SPSS Statistical Software

SPSS utilizes 2 windows while you're working in it. The first window is the Data file window and the other is the Output file window. These 2 windows work in tandem and you'll generally have them both opened as you work.

The Data file contains your data and allows you to perform different actions and analyses with the data. The Output file keeps a running record of everything you do to the data, and is also where any analysis results and visuals (e.g., charts, graphs, tables) will display.

Let's go through the specifics of both of these windows and their interfaces!

The Data File window contains your data and looks kind of like a spreadsheet. You can see your variable names at the top of each column and the values of each variable within those columns. Depending on what your data is, you may see numbers and/or text as the values.

The Data File window has 2 different views. You will notice 2 tabs at the bottom of the screen. One is labelled Data View, which is what we are currently in by default. The other tab is labelled Variable View. If you click on Variable View, you will see the following interface:

What is the difference between Data View and Variable View?

Data View:

• Shows your dataset.
• Looks like an Excel/CSV data file.
• Variable names are at the top of each column, with the values of each variable go down each column.
• Case number is on the left, this is your sample. Each number represents an individual case/participant. 
• The values in the columns represent each case/participant's data for a particular variable.

Variable View:

• Shows more detailed info about each of your variables.
• Let's go through what each column is:

1. Name = your variable names.
2. Type = what type of data it is. Like Numeric (which is numbers), String (which is letters or words), Dates, Dollars, or others. Click on the right-hand side of a cell to see a little button with 3 dots appear. Click the 3 dots and then you'll see a popup window that allows you to change the variable's Type (if the Type currently shown is incorrect).
3. Width = the number of digits or characters possible for the variable. So for example, Work_Field allows for 6 characters. If we go back to Data View, we can see the Work Fields have at most 6 characters, like HEALTH, (but can have less than 6, like GOV or TECH).
4. Decimals = how many decimal places you want for each variable. 
5. Labels = The next column is Labels. In contrast to Variable names, which are more messy shorthand names, Labels are clean, nicer looking names you can give your variables. Labels are also what will display in any tables or graphs you create. So if your variables have very non-descript names, like Q1 or Q2 or a series of random letters, be sure to add Labels so you don’t get confused when you’re running analyses.
6. Values = these are important when your data consists of Likert-based questions or demographic questions where the participants had multiple options to choose from (like Race/Ethnicity or Gender). For example, if you have questions where the scale was Strongly Agree to Strongly Disagree, but the responses were all recorded as numeric values 1-5, it can get confusing trying to remember what each number meant. This is where Data Values comes into play. You can assign meaningful text to each number, so when you run analyses, the output will display these text values next to the results so it’s easier to interpret.
   • Click on the right-hand side of a cell to see a little button with 3 dots appear. Click the 3 dots and then you'll see a popup window that allows you to add text names for each of your numeric values. You have to manually input each numeric value and associated text.
7. Missing = if you coded your missing data a specific way when you were cleaning your data, here is where you can tell SPSS what the code or value is that you’ve assigned to missing data. This way SPSS will take this into account when running analyses and not just consider these values as valid responses from participants which could then skew or bias your results.
   • For example, let’s say our dataset was missing some responses to the AGE question. If we already cleaned our data and entered -1 for all of those individuals who didn’t answer this question, here in this Missing Data column, we can add that information.
8. Columns = the column width of your variables on the Data View. This is purely aesthetic and doesn’t affect your data or analyses. You generally don’t need to touch this.
9. Align = another completely aesthetic option; this just lets you choose the alignment of how your variables appear in their columns in the Data Viewer (Left, Right, or Center).
10. Measure = is the Measurement Level of the variable. It tells you how each variable was measured – nominal, ordinal, or scale. Nominal data is categorical in nature, so we see for example that Gender and Ethnicity are both nominal because these variables are just categorizing genders and ethnicities, they’re not measuring anything numerically.
    • Ordinal variables are similar to Nominal as they are just categories but these categories have a rank order to them, meaning there is hierarchy to the response options. An example could be Education Level because a Bachelor’s degree is a higher level of education than an Associate’s Degree which is higher than a High School Diploma, so there is a hierarchy to which is more or less Education.
    • Scale variables are variables that use numbers in the traditional sense versus just for categorizing data. So for example, Age is scale, and so is Years Employed because these numbers measure amount of time in years, both of which involve using numbers in quantitative manner.
11. Role = this is for defining what role each variable will play in your analyses. By default they should all be set to Input (which means that the variables will be used as a predictor or independent variable).
    • Target means the variable will be used as an outcome or dependent variables.
    • Both means the variables will be used as both a predictor and outcome.
    • None is if the variable has no role assignment.
    • Partition means the variable will partition the data into separate samples. You won’t typically need to use this.
    • Split is not used here, it’s used in the SPSS Modeler which is something separate that is not covered in this guide.
    • As a note, Typically, you do not change Role from Input on any variables because it’s not necessary - you can still run any analyses on any of your variables and SPSS does not care which is your predictor or outcome.

As a final note here, all of these options in the Variable View are adjustable. If you want change the number of decimals for a variable, you can do that here. If you imported data and it categorized some of the Measurement levels incorrectly, you can fix that here.

Coming soon!