CS1703(1805): Data and Information (2015/16)

Dr Timothy Cribbin, Brunel University London

Laboratory Tutorial 1-8: SPSS Command

Syntax
In this laboratory tutorial you will:

1. Learn how to extract SPSS Command Syntax for the functions that you use

2. Learn how to edit and run command statements in the Syntax Editor

This is not a mandatory tutorial but you are strongly advised to complete it as it will help you to
pass the coursework.

Preamble
In this module, you have learned about statistics using a package called SPSS. SPSS is a powerful
application but, so far, you have only accessed this functionality through the graphical user interface
(GUI).

One particularly useful feature of SPSS is the facility to access commands using written statements.
This can be useful if you need to run a particular command or sequence of operations more than
once. SPSS command syntax is a type of programming language. It is similar, in principle, to SQL in
that it is a declarative language – statements describe what operations should be performed, rather
than how they should be performed. The underlying SPSS ‘engine’ translates these statements into
the relevant computations and outputs.

It is possible to combine command statements into ordered sequences or scripts. For instance, you
might want to create the same set of statistics for more than one dataset, monitor the change in
statistical outcomes as a dataset grows, or perhaps explore the effects of different transformations
or test configurations on one or more statistical tests. Using syntax in this way is much more efficient
than revisiting the GUI dialogues each time.

For the coursework, you will use SPSS command syntax as the way to communicate how you
attempted to answer questions using SPSS. This will apply principally to the Higher Questions. When
answering these questions you will be asked to describe your method as well as the actual answer. If
you used SPSS to arrive at your answer, then paste the command(s) that you used into this field.
Make sure that you paste multiple commands in the correct order in which they were executed
(run).

In the following exercises, you will be using the Epipen Tweets data table, as used earlier in Lab 5.

Exercise 1: Extracting commands from the GUI or Output

Viewer
There are three ways you can generate commands. The first is to type them by hand. Of course this
requires some working knowledge of the command language. For beginners, it is easier to work

1
CS1703(1805): Data and Information (2015/16)
Dr Timothy Cribbin, Brunel University London

using a copy/paste/edit approach. There are two ways to do this. The first is to paste commands
from the function configuration dialogue. The second is to copy and paste the command from the
Output window. We’ll look at the former method first in this section.

You may have noticed when configuring a function (e.g. Recode, Descriptives, t-test etc.) that next to
the OK button on the dialogue is another button labelled Paste. If you click this button it will paste
the command syntax, corresponding to the configuration you have set, into a Syntax Editor window.
If you don’t have a Syntax Editor open, one will open automatically.

Let’s try a simple example using the Frequencies function. Suppose we want to compute the
frequency distribution of positive sentiment scores across Tweets. Go to Analyze  Descriptive
Statistics  Frequencies to bring up the Frequencies dialogue and move SentPos into the Variables
list as shown below.

Let’s run this command. Now, instead of clicking OK, as you have done before, click the Paste
button. You should find that a Syntax Editor window opens and that it contains some text as shown
below:

2
CS1703(1805): Data and Information (2015/16)
Dr Timothy Cribbin, Brunel University London

Take a moment to examine the syntax statement. We can see the following keywords/clauses:

 FREQUENCIES – this is the name of the function/operation that we wanted to call

 VARIABLES – this keyword prefixes the items in the Variables list. In our case it’s SentPos, but
now that we are in the editor, we could edit this to any valid variable name. We could also
add additional items, simply by typing their names, separated by a space (e.g. ‘SentPos
SentNeg RealName’)

 /ORDER – this specifies the order in which results are output. This is only relevant if you have
multiple variables in the list. The default is ANALYSIS which means that each analysis is
presented once, with the results for each variable aggregated together. The alternative is
VARIABLE which would deal with one variable at a time, outputting the results for each
analysis, before repeating the output for the next variable.

 The statement terminates with a full-stop.

