SPSS stands for Statistical Package for the Social Sciences. It is a general purpose statistical and data management program. There are many different things that SPSS can be used for of which we will do just a few in the class. This document covers a few of the basics. If you go on in the social sciences you will almost certainly learn more of either SPSS or SAS another popular statistical package. At some point you may want to start the SPSS tutorial; this page will focus on some more specific issues.
SPSS is a windows program. When you go into windows you may see an SPSS icon. If you don't see it click on Start, then Programs, then SPSS. You will see a list of SPSS items. Click on the one that says "SPSS 10.0 for Windows"
When you open SPSS the first thing you will get a dialogue box. The simplest thing is to click cancel and then follow the instructions below. (After you use SPSS for a while you can play around with the dialoguebox.)
Next you will see a blank spreadsheet (a grid of rectangles). The spread sheet is where the data to be analyzed are stored. It is blank because you have not yet put any data in. It will say "Name" in the upper left hand corner. At the bottom of the spread sheet you will see two tabs, "variable view" and "data view."
You will probably initially be in the "variable view."
Along the top of the screen you will see a series of works:
File, Edit, View, Data, Transform, Analyze, Graphs, Utilities, Window, Help
This will be referred to as the main menu.
If you want to use a pre-existing dataset you click on File on the main menu, then select Data and Open. This will give you a list of files called XXX.sav with the XXX a real name. Usually it will be the list of files on the k drive. If you see the file you want, double click on it.
The general social survey data set stored on the l drive is gss96.sav
To run simple statistics Click on analyze on the main menu, then on Descriptive Statistics. Within the Descriptive Statistics menu, Frequencies will give you simple frequency distribution, Descriptives will give you simple means, Crosstabs will give you tables.
All of the procedures follow the same basic pattern. On the left you will see a box with a list of variables. On the right you will see an empty box or boxes (unless you have already run a procedure, in which case you will see what ever you did before). Between them you will see a gray square with a triangle. This is supposed to represent an arrow. If you click on a variable name in the left box and then click on the arrow, that variable will "move" to the right box (and vice versa).
If you are running a Cross Tabulation (crosstab) you will need to move a variable to the row box and another to the column box. (Conventionally, the independent variable goes in the columns and the dependent variable goes in the rows).
All of the procedures have various options that may be selected by clicking on various buttons. For cross tabulations you will probably want to use the Cell button to select column percentaging. For other procedures you probably do not need to worry about the options, although they are worth learning about.
When you are finished selecting the variables and any options,
click on continue. The output from the program will appear
in the "output window." To print the output file click
on the picture of the printer.
To create a new data set you work in the black spread sheet that appears on the screen when you first open SPSS. There are several steps: Defining variables, entering the data, and saving the data.
The first step in inputting your new data set is to define each variable. In general, each question in your questionnaire will correspond to a variable in SPSS. If you are entering official statistics, each type of statistic (e.g. the monthly unemployment rate) will correspond to one variable in SPSS. Each variable is a column in the grid.
Generally, before starting to input your data you should go through your survey and number the questions and within each question number the answers. It is also possible to use letters, but most people find numbers easier.
Go to the Variable View.
Each row will be one variable. Type the name of the variable (8 or fewer variables) under Name. . For example AGELAST would be the variable name for age at last arrest. Or, EDUC would be the variable name for years of education.
Adding labels
Next, add labels by typing the label in the column named "label." The variable label is a longer description of what the variable is about. "Years of education" would be the variable label for EDUC. "Age at last arrest" would be the variable label for AGELAST.
In many cases you will want to add value labels. These are descriptions of what each of the answers to the survey are. For example, if a person answers "strongly agree" that answer may be coded as a 5. However, you would still like the information about what 5 means to show up when you do statistical analysis, such as making tables. In general you will not do variable labels for interval variables such as a person's years of age or number of children because the label would not provide any additional information. To enter value labels, the column called Values. When you click onthebox, a small gray boxwill appear. Click on that box and a dialogue box will appear. First enter the value (the number corresponding to the answer) in the box marked value, then enter the label (what the answer means) in the box marked label. So the value of 5 would correspond to a value label on strongly agree. Then click on the Add button. The label you entered will move into the box, and you can repeat the process until all of the values have labels.
What if I have variables with the same value labels or missing values?
You can copy and past value labels, missing values information within the variable view. The only thing this will not work with is the variable name, which must be unique.
Once your variables are defined, switch to the "Data View."
Each row in the grid will generally represent one observation, for example a single survey or information on a single state. After you have defined the variables you simply type in the value for each variable in the appropriate row and column. You can either start with the first variable of the observation and work your way across the page to the last observation, then go to the next row and do the same for the second observation. Or you can type in the first variable for every observation, then the second and so on. Do which every is easier or more convenient.
Don't forget to proof read after enetering data.
When you have finished entering data, you can save the data set by clicking File/Save As and then entering a data set name. This should be xxxx.sav where you fill in the xxxx with something useful. If you are working in the lab, make sure that you point the file to your floppy disk (the A:\).