introductory Statistics I
 $269
Introductory Statistics II
 $269
Theory of Sample Surveys
 $289

For more information:

Call Evening Programs Course Information: (202) 314-3650.

Contact Patricia DeFerrari, Mathematics & Statistics Program Manager:
voice:  (202) 314_3656
e_mail: patricia_deferrari@grad.usda.gov

Visit the Graduate School, USDA Web Site: www.grad.usda.gov.

Return to top

Course Announcement

Graphical Exploratory Data Analysis (EDA)
(Data Analysis for Computers - STAT2201)

Graduate School, U.S. Department of Agriculture
Fall 2001 Semester

Class Dates, Place, and Times:
Capital Gallery, 600 Maryland Ave., SW (L'Enfant Plaza Metro Station) Thursday 6_9 p.m.
9/20-11/29

Instructor:
Dave DesJardins, Statistical Research Division, U.S. Census Bureau

Course Description:

Graphs, a natural human language, often make even complex statistical concepts easy to understand. This course provides subject matter specialists with a number of very powerful graphical techniques for the analysis of their data. It is a practical application course - designed first to provide students with a good basic understanding of the concepts of EDA and then to teach them how to apply these concepts. This is a hands-on, CPU course, teaching the fundamentals of JMP-IN and/or Insight - SAS Institute's powerful new easy-to-use (point-and-click) 4th generation interactive graphical data analysis software packages. This course requires only a basic understanding of statistical techniques.

Graphical Exploratory Data Analysis will focus on easily implemented, practical applications. It will use a variety of sample industry and scientific data sets to illustrate a number of general-purpose data editing and analysis techniques. It will include a number of hands-on/practice CPU sessions where students will learn how to apply these basic EDA techniques to their own data analysis needs. It will include a Practicum - where students will be required to apply these tools and techniques to the analysis of their own data sets. By the end of the course, students should see a marked improvement in their data analysis/editing capability.

Prerequisites:

This course is designed for subject matter specialists, individuals closest to the data. It requires only a basic understanding of mathematics and statistical techniques.

Students will need to purchase, or have available for their use, the SAS Institute's JMP-IN software with its reference text JMP Start Statistics (approx. total cost: $60) prior to the first class. (NOTE: The commercial (non-student) version of this software costs approximately $1,000.)

Visit www.grad.usda.gov or Call 202/314-3320 to register for STAT2201 today!

Return to top

2001 FCSM Research Conference
November 14-16, 2001

The Federal Committee on Statistical Methodology (FCSM) Research Conference will be held on November 14-16, 2001 at the Key Bridge Marriott in Arlington, Va. The conference will open with a plenary session on "The Interplay Between Research Innovation and Federal Statistical Practice," featuring Stephen Fienberg, Carnegie Mellon University. Kenneth Prewitt of The New School will deliver the Thursday evening banquet address on "Federal Statistics: The 'Right' Balance Between Accountability and Autonomy."

Contributed paper and technical demonstration sessions will provide platforms for discussion and exchange of a wide range of current research and methodological topics relevant to Federal government statistical programs. Topics include: the use of advanced technologies for data collection, processing, and dissemination, data mining, data warehousing and metadata, treatment of missing data, improving coverage and response rates, confidentiality and disclosure, sample design and estimation, cognitive research and usability testing, and data quality. Papers will be made available at the conference and posted to the FCSM web site following the conference.

Conference registration and the $150 fee are due by September 14, 2001. The registration form and periodic updates concerning the conference may be found on the website, www.fcsm.gov. Additional information may be obtained from Stephen H. Cohen, Office of Survey Methods Research, Bureau of Labor Statistics, Room 4915, Washington, DC 20212; Phone: 202-691-7400, Fax: 202-691-7426, and E-mail: fcsm@census.gov.

The 2001 FCSM Research Conference is being sponsored by various Federal government statistical agencies and hosted by the Council of Professional Associations on Federal Statistics.

