STOR 940, FALL 2009
Course Coordinator: Richard L. Smith
Email: rls at

Richard Smith's presentation of March 3, 2010

This page was last updated December 8, 2009.


Copy of email sent to class, October 13 2009:

It is time to start planning for the presentations at the end of the semester. Just to remind you, if you are registered for credit you are expected to make a presentation to the class, together with a short written report, on a topic of your choice related to the class. This may consist of a presentation of some paper from a journal (outside those discussed directly during the classes), or may be a short data analysis of your own. Independent research or a lengthy report are not required.

By the time I give my own class (two weeks today - Oct 27) I'd like to have a schedule worked out - from each student, I'd like to hear the topic of your proposed presentation so that I can formally approve it. The presentations themselves are tentatively scheduled for the regular class time of Tuesday, December 8, though we could arrange some alternative time if that does not work for everybody.

If you are taking the class for credit, please respond to this and let me know your choice of project. If you need some advice or suggestions, I'll be happy to assist.

Richard Smith (rls at, Course Coordinator


This is one of three courses being taught as part of SAMSI's 2009-10 Program on Space-time Analysis for Environmental Mapping, Epidemiology and Climate Change.

This course will take place at SAMSI (driving directions) on Tuesdays, 4:30-7:00 pm, beginning August 25.


Much of modern epidemiology is concerned with relationships between environmental factors and various types of human health outcome. When data are collected at many spatial locations, we may refer to the problem as one of spatial epidemiology. However in most cases, this includes a temporal component as well. Since modeling spatial dependence is often critical to the method of statistical inference, it is necessary to use methods from spatial or spatio-temporal statistics. Very often health data are aggregated (e.g. into zip code or county totals) so models for data at discrete spatial locations, such as Markov random fields, are more appropriate than geostatistical methods. Another kind of problem is exemplified by the NMMAPS study ( an air pollution-mortality relationship is developed initially for many time series at individual cities, but inferences are then drawn by combining data across spatial locations. A third kind of problem is when there is uncertainty about the pollution field itself, for example, when data collected at monitors are interpolated to other locations. Sometimes this interpolation is performed by spatial statistics methods, but there is a growing trend to use air pollution models such as CMAQ (the EPA's Community Multiscale Air Quality model).

Specific topics (tentative): Models for spatially distributed health data. Markov random fields; extensions to spatial-temporal processes. Multi-city time series studies; combining data across multiple studies at different spatial locations. Measurement error problems that involve spatial interpolation; use of air quality models.


Students affiliated with any of the three Triangle universities may register in this course for academic credit. There are no specific prerequisites, but it is assumed that students have a general background equivalent to a typical first-year graduate program in statistics or biostatistics. Students should register through the appropriate course number at their university:

Duke: 294-02
NCSU: MA/ST 810.003
UNC: MATH 891-002/STOR 940-001

Auditors are permitted, and there are no prerequisites for being an auditor.


The course will be taught by multiple instructors. The instructors may, at their discretion, assign exercises and homeworks. In addition, each student will be expected to prepare a written and/or verbal paper to be presented at the end of the course. The final session of the course, December 8, has been provisionally set aside for these presentations. The exact nature of these presentations will be determined after the course has started. There will be no formal exams in this course.


The current schedule for the course is as follows. Please note this is subject to change!

August 25, Class 1: Howard Chang (Duke). Multi-city time series analysis. Presentation 1 - Updated 08/31 -- Presentation 2
September 1, Class 2: Howard Chang Presentation 3 -- Presentation 4 -- Presentation 5
September 8, Class 3: Howard Chang Presentation 6 -- Presentation 7
September 15, No class (Opening workshop)
September 22, Class 4: Sudipto Banerjee (University of Minnesota, visiting SAMSI). Markov random fields, hierarchical models and epidemiological applications. Presentation 1
September 29, Class 5: Sudipto Banerjee
October 6, Class 6: Sudipto Banerjee Presentation
October 13, Class 7: Brian Reich (NCSU) - exposure models Presentation
October 20, No class (UNC Fall Break)
October 27, Class 8: Richard Smith (UNC-STOR) - Bayesian methods for incorporating exposure measurement error Presentation -- Crooks et al. paper
November 3, Class 9: Sudipto Banerjee Presentation
November 10, Class 10: Amy Herring (UNC Biostatistics) - Case Study: Characterizing the Neighborhood Environment
November 17, Class 11: Sudipto Banerjee Presentation
November 24, Class 12: Sudipto Banerjee --- GMCAR.txt --- LiBanerjee.pdf --- Banerjee.pdf
December 1, Class 13: Murali Haran (Penn State, visiting SAMSI) - infectious disease modeling --- Presentation
December 8, Class 14: Student presentations, Schedule:

4:30 Yingqi Zhao, "Power Analysis in Association between Source-Specific Swine Markers and Acute Changes in Health Status Measures." Presentation - Write-up
4:50 Jingwen Zhou, "Spatial-temporal model development for NMMAPS data" Presentation
5:10 Rob Erhardt, "Spatial Pattern of the Dengue Vector in Iquitos, Peru" Presentation - Write-up
5:30 Laura Boehm, "Analysis of Cardiac Birth Defects and Air Pollution in Texas" Presentation - Write-up
5:50 break
6:00 David Vock, "Using Estimating Equations for Spatially Correlated Areal Data" Presentation
6:20 Eric Kalendra, "Extending the CAR Model to Account for General Temporal Neighborhood Structures" Presentation - Write-up
6:40 Soyoung Jeon, "Measurement error caused by spatial misalignment in environmental epidemiology" Presentation - Write-up


Please address to the course coordinator, rls (at)

Return to Richard Smith's page