In this session the following will be covered:
– Introduction to SPSS
– Management of data in the SPSS environment
– Graphs etc
1. Introduction
To start SPSS locate and click on the SPSS icon. Retrieve an SPSS data file by clicking on the button in the toolbar.
Once you select the appropriate file click Open. The Data Editor window will open and you will see the dataset. The variable names appear in the boxes along the top and the cases appear down the left hand side, numbered sequentially. You can navigate using the arrows to see more variables and cases.
By clicking on the variable view tab at the bottom of the Data Editor window you can see the details of the variables in the data view. Each row represents one variable in the order they appear in the Data View. Here you can see and change the characteristics of the variables (e.g. add which question each variable relates to). It is important to keep this section up to date and organised as the more variables you have (or new ones you create) the bigger the dataset and the more confusing it becomes to remember and interpret each variable when you begin your analysis. Please note that any new variables that you add will automatically appear here.
2. Adding Labels
As mentioned above, it is important to keep your dataset tidy, especially if you are dealing with large datasets. In the variable view box, you can name (and change the name) of your variables anytime. There are two kinds of label that can be applied to each variable: variable labels and value labels. Variable labels expand on the variable name (e.g. PSM: Public Service Motivation, or ConAtt: Consumer attitudes) and they can tell you what questions were asked. Value labels can tell you what the numerical code given to each response means.
Example: Sex is the label given for ‘Sex of Respondent’ and ‘Female’ and ‘Male’ could be the label given to code 1 and 2
Variable labels: just click on the Cell and you will be given the option to type in a name of your choice
Value labels: This column will initially have the entry None. To add Value Labels, select the Values cell and click on the dotted button to open the Value Labels dialog box. Each value is typed into the Value by, followed by its label in the Value Label box. Finally, the Add button is clicked. When you finish adding the value labels, click on OK.
3. Missing values
Your next task must be to check your dataset for missing values. Missing values result when data has not been input correctly (when you are manually inputting the data) or when respondents have not answered a question. In the first case, go back to your questionnaires and check if you have a response for that question and if so, add it. In the second case, you can ask SPSS to code specify that the response is missing and assign a specific number to the missing response. To add Missing Values for a variable select the Missing cell for the variable in the Variable View and click on the dotted button. The missing values dialog box appears (as below).
You can specify up to three separate missing values, for example the values missing could be
0 = Not Applicable
8= Don’t know
9= No response
4. Recoding Data
One of the most common reasons to recode is when you have revered score items, for example:
Please indicate the extent to which you agree with the following statements Strongly disagree Somewhat disagree Neutral Somewhat Agree Strongly Agree
AC1 I would be very happy to spend the rest of my career in this organisation
AC2 I really feel as if this organisation’s problems are my own
AC3 I do not feel like ‘part of the family’ at my organisation
AC4 I do not feel ‘emotionally attached’ to this organisation
AC5 This organisational has a great deal of personal meaning to me
AC6 I do not feel a strong sense of belonging to my organisation
Statements AC3 and AC6 are reverse statements so should be re-coded to reflect this.
Non-Reveresed Items Reversed items
Strongly disagree= 1 1=5
Somewhat disagree= 2 2=4
Neutral= 3 3=3
Somewhat agree= 4 4=2
Strongly agree= 5 5=1
There are two ways to recode in SPSS
1. Recode into the same variable
2. Recode into a different variable
The first option has the advantage of being more simple, but you lose the original values whereas the second option you can retain the old values and create a new variable with the recoded values.
Procedure:
• Select: Transform ——- Recode—– Into Same Variable
• Select the Variable to recode
• Then Click on the Old and New Values button
• Insert the Old Value e.g 1 as in the table above
• Insert the New Value e.g. 5 as in the table above
• Click Add
• Repeat the process for the rest of the values (see table above)
• Click Continue
• This returns you to the Recode into Same Variables dialog box
• To Finish click OK
5. Selecting Cases
This is a useful tool if you wish to perform an analysis on a subset of cases, e.g. only women, or only EU students. To select a subset of cases you select: Data and then Select Cases
The following dialog window appears:
The default selection is All Cases. To select a subset of cases, select the If condition is satisfied option and click on the If button. The following box will appear:
What you need to do then is create a condition. You can do this based on how you coded your variables.
