Washington Statistical Society Seminars: 2010
Title: Statistical Issues Arising in Jury Discrimination Cases: A Reanalysis of the Data in Berghuis v. Smith
- Speaker: Joseph L. Gastwirth, Department of Statistics, George Washington University
- Date/Time: Friday, January 15, 2010, 11:00-12:00pm
- Location: Monroe Hall, Room 113 (2115 G Street, NW, Washington, DC 20052)
- Sponsor: The George Washington University, Department of Statistics
Abstract:
The measures and statistical tests used in the examination of the demographic composition of venires for under-representation of women or minorities will be described, along with their advantages and limitations. These cases can be brought under either the "equal protections" clause or Sixth Amendment right to a fair trial from a jury that is a fair cross section of the community. While similar statistical evidence plays a major role in both types, the legal burdens of proof on a defendant challenging a jury selection system are somewhat different. In cases involving the "equal protection" clause one needs to provide evidence that is strong enough to support an inference that the state intentionally or purposely discriminated. After illustrating the use of statistical procedures in the Castaneda v. Partida, equal protection case, and the Duren v. Missouri, equal representation case, the talk will focus on the statistical issues in the Berghuis v. Smith case that the U.S. Supreme Court will hear this term. The data in Berghuis is unusual as the racial composition of the venires was not available; so defendant's expert estimated the race of members of the panel from their address. Thus, he demonstrated that the census tracts with a high proportion of African Americans were underrepresented but did not apply a formal statistical test. One judge of the Michigan Supreme Court tried to follow the formulas in Castaneda but made a minor error. The correct large sample theory of the statistic used by the expert will be presented and applied to the data. It provides stronger evidence of the underrepresentation of African Americans on the venires than is in the record. On the other hand, it will be seen that adjusting the Census data to exclude individuals legitimately excused from jury duty due to child care obligations might well "explain" much of the shortfall. Because neither party presented detailed data enabling one to examine the effect of the excusal process on the composition of the venires, a definitive statistical analysis of that issue is not possible. Note: The talk is based on joint work with Prof. Qing Pan.
Return to topTitle: Survey Redesign Panel
- Date/Time: January 20, 2010, 1:00 to 3:00
- Location: Bureau of Labor Statistics, Conference Center Room 1. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station.
- Presentation material:
Redesign of the Consumer Expenditure Surveys (pdf, ~68kb)
Overview of the National Crime Victimization Survey Redesign (pdf, ~2.3mb)
Redesigning the National Household Education Survey (NHES) (pdf, ~148kb)
Redesign Considerations for the National Survey on Drug Use and Health (NSDUH) (pdf, ~276kb)
Re-engineered SIPP - Progress Update (pdf, ~38okb)
Abstract:
To learn about the survey redesign process from a variety of large scale surveys, we will have presentations from five major government surveys discuss their redesign process. Panelists will represent the Consumer Expenditure Survey, the National Household Education Survey, the National Crime and Victimization Survey, the Survey of Income and Program Participation, and the National Survey on Drug Use and Health.
Panelists will give a brief overview of their redesign process, including discussion of redesign motivations, challenges faced, testing done or planned, evaluation process and current status. Following the presentations, the audience and panelists will participate in an in-depth question and answer session focused the survey redesign process.
Return to topTitle: The Challenges of Conducting the Census 2010
- Speaker: Dr. Robert Groves, Director, U.S. Census Bureau
- Date/Time: Thursday, January 21st, 2010, 9:30-11am
- Location: Pew Research Center 1615 L Street, NW, Suite 700 Washington, DC
- Presentation material:
Transcipt
Abstract:
The Pew Research Center, DC-AAPOR, and the Washington Statistical Society are pleased to announce an event discussing the 2010 Census, how it will be conducted and what it means for researchers. Census Director Robert Groves describes methodological and other challenges the bureau faces. Connie Citro, who has directed National Academies panels that have evaluated the Census; Jeff Passel, a leading expert on demography and a former Census Bureau researcher; and Joe Salvo, who heads up the population office for the nation's largest city, respond from their unique perspectives. The event will be moderated by Pew Research Center Director of Survey Research Scott Keeter.
Seatingis limited. Since an RSVP is required, please visit the DC-AAPOR website: http://www.dc-aapor.org/upcomingevents.php to do so.
Return to topTitle: Developing a Data Analysis System for Categorical Survey Data
- Speaker: Phillip S. Kott, Senior Research Statistician, RTI International
- Discussant: TBD
- Chair: Brian Meekins, BLS
- Date/Time: Thursday, January 21, 2010 / 12:30 - 2:00 p.m.
- Location: Bureau of Labor Statistics Conference Center. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station.
- Video Link: Westat,Rockville Offices. On a trial basis, Westat is opening up its conference center for watching the lecture remotely. Reservation required. Call Fran Winter,301-294-4419.
- Sponsor: Methodology Program, WSS
Abstract:
Many government statistical agencies are either thinking about developing a data analysis system (DAS) to display interactively the results of their surveys or already have one in place. A DAS can be used to generate tables at the user's request and may even be able to conduct more sophisticated (but still limited) statistical analyses. Before constructing such a system, there are a number of questions the agency must address. Two in particular are of concern here for categorical data derived from a sample survey: How is the anonymity of the survey respondent to be protected given that the same user can make multiple requests of the system; and should public users be protected from the release of estimates with overly large coverage intervals? We argue that the users themselves can decide whether estimates are accurate enough for their purposes, but to do that there need to be well-behaved coverage intervals for those estimates. It turns out that the rule needed to construct a good coverage interval for the estimated target is very similar to that needed to assure data confidentiality.
Return to topTitle: Data, Information and Interpretation in Assessing the Sustainability of the Nation's Forests
- Speaker: Guy Robertson, Ph.D., Sustainability Program Lead, U.S. Forest Service
- Chair: Mike Fleming
- Date/time: Wednesday, January 27, 2010 / 12:30 - 1:30 p.m.
- Location: Bureau of Labor Statistics Conference Center. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station.
- Sponsor: WSS Agriculture and Natural Resources Section
- Presentation material:
Slides (pdf, ~628kb)
Abstract:
The Montreal Process Criteria and Indicators for Forest Sustainability (MP C&I) provide the foundation for the 2010 National Report on Sustainable Forests, a major Forest Service reporting effort currently underway. The processes through which the MP C&I were derived and applied as well as the specific content of selected indicators will be the focus of this talk. The MP C&I include 64 indicators spanning ecological, economic and social dimensions associated with the sustainability of forest ecosystems, and they entail a host of technical and conceptual issues related to data gathering, reporting and interpretation. Moreover, the underlying concept of sustainability presents various challenges both when considered generally and within the context of specific indicators. These topics and others will be discussed within the general context of presenting the overall findings of the 2010 Report.
Point of contact e-mail: grobertson02@fs.fed.us
Return to topTitle: Novel Analytic Methods To Estimate Physical Activity From An Accelerometer
- Speaker: John Staudenmayer, PhD, University of Massachusetts, Amherst
- Time/Place: February 3, 2010 @ 11:00 a.m. / Conference Room G, 6130 (EPN) Executive Boulevard, Rockville, Maryland
- Sponsor: NCI Cancer Prevention Fellowship Program, the Division of Cancer Control and Population Science and the Washington Statistical Society Section on Public Health and Biostatistics
Abstract:
Dr. Staudenmayer'sinterests are in measurement error and nonparametric estimation methods. He will focus on the application of these methods towards obtaining estimates of usual level of physical activity from accelerometer measurements. He will demonstrate ways in which the new analytic methods provide more accurate estimates of physical activity than established methods. An important take home message is that the use of the new methods require different data collection techniques.
SecurityInfo: An NIH badge is required to freely enter the building unescorted. Visitors with photo ID only will need to be escorted by an NIH employee. I will be around to help out in this capacity.
Map: http://dceg.cancer.gov/images/localmap.gif
For Additional Information
Cancer Prevention Fellowship Program - (301) 496-8640
http://www.cancer.gov/prevention/pob
If you are a person with a disability and require any assistive device, services or other reasonable accommodation to participate in this activity, please contact the Cancer Prevention Fellowship Program at 301-496-8640 at least one week in advance of the lecture date to discuss your accommodation needs.
Return to topTitle: Defining Success in Oncology Drug Development
- Speakers:
Richard Pazdur, M.D., Director, Office of Oncology Drug Products,
Rajeshwari Sridhara, M.D., Acting Director, Division of Biometrics V, OB, CDER, Food and Drug Administration - Discussant: TBD
- Chair: David Judkins, Westat
- Date/Time: Wednesday, February 17, 12:30-2:00 p.m.
- Location: Bureau of Labor Statistics, Conference Center. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station.
- Video Link: Westat,Rockville Offices. On a trial basis, Westat is opening up its conference center for watching the lecture remotely. Reservation required. Call Fran Winter, 301-294-4419.
- Sponsor: Methodology Program, WSS
Abstract:
To be announced in February newsletter, but touching generally on issues of interest to statisticians from the perspectives of physicians.
Return to topTitle: Introduction to Online Mapping Using Google Earth
- Speaker: Anders Longthorne
National Center for Statistics and Analysis
National Highway Traffic Safety Administration - Chair: Mel Kollander
- Date/time: Wednesday, February 24, 2010 / 12:30 - 1:30 p.m.
- Location: Bureau of Labor Statistics Conference Center. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station.
- Sponsor: Agriculture and Natural Resources and Data Collection Methods
- Presentation material:
Slides (pdf, ~1.2M)
Getting Started with Google Online Maps (pdf, ~48kb)
Abstract:
The National Highway Traffic Safety Administration's (NHTSA) Fatality Analysis Reporting System (FARS) contains data on a census of fatal motor vehicle traffic crashes within the USA. With the release of the 2008 FARS data, NHTSA added fatal crash maps to its State Traffic Safety Information (STSI) website.
These maps were produced by generating Keyhole Markup Language (KML - this is an XML-based language that is used to express geographic information) files from the FARS data and then viewing these files via the Google Earth browser plug-in. This talk will primarily focus on the use of KML files in combination with the Google Earth software, in order to produce cost effective applications to spatially display geocoded databases.
Point of contact e-mail: anders.longthorne@dot.gov
Return to topTitle: Modeling Without Measurement: Why Risk Analyses Fail
- Speaker: Douglas A. Samuelson, Ph.D., President and Chief Scientist, InfoLogix, Inc
- Time: 11:00 a.m. Colloquium Talk
- Date: April 30, 2010
- Location: Engineering Building, Room 1107, Fairfax Campus, George Mason University, 4400 University Drive, Fairfax, VA 22030
- Sponsor: George Mason University CDS/CCDS/Statistics Colloquium
Abstract:
Recent events including terrorist attacks, Hurricane Katrina and the financial crisis have focused much attention on risk management. Did risk analysis itself fail? Were the analyses right, but not heeded? Were the events beyond anyone?s ability to predict? We examine a number of factors that contributed to these widely publicized failures:
- Unduly limiting assumptions in risk models. In particular, crises cause not only excursions of data beyond what we encountered when estimating the model, but also changes in the relationships among variables.
- Underestimation of the probability and consequences of rare events. Standard methods for estimating and reporting uncertainty tend to overstate precision, promoting overconfidence when conditions remain stable and excessive pessimism about all modeling when things change.
- Over-reliance on markets with respect to longer-term events. Markets are well-known to be myopic, even with respect to longer-term futures with incomplete information.
