Suchergebnisse

Three basic tasks were designed to measure how fast a user could position a cursor at various locations on an electronic display. The tasks were arbitrary cursor positioning, sequential cursor positioning, and check-reading for errors. Three cursor positioning devices were tested. Both lightpen and lightgun permitted faster cursor positioning than a poorly designed keyboard. These pointing devices were four to five times faster than the keyboard for the arbitrary positioning task, and about twice as fast for the other tasks.

Twelve subjects scanned displayed prose passages to detect randomly inserted letter substitution errors. In comparison with steady displays, this check-reading task was performed 10% more slowly but with equal accuracy when the displayed material blinked at a 3-Hz rate. Although the potential value of blink coding confirmed in previous research is not questioned, the reduction in readability of blinking displays demonstrated in the present study suggests that some precautions must be taken in the practical application of this display coding technique.

Ten men scanned cathode-ray-tube displays to search for designated target items. Search time increased with increasing display density, and was 50% faster when a 3-cps blink was imposed on items of the target class. The blinking of nontarget items was almost equally effective, indicating that blink can be used either as an inclusion or exclusion code. In comparison with steady, nonblinking displays, search time was not significantly different for displays in which all items blinked, or in which a random selection of items blinked. It is concluded that for visual search tasks of the kind reported here, there is no marked deleterious effect of blink coding on symbol legibility, and no marked interference attributable to irrelevant blink.

Computer analysis indicates the feasibility of single-stroke keying for data entry of alphabetic names using the standard touch-tone telephone pad. In many practical job situations little ambiguity will result from the single-keying technique, and this ambiguity can be resolved under computer control.

Large-scale use of talking computers may be anticipated for reasons of general convenience and special effectiveness. A telephone link to a computer, with touch-tone inputs and stored speech outputs, poses a number of design problems of interest for the human factors specialist. Recommended features include (1) user pacing and optional repetition of speech output in a transactional sequence, (2) use of different voices and other auditory coding to distinguish types of output, and (3) use of tone codes to indicate required input.

An evaluation of on-line computer-aided instruction within a management information system compared typewriter and lightpen input modes as students learned to use the system. The following conclusions were supported: 1. Computer-aided instruction is feasible in a general-purpose management information system; 2. It is also feasible to demonstrate retention of learned material through computer-administered tests; 3. Professional typing skills are not necessary to use the typewriter input mode effectively, provided that the inputs required are short and direct; 4. In this particular setting there seemed to be evidence that the typewriter was a more effective input device than the lightpen during the instructional sequence, but this evidence is questionable in view of considerable individual differences among the students; 5. Regardless of performance, students responded favorably to their experience with computer-aided instruction.