Article published in:How to do Linguistics with R: Data exploration and statistical analysis
[Not in series 195] 2015
► pp. 333–350
Chapter 17. Language and space
Dialects, maps and Multidimensional Scaling
This chapter introduces another popular method that deals with distance matrices. This method is called Multidimensional Scaling. It is a dimensionality reduction technique that represents distances between objects in a low-dimensional space. You will learn how to perform different types of metric and non-metric scaling and carry out the diagnostics of solutions by using the scree plot, the Shepard plot and goodness-of-fit measures. The chapter also shows how one can use R for creation of geographical maps with points and text labels. Finally, you will learn how to measure the correlation between two distance matrices with the help of the Mantel test. The case studies are based on geographic coordinates and several linguistic features of varieties of English all over the world.