I am looking for a statistical test procedure to check whether within subject reproducibility is higher for a certain condition compared to control condition, and if a certain kind of subject has a higher reproducibility in a repeated measure design compared to another group of subjects? As I know there is Cronbach's Alpha as a coefficient for internal consistency. As I know it has rather been used for validating questionnaires, but can it also been used as a method to proof consistency within and between subject  consistency in interval scaled data sets (e.g. maximum torque tested in one joint prior and following to a surgery)?