SPSS
Session 1: Assignment 1: SPSS, Syntax and Descriptive Statistics
Based on the measurement level of the variables, which method (that was discussed in the lectures), would be most
appropriate to analyze the relation between Individual Innovative Performance (Y) and the Company (X) that someone
works for? Answer: ANOVA
Syntax
The benefit of using syntax is that it provides a precise history of how the data was analyzed. Always aim to keep your
data file and your syntax file well-organized. This way you never have to save any output, because you can always recreate
it as needed by having SPSS perform all the syntax that you saved.
There are two ways to create a new syntax file:
1. In SPSS go to File → New → Syntax
2. Before performing any analysis instead of clicking the Ok button, click the Paste button
After starting a new syntax file, using the paste button for subsequent analyses will add the commands to the bottom of
the original file. You do not have to make a new file for each analysis. Clicking Paste will often close the dialog window.
Two ways of running the code is opening the same menu and clicking Ok this time (It will remember all your settings).
Another way is placing your cursor anywhere in the generated syntax and pressing Ctrl + R.
Frequencies
Analyze > Descriptive statistics > Frequencies > Select variable > Paste
The SPSS syntax for the Frequencies function looks like this:
FREQUENCIES VARIABLES = x
/ORDER = ANALYSIS.
So, for this step it should look like this:
FREQUENCIES VARIABLES=company
/ORDER=ANALYSIS.
SPSS does allow abbreviated functions, and works with any command shorter than its original as long as it is unique. This
means that this will also work:
FREQ VAR = x.
Descriptives
For continuous variables, the Frequency function is often not very useful, because many different scores are possible. If
many different values are possible, or the sample is not very big, each respondent will have its own unique value with a
frequency of one. A better alternative is using summary statistics for these variables, such as the mean, median, and
variance. Many different summary statistics for variables can be obtained using the Descriptives function.
Analyze > Descriptive statistics > Descriptives > Options > Tick boxes > Continue > Paste
The SPSS syntax for the Descriptives function looks like this:
DESCRIPTIVES VARIABLES=x
/STATISTICS=MEAN STDDEV VARIANCE MIN MAX.
SPSS does allow abbreviated functions, and works with any command shorter than its original as long as it is unique. This
means that this will also work:
DESC VAR=x.
Mean
Using the Descriptives function of SPSS we saw one way to obtain the mean for a variable. A second way is by using the
Means function. Use the Means function located under:
Analyze > Compare Means > Means
1
, SPSS
The SPSS snytax for the Means function looks like this:
MEANS TABLES=x
/CELLS=MEAN COUNT STDDEV.
SPSS does allow abbreviated functions, and works with any command shorter than its original as long as it is unique. This
means that this will also work:
MEAN VAR=x.
The Means function is not only there to provide the mean for one variable. It is possible to provide an additional
independent variable. Keep the variable Y in the list for the dependent variables and add the variable X to the list of
independent variables by selecting it in the left panel and clicking the arrow next to the list of independent variables. As
you can see, for a categorical independent variable, it now computes the mean for each variable X separately.
The SPSS syntax for the Means function using an independent variable looks like this:
MEANS TABLES=y by x /CELLS=MEAN COUNT STDDEV.
SPSS does allow abbreviated functions, and works with any command shorter than its original as long as it is unique. This
means that this will also work:
MEAN VAR=y by x.
Can we, based on this information, say anything about the differences in the entire population under study? (Assume that
these are very big companies and we have only observed a small part of their employees.) Answer: No - We are only
looking at sample data, and we need further tests to say anything about the population
Assignment 2: ANOVA
Before performing any analysis on the data, several assumptions (properties that the data should adhere to) must be
checked. This to see whether the analysis method will provide the correct results and we are not analyzing data that we
may not apply ANOVA to. The assumptions for an ANOVA are:
1. Correct measurement level of the X and Y variables; quantitative dependent variable of interval/ratio (continuous)
measurement level, independent variable of nominal measurement level (categorical).
2. Normal distribution of scores on the Y variable (for the full dataset, and within each group)
3. No outliers
4. Approximately equal variances in all the groups
5. The observations should be independent (within- and between groups)
Normality assumption of the Y variable
Checking the normality assumption of the Y variable is often done visually through a histogram. A histogram is a bar-graph
where the range of continuous scores on a variable, are grouped into (small) categories on the X-axis (horizontal). The
frequency count of these categories are plotted on the Y-axis (vertical) to see how often different values occur. In new
versions of SPSS graphs can be built entirely from scratch to suit your specific needs, but here we will use the older built-
in default settings:
Graphs > Legacy Dialogs > Histogram > Select variable > To get a reference line, tick Display normal curve > Paste > Run
The SPSS snytax for a Histogram looks like this:
GRAPH
/HISTOGRAM = y.
Normality assumption of the Y variable per group
As the assumption about normality states, before we can perform the actual ANOVA, we should also inspect whether Y is
normally distributed within the different groups.
Graphs > Legacy Dialogs > Histogram > Select variable > To get a reference line, tick Display normal curve > and add variable
X in the Columns box by selecting the variable and clicking the correct arrow button.
