Dependencies Across Time and Space (Salt Lake City, UT)


  • Simon Brewer, University of Utah

Research in the social and physical sciences increasingly generates large and complex datasets with observations that are indexed in time and/or space. These datasets are characterized by autocorrelation, implying that the closer two samples are in time or space, the more similar they are likely to be, impacting in turn the estimate of standard regression models.

This 5-day workshop is designed to provide an overview of these methods, building from simple spatial models to incorporate the time dimension in more complex spatio-temporal models. The week will start with the basics of managing and exploring spatial and temporal data, including tests for the presence and strength of autocorrelation. The second part of the workshop introduces spatial econometric methods and INLA as a method to implement full spatio-temporal models with large datasets. Workshop sessions will include lectures and hands-on applications, as well as time for students to apply the methods covered to their own data. Additional consultation time with the course instructors will be available.

Prerequisites: No specialized statistical knowledge or training is needed beyond a basic understanding of linear regression.

Software: Examples of all types of analysis will be provided using R software packages.

NOTE: The course instructor welcomes questions from potential participants regarding course appropriateness for potential student’s research designs/questions, readiness for the course, additional specifics of course content, etc.

Fee: Members = $1700; Non-members = $3200

Tags: spatio-temporal models, spatial econometrics, autocorrelation

Course Sections

Section 1

Location: University of Utah -- Salt Lake City, Utah

Date(s): May 14 - May 18

Time: 9:00 AM - 5:00 PM


  • Simon Brewer, University of Utah