Selection bias from sampling frames: telephone directory and electoral roll compared with door—to—door population census: results from the Blue Mountains Eye Study

Abstract
No abstract available