2
Session 1: Assignment 1: SPSS, Syntax and Descriptive Statistics
Based on the measurement level of the variables, which method (that was discussed in the lectures), would be most
appropriate to analyze the relation between Individual Innovative Performance (Y) and the Company (X) that someone
works for? Answer: ANOVA
Syntax
The benefit of using syntax is that it provides a precise history of how the data was analyzed. Always aim to keep your
data file and your syntax file well-organized. This way you never have to save any output, because you can always recreate
it as needed by having SPSS perform all the syntax that you saved.
There are two ways to create a new syntax file:
1. In SPSS go to File → New → Syntax
2. Before performing any analysis instead of clicking the Ok button, click the Paste button
After starting a new syntax file, using the paste button for subsequent analyses will add the commands to the bottom of
the original file. You do not have to make a new file for each analysis. Clicking Paste will often close the dialog window.
Two ways of running the code is opening the same menu and clicking Ok this time (It will remember all your settings).
Another way is placing your cursor anywhere in the generated syntax and pressing Ctrl + R.
Frequencies
Analyze > Descriptive statistics > Frequencies > Select variable > Paste
The SPSS syntax for the Frequencies function looks like this:
FREQUENCIES VARIABLES = x
/ORDER = ANALYSIS.
So, for this step it should look like this:
FREQUENCIES VARIABLES=company
/ORDER=ANALYSIS.
SPSS does allow abbreviated functions, and works with any command shorter than its original as long as it is unique. This
means that this will also work:
FREQ VAR = x.
Descriptives
For continuous variables, the Frequency function is often not very useful, because many different scores are possible. If
many different values are possible, or the sample is not very big, each respondent will have its own unique value with a
frequency of one. A better alternative is using summary statistics for these variables, such as the mean, median, and
variance. Many different summary statistics for variables can be obtained using the Descriptives function.
Analyze > Descriptive statistics > Descriptives > Options > Tick boxes > Continue > Paste
The SPSS syntax for the Descriptives function looks like this:
DESCRIPTIVES VARIABLES=x
/STATISTICS=MEAN STDDEV VARIANCE MIN MAX.
SPSS does allow abbreviated functions, and works with any command shorter than its original as long as it is unique. This
means that this will also work:
DESC VAR=x.
Mean
Using the Descriptives function of SPSS we saw one way to obtain the mean for a variable. A second way is by using the
Means function. Use the Means function located under:
Analyze > Compare Means > Means
1
, SPSS
The SPSS snytax for the Means function looks like this:
MEANS TABLES=x
/CELLS=MEAN COUNT STDDEV.
SPSS does allow abbreviated functions, and works with any command shorter than its original as long as it is unique. This
means that this will also work:
MEAN VAR=x.
The Means function is not only there to provide the mean for one variable. It is possible to provide an additional
independent variable. Keep the variable Y in the list for the dependent variables and add the variable X to the list of
independent variables by selecting it in the left panel and clicking the arrow next to the list of independent variables. As
you can see, for a categorical independent variable, it now computes the mean for each variable X separately.
The SPSS syntax for the Means function using an independent variable looks like this:
MEANS TABLES=y by x /CELLS=MEAN COUNT STDDEV.
SPSS does allow abbreviated functions, and works with any command shorter than its original as long as it is unique. This
means that this will also work:
MEAN VAR=y by x.
Can we, based on this information, say anything about the differences in the entire population under study? (Assume that
these are very big companies and we have only observed a small part of their employees.) Answer: No - We are only
looking at sample data, and we need further tests to say anything about the population
Assignment 2: ANOVA
Before performing any analysis on the data, several assumptions (properties that the data should adhere to) must be
checked. This to see whether the analysis method will provide the correct results and we are not analyzing data that we
may not apply ANOVA to. The assumptions for an ANOVA are:
1. Correct measurement level of the X and Y variables; quantitative dependent variable of interval/ratio (continuous)
measurement level, independent variable of nominal measurement level (categorical).
2. Normal distribution of scores on the Y variable (for the full dataset, and within each group)
3. No outliers
4. Approximately equal variances in all the groups
5. The observations should be independent (within- and between groups)
Normality assumption of the Y variable
Checking the normality assumption of the Y variable is often done visually through a histogram. A histogram is a bar-graph
where the range of continuous scores on a variable, are grouped into (small) categories on the X-axis (horizontal). The
frequency count of these categories are plotted on the Y-axis (vertical) to see how often different values occur. In new
versions of SPSS graphs can be built entirely from scratch to suit your specific needs, but here we will use the older built-
in default settings:
Graphs > Legacy Dialogs > Histogram > Select variable > To get a reference line, tick Display normal curve > Paste > Run
The SPSS snytax for a Histogram looks like this:
GRAPH
/HISTOGRAM = y.
Normality assumption of the Y variable per group
As the assumption about normality states, before we can perform the actual ANOVA, we should also inspect whether Y is
normally distributed within the different groups.
Graphs > Legacy Dialogs > Histogram > Select variable > To get a reference line, tick Display normal curve > and add variable
X in the Columns box by selecting the variable and clicking the correct arrow button.
2
