SPSS: Difference between revisions

The educational technology and digital learning wiki
Jump to navigation Jump to search
Line 47: Line 47:
* [http://www.indiana.edu/~statmath/stat/spss/ SPSS] (Indiana). Includes some getting starting tutorials plus links.
* [http://www.indiana.edu/~statmath/stat/spss/ SPSS] (Indiana). Includes some getting starting tutorials plus links.
* [http://www.lrz-muenchen.de/~wlm/wlmspss.htm Internet Guide to SPSS for Windows]. Useful tutorials for command syntax, data transformation and data analysis.
* [http://www.lrz-muenchen.de/~wlm/wlmspss.htm Internet Guide to SPSS for Windows]. Useful tutorials for command syntax, data transformation and data analysis.
* [http://www.unt.edu/rss/class/Jon/SPSS_SC/ SPSS Short Course] (UNT)


=== SPSS command line ===
=== SPSS command line ===

Revision as of 16:00, 3 December 2013

Draft

<pageby nominor="false" comments="false"/>

Definition

SPSS stands for Statistics Package for the Social Sciences and is probably the most popular general statistics packages in academia.

This page just contains a few links we found useful ....

Important tips for beginners

Before you start analysing anything:

(1) Verify the data, e.g.

  • list minima and maxima for each variable or directly look at the raw data for small samples
  • Use for example: Analyze -> Descriptive Statistics -> Descriptives
  • Descriptives also may help you getting an idea about distributions: In the Options, tick all...

(2) Document the data and declare missing values (if needed)

  • Open the data editor -> variable view
  • Verify the type for each variable you plan to use (mostly it should just be "number")
  • Important: Declare measure, i.e. either nominal, Scale or ordinal. If you get this wrong, SPSS will not let do certain types of analysis or let you do it when it should not ...
  • Define a name (usually this should match questionnaire numbers), and label that defines the description
  • Important: Declare missing values. These will exclude a case from analysis. By default empty cells are already declared missing. Sometimes people use 999 as missing values and that should be declared !

(3) Configure working directory and turn on logging

  • Open Edit -> Options -> File Locations tab
  • Important: Tick Record syntax in journal. This will allow you to redo some analysis (and you can gain days of work if you understand how to do this).
  • Open the Output Labels tab and define how you would SPSS like to display the results. I usually want both variable (short) names plus the labels.

(4) Change language to writing language

  • Open Edit -> Options -> General tab

Links to tutorials and reference

How to use SPSS

At SPSS

These tutorials also may teach statistics.

SPSS command line

Statistics tutorials that include some SPSS hints

These are mainly statistics tutorials, but come with an explanation (i.e. the command line sequence) on how to do it with SPSS.

Links of links

Links