|
|
|
Module IV: Data Analysis
| If you are familiar with navigation and data entry in SPSS and want to move directly to data analysis, you are in the right place. If you have not worked with SPSS before, it is recommended that you learn some navigation and data entry techniques before proceeding (click here to do so). For all users, it is suggested that you have an active SPSS data set to use while working through this module. It can be your actual data or one of the data sets available in SPSS (to learn how to access them click here). It is further suggested that you run descriptives on your data yet before continuing. To refresh your memory on how to run descriptive statistics, click here. |
The purpose of this module is to help you determine which statistical test might be best suited for your data and research question, as well as to give you some guidance on performing that test.
Many considerations go into choosing the appropriate statistical test. One of the most fundamental is the type of data to be analyzed. Not all statistical tests can be run on all types of data. To determine the type of data you have--which will in turn determine the statistical test you are likley to run--consider the following:
- Are you interested in comparing mean scores on a dependent variable between two or more groups? Do you want to know if there is a (statistically significant) difference between or among groups on a particular measure? Does your study design include at least two groups, each of which is treated differently in some fashion than the other(s)? If you answered yes to these questions you probably have experimental data and will likely use t-tests or analyses of variance (ANOVAs) to analyze your data. Click here to learn more about these tests.
- Are you interested in subjects' opinions, perceptions, or feelings about something? Does your study design include a survey or Likert scale response items? Do you suspect there is a relationship between two or more variables that you wish to investigate? Answering yes to these questions suggests that you have non-experimental data. In this case you will probably want to compute correlation or regression coefficients. Learn to do this by clicking here.
- Is one or more of your variables nominal (categorical) or ordinal (rank ordered) in nature? Do you have a small number of subjects or cases, say, 20 or less? If so, you will probably run non-parametric statistics. Click here to learn how this is done.
|
|
|
