How-to Guide for IBM® SPSS® Statistics Software
Introduction

In this guide you will learn how to produce a time series plot for a single variable in IBM® SPSS® Statistical Software (SPSS) using a practical example to illustrate the process. You are provided with links to the example dataset and you are encouraged to replicate this example. An additional practice example is suggested at the end of this guide. The example assumes you have already opened the data file in SPSS.

Contents
• Time Series Plot
• An Example in SPSS: Global Ocean Temperatures, 1910–2015
• 2.1 The SPSS Procedure
• 2.2 Exploring the SPSS Output
1 Time Series Plot

A time series plot is a particular kind of two-way scatter plot where time is plotted on the horizontal axis and values of the variable in question are plotted on the vertical axis. Such a plot allows researchers to explore graphically any trends or other over-time dynamics that might be present in a particular variable. Typically the variable of interest is a continuous variable. Time is generally measured in equidistant units – days, months, years – though that is not required to produce a time series plot.

2 An Example in SPSS: Global Ocean Temperatures, 1910–2015

This example explores the average global ocean temperature measured annually from 1910 to 2015 as reported on the NOAA Climate at a Glance website. There are 106 observations in the dataset. The two variables we examine are:

• The year in which the average temperature was recorded (year), measured as a numerical indicator of each year from 1910 to 2015.
• Average annual ocean temperature deviation from the global mean in a given year (globalocean), measured in degrees on the Celsius scale.
• Average annual ocean temperature deviation is a continuous variable measured every year, making this an appropriate variable to plot as a time series.
2.1 The SPSS Procedure

You can produce a two-way scatter plot in SPSS using the Chart Builder menu. Begin by selecting from the menu:

Graphs → Chart Builder

This opens a Chart Builder dialog box. Figure 1 shows what this looks like in SPSS.

(Note: you may get a message asking you to define the level of measurement for variables in your dataset before you can proceed to building the chart. If you get this message, typically you can select “OK” and then “Scan Data” and SPSS will execute this task for you automatically.)

Figure 1: Selecting Chart Builder from the Graphs menu in SPSS.

In the lower half of the Chart Builder dialog box, select from the lower left menu “Line”. Two icons for line plots appear. The one on the left is for a simple line plot. Drag and drop that icon up into the open window in the upper half of this dialog box where the instructions say “Drag a Gallery chart here …”.

Next, drag and drop the variable you want plotted on the x-axis into the text box labeled “X-Axis?” For this example, the variable is year. Next, drag and drop the globalocean variable into the text box labeled “Y-Axis?” Note that SPSS will display the variable label for the ocean temperature variable (e.g. “Global Ocean”) and not the actual variable names if such labels are available. Figure 2 shows what this looks like in SPSS.

Figure 2: Building a line plot in the Chart Builder dialog box in SPSS.

Once done, click OK to produce the plot.

Executing the procedure above produces the time series plot shown in Figure 3.

Figure 3: Time series plot of global ocean temperature deviations from the 20th century mean, measured in degrees Celsius, from 1910 to 2015, NOAA Climate at a Glance.

Notice that Figure 3 only shows a line moving up and down from year to year. If you want to add markers at each point, you need to click on the chart as produced in SPSS to select it, then follow the menus:

Edit → Edit Content → In Separate Window

This will open the chart in a Chart Editor window, as shown in Figure 4.

Figure 4: Opening a chart in the Chart Editor window in SPSS.

To add markers for each data point, you can either click on the “Elements” menu item and select “Add Markers” or you can click on the fourth icon from the left in the row of icons just above the chart itself. Either way, a Properties dialog box will open as shown in Figure 5. By default, SPSS will select empty circles as markers for the data points. This can be changed by selecting the “Marker” button at the top of the Properties dialog box and making new selections. Figure 5 shows what this looks like:

Figure 5: Selecting Marker in the Properties dialog box from the Chart Editor in SPSS.

Once you have configured the markers as desired, click the “Apply” button in the Properties dialog box and close the Chart Editor window. The graph you originally created will be updated automatically, and in this case will look like Figure 6:

Figure 6: Time series plot of global ocean temperature deviations from the 20th century mean (including markers), measured in degrees Celsius, from 1910 to 2015, NOAA Climate at a Glance.

2.2 Exploring the SPSS Output

Figure 6 shows a fairly steady increase in global ocean temperature deviations over time, which is equivalent to saying that there has been a steady increase in global ocean temperatures from 1910 to 2015. The figure shows some volatility in the series over time, including a few large positive deviations in the early 1940s, but overall, the trend over time looks fairly steady and linear.

Two common statistical measures of correlation – the Pearson correlation coefficient and the Spearman rank-order correlation coefficient – can be computed to measure the association between global ocean temperatures and time (see the SAGE Research Methods Datasets modules on these methods). The Pearson correlation coefficient assumes a linear relationship between the two variables in question while the Spearman rank-order correlation coefficient only assumes a monotonic relationship. Figure 6 suggests that the Pearson correlation coefficient would be appropriate given the largely linear appearance of the trend in global ocean temperatures over time. When plotting a single time series, researchers might also explore fitting a trend line or a non-linear curve to the data (see the SAGE Research Methods Datasets module on Time Series Fitted Lines).