Effects of Task Probability on Integral and Separable Task Performance

Abstract

Relative to bar displays, object displays have repeatedly elicited superior performance for tasks involving information integration. This has been attributed to the emergent feature that is present in such displays; however, the attentional strategy used by the subjects may also playa role. Most previous research has had subjects perform the information integration task much more frequently than the separable tasks. Therefore, subjects may have learned to attend to the information integration task more, and thus they performed it better than they did the separable tasks. The present experiment manipulates the probability of the occurrence of the information integration and separable tasks. As the probability of the integral task increased, absolute error increased for the separable task but remained constant for the integral task. Also, reaction time dropped sharply for the integral task as the probability of the integral task increased, but it rose slowly for the separable task. Before using object displays, designers of visual displays should consider the attentional strategy of the operator.

Depth of processing and design-assessment of ecological interfaces: Task analysis

“…This comes across clearly as soon as authors present their reasons for proposing several interfaces and display techniques: the Configural Display, which ''represents high-level constraints of the domain through the relationship among the low-level data that define the constraint'' (Bennett et al, 1993, p. 72). Thus, the polar star (Woods, Wise & Hanes, 1981) integrates the values of more than 100 sensors to represent the state of the process; the Representation Aiding Technique which, to assist anticipation behaviour, proposes helping the operator to assess ''where the level is, what factors are influencing the level (Shrink/swell vs. changes in the mass balance), and where the level will go given these influences and possible interventions'' (Bennett et al, 1993, p. 77); the Compensated Steam Generator Level put forward by Bennett, Woods and Haley (1986) is closely related to the higher levels of the abstraction hierarchy, although this technique is traditionally considered as a product of the representation aiding technique; the Integral Display or Object Display is also an example of this logic (Carswell Adapthya, Klaver, Kancler, Dolan & Wickens, 1987;Jones, Wickens & Deutsh, 1990;Elvers, et al 1993).…”

Human-machine system design optimization for nondeterministic spacecraft anomaly determination/resolution

“…Additionally, there was evidence o f what the researchers described as "cognitive lockup": once complex diagnosis began, operators ignored the opportunity to stabilize additional, simpler system faults and tended not to interrupt the original troubleshooting process. Elvers et al (1993) produced a factorial experiment similar to the one presented in Chapter 5 (Experiment Design), measuring reaction time and absolute error with students of the same level of knowledge. This variation is similar to that of Kerstolt's (1996) study, concentrating on overall performance with no variation of operator knowledge.…”

“…The current screens by which data is viewed are cumbersome, hard to read, and over saturates the operators with unnecessary data. According to Elvers, et al (1993), "the role of the human in many complex systems is as a system monitor. Rather than providing a single overall indication of their status, these systems tend to present the operator with numerous separate pieces o f information, none of which reflects overall status."…”