The program will consist of eight invited one-hour talks designed to be accessible to probabilists, statisticians, and others not already expert in this area. Four of the talks will stress analytical methods, while the other four will deal with issues of simulation and graphical visualization. Each day's program will feature two talks of each sort, as well as ample time for informal discussion, computer demos, and perhaps a roundtable or two.
Confirmed speakers are: "Theory": J. T. Cox (Syracuse), D. A. Dawson (Carleton), G. Giacomin (Rutgers), R. Schonmann (UCLA); "Simulation": P. Glynn (Stanford), D. Griffeath (Wisconsin), A. Lawniczak (Guelph).
Registration fees are: SSC or IMS Members: C$30/US$25; non-members: C$40/US$35; Students: C$15/US$15. These rates are valid through May 15, 1995. A late surcharge of C$10/US$10 will apply thereafter for all non-student participants.
Requests for more information concerning the scientific program, as well as any organizational questions, should be directed to: David Griffeath, Dept. of Mathematics, University of Wisconsin-Madison, Van Vleck Hall, 480 Lincoln Drive, Madison, WI 53706-1388, USA; FAX (1-608) 263-8891, GRIFFEAT@MATH.WISC.EDU (e-mail preferred).
(1) Nonparametric (randomization) methods (e.g., Mantel-Haenszel tests) for testing hypotheses of no association under minimal assumptions.
(2) Regression methods for fitting statistical models to describe multivariate relationships (e.g., logistic regression, Poisson regression, weighted least-squares regression).
Consideration of these strategies will be motivated through examples from clinical trials, observational studies, and sample surveys. For these examples, the roles for alternative methods and aspects of application will be discussed from the perspective of the questions which statistical analyses are to address, the posture of inference, the sampling process, and the data structure.
The specific topics for this Workshop are: Structure of categorized data and contingency tables; Mantel-Haenszel procedures for 2x2 contingency tables and extensions of logistic regression for ordinal data; logistic regression; proportional-odds model extension of logistic regression for ordinal data; Poisson regression for incidence densities; methods for studies with repeated measures and/or matching; computing procedures for implementing methods.
The primary references are:
G. G. Koch, G. J. Carr, I. A. Amara, M. E. Stokes & T. J. Uryniak (1990). Categorical data analysis. In Statistical Methodology in the Pharmaceutical Sciences (D. A. Berry, ed.), Marcel Dekker, New York, 389-473 (Chapter 13).
G. G. Koch & S. Edwards (1988). Clinical efficacy in trials with categorical data. In Biopharmaceutical Statistics for Drug Development (K. E. Peace, ed.), Marcel Dekker, New York, 403-457.
G. G. Koch; J. E. Schmid, J. M. Begun & W. C. Maier (1993). Meta-analysis of drug safety data. In Drug Safety Assessment in Clinical Trials (G. S. Gilbert, ed.), Marcel Dekker, New York, 285-310.
L. M. LaVange; L. L. Keyes; G. G. Koch & P. A. Margolis (1994). Application of sample survey methods for modelling ratios to incidence densities. Statistics in Medicine, 13, 343-355.
LogExact: Software for Exact Logistic Regression. Cytel Software Corp., Cambridge, Mass., 1992.
StatXact: Statistical Software for Exact Nonparametric Inference. User Manual, Version 2. Cytel Software Corp., Cambridge, Mass., 1991.
Registration fees are: SSC or IMS Members: C$90/US$80; non-members: C$110/US$100; Students: C$35/US$30. These rates are valid through May 15, 1995. A late surcharge will apply thereafter for all non-student participants: C$20/US$20 for SSC or IMS members and C$25/US$25 for non-members.
For further information about this Workshop contact: Claudine Legault, Dept. of Public Health Services, Bowman Gray School of Medicine, Medical Center Boulevard, Winston-Salem, NC 27157-1063, USA; e-mail: LEGAULT@PHS.BGSM.WFU.EDU, tel. (1-910) 716-7467, FAX (1-910) 716-5425.
Organized by David Binder and Pierre St-Martin (Statistics Canada) this Workshop will review the various issues and techniques for analysis of complex survey data with some applications of those special techniques being illustrated based on some complex surveys conducted mainly by Statistics Canada. The themes being covered could be summarized as follows:
What is a complex survey design?
What can go wrong when applying classical methods to complex survey data?
What adjustments can be made to classical methods to accommodate for complex designs?
The typical participants in mind for this workshop are:
(1) University students who had undergraduate courses on statistical methods such as multivariate regression analysis, analysis of variance, and categorical data analysis and wish to apply those methods to survey data, without necessarily having a good background in survey sampling;
(2) Researchers such as statisticians, econometricians, sociologists or epidemiologists, mainly secondary analysts of survey data collected by government agencies, who wish to apply statistical methods to complex survey data.
List of topics: Simple and Complex Sample Designs, Target Population, Randomization Distributions: Model-Based vs. Design-Based Estimation, Non-Response Issues, Examples. Variance Estimation under Complex Designs: Design Effects, Variance Estimation with a Small Number of PSU's, Confidence Intervals and Hypothesis Testing, Examples. Classical Regression Assumptions and Methods: To Weight or Not To Weight When Fitting Models to Survey Data, Examples. Tests of Independence and Homogeneity, Multivariate Design Effects, Examples. A few words on Other Related Topics: Logistic Regression, Case-Control Study, Statistical Packages.
