Effects of Topical Interest and Mode of Address on Response to Mail Survey