Statistics for Geography and Environmental Science: Correlation and Regression

Relational statistics ask whether two or more sets of data – variables – are related. Scatter plots are a simple and effective way of detecting a relationship and if it is linear. If it is, a measure of linear correlation can be calculated. A regression model produces a line of best fit summarising the relationship. Multiple regression extends the approach to allow more that one variable to explain another.\ud In this session you will 1) Learn how to calculate correlations in R. 2) Learn how to fit regression models in R and about the various tests and visual diagnostics that support the decision-making process