I am conducting some analyses with a composite variable constructed by averaging 4 separate measures. I would like to know the reliability of the composite variable. Is Cronbach's alpha appropriate in this case (even though I have entire scale scores instead of item scores), or is a different measure more appropriate?