Data Visualisation

Charts make your data look fancy and also help you understand what your data is showing. In this video I introduce the most useful kinds of charts for your data.

Eleven-minute video

You can also view this video on YouTube

You can find the slides here and also as .odp.


Key Points

Data visualisation is important to help other people (and yourself) understand your data. In the academic literature there are standard ways certain types of data are represented.

There are loads of different types of plot. Here are some of the most useful to know.

Bar charts

  • Compares values on a common scale
  • Shows the ratio between values
  • Appropriate for categorical variables

Pie charts

  • Shows proportions of a whole
  • Looks nice
  • Hard to visually compare values and generally avoided in the research literature

Histogram

  • Like a bar chart, but continuous values are grouped into bins
  • Appropriate for continuous variables

Density Plots

Density plots show how many observations of a variable exist in a particular range.

Single-variable density plots are often used to show probability distributions. They show the relative liklihood of getting data in a particular range. For example, the density plot below shows that our data approximates a normal distribution due to its characteristic bell curve shape.

You can also produce 2D density plots, which shows clusters of data points.

Box plots

  • Visualises distribution of data
  • Shows median, upper quartile, lower quartile, and highest and lowest values
  • Commonly used for illustrating hypothesis tests

Heat Map

Heatmaps use colour to show mangnitude of a variable.

For example, a heatmap might represent a level in a video game and show through colour the amount of time a player spends at different places.

Scatterplot

  • Shows dots representing data points for two variables x and y.
  • Independant varaible (IV) placed on x axis
  • Dependant variable (DV) on y axis

Questions

1. Check your understanding

Does this show a correlation between the two variables?

Generate


Summary

In this section we have learned about data visualisation.

  • You should be able to draw bar charts, histograms, box plots, and scatter plots
  • You should be able to interpret density plots and heatmaps

You may now move on to the descriptive statistics challenges