Wed, 19 Apr|
Roger Bivand on the use of R for spatial econometrics
In his EMOS webinar, Roger Bivand will be giving insights into "The use of R for spatial econometrics". His focus will be on the question "What is spatial econometrics?" and estimation methods as well as the representation of spatial data in R packages, and other important aspects.
Time & Location
19 Apr, 16:00 – 17:00 GMT+2
Roger Bivand is an academic geographer, and has been a professor since 1996, retiring in 2021, after joining the Norwegian School of Economics, Bergen, Norway in 1988 from previous academic positions in Bodø, Norway, and in Poznań, Poland. His current research interests are in supporting the development of open source software for analysing spatial data, including spatial econometrics, especially using R.
- What is spatial econometrics? How does it relate to econometrics and to other fields of modelling with spatial data?
- Which estimation methods are used in spatial econometrics, which are specific to spatial econometrics, and which share with proximate fields?
- How are spatial (and spatio-temporal) data represented in R packages, and which packages provide implementations of relevant estimation methods?
- Boston housing value data set: caa se of trying to study a problem when the support of the data probably does not match the problem.
Webinar learning outcomes:
- Being able to place spatial econometrics in a broader context of modelling with spatial data.
- Knowing the most common models proposed by spatial econometrics.
- Knowing which R packages provide these models.
- Understanding the concepts of support, spatial autocorrelation, and how they may interact when modelling with spatial data.
To provide participants with an overview of one of the three kinds of modelling with spatial data, namely areal or lattice data modelling. It will become clear that these kinds of data may be characterised by spatial autocorrelation. On the one hand, information leaking between neighbouring spatial entities needs to be taken into account. On the other, we will see that looking carefully at spatial entities, and understanding that such spillovers can occur, may lead us to a clearer analysis.
Prerequisites: Some familiarity with geo-spatial data and R
- Chapters 16 and 17 of Pebesma and Bivand (forthcoming) Spatial Data Science with applications in R: Click here
- Bivand (2017) Revisiting the Boston data set – Changing the units of observation affects estimated willingness to pay for clean air: Click here
Webinar materials: Github