All you have done at this point, is pasted the command syntax into the editor. In order to execute
the command you need to select the statement and click the green ‘play’ arrow. The results of the
operation will appear within the output window, just as they would have done had you clicked OK
from the dialogue.

3
CS1703(1805): Data and Information (2015/16)
Dr Timothy Cribbin, Brunel University London

If you look at the output that appears, you’ll notice that the statement that you just executed is also
displayed prior to the results. Hence this is another, but less tidy, way of accessing syntax commands
i.e. you could manually copy and paste this text into the editor.

Exercise 2: Editing and running commands from the Syntax

Editor
The above command statement example is quite minimal. We just entered one variable and made
no other changes. However, most functions in SPSS have a range of different, optional, parameters
which can be accessed and set using keywords. Some of these are accessible from the GUI but
others can only be accessed using typed commands. Hence, learning command syntax provides you
with full access to and control over the more powerful features and functionality of SPSS.

Let’s take another look at our Frequencies statement. If you revisit the GUI dialogue, you’ll notice
some buttons on the right hand side. Let’s start by clicking on Statistics. This allows you to select
from a range of averages, dispersion measures and other statistics. Let’s see what happens to the
Frequencies statement if we select one of these options.

Select Mean from the Central Tendencies frame and Skewness from the Distribution frame as
shown below. Click Continue to go back to the main dialogue. Then click Paste and go back to the
Syntax Editor.

You should see a new section of command syntax below your earlier statement. Take a moment to
examine the differences. You’ll see that there is one extra line beginning with the keyword
/STATISTICS=. This keyword is followed by a list of the statistical measures that you selected. There
4
CS1703(1805): Data and Information (2015/16)
Dr Timothy Cribbin, Brunel University London

are actually three parameters – SPSS has assumed you want to compute the Standard Error of
Skewness (SESKEW) as well as Skewness. You could delete that if you wanted to; this wouldn’t of
course be possible if you were just using the GUI.

You can add further statistics without returning to the GUI. You might be wondering how to find out
what measures are available and the keywords that describe them. The editor offers a form of
intelli-sense (context sensitive suggestions), similar to an IDE like Eclipse or Visual Studio. It’s a little
bit flaky but is helpful if you know how to use it.

Try deleting and replacing the ‘=’ character following /STATISTICS. You’ll see a menu appear with a
list of keywords in it. In this case, the set of possible measures is pretty much the same as what we
could access from the GUI. However, note that it is possible specify ALL measures, which saves you
from having to type every keyword into the command line. Try adding some keywords and see how
this affects the output when you run the statement.

In addition to statistics there are other sub-commands accessed using keywords prefixed with a
forward slash. If you create a new line (before or after the /STATISTICS line) and type a forward
slash, a menu should appear showing you all the options.

Try adding sub-commands to your statement to do the following:

 Display a histogram

 Display a pie chart

 Sort the rows of the frequency table in descending order of frequency

The first two should be obvious, but the keyword for the last task is less so. Go back to the GUI if
necessary and find out how this option is referenced (labelled) there. Another tip is that the first
time you select a sub-command keyword, you should always type an equals character to see if there
are any parameters that you can set.

Finally, you can use the inline Help function to access a full reference guide to the command syntax.
You can search by topic or if you right click on a command statement in the syntax editor and select
Help it should take you directly to the relevant web page.

5
CS1703(1805): Data and Information (2015/16)
Dr Timothy Cribbin, Brunel University London

Summary and further work

In this tutorial you have learned how to access, edit and execute SPSS command syntax statements.
So far you have only focused on the Frequencies command. You are encouraged to explore the
command sets for all of the functions you have learned about during the module, including t-tests,
correlations and also Compute and Recode. Remember that during the final coursework task you will
need to communicate all of your SPSS methods using the relevant command syntax. After making
any changes, check that the command statement runs correctly on your data before you copy it into
the coursework submission form. As always, remember to save any important syntax files before
exiting SPSS.