Mini-course 5: Scatterplots - Health and Growth

This mini-course from the Nuffield Exploratory Data Skills Project: Making Sense of Data is about relationships between two, or more, quantities, looked at by plotting graphs of one against the other. As a result of this mini-course students should:

• look at the shape of the scatterplot before making any calculations

• pay attention in a scatterplot to outlying points, curvature, clustering in one region of the plot and differences across the plot in the variability of a variable

• know that a relationship is not a cause

• understand what different correlations look like on a scatterplot

• know how to fit a resistant straight line

• look at residuals from a fit

• know how to look for a re-expression to straighten a curved plot

• try plotting ranks if either variable is unevenly distributed

• allow for the effect of a variable by fitting and removing it

• know that selecting one group of cases may very much alter the correlation

The mini-course consists of five units

Health and medical provision forms an introduction to looking at scatterplots, and could be used on its own for that purpose.

Diet: quantity and quality covers fitting a resistant straight line and using residuals from the fit to introduce correlation.

Bodies: how to re-express variables to straighten a plot.

Health care provision considered will help those who need to consider ranking variables.

Health and diet concerns the very important issues which arise when one has more than two variables.

The teacher guide gives an overview of the mini-course.