Cronbach's alpha

What is and when do you use Cronbach’s alpha?

Cronbach's alpha is a tool to establish if a scale can contain more than one item. This it tested by mutual correlation of the different items.


Example of Cronbach’s alpha

Is it possible to make a combination of item 1 (how happy are you in the morning?), item 2 (how happy are you in the afternoon?) and item 3 (how happy are you in the evening?) in one item; 'happiness'?


What's important when considering Cronbach’s alpha?

Cronbach's alpha only tests mutual correlation and is therefore only giving information on reliability of the scale, not about validity (do you really measure, what you claim to measure?). Cronbach's alpha gives a score between 0 and 1 and the tipping point is approximately 0.7. Always use the function 'Cronbach’s alpha if item deleted' to examine whether deleting an item gives a better scale. This way you can spot if you've forgotten to change a negative formulated item into a positive formulated item. Furthermore don't draw conclusions from a Cronbach's alpha from former research; it's possible your research group didn't understand a question while in former research they did understand this question.