INCORPORATING CHECKS AGAINST OBSERVER ERROR INTO IDENTIFICATION KEYS

Abstract

A method is suggested for checking for errors that may be made when an identification (or diagnostic) key is used. Such errors may arise either from tests being incorrectly used or observed, or because the specimen examined is aberrant or incomplete. The method involves constructing a check key for each endpoint of the main key. Use of the appropriate check key confirms or corrects the initial identification. Copyright © 1977, Wiley Blackwell. All rights reserved