Advanced statistical methods in geography

Teacher(s)

Hafner Christian;

Language

French

Content

Summary The econometric techniques used in economic geography have dramatically improved during the last decade. Moreover, similar statistical problems arise in the various fields of physical geography. The objective of the course is to allow a geographer acquainted with a preliminary background in statistics to meet the level of statistical requirements for understanding articles and for publishing in high-ranking journals. The course focuses on linear models. A great deal of efforts is placed on the statistical validation of these models: selection of variables, functional form, endogeneity problems, temporal and spatial autocorrelation in errors terms, selection biases, etc. To ignore misspecifications of the model may entail spurious interpretations and unreliable predictions. Several techniques that allow to overcome some of these issues are studied: weighted least-squares, feasible generalized least squares, instrumental variables, autoregressive and correlated error models, etc. An initiation to the R language is also provided, as well as an exploration of the related spatial statistics libraries. Methods Theoretical developments and practical illustrations on the computer alternate during the class.

Teaching methods

Due to the COVID-19 crisis, the information in this section is particularly likely to change.

lectures on theory and illustrations on computers

Evaluation methods

Due to the COVID-19 crisis, the information in this section is particularly likely to change.

The evaluation is based on a test of knowledge (short-answers questions), a personal project on a realistic dataset, the presentation of selected recent articles drawn high-ranked journals (Regional Science and Urban Economics, Journal of Urban Economics,…).

Other information

Prerequisite GEO1341 Modélisation statistique en géographie (or similar).

Online resources

Every note, chunk of R code, or dataset used in the course is available on the moodle site associated with this course.

Bibliography

• R Bivand, E Pebesma and V Gómez-Rubio, Applied Spatial Data Analysis with R, Springer, New York, 2008.
• MJ Crawley, Statistics: An Introduction Using R, John Wiley, 2005.
• MJ Crawley, The R Book, John Wiley, 2007.
• O Schabenberger and C Gotway, Statistical Methods for Spatial Data Analysis, Chapman & Hall, 2005.
• WN Venables and BD Ripley, Modern Applied Statistics with S (4th edition), Springer, 2002.
• M Verbeek, A Guide to Modern Econometrics, John Wiley, 2000.

Faculty or entity

GEOG