- Over-specialization and resistance among disciplines. We all have our preconceptions. Well-educated professionals tend to have more of them and to be more reluctant to change them, especially when they have large personal investments (not necessarily financial) in a way of doing things.
- Insufficient empiricism in assessing emplrical methods. Recent studies have shown that many popular approaches to improving decision-making are better at making the decision-makers feel good than at improving the quality of decisions.
We will discuss these findings and observations and offer some suggestions for improvement.
The CDS/CCDS/Statistics Colloquium Series is co-sponsoring this seminar that has been organized as a part of the C4I Seminars/SEOR Seminar Series/CS Seminar Series.
Return to topTitle: A Combinatorial Approach To The Interpolation Method And Scaling Limits In Sparse Random Graphs
- Speaker: Prof. David Gamarnik, MIT Sloan School of Management, Massachusetts Institute of Technology
- Date/Time: Tuesday, March 2, 2010, 3:30pm (Note Time and Room Change)
- Location: Room 1313, Math Bldg, University of Maryland College Park (directions).
- Sponsor: University of Maryland, Statistics Program (seminar updates).
Abstract:
We establish the existence of scaling limits for several combinatorial optimization models on Erdos-Renyi and sparse random regular graphs. For a variety of models, including maximum independent sets, MAX-CUT, coloring and K-SAT, we prove that the optimal value appropriately rescaled, converges to a limTITLE: A combinatorial approach to the interpolation method and scaling limits in sparse random graphs. ABSTRACT: We establish the existence of scaling limits for several combinatorial optimization models on Erdos-Renyi and sparse random regular graphs. For a variety of models, including maximum independent sets, MAX-CUT, coloring and K-SAT, we prove that the optimal value appropriately rescaled, converges to a limit with probability one, as the size of the underlying graph diverges to infinity. For example, as a special case we prove that the size of a largest independent set in these graphs, normalized by the number of nodes converges to a limit with probability one, thus resolving an open problem. Our approach is based on developing a simple combinatorial approach to an interpolation method developed recently in the statistical physics literature. Among other things, the interpolation method was used to prove the existence of the so-called free energy limits for several spin glass models including Viana-Bray and random K-SAT models. Our simpler combinatorial approach allows us to work with the zero temperature case (optimization) directly and extend the approach to many other models. Additionally, using our approach, we establish the large deviations principle for the satisfiability property for constraint satisfaction problems such as coloring, K-SAT and NAE(Not-All-Equal)-K-SAT. The talk will be completely self-contained. No background on random graph theory/statistical physics is necessary. Joint work with Mohsen Bayati and Prasad Tetali it with probability one, as the size of the underlying graph diverges to infinity. For example, as a special case we prove that the size of a largest independent set in these graphs, normalized by the number of nodes converges to a limit with probability one, thus resolving an open problem. Our approach is based on developing a simple combinatorial approach to an interpolation method developed recently in the statistical physics literature. Among other things, the interpolation method was used to prove the existence of the so-called free energy limits for several spin glass models including Viana-Bray and random K-SAT models. Our simpler combinatorial approach allows us to work with the zero temperature case (optimization) directly and extend the approach to many other models. Additionally, using our approach, we establish the large deviations principle for the satisfiability property for constraint satisfaction problems such as coloring, K-SAT and NAE(Not-All-Equal)-K-SAT. The talk will be completely self-contained. No background on random graph theory/statistical physics is necessary.Joint work with Mohsen Bayati and Prasad Tetali
Return to topTitle: Multi- and Matrix-variate Times Series & Graphical Models
- Speaker: Mike West, Department of Statistical Science, Duke University
- Time: Friday, March 5, 2010, 4:00-5:00 pm
- Location: The George Washington University, Duques
553 (2201 G Street, NW)
(Followed by wine & cheese reception) - Sponsor: The George Washington University, The Institute for Integrating Statistics in Decision Sciences and Departments of Decision Sciences and Statistics
Abstract:
I will review some recent and current developments in Bayesian modelling of multi- and matrix-variate time series, all involving the integration of graphical modelling ideas and methods with dynamic models. This includes graphical models to constrain multivariate stochastic volatility models in financial applications and extensions to matrix-variate times series with economic examples. Stochastic simulation and search for Bayesian computations in these models are key and will be discussed, as will some current research frontiers. The talk covers developments from projects in collaborations with current and past students Carlos Carvalho, Craig Reeson and Hao Wang.
Return to topTitle: Special Topics in Propensity Scoring
- Speaker: David Judkins, Senior Scientist, Westat
- Chair: Brian Meekins, BLS, WSS Methodology Section Chair
- Date/time: Wednesday, March 10, 2010 / 12:30-2:00 p.m.
- Location: Bureau of Labor Statistics, Conference Center. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station.
- Video Link: Westat, Rockville Offices. On a trial basis, Westat is opening up its conference center for watching the lecture remotely. Reservation required. Call Fran Winter, 301-294-4419.
- Sponsor: Methodology Section, WSS
Abstract:
Propensity scoring is probably the preferred technique for causal inferences from observational studies when outcome distributions do not follow any of the standard parametric forms. It can also be a labor-saving device when the number of potential outcome variables is larger than the number of putative causal agents. It can be combined with outcome modeling for doubly robust inference. The set of conditions for when to prefer propensity scoring is the first topic of this presentation. As part of this topic, there will be a report on an application where both propensity scoring and traditional ANCOVA methods were applied. Then, some special application topics will be discussed. If the covariate space is very large, how rich should one allow the propensity models to become? How to test for balance (the adequacy of covariate control)? How to remedy balance failure? Finally, there will be a discussion of how to use propensity scoring to reconcile dose-response analysis with temporal trend analysis when both are available for a program evaluation.
Return to topTitle: Small Area Estimation Using Linking Models with Corrections for Specification Errors
- Speaker: P. A. V. B. Swamy, Federal Reserve Board (retired)
- Time: 10:30 a.m. Refreshments, 10:45 a.m. Colloquium Talk
- Date: March 19, 2010
- Location: Research 1, Room 301, Fairfax Campus, George Mason University, 4400 University Drive, Fairfax, VA 22030
- Sponsor: George Mason University CDS/CCDS/Statistics Colloquium
Abstract:
When a sample estimator for a domain is based on a very small sample, attempts were made to improve the precision of the estimator using some linking models. This is known as small area estimation. Sufficient care should be exercised in small area estimation because misspecifications of linking models used in this estimation can introduce specification biases into the sample estimators. In practice, different types of misspecifications occur: (i) Linking models have incorrect functional forms, (ii) certain explanatory variables that belong in these models are omitted and the effects of this omission are serious, (iii) data on the variables of these models contain measurement errors, and (iv) the explanatory variables of these models are correlated with their error terms. Substantial improvements in small-area sample estimators are achieved when the linking models used to improve the precision of these estimators are corrected for the specification errors. In this talk, how such corrections can be made will be shown. A method of simultaneously estimating the true value and sampling and non-sampling error components of a sample estimator will be shown to yield consistent estimators. It will also be shown that a twostep method of estimating the components of sample estimators leads to inconsistent estimators. Return to topTopic: A Modified Adaptive Accept-Reject Algorithm for Univariate Densities with Bounded Support
- Speaker: Dr. Carsten Botts, Williams College, Department of Mathematics and Statistics
- Time: 2:15 PM, Friday, March 19, 2010. Refreshments will be served after the talk.
- Location: Conference Room 326 St. Mary's Hall, 3700 Reservoir Road, NW (between 37th and 38th Streets, NW). Building #16 on Campus Map at http://maps.georgetown.edu/index.cfm?Action=View&MapID=3 or Google Maps: 3700 Reservoir Rd NW, Washington, DC 2000
- Sponsor: Georgetown University, Department of Mathematics and Statistics
Abstract:
The need to simulate from a univariate density arises in several settings, particularly in Bayesian analysis. An especially effcient algorithm which can be used to sample from a univariate density, fX , is the adaptive accept-reject algorithm. To implement the adaptive accept-reject algorithm, the user has to envelope T ° fx, where T is some transformation such that the density g(x) α T-1(α + βx) is easy to sample from. Successfully enveloping T ° fx, however, requires that the user identify the number and location of T ° fx's inflection points. This is not always a trivial task. In this paper we propose an adaptive accept-reject algorithm which relieves the user of precisely identifying the location ofT ° fx's inflection points. This new algorithm is shown to be effcient and can be used to sample from any density such that its support is bounded and its log is three-times differentiable. Return to topTitle: Creating and Evaluating a New Method for Collecting Survey Data via the Internet
- Speaker: Dr. Jon Krosnick, Stanford University
- When: March 26, 2010
- Time: 10:30am
- Where: Room 375, National Science Foundation,4201 Wilson Blvd. Arlington, VA. RSVP is REQUIRED for NSF visitors. To RSVP, contact: Ron Bello at rbello@nsf.gov or 703-292-8780
Abstract:
Collecting high-quality data from a representative sample of Americans using traditional face-to-face and telephone surveys has become increasingly difficult and expensive. Dr. Jon Krosnick's NSF-funded project recruited a scientifically selected sample of survey respondents through:
- face-to-face contacts
- offers of free computers, high-speed Internet access, and cash in exchange for completing monthly online questionnaires
Come hear about:
- the many interesting events that occurred
- lessons learned in the recruitment process
- the quality of the data collected, especially compared with other major national surveys
Title: A Unified Approach to Variations of Ranked Set Sampling
- Speaker: Prof. Kaushik Ghosh, University of Nevada, Las Vegas
- Date/Time: Thursday, April 1, 2010, 3:30pm
- Location: Room 1313, Math Bldg, University of Maryland College Park
- Sponsor: University of Maryland, Statistics Program
Abstract:
This talk will develop a general theory of inference using data collected from several variations of ranked set sampling. Such variations include balanced and unbalanced ranked set sampling, balanced and unbalanced k-tuple ranked set sampling, nomination sampling, simple random sampling, as well as a combination of them. Methods of estimating the underlying distribution function as well as its functionals, and asymptotic properties of the resulting estimators will be discussed. The results so obtained will be used to develop nonparametric procedures for one- and two-sample problems. The talk will conclude with a study of the small-sample properties of the estimators and an illustrative example.
Please check for seminar updates at: http://www.math.umd.edu/statistics/seminar.html
Directions to Campus: http://www.math.umd.edu/department/campusmap.shtml
Return to topTitle: ROC Curves For Biomarkers Affected By Measurement Error
- Speakers:
Neil Perkins, Ph.D. Staff Scientist, Division of Statistics, Epidemiology and Prevention Research
Eunice Kennedy Shriver National Institute of Child Health and Human Development - Date/Time: Friday, April 2nd, 2010 / 11:00a.m. - 12:30 p.m.
- Location: Executive Plaza North, 6130 Executive Boulevard, Conference Room H, Rockville MD. Photo ID and sign-in required.
- Metro: Get offat the White Flint stop on the Red Line, and take Nicholson Lane to Executive Blvd. Make a Right and continue, crossing Old Georgetown Rd. When you reach what is more or less tree foliage and the road begins to bend to the right you enter the Executive Plaza Complex parking lot. EPN will be the right most of the two twin buildings.
- Map: http://dceg.cancer.gov/images/localmap.gif
- Sponsor: Public Health and Biostatistics Section, WSS and the NCI
Abstract:
Biomarkers are of increasing importance in clinical research and epidemiology. The receiver operating characteristic curve is a statistical tool that helps investigators identify biomarkers which can effectively, for a given criteria, distinguish between two populations, often those with or without a particular disease. These methods are largely well established for true biomarker levels, however it is rarely the case in which biomarkers are measured without error. In this talk, methods to estimate the ROC curve and its summary indices while adjusting or accounting for various types of measurement error will be presented.
