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.
Gross Domestic Product in the USA
The time series plot below shows how Real GDP changed between 1960 and 2013 in the USA. (The GDP values have been corrected for inflation — GDP is measured in billions of 'chained 2006 dollars'.)
Click on crosses to read off the GDP to 1 decimal place.