Time-ordering of univariate data

Some data sets are apparently univariate, but the measurements are made sequentially in time. For example, many demographic and economic measurements are recorded monthly or annually. A data set of this form is called a time series.

The time at which each measurement was made may be treated as an additional numerical variable, and the measurement can then be plotted against time. This type of scatterplot is often called a time series plot.

Time series data can be displayed and analysed with many of the techniques that are used for bivariate data.

Additional methods of analysis that are specific to time series data will be presented in a later chapter.

World population increase

The time series plot below shows the percentage increase in the world population between 1950 and 2013.

Click on crosses to read off the years and percentage increases in population.