Return to topTitle: SCRATCHing my Head — From Tokens to Virtual Manipulatives: A 12,000 Year History of Numbers and Mathematical Concepts
- Speaker: Behrouz Aghevli, Ph.D., a.k.a. Dr. Super, World Bank
- Time: 10:30 a.m. Refreshments, 10:45 a.m. Colloquium Talk
- Date: April 2, 2010
- Location: Research 1, Room 301, Fairfax Campus, George Mason University, 4400 University Drive, Fairfax, VA 22030
- Sponsor: George Mason University CDS/CCDS/Statistics Colloquium
Abstract:
On-line games and activities are fast becoming one of the best ways for young people to learn math and science. In his talk "Scratch-ing my Head," Dr. Super traces the 12,000 year history of numbers and mathematical concepts! Dr. Super's whirlwind tour shows how sedentary living in the Iranian Plateau and Mesopotamian Regions 12,000 years ago gave rise to "tokens," and how the use of these tokens lead to the invention of writing. Tokens were the first math manipulative?the rocks upon which the foundation of our civilization was built. Fast-forward to the mid-nineteenth century. Civilizations continued to use math manipulatives such as the abacus, and Froebel, who invented Kindergarten, called manipulatives his "gifts." Dr. Super also highlights the works Montessori and Piaget, who reintroduced manipulatives to teach math concepts. While examining more recent uses of math manipulatives by Sultan Dienes, Dale Seymour and others, Dr. Super introduces a few of his own manipulatives. He then delves into "virtual manipulatives," and tools to create games and activities using virtual manipulatives. He also covers some of the original work conducted at GMU, the National Library of Virtual Manipulatives, and NCTM's Imagination site. As the tour concludes, Dr. Super highlights "Scratch," which is a simple animation and game-making tool from MIT's Life Long Kindergarten Group. He demonstrates several Scratch games and activities which illustrate why the tool is so attractive to educators and such a popular way for young people to learn math and science skills.
Return to topTitle: Human Resources in Science and Technology: Surveys, Data, and Indicators from the National Science Foundation
- Speaker: Nirmala Kannankutty, PhD, National Science Foundation, Science Resources Statistics, http://www.nsf.gov/statistics/
- Time: Tuesday, April 6, 2010, 1-3pm EST, http://www.amstat.org/sections/SRMS/webinar.cfm
- Sponsors: ASA Survey Research Methods Section &Amp; AAPOR
Abstract:
The Division of Science Resources Statistics (SRS) is a federal statistical agency housed at the National Science Foundation (NSF). SRS's role within NSF is to "provide a central clearinghouse for the collection, interpretation, and analysis of data on scientific and engineering resources, and to provide a source of information for policy formulation by other agencies of the Federal Government…" Within this mandate SRS is involved in collecting and disseminating information on R&D expenditures and activities and on human capital issues. The United States is unique among major industrialized nations in that it has directly invested in collecting detailed data from a variety of sources on the entire science and engineering pipeline. Each of the data sources came about from U.S. federal administrative needs. The sources have evolved into important elements for the study of higher education and the scientific workforce. In this webinar, these surveys and data sources are described. Key indicators regarding trends in U.S. science and engineering degree production, enrollments, and workforce are defined and described. The Science and Engineering Indicators: 2010 and Women, Minorities and Persons with Disabilities in Science and Engineering reports will be used as examples for these indicators. At the end of the webinar participants should be aware of data sources and how data are collected, indicators and reports from the NSF, and where to find more information from the NSF.
For each webinar, participants register for a modest fee. Fees may vary from webinar to webinar depending on the length of the presentation and expected audience. Each registration is allowed one web connection and one audio connection. The section encourages multiple persons to view each registered connection.
If you have any questions, please feel free to contact Rick Peterson at the ASA office using the below information.
Rick Peterson
Education Programs Associate
American Statistical Association
732 North Washington Street
Alexandria, VA 22153
(703) 684-1221 ext. 1864
FAX: (703) 684-3768
rick@amstat.org
www.amstat.org
Title: Overview of the Department of Energy's Office of Science
- Speaker: Laura J. Biven, Ph.D., Office of Science, US Department of Energy
- Date/Time: 3:35pm, Tuesday, April 6, 2010
- Location: Bentley Lounge, Gray Hall 130, American University
- Directions: Metro RED line to Tenleytown-AU. AU shuttle bus stop is next to the station. Please see campus map on http://www.american.edu/media/directions.cfm for more details
- Contact: Stacey Lucien, 202-885-3124, mathstat@american.edu
- Sponsor: American University Department of Mathematics and Statistics Colloquium
Abstract:
The Office of Science in the Department of Energy (DOE) is one of the nation's largest supporters of peer-reviewed basic research. We support over 25,000 researchers at more than 300 institutions including all 17 DOE national laboratories. Our budget of nearly $5 Billion supports science for discovery, science for national need, and our national scientific user facilities-- 21st century tools of science. I will begin with an overview of the Office of Science and how we are integrated into the DOE, both organizationally and in terms of our mission. I will highlight the important roles that the academic and research community play in generating new areas of inquiry, ensuring that we fund high-quality projects, and advising us on scientific opportunities and strategic planning. Last, but not least, I hope to excite you with an abridged survey of the research we support and opportunities to become a part of our efforts.
Return to topTitle: SAS/STAT 9.22: The Next Generation
- Chair: Brian Meekins, BLS & WSS Methodology Section Chair
- Speaker: Robert N. Rodriguez, SAS Institute
Robert N. Rodriguez is Senior Director of Statistical R&D at SAS Institute. Bob joined SAS in 1983 and received his Ph.D. in statistics from the University of North Carolina at Chapel Hill in 1977. He is a Fellow of the American Statistical Association and is a candidate for ASA president-elect in the 2010 election. - Date/Time: Wednesday, April 21, 12:30-2:00pm
- Location: Bureau of Labor Statistics, Conference Center. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station.
- Video Link: Westat, Rockville Offices
On a trial basis, Westat is opening up its conference center for watching the lecture remotely. Reservation required. Call Fran Winter, 301-294-4419. - Sponsor: Methodology Section, WSS
Abstract:
This presentation will begin with a review of SAS/STAT 9.2 and then describe new statistical functionality in SAS/STAT 9.22, the next major release of SAS/STAT, which will be coupled with the third maintenance release of SAS 9.2. In SAS/STAT 9.22, the new SURVEYPHREG procedure fits Cox regression models to survey data. Many linear models procedures are now equipped with comparable facilities for postfitting inference, including recent techniques for testing complex research hypotheses. The new PLM procedure performs postfitting inference with model fit information saved from a number of modeling procedures. The EFFECT statement, which defines a richer family of linear models, is implemented in a large number of procedures. The EFFECTPLOT statement, which uses ODS Graphics to visualize the influence of model effects, is available in the SURVEYLOGISTIC, GENMOD, and LOGISTIC procedures. The GENMOD procedure provides zero-inflated negative binomial models as well as zero-inflated Poisson models for dealing with overdispersion, and it provides exact Poisson regression. The CALIS procedure provides more powerful features for structural equations modeling.
Return to topTitle: Is it Culturally Ethical? Human Rights Challenges in International Survey Research
- Speaker: Safaa Amer, Senior Statistician, NORC at the University of Chicago
- Discussant: Mary Gray, Professor of Mathematics and Statistics, American University
- Chair: Michael P. Cohen, Senior Consultant, NORC at the University of Chicago
- Date/Time: Monday, April 26, 2010 / 12:30 - 2:00
p.m.
(Rescheduled from February 5, 2010) - Location: PLEASE NOTICE CHANGE FROM USUAL
LOCATION.
Pew Research Center 1615 L Street, NW, Suite 700, Washington, DC 20036-5621 - RSVP Instructions: If you are planning on attending this seminar, please email Carol Joyce Blumberg at carol.blumberg@eia.doe.gov by Monday, April 19.
- Sponsors: WSS Data Collection Methods, WSS Human Rights, DC-AAPOR, and Capital Area Social Psychological Association
- Presentation material:
Is It Culturally Ethical? (Amer, pdf ~376kb)
Ethical considerations for the role of a statistician in international survey research (Gray, pdf ~8mb)
Abstract:
In a global research world the need continues to arise for an umbrella of standards to shield human participants' rights and research ethics. When a team of survey researchers works on an international project where a mix of cultures, languages, or regions is handled, careful eyes should look at ethical and human rights issues. Although the basics of research ethics tend to seem simple, cultural considerations, differences of laws and regulations, level of vulnerability of the population, and many other factors add a twist to the situation. This calls for special handling due to inherent differences in the population of interest and their surrounding environment. This talk presents examples from around the world of challenges faced in social and behavioral projects; use and abuse of cross-cultural ethics; some lessons learned, and future considerations.
Important Note: Videoconferencing will not be available for this seminar.
Forfurther information contact Carol Joyce Blumberg at carol.blumberg@eia.doe.gov or (202) 586-6565
Title: A Stochastic Search Approach to Solving the Cell Suppression Problem for 3-Dimensional Hierarchical Tables
- Speaker: Matt Fetter, National Agricultural Statistics Service
- Chair: Mel Kollander
- Date/time: Wednesday, April 28, 2010 / 12:30 - 1:30 p.m.
- Location: Bureau of Labor Statistics, Conference Center. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station.
- Sponsor: Agriculture and Natural Resources
Abstract:
Cell suppression is one method that is commonly used to reduce disclosure risk when data are published in hierarchical tables. A form of optimality is achieved for 2-dimensional tables by formulating the cell suppression problem as a minimum cost flow problem. There are issues with this approach in general, and for its application to 3-dimensional tables in particular. First, cell suppression is fundamentally an integer programming problem with a non-smooth cost function. Secondly, the minimum cost flow approach is not directly applicable to 3-dimensional tables. A stochastic search approach is presented that is guaranteed to generate closed paths in 3-dimensional tables. Although no claim of optimality can be made, this method is capable of finding good solutions significantly faster than a blind random search.
Point of contact e-mail: Matt_Fetter@nass.usda.gov
Return to topTitle: A Journey to the Center of the Earth
- Speaker: Prof. Ping Ma, University of Illinois at Urbana-Champaign
- Date/Time: Thursday, April 29, 2010, 3:30pm
- Location: Room 1313, Math Bldg, University of Maryland College Park
- Sponsor: University of Maryland, Statistics Program
Abstract:
At a depth of 2890 km, the core-mantle boundary (CMB) separates turbulent flow of liquid metals in the outer core from slowly convecting, highly viscous mantle silicates. The CMB marks the most dramatic change in dynamic processes and material properties in our planet, and accurate images of the structure at or near the CMB--over large areas--are crucially important for our understanding of present day geodynamical processes and the thermo-chemical structure and history of the mantle and mantle-core system. In addition to mapping the CMB we need to know if other structures exist directly above or below it, what they look like, and what they mean (in terms of physical and chemical material properties and geodynamical processes). Detection, imaging, (multi-scale) characterization, and understanding of structure (e.g., interfaces) in this remote region have been--and are likely to remain--a frontier in cross-disciplinary geophysics research. I will discuss the statistical problems and challenges in imaging the CMB through generalized Radon transform.