For example, say you only want to include women and you have coded 2= Female and 1= Male. The variable name is RSEX. Your condition is then rsex=2. So you are asking SPSS to keep those observations (or cases, or respondents) that are female, i.e. they were coded as 2.
You can have more than only conditions: e.g.
rsex=2 and feestatus= 2 (here for example you have selected female students that pay EU fees)
Here is what your dataset will look like if you only keep female respondents. As you can see these cases have been crossed out and will not be taken into account in any analysis that follows but remain in the dataset.
You can unselect the cases by clicking in the Reset button at the Select Cases tab.
6. Descriptive statistics
Select: Analyze, Descriptive Statistics, Descriptives
This produces the Descriptives dialog box (below). Scroll down the variables list on the left-hand side, select it and move it to the right-hand side.
Click on Options to get the options dialog box (below). The default options are marked with a tick and they can be removed and others added by ticking and unticking. Once you have selected the descriptives you want to use then you click Continue and return to the Descriptives dialog box and click OK.
Here is what your output will look like:
N is the total number of cases (or observations) you have in your dataset (or that have been analysed), and you can also see min & max values, the mean and standard deviation because this is what we chose in the dialog box above.
7. Graphs
Graphs on SPSS are thoroughly and very clearly covered in the Field textbook, so please refer to that. Please note that there is a dedicated Graphs tab on the main SPSS window. The entire selection of graphs you can product can be found under Legacy Dialogs.
8. Crosstabulations
This is a particularly useful way to describe the relationship between two variables.
Click Analyze, Descriptive Statistics, Crosstabs. Select the variables that you would like to crosstabulate and move them from the left to the right boxes (and define which one you want in row and which in the column). Click OK. (PS: under the statistics tab one can ask SPSS to produce other tests such as correlations and Chi-square, but more of these next week)
The resulting output will look like this:
9. Averaging different questions into one variable (=scale)
In some disciplines, mainly those that measure people’s attitudes in their research (e.g. marketing, human resource management) involve concepts that are assessed using scales. These scales involve different questions that are measuring the same variable. Below is an example of the scale used to measure organisational commitment. In your dataset, you will have responses for each question, but then you need to aggregate these responses into one variable (organisational commitment in the example below). We aggregate by taking the mean of all the sub-variables.
Organisational Commitment
Please indicate the extent to which you agree with the following statements Strongly disagree Somewhat disagree Neutral Somewhat Agree Strongly Agree
C1 I would be very happy to spend the rest of my career in this organisation
C2 I really feel as if this organisation’s problems are my own
C3 I do not feel like ‘part of the family’ at my organisation
C4 I do not feel ‘emotionally attached’ to this organisation
C5 This organisational has a great deal of personal meaning to me
C6 I do not feel a strong sense of belonging to my organisation
What In SPSS this is done in the following way:
Click Transform, Compute Variable
In the Target variable box enter the name of the new variable: e.g. Commitment
In the numeric expression box type: mean and then manually enter the variables from the left hand side in brackets, separated by commas. Then click OK. In the above example these would be
mean(C1, C2,C3, C4, C5, C6)
The table below illustrates this.
To ensure that this has worked, go to the Variable View tab and at the bottom you should see your new variable.
10. Scale reliability
Scales usually measure attitudinal constructs and they are developed using statistical procedures such as factor analysis. If you are using such scales, it is important to check its reliability. Reliability refers to the ability of a measure to produce consistent results when the same subjects are measured user the same conditions. This means that other things equal, a person should get the same score on a questionnaire if they complete it at two different points in time (known as test-retest reliability). Cronbach’s Alpha is the test that you need to use and SPSS does all the work for you. All you need to do is interpret the findings. We will not cover this today, but for more information and step-by-step guidance on how to run the test please consult Field pp. 666-677
