I think it depends on what kind experiment are u doing. If it's perception experiment which will not be influenced by subject's background, it's ok to have psy students. If their expectation resulting from their knowledge about experiment psychology is possible to affect ur result, it's better to have some "fresh subjects".