Please check for seminar updates at: http://www.math.umd.edu/statistics/seminar.html
Directions to Campus: http://www.math.umd.edu/department/campusmap.shtml
Return to topTitle: Prediction of Ordered Random Effects in a Simple Small Area Model
- Speaker: Dr. Yaakov Malinovsky, NICHD, Rockville, MD
- Date/Time: May 6, 2010, 2010, 3:30pm
- Location: Room 1313, Math Bldg, University of Maryland College Park
- Sponsor: University of Maryland, Statistics Program
Abstract:
Prediction of a vector of ordered parameters, or part of it, arises naturally in the context of Small Area Estimation (SAE). For example, one may want to estimate the parameters associated with the top ten areas, the best or worst area, or a certain percentile. We use a simple SAE model to show that estimation of ordered parameters by the corresponding ordered estimates of each area separately does not yield good results with respect to Mean Square Error. Shrinkage-type predictors, with an appropriate amount of shrinkage for the particular problem of ordered parameters, are considerably better, and their performance is close to that of the optimal predictors, which cannot in general be computed explicitly.
Please check for seminar updates at: http://www.math.umd.edu/statistics/seminar.html
Directions to Campus: http://www.math.umd.edu/department/campusmap.shtml
Return to topTitle: Use of Data Mining Methods For Survey Data
- Speakers: Thomas Jacob, Jaki McCarthy, and Dan Beckler National Agricultural Statistics Service
- Chair: Mel Kollander
- Date/time: Wednesday, May 12, 2010 / 12:30 - 1:30 p.m.
- Location: Bureau of Labor Statistics Conference Center. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station.
- Sponsor: Agriculture and Natural Resources
Abstract:
Data mining is the process of discovering meaningful correlations, patterns and trends in large databases. Data mining methods can be very effective at improving survey data quality during data collection, editing, and at summarization. This presentation reports the results of applying some data mining methods to survey data. The methods include anomaly detection, exploratory analysis, clustering, and data visualization.
Point of contact e-mail: Thomas_Jacob@nass.usda.gov
Return to topTitle: Analysis of Case-Control Studies in Genetic Epidemiology
- Speaker: Nilanjan Chatterjee, Ph.D., Senior Investigator and Chief, Biostatistics Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute
- Date/Time: Friday, May 21st, 11:00am-12:30
- Location: Executive Plaza North, 6130 Executive Boulevard, Conference Room(s) C,D,E & F, Rockville MD. NIH badge required for un-escorted entry. All others must sign in and need Photo ID and an NIH employee to escort them in. Someone will be around to assist in this capacity until the lecture begins.
- Metro: Get off at the White Flint stop on the red line, and take Nicholson lane to Executive blvd. Make a Right and continue, crossing Old Georgetown road. When you reach what is more or less tree foliage and the road begins to bend to the right, make a left into the executive plaza complex. EPN will be the right most of the two twin buildings.
- Map: http://dceg.cancer.gov/images/localmap.gif
- Sponsor: Public Health and Biostatistics Section, WSS and the NCI
Biographical Sketch:
Nilanjan Chatterjee, Chief and Senior Investigator of the Biostatistics Branch of the Division of Cancer Epidemiology and Genetics, National Cancer Institute, was named to be the winner of the Mortimer Spiegelman Award for 2010. This award, presented annually since 1970 by the American Public Heatlh Association (APHA) statistics section, recognizes outstanding contribution by a statistician under age 40 to public health. Demographer, actuary, and biostatistician Mortimer Spiegelman (1901-1969) made exceptional contributions to public health statistics. The award serves three purposes: to honor the outstanding achievements of both the recipient and Spiegelman, to encourage further involvement in public health of the finest young statisticians, and to increase awareness of APHA and the Statistics Section in the academic statistical community. The distinction associated with the award has increased over time with the extraordinary continuing accomplishments of its recipients. Dr. Chatterjee received his PhD in Statistics from the University of Washington, Seattle in 1999. His research focuses on statistical methods for modern genetic and molecular epidemiologic studies. He also actively collaborates in design and analysis of a variety of major cancer epidemiologic studies at NCI.
Return to topTitle: The Concept of Human Exposure Assessment
- Speaker: Mel Kollander
- Chair: Barry Nussbaum
- Date/time: Wednesday, May 26, 2010 / 12:30 - 1:30 p.m.
- Location: Bureau of Labor Statistics Conference Center. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station.
- Sponsor: Agriculture and Natural Resources
Abstract:
The U.S. Environmental Protection Agency recognized that exposure to environmental pollutants can lead to adverse health effects. Exposures to a given chemical can be occupational and/or non-occupational, and they can take place in indoor and outdoor microenvironments. Exposures can occur via different routes, such as inhaled air, drinking water, ingested food and dermal contact, that need to be added together to estimate a subject's total exposure, that determines their health effect. EPA scientists and survey experts by working together were the leaders in developing scientifically defensible methodologies to characterize human exposures.
A number of landmark studies were sponsored by EPA and for the first time quantitative evidence of exposures to pollutants such as Carbon Monoxide (CO) were developed and used to support regulatory development. The presentation will include descriptions of how personal exposures are measured using personal monitors, collections of body fluids, the survey techniques used, and examples of results from such studies.
Point of contact e-mail: mellk@erols.com
Return to topTitle: Trends in Income Nonresponse Over Two Decades
- Organizer: David Judkins, Westat, WSS Methodology Program Chair
- Chair: Brian Meekins, BLS, WSS Methodology Section Chair
- Speaker: Ting Yan, NORC at the University of Chicago
- Date/Time: Tuesday, June 8, 2010, 12:30-2:00pm
- Location: Bureau of Labor Statistics, Conference Center. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station.
- Sponsor: Methodology Section, WSS
Abstract:
Survey data on personal and household income is usually associated with a large amount of item nonresponse. This talk is adapted from a recent published article by Yan, Curtin, and Jans (2010) that focuses on trends in income nonresponse over the past two decades in the Surveys of Consumers (SCA) conducted in the U.S. SCA asks respondents first to report their income in dollar amounts with an open-ended question. Those who do not provide an answer are followed up with a closed-ended question with income brackets. Analyses indicate that missing data on income has decreased over time, and the decline is related to respondents' overall willingness to participate in the survey and to answer survey questions once in the survey. The results suggest that, for questions on household income, there exists a trade-off between unit and item nonresponse, which may have implications for income estimates and economic modeling.
Return to topTitle: Expeditiousnesses and Delay in State Courts: An Exploration of Case Processing Time in Civil Trials
- Chair: Michael L. Cohen, CNSTAT
- Speaker: Thomas H. Cohen, Bureau of Justice Statistics
- Discussant: Nicole Waters, National Center for State Courts
- Date/Time: Wednesday, June 16, 2010 / 12:30 pm - 2:00 pm
- Location: Bureau of Labor Statistics Conference Center. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Ave., NE. Take the Red Line to Union Station.
- Sponsor: Washington Statistical Society's Section on Public Policy
Abstract:
The issue of how long it takes for America's courts to process civil cases dates back to colonial times and has been a topic of concern going back into antiquity. Although civil case delay has garnered a great deal of attention, most empirical efforts to analyze this topic occurred over twenty years ago and were limited in geographic scope. This paper attempts to update and enhance our understanding of civil case processing time by exploring factors related to disposition time in civil cases concluded by trial in a national sample of courts located in 156 urban, suburban and rural jurisdictions. The article applies multilevel models to examine both the effects of case and litigant characteristics as well as the influence of locale on case processing time. Results show that the legal issues adjudicated at trial, the type of trial (bench/jury), litigant characteristics, post-trial activity, and surprisingly, alternative dispute resolution referrals, are related to longer case processing times. The models also show locale exerting a significant influence on case processing time irrespective of a jurisdiction's case composition characteristics. This effort to empirically explore disposition time provides scholars, policymakers, and court personnel with information that could be used to assess what types of reform efforts might result in more expeditious case resolution and hence, speedier justice.
Return to topTitle: The OMB Clearance Process for Federal Statistical Surveys
- Speakers: Paul Bugg, Brian Harris-Kojetin, and Shelly Wilke Martinez, Office of Management and Budget
- Date/ Time: Wednesday, June 30, 2010/ 12:30-2:00 pm
- Location: Bureau of Labor Statistics Conference Center. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Take the Red Line to Union Station.
- Sponsors: WSS Data Collection Methods and DC-AAPOR
- Presentation material:
Slides (pdf, ~332kb)
Abstract:
The Paperwork Reduction Act of 1995 (PRA) requires that Federal agencies submit proposed information collections to the Office of Information and Regulatory Affairs (OIRA) in the Office of Management and Budget (OMB) before imposing information collection burdens on the public. In this presentation the speakers will describe the relevant aspects of the PRA for agencies and contractors that conduct statistical surveys and note common misconceptions about what is required under the PRA. The presentation will cover the process of an OMB review for surveys and highlight the information required for an OMB submission. Key subsections of agencies' information collection request supporting statement will be discussed including the justification of the need for and utility of the collection, and focusing on the information required for collections that employ statistical methods. The speakers will also address some of the common issues that arise in OMB reviews, and what agencies and contractors can do to anticipate and address these issues.
For further information contact Grace O'Neill at grace.o'neill@eia.doe.gov or 202.586.6485
Return to topTitle: Basics of Cryptology and Its Interplay with Statistics
- Speaker: Dr. Bimal Roy, Director, Indian Statistical Institute, Kolkata, India.
- Time and Date: 3-4 pm, Monday, Aug. 23, 2010
- Location: Monroe Hall 113 (2115 G Street, NW, Washington, DC 20052)
- Sponsor: The George Washington University, Department of Statistics
Abstract:
Basics concepts of Cryptology are introduced. A specific model of encryption, viz. Non-linear Combiner Model is described in details. A cryptanalysis, i.e., "breaking the code" is carried out on a specific instance of this model. The method is a sub exponential search aided by testing of hypothesis and maximum likelihood estimation.
Return to topTitle: Statistics Without Borders Post-Earthquake Efforts in Haiti
- Speakers: James D. Ashley and Justin S. Fisher, U.S. Government Accountability Office
- Discussant: Fritz Scheuren, NORC
- Chair: Scott Clement, Pew Forum on Religion & Public Life
- Date/Time: Tuesday, September 14, 12:30 - 2:00 p.m.
- Location: Pew Research Center 1615 L Street, NW, Suite 700 Washington, DC 20036-5621
- RSVP Instructions: If you are planning on attending this seminar, please email Michael P. Cohen at mpcohen@juno.com as soon as possible (deadline: Monday, September 13).
- Sponsors: WSS Human Rights, DC-AAPOR, and Capital Area Social Psychological Association
Abstract:
Considerations in the Study Design of a Mobile Phone Survey of the Haitian Population,
Presented by James D. Ashley, U.S. Government
Accountability Office
We outline the feasibility and benefits of conducting
telephone interviews to measure the economic impact of the
January 2010 Haitian earthquake using a random digit dial
(RDD) sample of mobile phone numbers. Traditional methods
of sampling and data collection after the earthquake would
have been difficult, time consuming, and costly. The
logistical challenges caused by the destruction of
buildings and roads, the large amount of internal
displacement, migration away from the Port-au-Prince
population center, and the increasing use of mobile phones
in Haiti made this kind of survey an attractive option.
Moreover, mobile technology in Haiti allowed us to deliver
instant incentives by adding pre-paid minutes to the
respondents' cell phones. Our presentation will focus on
coverage issues, contact and response rates, defining
households, and the construction of appropriate household
weights. This survey has provided invaluable information
about the potential of conducting mobile phone interviews
after a natural disaster and may help shape methodological
directions for collecting data in developing countries
around the world.
