Mary Jackson Pitts, Ph.D.

mpitts@astate.edu

[ Home ] [ Up ] [ getting started with spss ] [ frequencies and percentages ] [ chi square ] [ recoding data ]

Using SPSS with a PC for version 12

SPSS is a valuable tool to assist research with the data crunching.   SPSS for PC allows for quick assimilation of data and should be used to help the researcher interpret the results from the data collection instrument (survey or coding instrument or experimental tools.

To begin use of SPSS one should have the variables that are to be examined in the study available.  Having your survey or coding instrument makes it easy to keep up with your variables.  You should have variable labels and value labels for the scale of measurement.  I like to have variables 1X to whatever and then each is assigned a variable label. 

When you open the software for SPSS, a screen will appear which asks “what would you like to do?”   You should at that point check the box that says type in data.  Since the computer software package knows nothing, you should begin by defining your variables. 

Once you click on Type in data, you will see a window called data editor.  Below you will see a table which helps you define the variables.  You will be interested in providing name, label values to this window.  Each of your variables should be identified on this screen.   There will be a tab at the bottom of the screen that will say data view and variable view.  The variable view will allow you to identify each variable, provide a label for the variable and provide the values for the scale of measurement.   The data view window allows should be used when you get ready to enter your data.

Every thing that you measure in the survey or content analysis should be assigned a number.  The numbers typify the main reason we call this quantitative data.  We are counting frequencies of occurrences.  Each number is assigned a meaning.    We use mostly nominal and ordinal data in our research for this class.  You will recall nominal means we assign something an arbitrary number.  Ie 1= male and 2=female.    Ordinal data is found within continuums, ie 1=small 2=medium, 3=large.    While interval data is often used in the case of Likert statements .  It is important to understand the difference in how the data was collected.  Some statistical measurements are determined by the type of data collected.

For the variable you should define the variable by providing an eight character name for the variable.  You can use a number in the variable label, but the numeral should never start the label.    Variable labels should attempt to describe the variable.  Ie  location of a newspaper….”location” can be used to provide a variable name.   The exact purpose of the variable can be labeled in the label section on the data editor screen.   Values refers to the scale used to measure.     Ie   1= small, 2=medium 3=large.

When working with the value labels a screen will appear that has

Value label 

Value                                 

Value label                      

Add                                 

Change

Remove

Once variables are named, defined, and value labels assigned, the researcher can begin entering data into the data editor.   Click on the tab that shows data view and you are ready to type in your numbers.  

Type in your data according to your survey or coding instrument.  To save the file save the data by going to file in the drop down menu at the top of the screen.   Give the file a name.  Make sure to attach a      .sav extension to your file name.   Press enter.   The data is now saved.  Always remember to save data to your disk, too.  When you are working with the data be sure to also save your data to a disk.  A zip disk is best.   When coming back to the data, always transfer data to the hard drive before running the statistical measurement.

The joy of SPSS is that you don’t have to do all the calculations by hand, many of which could take days to calculate.  There are formulas for all the statistical measurements, but you do not have to know them all. 

We are most interested in this class at running frequencies, percentages and chi-square.   

The next phase of analysis comes from running the data. 

In the data editor screen, there are several pull down menus.  Go to the analyze pulled down.  Click and hold.  You will see a list of possible analysis which can be used.  Drag your cursor to the frequencies slug and hold.  

Within frequencies, find central tendencies and check on mean, median, mode, and sum.

Within the same screen, within dispersions, check st. deviation, variance, range, minimum, maximum, se mean. 

Alright, you are now ready to click ok and run your data.   Run your frequencies at this point.  You will get output that you should examine for typos and so forth.   Missing numbers, values typed in that do not exist.  This is essentially called cleaning up your data.  Once you have done so you need to save your data file again, get rid of the previous output then run the frequencies over again.   Either print the output or save to a disk and examine.   

Crosstabs….   Those of you doing comparisons between groups will use the crosstabs feature to examine your data.   Under the pull down menu Analyze you will find within the choices a link to crosstabs.  Click on crosstabs, a screen will appear that ask you what  variables you will be examining.  You will identify independent variables which will often go in the row area of the screen, while dependent variables often go in the column section of the screen.

     Once your variables are chosen, click on the statistics link, a screen will appear.  At the top of that screen check the chi-square box, at that point, click continue,.   Now click on cells, a new screen will appear, check all the boxes in counts and percentages, then click continue.

Now click on the box that shows OK.   Your output will appear.  

Run frequencies first.   Examine all data for typos and the like, clean data up, run frequencies again.  Do cross tabs.