I am want to find out how certain variables can produce an effect on the internal consistency (reliability) of the 36-Item short form survey (SF-36). I have chosen this survey since it is one of the most used for HRQL studies, and there is already an amount of studies describing its consistency.

However, it is not described how do you go about calculating Cronbach's alpha por each item and for the whole scale.

Can anybody explain to me the process of doing so? or know of a paper describing the method?

Thank you for your help.