Survey Administration in the Wake of a Natural Disaster, Presented by Justin S. Fisher, U.
S. Government Accountability Office
Only two months after the devastating earthquake in Haiti,
our goal was to collect quality data to inform aid
decisions by measuring the earthquake's economic impact and
any changes in household composition. Because of the extent
of the damage, and the slow nature of recovery work, the
physical infrastructure and living conditions of many
Haitians presented special challenges for our data
collection efforts. Our initial objective of conducting an
area sample proved unrealistic, given the situation on the
ground. Consequently, we decided to conduct a survey
entirely by mobile phone, which required special
consideration be given to minimizing non-sampling errors.
We will focus on questionnaire development, cooperation
incentives, and supervisor and interviewer training (as
well as logistical considerations) in light of the
challenges that this collection mode poses for measurement,
non-response, coverage, and processing errors.
For further information contact Michael P. Cohen, mpcohen@juno.com or (202) 232-4651.
Return to topTitle: Child Health and the Environment: Do Epidemiology and Survey Statistics Make a Difference Globally?
- Speaker: Ruth H. Allen, PhD, MPH, US EPA (retired) and Visiting Scientist at Johns Hopkins University
- Chair: Mel Kollander
- Date/Time: Thursday, September 16, 2010 12:30 - 1:30 p.m.
- Location: Bureau of Labor Statistics Conference Center. To be placed on the seminar list attendance list at the Bureau of Labor Statistics you need to e-mail your name, affiliation, and seminar name to wss_seminar@bls.gov (underscore after 'wss') by noon at least 2 days in advance of the seminar or call 202- 691-7524 and leave a message. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station.
- Point of Contact: ruth.allen@gmail.com
- Sponsor: Agriculture and Natural Resources
Abstract:
The state of the world's children is a sensitive marker of global health and environmental progress. Statistics are the core element in any reassessment of progress over the last several decades. Concepts of survey statistics and epidemiology provide a framework for a reexamination of child mortality, morbidity, including birth defects and childhood brain cancer, education and special environmental circumstances in the last forty years. Data come from published sources including the UN, World Bank, public and private sector. The objective of this presentation is to evaluate what happened and to recommend next steps to several audiences of scientific, governmental, private sector and ordinary people.
Return to topTitle: Reconstruction of Distributions via Moments
- Speaker: Robert Mnatsakanov, Department of Statistics, West Virginia University
- Date/Time: Friday, September 17, 2010, 4-5pm
- Location: Monroe Hall, Room 113 (2115 G Street, NW, Washington, DC 20052)
- Sponsor: The George Washington University, Department of Statistics
Abstract:
The problem of recovering an unknown cumulative distribution function F and its density function f from its moments will be discussed. Under the mild conditions on the target distribution F the rate of approximations for proposed constructions are derived. The extension of the results to the multivariate Hausdorff moment problem will be presented as well. Several indirect models, e.g., the mixtures and the image reconstruction from its Radon projections will be considered. Finally, the comparison of theoretical and approximated curves will be illustrated via the simulations.
Return to topTitle: Implicit Translation: An Experiment in English and Farsi
- Speaker: David J. Marchette, Naval Surface Warfare Center, Dahlgren Division and The Johns Hopkins University
- Time: 10:30 a.m. Refreshments, 10:45 a.m. Colloquium Talk
- Date: September 17, 2010
- Location: Research 1, Room 301, Fairfax Campus, George Mason University, 4400 University Drive, Fairfax, VA 22030
- Sponsor: George Mason University CDS/CCDS/Statistics Colloquium
Abstract:
Implicit translation involves the association of an article in one language with another on the same topic in another. Thus a Wikipedia article in English on the Eiffel tower is associated with the French Wikipedia entry "Tour Eiffel". Neither is a translation of the other, but in some sense they are "the same". We discuss a method of fusing disparate information that allows the embedding of documents in different languages in order to make the proper associations. This method is completely general, and can be applied to any problem involving paired data in two or more dissimilarity spaces. The algorithm is illustrated on the implicit translation problem using documents in English and Farsi.
Return to topTitle: Conducting Nonresponse Bias Analysis for Business Surveys
- Speaker: Joanna Fane Lineback, Mathematical Statistician, U.S. Census Bureau
- Discussant: Clyde Tucker, Senior Survey Methodologist, Bureau of Labor Statistics
- Chair: Maura Bardos, Research Assistant/Programmer, Mathematica Policy Research
- Date/Time: Wednesday, September 22, 2010 / 12:30 - 2:00 p.m.
- Location: Bureau of Labor Statistics Conference Center. To be placed on the seminar attendance list at the Bureau of Labor Statistics you need to e-mail your name, affiliation, and seminar name to wss_seminar@bls.gov (underscore after `wss') by noon at least 2 days in advance of the seminar or call 202-691-7524 and leave a message. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Take the Red Line to Union Station.
- Point of Contact: Grace O'Neill at grace.o'neill@eia.doe.gov or (202) 586-6485
- Sponsors: WSS Data Collection Methods and DC-AAPOR
- Presentation material:
Conducting Nonresponse Bias Analysis for Business Surveys (Lineback slides, pdf ~372kb)
Conducting Nonresponse Bias Analysis for Business Surveys (Lineback paper, pdf ~176kb)
Moving Outside the Box (Tucker slides, pdf ~304kb)
Abstract:
Office of Management and Budget Statistical Standards require survey programs to conduct a nonresponse bias analysis when response rates fail to meet target values. The literature focuses largely on nonresponse bias analysis methods for demographic surveys. Such surveys are generally characterized by multi-stage designs with heterogeneous populations within selected clusters. In contrast, business surveys are characterized by single-stage designs with highly skewed populations.
In this presentation, I discuss nonresponse bias analysis methods for business surveys, including response rate analysis, the use of frame data, and the examination of the response prediction and propensity models. I illustrate these methods with examples from ongoing economic programs conducted by the U.S. Census Bureau.
Return to topTitle: On Estimating the Multinomial Parameter
- Speaker: Prof. Abram Kagan , University of Maryland, College Park, MD
- Date/Time: September 23, 2010, 2010, 3:30pm
- Location: Room 1313, Math Bldg, University of Maryland College Park
- Sponsor: University of Maryland, Statistics Program
Abstract:
If a random vector X = (X1, … ,Xm) has a multinomial distribution with parameters (n;θ = (θ1, …, θm)), the matrix of Fisher information on θ in X does not exist due to the constraint θ1 + … + θm = 1 (the definition of the information matrix requires that the parameter set be an open subset of Rm ). However, we show that the inverse of the information matrix is well defined and the standard estimator (X1/ n,…, Xm/n) of θ is not only UMVUE but also Cramer-Rao efficient, a stronger property. Cases when the components of θ are subject to extra constraints are also considered.
Return to topTitle: Wikipedia as a Testbed for Implicit Translation
- Speaker: Kristin Ash, Naval Surface Warfare Center, Dahlgren Division
- Time: 10:30 a.m. Refreshments, 10:45 a.m. Colloquium Talk
- Date: September 24, 2010
- Location: Research 1, Room 301, Fairfax Campus, George Mason University, 4400 University Drive, Fairfax, VA 22030
- Sponsor: George Mason University CDS/CCDS/Statistics Colloquium
Abstract:
I present an overview of the Wikipediae as a testbed for methods such as implicit translation. Discussion includes the kinds of data available within the Wikipediae (i.e. article text, inter-article links, etc.), the linguistic makeup of the Wikipediae, and how the Wikipedia data may be obtained and processed. Because a random graph embedding is used to perform implicit translation, I give particular attention to the Wikipediae as graphs.
Return to topTitle: Modeling Log-Linear Conditional Probabilities for Prediction in Surveys
- Speaker: Yves Thibaudeau, Principal Researcher, U.S. Census Bureau
- Organizer: David Judkins, Westat & WSS Methodology Section Chair
- Chair: Adam Safir, BLS & WSS Methodology Program Chair
- Time and Date: Wednesday, September 29, 2010, 12:30-2:00pm
- Location: Bureau of Labor Statistics, Conference Center. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station
- Sponsor: Methodology Section, WSS
Abstract:
It is shown how to model conditional probabilities subject to log-linear constraints and construct estimators that are more efficient than the Horvitz-Thompson weighted sum estimator for estimating population sizes in the context of a poststratified survey with unknown stratum sizes. Our approach rests on computing an hybrid predictor, similar to that of Pfeffermann et al (1998), and the expansion of a specific parameterization for conditional probabilities restricted by log-linear constraints, as proposed by Thibaudeau (2003). This parameterization facilitates the computation of MLE's and makes it possible to apply the method of Laplace for variance estimation. Calibration estimators will also be considered as an alternative to our approach.
Return to topTitle: Symmetric "Rejective" Probability Proportional to Size Sampling
- Speaker: Eric Slud, University of Maryland, College Park, MD
- Date/Time: September 30, 2010, 2010, 3:30pm
- Location: Room 1313, Math Bldg, University of Maryland College Park
- Sponsor: University of Maryland, Statistics Program
Abstract:
This talk will first introduce the topic of Probability Proportional to Size (PPS) sampling in surveys, and explain why PPS sampling without replacement in a way which is at the same time tractable to implement and has good theoretical properties, is still a topic of research.
The PPS method and results from a classical survey sampling paper of Hajek (1964) will then be described, and the rest of the talk will show the solution of a general existence problem related to such designs, and describe some computational developments which make rejective PPS sampling genuinely usable in practice.
Return to topTitle: Obesity Index
- Speaker: Mike West, Department of Statistical Science, Duke University
- Time: Friday, October 1, 2010, 3:30-4:30 pm
- Location: The George Washington University, Duques
553 (2201 G Street, NW)
(Followed by wine & cheese reception) - Sponsor: The George Washington University, The Institute for Integrating Statistics in Decision Sciences and Departments of Decision Sciences and Statistics
Abstract:
Current notions of tail fatness or tail obesity rely on estimates of the density for extreme values. For example the index of regular variation requires that, after an initial segment, the distribution is approximately Pareto, and the mean excess function is approximately linear. Loss data we have studied are (a) very rich, (b) very fat tailed and (c) not remotely Pareto. This paper explores a measure of tail obesity for positive random variables which characterizes tail obesity in samples, and can be computed for familiar classes of distributions.
If X1, …, X4 are independent samples of positive
random variable X, define
Obx(X) = P{X1 + X4 > X2 + X3|X1 > X2 > X3 > X4 },
capturing the intuition, "the fatter the tail, the more the
sum behaves like the max". Properties of Obx will be
described in the talk.