Related book: Analysis of Complex Surveys, edited by C. J. Skinner, D. Holt and T. M. F. Smith, John Wiley & Sons, New York, 1989.
Registration fees are: SSC or IMS Members: C$80/US$75; non-members: C$100/US$90; Students: C$40/US$35. These rates are valid through May 15, 1995. A late surcharge will apply thereafter for all non-student participants: C$20/US$15 for SSC or IMS members and C$25/US$25 for non-members.
For further information about this Workshop contact: Charles A. Patrick, International and Professional Relations Division, Statistics Canada, RH Coats Building 25-F, Tunney's Pasture, Ottawa, Ontario, Canada K1A 0T6; FAX (1-613) 951-1231; PATRCHA@STATCAN.CA.
The purpose of this Workshop is to stimulate research and, in an informal setting, to foster the interaction of researchers in the interface between matrix theory and statistics. This Workshop will provide a forum for statisticians and mathematicians working in the areas of linear algebra and matrix theory to become better informed of the latest developments and newest techniques and to interchange ideas with researchers from a wide variety of countries. Invited guest speakers are expected to come from these countries: Canada, Chile, China, Croatia, Czech Republic, Denmark, Estonia, Finland, Germany, India, Japan, The Netherlands, New Zealand, Poland, Portugal, Slovenia, Sweden, Turkey, United Kingdom, and the United States. Funding for non-Canadian guest speakers' travel and local expenses has been requested from the Natural Sciences and Engineering Research Council of Canada (NSERC).
The following persons have been invited and are expected to participate: Fikri Akdeniz (Adana), S. Ambikkumar (McGill), T. W. Anderson (Stanford), Nátalia M. Bebiano (Coimbra), James V. Bondar (Carleton), S. W. Drury (McGill), Morris L. Eaton (Minnesota), R. William Farebrother (Manchester), Miroslav Fiedler (Prague), Luis S. Firinguetti (Santiago), Ali S. Hadi (Cornell), Jeffrey J. Hunter (Massey), Alexander KovaO([[hungarumlaut]],c)ec (Coimbra), Augustyn Markiewicz (Poznan;), Thomas Mathew (Baltimore), Jorma Kaarlo Merikoski (Tampere), Renate Meyer (Auckland), Sujit Kumar Mitra (Delhi), Kenneth Nordström (Helsinki), Ingram Olkin (Stanford), Christopher C. Paige (McGill), Josip E. PeO([[hungarumlaut]],c)aric;(Zagreb), Michael D. Perlman (Washington), Serge Provost (Western Ontario), Simo Puntanen (Tampere), C. Radhakrishna Rao (PennState), Shayle R. Searle (Cornell), Peter SO([[hungarumlaut]],e)mrl (Maribor), Alastair J. Scott (Auckland), Gerald E. Subak-Sharpe (New York), Yoshio Takane (McGill), Götz Trenkler (Dortmund), Song-Gui Wang (Beijing), Hans Joachim Werner (Bonn), Henry Wolkowicz (Waterloo) & Haruo Yanai (Tokyo).
Talks are expected to cover the following topics: biased estimation in linear models, bounds for matrix sums, bounds for singular values, C-S decomposition, Campbell-Youla inverse, canonical correlations, Cochran's Theorem, control theory, convex matrix functions, matrix convexity, econometrics, eigenvalues & optimization, generalized inverses, generalized least squares, Kantorovich inequality, Kiefer ordering, least squares with missing observations, Marcus-de Oliveira conjecture, matrix inequalities, matrix special functions, mixed model of the analysis of variance, modified eigenvalue problems, multidimensional scaling, multivariate statistical analysis, orthogonal projectors, partially generalized least squares, positive definite matrices, predictive g-inverse, seemingly unrelated regressions, sexy matrices, shorted operators, stochastic matrices, and tensor products.
This Workshop is the fourth in a series. The previous three Workshops were held in Tampere, Finland, in August 1990, in Auckland, New Zealand, in December 1992, and in Tartu, Estonia, in May 1994. The 5th Workshop in this series is scheduled for Dunedin, New Zealand, in August 1995 (as a satellite to the A. C. Aitken Centenary Conference) and the 6th in Shrewsbury, England, in August 1996.
It is expected that papers from this Fourth International Workshop on Matrix Methods for Statistics will be published in the Sixth Special Issue on Linear Algebra and Statistics of Linear Algebra and Its Applications.
Registration fees are: SSC, IMS or ILAS Members: C$50/US$45; non-members: C$65/US$60; students: C$15/US$15. These rates are valid through May 15, 1995. A late surcharge of C$15/US$15 will apply thereafter for all non-student participants.
Contributed papers are welcome! Submit a one-paragraph abstract, in English or French: no more than 10 lines please! If at all possible use plain TEX, emTEX, LATEX or Microsoft Word and send the abstract by e-mail (as a regular ASCII text file or in Binhex) to George Styan at MT56@MUSICA.MCGILL.CA or STYAN@MATH.MCGILL.CA. All abstracts should arrive by 15 May 1995.
Requests for more information concerning the scientific program, as well as any organizational questions, should be directed to: George P. H. Styan, Dept. of Mathematics and Statistics, McGill University, Burnside Hall, 805 ouest rue Sherbrooke, Montréal, Québec, Canada H3A 2K6; MT56@MUSICA.MCGILL.CA or STYAN@MATH.MCGILL.CA, FAX (1-514) 398-3899.