ARCHIVED: In SPSS, how do I compute Cronbach's alpha statistic to test reliability?

Suppose you wish to give a survey that measures job motivation by asking five questions. In analyzing the data, you want to ensure that these questions (q1 through q5) all reliably measure the same latent variable (i.e., job motivation). To test the internal consistency, you can run the Cronbach's alpha test using the reliability command in SPSS, as follows:

RELIABILITY /VARIABLES=q1 q2 q3 q4 q5.

You can also use the drop-down menu in SPSS, as follows:

1. From the top menu, click Analyze, then Scale, and then Reliability Analysis.
2. Transfer variables q1 through q5 into the Items, and leave the model set as Alpha.
3. In the dialog box, click Statistics.
4. In the box description, select Item, Scale, and Scale if item deleted. In the inter-item box, select Correlation.
5. Click Continue and then OK to generate the output.

To interpret the output, you can follow the rule of George and Mallery (2003):

 > .9 (Excellent), > .8 (Good), > .7 (Acceptable), > .6 (Questionable), > .5(Poor), and < .5 (Unacceptable)

Notes:

• Cronbach's alpha reliability coefficient normally ranges between 0 and 1.
• The closer the coefficient is to 1.0, the greater is the internal consistency of the items (variables) in the scale.
• Cronbach's alpha coefficient increases either as the number of items (variables) increases, or as the average inter-item correlations increase (i.e., when the number of items is held constant).

To run the same test in Stata or SAS, see ARCHIVED: In Stata, how do I compute Cronbach's alpha statistic to test reliability? or ARCHIVED: In SAS, how do I compute Cronbach's alpha statistic?