Topic: Hybrid Dirichlet Mixture Models for Functional Data
- Speaker: Michele Guindani, Department of Biostatistics, University of Texas MD Anderson Cancer Center
- Time: riday, October 1, 2010, 3:00-4:00 PM
- Location: Conference Room 326 St. Mary's Hall, 3700 Reservoir Road, NW (between 37th and 38th Streets, NW). Building #16 on Campus Map at http://maps.georgetown.edu/index.cfm?Action=View&MapID=3 or Google Maps: 3700 Reservoir Rd NW, Washington, DC 2000
- Sponsor: Georgetown University, Department of Mathematics and Statistics
Abstract:
In functional data analysis, curves or surfaces are observed, up to measurement error, at a finite set of locations, for, say, a sample of "n" individuals. Often, the curves are homogeneous, except perhaps for individual-specific regions that provide heterogeneous behaviour (e.g. 'damaged' areas of irregular shape on an otherwise smooth surface). Motivated by applications with functional data of this nature, we propose a Bayesian mixture model, with the aim of dimension reduction, by representing the sample of "n" curves through a smaller set of canonical curves. We propose a novel prior on the space of probability measures for a random curve which extends the popular Dirichlet priors by allowing local clustering: non-homogeneous portions of a curve can be allocated to different clusters and the "n" individual curves can be represented as recombinations (hybrids) of a few canonical curves. More precisely, the prior proposed envisions a conceptual hidden factor with "k"-levels that acts locally on each curve. We discuss several models incorporating this prior and illustrate its performance with simulated and real data sets. We examine theoretical properties of the proposed finite hybrid Dirichlet mixtures, specifically, their behaviour as the number of the mixture components goes to ∞ and their connection with Dirichlet process mixtures. Return to topTitle: Bayesian Grouped Factor Models
- Speaker: Merrill Liechty, LeBow College of Business, Drexel University
- Time: Friday, October 8, 2010, 3:30-4:30 pm
- Location: The George Washington University, Duques
553 (2201 G Street, NW)
(Followed by wine & cheese reception) - Sponsor: The George Washington University, The Institute for Integrating Statistics in Decision Sciences and Departments of Decision Sciences and Statistics
Abstract:
Firms that are publicly traded are classified based on their business models (i.e., how they make money) through industry classifications and based on their financial strength through debt ratings. As these classifications are based on the judgment of experts, it is an interesting question to determine the extent to which these classifications could be used to form prior distributions for correlation structures. Using a variable dimension Bayesian grouped factor model and standard classification schemes, we explore the value of these schemes with respect to model fit criteria, variance estimates of a tangency portfolio and value at risk calculations. In addition we demonstrate how this modeling framework can be used to include a firm which has just transitioned from being a privately held company to a publicly traded company with regards to asset allocation and risk assessment system.
Return to top20th ANNUAL MORRIS HANSEN LECTURE
On the Occasion of His 100th Birthday
Title: Dealing with Survey Nonresponse: In Data Collection, in Estimation
- Speaker: Carl-Erik Särndal, Statistics Sweden
- Discussants
J. Michael Brick, Westat
Roger Tourangeau, University of Michigan & Joint Program in Survey Methodology, University of Maryland - Date: Tuesday, October 12, 2010
- Time: 3:30-5:30 p.m.
- Location: Jefferson Auditorium of the U.S. Department of Agriculture's South Building (Independence Avenue, SW, between 12th and 14th Streets); Smithsonian Metro Stop (Blue/Orange Lines). Enter through Wing 5 or Wing 7 from Independence Ave. (The special assistance entrance is at 12th & Independence). A photo ID is required.
- Sponsors: The Washington Statistical Society, Westat, and the National Agricultural Statistics Service.
- Download the poster
Abstract:
In dealing with survey nonresponse, statisticians need to consider (a) measures to be taken at the data collection stage, and (b) measures to be taken at the estimation stage. One may practice a form of responsive design: In the later stages of the data collection, one tries to realize an ultimate response set that is better balanced or more representative than if no special effort is made. When the data collection is terminated, one still faces, at the estimation stage, the question of how to achieve the best possible reduction of nonresponse bias in the estimates. This effort is aided by a bias indicator, constructed as a function of selected powerful auxiliary variables. The concept of balanced response set extends the well known idea of balanced sample. It has a bearing on both aspect (a) and aspect (b). The statistical properties of a proposed measure of lack of balance are explored and illustrated.
A reception will follow at 5:30pm in the Patio of the Department of Agriculture Jamie L. Whitten Building. Please pre-register for this event to help facilitate access to the building online at http://www.nass.usda.gov/morrishansen/.
Return to topTitle: Contingency Tables from the Algebraic Statistics Viewpoint
- Speaker: Prof. Gérard Létac, Université Paul Sabatier, Toulouse, France
- Date/Time: Thursday, ctober 14, 2010, 3:30 PM
- Location: Room 1313, Math Bldg, University of Maryland College Park
- Sponsor: University of Maryland, Statistics Program
Abstract:
Contingencytables are governed by a hierarchical model. We explain how to choose between two hierarchical models by the method of Bayes factor. With this aim we consider each model as an exponential family on a certain polytope and take as a prior distribution a member of the Diaconis-Ylvisaker conjugate family with a probability element
| dπ(θ; α, m) = exp(α(θ, m) k(θ)) | dθ |
| I(m, α) |
where k is the cumulant function and II(m, α) the crucial normalizing constant. We study the behavior of I(m, α) and of the Bayes factor when α → θ. If N is the number of facets of the polytope containing the sample mean, we give an explicit formula in terms of N for the asymptotic behavior for the Bayes factor.
Please check for seminar updates at: http://www.math.umd.edu/statistics/seminar.html
Directions to Campus: http://www.math.umd.edu/department/campusmap.shtml
Return to topTitle: Workshop on Understanding Presidential Elections: 2008 and Beyond
- Speakers: Vincent L. Hutchings, Josh Pasek, Michael Mokrzycki, David Rothschild, Michael W. Traugott, Tasha S. Philpot, Seth E. Masket, and moderated discussions by D. Sunshine Hillygus, Nancy Mathiowetz, & Michael McDonald.
- Time: Friday, October 15, 2010, 9:00 a.m. - 4:00 p.m.
- Location: KFF Barbara Jordan Conference Center 1330 G Street, NW, Washington, DC 20005
- Sponsors: Abt SRBI, Pew Research Center, Westat
Abstract:
The authors of the latest special issue of Public Opinion Quarterly will present up-to-date research on presidential elections and the quality and role of polling during campaigns. This workshop focuses on the future of non-coverage bias in pre-election polls, poll performance, and the dynamics of racial prejudice and turnout. Contributing authors from the special issue will summarize recent findings and present up-to-date information on their election research and engage one another in an engaging workshop forum.
Please follow this link to view the workshop web page where you can find the workshop agenda as well as RSVP and payment instructions.
Return to topTitle: Inferring Social Network Structure from Incident Size Distribution in Iraq
- Speaker: Professor Tim Gulden, Center for Social Complexity, George Mason University
- Time: 0:30 a.m. Refreshments, 10:45 a.m. Colloquium Talk
- Date: October 15, 2010
- Location: Research 1, Room 301, Fairfax Campus, George Mason University, 4400 University Drive, Fairfax, VA 22030
- Sponsor: George Mason University CDS/CCDS/Statistics Colloquium
Abstract:
The violence in Iraq between 2003 and the present has been multi-faceted and extremely hard to characterize in terms of its motivations, participants and even overall scale. This work identifies a remarkably stable truncated power-law pattern in the size distribution of violent incidents in the Iraq Body Count (IBC) database and seeks to explain it in terms of social networks and the US role in breaking up and repressing major violent groups. While the work is preliminary, it offers a potentially useful metric in assessing the progress of counterinsurgency operations.
Return to topTitle: Small Area Estimation
- Speaker: Partha Lahiri, PhD, Joint Program in Survey Methodology (JPSM) at the University of Maryland
- Time: Tuesday, October 19, 2010, 1-3pm EST, http://www.amstat.org/sections/SRMS/webinar.cfm
- Sponsors: ASA Survey Research Methods Section &Amp; AAPOR
Abstract:
Direct survey estimates of various socio-economic, agriculture, and health statistics for small geographic areas and small domains are generally highly imprecise due to small sample sizes in the areas. To improve on the precision of the direct survey estimates, small area estimation techniques are often employed to borrow strength from related information that can be extracted from one or more existing administrative and/or census databases. In this talk, I will first discuss the main concepts and issues in small area estimation and then illustrate the effectiveness of small area estimation techniques in different applications. The talk will be presented at a level appropriate for individuals who are new to small area estimation, but also include discussion of research topics of interest to more experienced researchers.
Biosketch:
Partha Lahiri is a Professor of the Joint Program in Survey Methodology (JPSM) at the University of Maryland, College Park, and an Adjunct Research Professor of the Institute of Social Research, University of Michigan, Ann Arbor. Professor Lahiri's research on small-area estimation has been widely published in leading journals such as Biometrika, the Journal of the American Statistical Association, the Annals of Statistics and Survey Methodology. Professor Lahiri has served as member, advisor, or consultant to many organizations, including the U.S. Census Advisory committee, a National Academy of Science panel, the United Nations, the World Bank, and the Gallup Organization. He has served on the Editorial Board of many international journals, including the Journal of the American Statistical Association and Survey Methodology. Dr. Lahiri has been honored by being made a Fellow of the American Statistical Association and the Institute of Mathematical Statistics and an elected member of the International Statistical Institute.
Registration is now open!!!
http://www.amstat.org/sections/SRMS/webinar.cfm
For each webinar, participants register for a modest fee. Fees may vary from webinar to webinar depending on the length of the presentation and expected audience. Each registration is allowed one web connection and one audio connection. The section encourages multiple persons to view each registered connection.
If you have any questions, please feel free to contact Rick Peterson at the ASA office using the below information.
Rick Peterson
Education Programs Associate
American Statistical Association
732 North Washington Street
Alexandria, VA 22153
(703) 684-1221 ext. 1864
FAX: (703) 684-3768
rick@amstat.org
www.amstat.org
Title: The Survey Methodology Pipeline—Providing Needed Expertise For The Federal Statistical System
- Speakers:
Welcome - Lawrence Brown, CNSTAT Chair and University of Pennsylvania
evelopments at the OMB Statistical and Science Policy Office - Katherine K. Wallman, Chief Statistician, OMB
Introduction to the problem - Sally Morton, University Of Pittsburgh
The survey methodology pipeline: A view from the academic sector - Roger Tourangeau, University Of Maryland
The survey methodology pipeline: A view from the private sector - Graham Kalton, Westat
The benefits of dedicated grants in enlarging the pipeline - Cheryl Eavey, National Science Foundation - Discussants:
John Eltinge, Bureau Of Labor Statistics
Rod Little, U.S. Census Bureau - Date: Friday, October 22, 2010
- Time: 1:30-5:30 p.m.
- Location: Terrell Conference Center, 575 7th St, NW, Washington, DC - The Capitol Room
- Reception: 3rd Floor Atrium, NAS Keck Center, 500 5th St. NW
- Sponsor: he National Academies Committee on National Statistics
Abstract:
The current pipeline of survey and other methodologists seems inadequate to enable the federal statistical agencies to respond to the continuing demand for innovative solutions to the increasingly complicated problems encountered in collecting high quality data from surveys and administrative records. This issue can be examined from several perspectives. There is the view from academia on what programs and degrees departments currently offer, why more students are not graduating higher education institutions with relevant degrees, and what can be done to increase the pool of candidates for hire by the federal statistical agencies. There is the view from private survey firms on what they offer in the form of mentoring and training, the percent of graduates that are non-citizens, and why people sometimes leave federal service to go to the private sector. There is also the perspective of grant-giving institutions on how directed research might attract more students to survey and related work. The seminar will examine the issue from these three perspectives, with discussion from federal statistical agency methodologists.
NOTE the different venue: the Terrell Conference Center is in the old Hecht department store building at 7th and F Sts. NW (entrance on 7th St), across from the arena exit at the Gallery Place Metro Station (red line).
To REGISTER, please respond to cnstat@nas.edu or call Bridget Edmonds at 202-334-3096.
