Cronbach's alpha coefficient can be used in a battery of tests to measure the extent to which the individual items show internal consistency. In the ideal situation the tests should all measure the same thing and they should be correlated with each other.


alpha =

= (number of test items in battery) / ((number of items in battery) – 1) * (1 – Y)


Y =

= SUM((standard deviation for each test)^2) / ((SUM(standard deviation for each test))^2)



• The number of items in the battery must be >= 2, or the value ((number of items) – 1) will be undefined.



• minimum alpha: 0, indicating that none of the items are related to another

• maximum alpha: 1, indicating that all of the items are identical

• An alpha coefficient of 0.7-0.8 is adequate for comparing groups.

• An alpha coefficient >= 0.90 is desirable for clinical applications.


To read more or access our algorithms and calculators, please log in or register.