Return to top

Call for Papers:
New Zealand Conference on Database Integration and Linked Employer-Employee Data

The Department of Labour, Statistics New Zealand, The Treasury, The State Services Commission, and the Ministries of Social Policy, Economic Development and Education are sponsoring a two day international conference on Data Integration (creating new datasets/ databases for research and policy use) to be held on March 21-22, 2002, at Te Papa in Wellington, New Zealand.

To have a paper considered for presentation send your one-page abstract by 30 September 2001 to:

The Data Integration Conference Organising Committee
c/o Julian Wood
Department of Labour
PO Box 3705
Wellington Fax: (04) 915 4040
New Zealand

Email: dileed@lmpg.dol.govt.nz

Abstracts should include a brief description of the paper, the presenter's name, affiliation, mailing address, telephone and fax number and e-mail address if available. Final papers will be due on 15 February 2002. A limited number of papers will be selected for publication in New Zealand Economic Papers.

Return to top

September SIGSTAT

SIGSTAT meetings will resume in September, 2001. The schedule & directions to the meetings can be found on the website:

www.cpcug.org/user/sigstat/

First-time attendees should contact Charlie Hallahan, 20 Announcements2-694-5051, hallahan@ers.usda.gov and leave their name to gain entry to the building.

SIGSTAT Topics for Fall-Winter 2001/2002

September 19, 2001: The Time Series Forecasting System in SAS/ETS

The Time Series Forecasting System (TSFS) is a component of SAS/ETS. that provides a menu-based front-end for forecasting activities. The tasks of creating a date variable, graphing a data series and quickly seeing the results of differencing and/or applying a log transformation, testing for unit roots, examining autocorrelation and partial autocorrelation plots, performing seasonality tests, and, finally, estimating models and producing forecasts are just a mouse click away. In this demo we'll first show how to access the system and generate forecasting models for several variables through automatic selection from a default list of models provided with the TSFS. Next we'll look at overriding the defaults of the TSFS and take control of specifying a model. Finally, we'll show how to add explanatory variables to the forecasting model and control the period of evaluation and fit of a model. Enhancements to the TSFS for Version 8 will be discussed.

October 10, 2001: PROC LOESS in SAS/STAT

The LOESS procedure implements a nonparametric method for estimating regression surfaces. The main features of the LOESS procedures include the fitting of nonparametric models, the use of multidimensional data, multiple dependent variables, both direct and interpolated fitting using kd trees, statistical inference, performs iterative reweighting to provide robust fitting when there are outliers in the data, and supports multiple SCORE statements

November 14, 2001: What's New in SAS 8.2

Release 8.2 of the SAS System is the third release of Version 8. This demo will feature some of the new enhancements that have been added, including updates to the Output Delivery System, Program and Enhanced Editors, SAS/ETS and SAS/STAT.

December 12, 2001: PROC VARMAX in SAS/ETS

The VARMAX procedure estimates multivariate time series models and generates forecasts associated with Vector AutoRegressive and Moving Average process with eXogenous regressors (VARMAX) models. Automatic model selection is available with several information criteria. The VARMAX procedure provides tests to aid in determining the presence of unit roots and cointegration.

January 9, 2002: Output Delivery System (ODS) in SAS

Procedure output became much more flexible with recent versions of SAS. The Output Delivery System (ODS) has been designed to overcome the limitations of traditional SAS output and to make it easy to make new formatting options available to users. ODS is a method of delivering output in a variety of formats and of making the formatted output easy to access.

Return to top

Note From The WSS NEWS Editor

Items for publication in the November 2001 WSS NEWS should be submitted no later than September 25, 2001. E-mail items to Michael Feil at michael.feil@usda.gov. Any items received after this date will appear in the next issue of the WSS NEWS.

Return to top

[an error occurred while processing this directive]

Return to top

WSS Home | Newsletter | WSS Info | Seminars | Courses | Employment | Feedback | Join!

First posted September 6, 2001
Last modified November 03, 2015

http://washstat.org/newsletters/wss0109.shtml

Webmaster: Dan Jacobs, Maryland Sea Grant