Return to topTitle: Choice-Based Revenue Management
- Speaker: Garrett Van Ryzin, Columbia University Graduate School of Business
- Time: Friday, October 22nd 2:30-3:30 pm
- Location: The George Washington University, Funger Hall 520 (2201 G Street, NW)
- Sponsor: The George Washington University, The Institute for Integrating Statistics in Decision Sciences and Departments of Decision Sciences and Statistics
Abstract:
Using consumer choice models as a basis for revenue management (RM) is appealing on many levels. Choice models can naturally model important buy-up and diversion phenomenon and can be applied to newer, undifferentiated low-fare structures and dynamic pricing problems. And recent research advances have now brought choice-based RM within striking distance of being truly practical. In this talk, we survey the recent research results in this area and discuss their implications for RM research and practice.
Return to topTitle: Simultaneous Calibration and Nonresponse Adjustment
- Speaker: Eric Slud, University of Maryland and U.S. Census Bureau
- Organizer: David Judkins, Westat, WSS Methodology Section Chair
- Chair: Adam Safir, BLS, WSS Methodology Program Chair
- Date/Time: Tuesday, October 26, 2010, 12:30-2:00pm
- Location: Bureau of Labor Statistics, Conference Center. To be placed on the seminar attendance list at the Bureau of Labor Statistics, you need to e-mail your name, affiliation, and seminar name to wss_seminar@bls.gov (underscore after 'wss') by noon at least 2 days in advance of the seminar or call 202-691-7524 and leave a message. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station.
- Sponsor: Methodology Section, WSS
Abstract:
Single and joint inclusion probabilities are generally known for complex survey designs up to the point where survey weights are modified due to nonresponse and population controls. Best practice by sophisticated survey practitioners generally includes weight modifications, first by calibration, ratio adjustment or raking to correct for nonresponse, next by further steps to impose population survey controls; and often, by final steps involving weight truncation or cell-collapsing to constrain the modified weights, usually so that the largest and smallest weights do not differ by more than a designated multiplicative factor. This talk presents a method for generalized-raking calibration in which all of these adjustments are accomplished in a single stage. Some large-sample superpopulation theory is described, which allows estimated or incorrect control totals, and which leads to linearization-based variance formulas for the resulting estimators. The method and estimators are illustrated in a moderate-sized hypothetical design and in the setting of the (Survey of Income and Program Participation) SIPP 1996 (Wave 1).
This talk is based on joint work with Yves Thibaudeau at the Statistical Research Division, of the Census Bureau.
Return to topTitle: Calculating a Symmetry Preserving Singular Value Decomposition
- Speaker: Mili Shah, Ph.D., Mathematical Sciences, Loyola University Maryland
- Date/Time: 3:35pm, Tuesday, October 26, 2010
- Location: Bentley Lounge, Gray Hall 130, American University
- Directions: Metro RED line to Tenleytown-AU. AU shuttle bus stop is next to the station. Please see campus map on http://www.american.edu/media/directions.cfm for more details
- Contact: Stacey Lucien, 202-885-3124, mathstat@american.edu
- Sponsor: American University Department of Mathematics and Statistics Colloquium
Abstract:
The symmetry preserving singular value decomposition (SPSVD) produces the best symmetric (low rank) approximation to a set of data. These symmetric approximations are characterized via an invariance under the action of a symmetry group on the set of data. The symmetry groups of interest consist of all the non-spherical symmetry groups in three dimensions. This set includes the rotational, reflectional, dihedral, and inversion symmetry groups. In order to calculate the best symmetric (low rank) approximation, the symmetry of the data set must be determined. Therefore, matrix representations for each of the non-spherical symmetry groups have been formulated. These new matrix representations lead directly to a novel reweighting iterative method to determine the symmetry of a given data set by solving a series of minimization problems. Once the symmetry of the data set is found, the best symmetric (low rank) approximation in the Frobenius norm and matrix 2-norm can be established by using the SPSVD. Applications of the SPSVD to protein dynamics problems as well as facial recognition will be presented.
Return to topTitle: Lower bounds for the Fisher information and the least favorable distributions
- Speaker: Professor Abram Kagan, UMCP
- Date/Time: Thursday, October 28, 2010, 3:30 PM
- Location: Room 1313, Math Bldg, University of Maryland College Park (directions).
- Sponsor: University of Maryland, Statistics Program (seminar updates).
Abstract:
On projecting the Fisher score into appropriately chosen subspaces, useful lower bounds are obtained for the Fisher information on a location parameter theta contained in an observation with pdf f(x-theta) for (more or less) natural classes of f. Also of interest are the least favorable distributions for which the bounds are attained. It is a joint work with Nicholas Henderson. Return to topTitle: Linear Regression Diagnostics for Survey Data
- Speaker: Dr. Richard Valliant, Professor, and U.MARYLAND/JPSM
- Discussant: James R. Knaub, Jr., Lead Mathematical Statistician, U.S. Energy Information Administration
- Date/Time: Wednesday, November 3, 2010 / 12:30 - 2:00 p.m.
- Location: Bureau of Labor Statistics, Conference Center. To be placed on the seminar attendance list at the Bureau of Labor Statistics, you need to e-mail your name, affiliation, and seminar name to wss_seminar@bls.gov (underscore after 'wss') by noon at least 2 days in advance of the seminar or call 202-691-7524 and leave a message. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station.
- Point of Contact: Linette Lanclos at Linette_Lanclos@nass.usda.gov (underscore after 'Linette') or (202) 720-2641 Sponsor: Economics Section, WSS
Abstract:
Diagnostics for linear regression models have been developed primarily for non-survey data. The models and sampling plans used for finite populations often entail stratification, clustering, and survey weights. I adapt some diagnostics for ordinary or weighted least squares for use with survey data. Statistics considered here include leverages, DFBETAS, DFFITS, and Cook's D. Differences in the performance of ordinary least squares and survey-weighted diagnostics are compared in an empirical study where values of weights, response variables, and covariates vary substantially. Another set of diagnostics are needed to identify collinearity among predictor variables. These also require some modification to be appropriate for survey data, which will be sketched in this talk. I also review how the forward search method for identifying groups of influential points may be applied to survey data.
Return to topTitle: House Price Index Methodology
- Speaker: Chaitra Nagaraja, U.S. Census Bureau
- Time: Friday, November 5, 4:00-5:00 pm
- Location: The George Washington University, Monroe Hall, Room 113 (2115 G Street, NW, Washington, DC 20052)
- Sponsor: The George Washington University, Department of Statistics
Abstract:
We examine what makes a house price index both practical and representative and focus on repeat sales indices. Two approaches are investigated: index structure (qualitative) and predictive ability (quantitative). We introduce an autoregressive index which utilizes repeat sales information and compare it to existing repeat sales indices. Each index is applied to data on single-family U.S. home sales from July 1985 through September 2004. We show the autoregressive model performs best.
Return to topTitle: A Unified Competing Risks Limited-Failure Model
- Speaker: Sanjib Basu, Northern Illinois University
- Time: Friday, November 5th 11:00-12:00 noon
- Location: The George Washington University,Duques 360 (2201 G Street, NW, Washington, DC 20052)
- Sponsor: The George Washington University, Department of Decision Sciences
Abstract:
A competing risks framework refers to multiple risks acting on a system. This can result from multiple components or multiple failure modes and are often conceptualized as a series system. A limited-failure model postulates a fraction of the systems to be failure-free and can be formulated as a mixture model, or alternatively by a bounded cumulative intensity model. We develop models that unify the competing risks and limited-failure approaches. We describe Bayesian analysis of these models, and discuss conceptual, methodological and computational issues related to model fitting and model selection. We compare the performances of the two limited failure approaches and illustrate in application.
Return to topTitle: Plans for the First Release of Small Area Data from the American Community Survey
- Speaker: Deborah Griffin, U.S. Census Bureau
- When: November 8, 2010
- Time: 2:00 pm - 4:00 pm
- Location: Bureau of Labor Statistics, Conference Room 2. To be placed on the seminar attendance list at the Bureau of Labor Statistics you need to e-mail your name, affiliation, and seminar name to dcaapor@gmail.com by noon at least 2 days in advance of the seminar. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Take the Red Line to Union Station.
Abstract:
In 2005 the U.S. Census Bureau began full sample data collection for the American Community Survey (ACS). With the completion of data collection in 2009, sufficient data were collected to produce small area estimates based on 5-year aggregations. This year the Census Bureau will meet a key ACS milestone with the release of the initial set of 5-year estimates. Debbie Griffin will provide background on the ACS and specifics about this year's release to provide DC-AAPOR members with the opportunity to learn more about this important new source of small area data. She will highlight educational resources available to ACS data users and answer questions about the survey and how data from the ACS compare with the data previously collected concurrent with the decennial census.
Debbie will discuss details on the schedule and content of the release and comment on some of the issues associated with this initial release. She will also acknowledge other Census Bureau data products related to the 2010 Census that will be released this fall and winter and explain how they will differ from the ACS.
The seminar is scheduled for one hour to be followed by a question and answer session
Return to topTitle: A Discussion of Nonresponse Bias Studies
- Speaker: J. Michael Brick, Westat and JPSM
- Date/Time: Wednesday, November 10, 2010 / 12:30 - 2:00 p.m.
- Location: Bureau of Labor Statistics, Conference Center. To be placed on the seminar attendance list at the Bureau of Labor Statistics, you need to e-mail your name, affiliation, and seminar name to wss_seminar@bls.gov (underscore after 'wss') by noon at least 2 days in advance of the seminar or call 202-691-7524 and leave a message. Bring a photo ID to the seminar. BLS is located at 2 Massachusetts Avenue, NE. Use the Red Line to Union Station.
- Sponsors: WSS Data Collection Methods and DC-AAPOR
- Presentation material:
Slides (pdf, ~92kb)
Abstract:
Nonresponse bias studies have been done for both household and establishment surveys over the years, but the number of these studies has greatly increased recently due to the 2006 OMB initiative requiring them when the response rate is too low. This session will begin by briefly reviewing methods used to conduct nonresponse bias. This review will describe some of the strengths and weaknesses of these methods, and the utility of using multiple methods in the same survey. Following this summary, the floor will be open for a moderated discussion of topics of nonresponse bias studies of interest to the audience. Come prepared with questions, comments, and points of interest to share!
For further information contact Grace O'Neill at Grace.O'Neill@eia.doe.gov or (202) 586-6485.
Return to topTitle: Semiparametric Estimation in Exponential Families
- Speaker: Abram Kagan, Department of Mathematics, UMCP
- Date/Time: Thursday, November 11, 2010, 3:30pm
- Location: Room 1313, Math Bldg, University of Maryland College Park (directions).
- Sponsor: University of Maryland, Statistics Program (seminar updates).
Abstract:
The talk deals with combining large samples of size n from the Natural Exponential Family (NEF) with an unknown generator F and of size m from population F to estimate the parameter of the NEF. All the cases m=cn(1+o(1)) with c>0, m=o(n) and n=o(m) are considered.
Return to topTitle: Exploring the World of Actuarial Science
- Speaker: William Shipley, American University Alumnus
- Date/Time: 3:35pm, Tuesday, November 16, 2010
- Location: Bentley Lounge, Gray Hall 130, American University. Metro red line to Tenleytown-AU. AU shuttle bus stop is next to the station. Please see campus map on http://www.american.edu/media/directions.cfm for more details.
- Contact: Stacey Lucien, 202-885-3124, mathstat@american.edu
- Sponsor: American University, Department of Mathematics and Statistics
Abstract:
This presentation will review some typical tasks of an actuary in various practice areas of actuarial science. The presentation will give suggested college courses to prepare for the actuarial work field and explain the actuarial examination process. Several hypothetical examples in various actuarial science practice areas will be presented.
