REVISING DOCUMENTS WITH TEXT EDITORS, HANDWRITING-RECOGNITION SYSTEMS, AND SPEECH-RECOGNITION SYSTEMS.

Twelve participants were given marked-up manuscripts to edit. Results showed that editing was done much faster (50% and 90%) with a simulated handwriting-recognition system and with a simulated speech-recognition system than with a full-screen text editor (Xedit) that participants had experience in using. With the text editor, participants spent considerable time deciding how to locate a revision on the screen, locating it on the screen, and sometimes rereading the document. These behaviors were not necessary with the simulated handwriting and speech systems, in large part because of differences in external feedback provided by the systems. Handwriting-recognition and speech-recognition systems, in addition to being useful principal-support workstations (as each is often envisioned), may also form the bases for potentially good text editors for typists and secretaries.