Disclaimer
The opinions or views expressed in this presentation are the opinions or views of the presenter only and do not represent the opinions or views of any employer, any other person, any educational institution, or any other entity. This presentation is solely for educational purposes and should not be considered actuarial advice.
Return to topTitle: U-statistics with side information
- Speaker: Dr. Ao Yuan, National Human Genome Center, Howard University
- Date/Time: Thursday, November 18, 2010, 3:30pm
- Location: Room 1313, Math Bldg, University of Maryland College Park (directions).
- Sponsor: University of Maryland, Statistics Program (seminar updates).
Abstract:
We study U-statistics with side information incorporated using the method of empirical likelihood. Some basic properties of the proposed statistics are investigated. We find that by implementing the side information properly, the new U-statistics can have smaller asymptotic variance than the existing versions. The proposed U-statistics can achieve asymptotic efficiency and their weak limits admit a convolution result. We also find that the corresponding U-likelihood ratio procedure, as well as the U-empirical likelihood based confidence interval construction, do not benefit from incorporating side information, a result consistent with that under the standard empirical likelihood ratio. The impact of incorrect side information in the proposed U-statistics is also explored. Simulation studies are conducted to assess the finite sample performance of the proposed method. The numerical results show that with side information implemented, the deduction of asymptotic variance can be substantial in some cases, and the coverage probability of confidence interval using the U-empirical likelihood ratio based method outplay that of the normal approximation based method, especially when the underlying distribution is skewed. (it is a joint work with Wenqing He, Binhuan Wang, Gengsheng Qin). Click here to see the slides.
Return to topTitle: Why Use The Nonhomogeneous Poisson Process-I (NHPP-I) Model? After All, It Has Some Serious Issues
- Speaker: Sudip Bose, Department of Statistics, George Washington University
- Date: Friday, November 19, 4:00-5:00pm
- Location: The George Washington University, Monroe Hall, Room 113 (2201 G Street, NW, Washington, DC 20052)
- Sponsor: The George Washington University, Department of Statistics
Abstract:
The nonhomogeneous Poisson process (NHPP) model is widely used for describing software failure processes and has also been applied to other processes, such as call center data, and bidding in auctions. NHPP models for which the expected number of events (observed over infinite time) is finite, are called NHPP-I models.
Nayak, Bose and Kundu (2008) and Kundu, Nayak and Bose (2008) discussed a major statistical limitation of NHPP-I models, namely inconsistency of parameter estimates. In other words, even if the process is observed for an arbitrarily long time one cannot estimate unknown features of the model beyond a certain level of accuracy.
Our earlier research provided a formal proof of the non-existence of consistent estimates for certain parametric functions and heuristic arguments for other functions. Newer research is presented that provides a simple proof that no non-constant parametric functions can be estimated consistently. The above inconsistency results have implications for a Bayesian analysis of NHPP models. It is interesting to think about the ramifications of inconsistency of this model.
Return to topTitle: Finding Anomalous Trajectories by Kernel Learning
- Speaker: Professor Daniel Barbara, Department of Computer Science, George Mason University
- Time: 10:30 a.m. Refreshments, 10:45 a.m. Colloquium Talk
- Date: Friday, November 19,2010
- Location: Research 1, Room 301, Fairfax Campus, George Mason University, 4400 University Drive, Fairfax, VA 22030
- Sponsor: George Mason University CDS/CCDS/Statistics Colloquium
Abstract:
Analysis of video sequences has become an important problem in applications ranging from surveillance to automatic video annotation. Thus, the discovery of anomalous trajectories is becoming an important task in data mining. We propose a novel algorithm to discover anomalous trajectories based in learning a distance function, or kernel that is able to compare pairs of trajectories. Leveraging our previous work on anomaly detection, the algorithm uses transduction and hypothesis learning to discover trajectories that are anomalous with respect to a historical baseline. No assumptions are made on the distribution of trajectories and no thresholds are employed. We show empirical evidence of the performance of the algorithm which exhibits large true positive rates, while keeping the false positives low. We also discuss ways of improving the running time of the entire technique.
This is work in progress, sponsored by ARMY-TEC and jointly performed with Sheri Williamson, James P. Rogers, and Kathlyn Winter.
Return to topTitle: Universal Compressor-based Statistical Inference
- Speaker: Prof. Mikhail Malioutov, Northeastern University
- Date/Time: December 2, 2010, 3:30pm
- Location: Room 1313, Math Bldg, University of Maryland College Park (directions).
- Sponsor: University of Maryland, Statistics Program (seminar updates).
Abstract:
New affordable convenient instruments often contribute to the development of science, music, etc.
My talk is on a new class of applications of a well-known instrument - UNIVERSAL COMPRESSORS (UC), such as zip, originated in the interface between the Information theory and Computer Science by the end of the last century.
In addition to their direct application, the lengths of compressed stationary ergodic strings give a good approximation to the likelihood of these strings which generally cannot be evaluated analytically. This fact is implied by the nearly minimal average length of compression which is achievable, if the lengths of compressed strings approximate their likelihood.
This was formulated as the MDL principle by J. Rissanen in 1984 aa a new principle of statistics since the Likelihood is the main tool of statistical inference.
My talk is on two celebrated implementations of the MDL-principle.
Literary texts are modeled as stationary ergodic sources after the groundbreaking Shannon work of 1949. I test homogeneity between literary texts thus contributing to their authorship attribution.
Another application is screening out active inputs of a general system modeled as a functional multivariate relationship disturbed by a 'colored noise'.
The latter topic is an alternative approach to the Compressed Sparse Sensing which became recently extremely popular.
Our theoretical results make feasible the following types of applications:
- 'Tagging Flags' to the users' accounts in a large computer network with abrupt change in users' profiles possibly caused by unauthorized intrusion into the system for their more detailed follow up study.
- monitoring large corpora of audio/text strings, e.g. on-line forums or the phone call traffic in some areas, for 'flagging' matches to specific profiles of interest.
Title: On the Tradeoff Between Remanufacturing and Recycling
- Speaker: Tharanga Rajapakshe, School of Management, The University of Texas at Dallas
- Date/Time: Monday, December 6th 11:15-12:20 pm
- Location: Duques 652 (2201 G Street, NW, Washington, DC 20052)
- Sponsor: The George Washington University, The Institute for Integrating Statistics in Decision Sciences & Department of Decision Science
Abstract:
For a firm, the dual goals - induced by the drive on Extended Producer Responsibility -of meeting environmental regulations and positioning itself as a socially-responsible entity, necessitate the understanding of supply- and demand-side implications as well as product design characteristics. These, in turn, result in a healthy tradeoff between feasible sustainability measures, thus making the implementation of an appropriate option critical for long-term survival. Motivated by our interactions with two Dallas based is that of a manufacturer who produces and markets a product with the objective of maximizing profit. A unit of the product consists of two modules - Module A and Module B - that could each be either remanufactured or recycled. Module B incurs a higher per-unit production cost and is also priced higher than Module A. Once a module is recovered via a take-back mechanism, it can be either used in a remanufactured unit or can be further disassembled and recycled to recover its raw material, which can then be used to produce (albeit with different yields) new units of either Module A or Module B. Any unused units of either the complete product, Module A, or Module B, can be disposed. Under this setting, we investigate three options: (i) recycling of Module A, (ii) remanufacturing of Module B, and (iii) recycling of Module A and remanufacturing of Module B.
We first provide a complete theoretical characterization of the regions of optimality of each option. Next, we study the impact of choosing an option in an ad-hoc manner on the manufacturer's profit and analyze the sensitivity of this impact to changes in the supply demand gap and the take-back fraction. Recognizing that emerging governmental regulations render the disposal cost particularly vulnerable to dis-economies of scale, we examine the impact of non-linear disposal cost on the (i) optimal amount recycled or remanufactured and (ii) choice of an optimal operational strategy. To obtain richer managerial insights, we introduce the concept of "ability of sustainability", defined as a joint measure of the fraction of green consumers in the market, the take-back fraction, and product design characteristics such as the degree of substitutability of material, and examine its influence on the optimal option. Useful insights are developed on the sensitivity of the optimal choice to the relative profitabilities of the remanufacturing and recycling operations.
Finally, based on the demand for the remanufactured product, we also analyze the cases when green consumers are flexible and when they are dedicated
Return to topJulius Shiskin Memorial Award Seminar
Title: Information Technology and U.S. Economic Growth:
Evidence from a Prototype Industry Production Account
- Speaker: Dale Jorgensen, Professor of Economics at Harvard University and 2010 Recipient of the Shiskin Award
- Title: Information Technology and U.S. Economic Growth: Evidence from a Prototype Industry Production Account
- When: December 8, 2010
- Time: 2:00 pm - 3:30 pm
- Where: Bureau of Economic Analysis (Second Floor, Conference Rooms A & B), 1441 L Street NW. To be placed on the seminar attendance list at the Bureau of Economic Analysis you need to e-mail your name and affiliation to Vicki.Bingham@bea.gov by Friday, Dec. 3rd. Bring a photo ID to the seminar. BEA is located at 1441 L Street, NW.
Abstract:
Professor Jorgenson will present a new data set on U.S. productivity growth by industry. This data set covers 70 industries for the period 1960-2007and uses the North American Industry Classification System (NAICS). An important advantage of NAICS over the SIC is the greater detail available on the service industries that make up a growing proportion of the U.S. economy. NAICS also provides more detail on industries that produce information technology hardware, software, and services.
The production of information technology equipment and software has proved to be highly volatile. The great IT investment boom of 1995-2000 was followed by the dot-com crash and the slow and painful recovery of 2000-2007. The boom of1995-2000 was generated by an unsustainable deluge of innovation in the production of semiconductors and semiconductor-intensive computers. By contrast, the wave of innovation that followed in 2000-2007 has spread across a broader spectrum of IT-using industries. This has created a diversified advance in the applications of information technology.
Return to topTitle: Government Statistics Research Problems and Challenges
- Speaker: Dr. Yang Cheng, U.S. Census Bureau
- Date/Time: December 9, 2010, 3:30pm
- Location: Room 1313, Math Bldg, University of Maryland College Park (directions).
- Sponsor: University of Maryland, Statistics Program (seminar updates).
Abstract:
The Governments Division in the U.S. Census Bureau has conducted many innovative research projects in the area of sample design, estimation, variance estimation, and small area estimation. In this talk, we first give the background of a particular government statistics challenge. Then, we introduce our solution to the problem, a unique modified cut-off sample design. This design is a new two-stage sampling method that was developed by combining stratified sampling with cut-off sampling based on the size of the unit. Next, we present our decision-based estimation methodology. This adaptive decision-based estimation method was introduced as a stratum-wise regression for strata defined first by cut-points for cut-off sampling and then through stratum collapsing rules determined from the results of a hypothesis test for equality of regression slopes. Also, we discuss the small area estimation challenges we face when we estimate functional level data, such as estimates for airports, public welfare, hospitals, and so on. Finally, we explore variance estimators for the decision-based estimation.
Return to topSeminar Archives
| 2024 | 2023 | |||
| 2022 | 2021 | 2020 | 2019 | |
| 2018 | 2017 | 2016 | 2015 | |
| 2014 | 2013 | 2012 | 2011 | |
| 2010 | 2009 | 2008 | 2007 | |
| 2006 | 2005 | 2004 | 2003 | |
| 2002 | 2001 | 2000 | 1999 | |
| 1998 | 1997 | 1996 | 1995 | |
