2008, Neth, Hansjörg, Khemlani, Sangeet S., Gray, Wayne D.

Objective:
We distinguish outcome feedback from control feedback to show that sub- optimal performance in a dynamic multitasking system may be caused by limits inher- ent to the information provided rather than human resource limits.

Background:
Tardast is a paradigm for investigating human multitasking behavior, complex system management, and supervisory control. Prior research attributed the suboptimal perfor- mance of Tardast operators to poor strategic task management.

Methods:
We varied the nature of performance feedback in the Tardast paradigm to compare continuous, cumulative feedback (global feedback) on performance outcome with feedback lim- ited to the most recent system state (local feedback).

Results:
Participants in both con- ditions improved with practice, but those with local feedback performed better than those with global feedback. An eye gaze analysis showed increased visual attention directed toward the feedback display in the local feedback condition.

Conclusion:
Pre- dicting performance in the control of a dynamic multitasking system requires under- standing the interactions between embodied cognition, the task being performed, and characteristics of performance feedback. In the current case, at least part of what had been diagnosed as a deficit caused by limited cognitive resources has been shown to be data limited.

Application:
Perfect outcome feedback can provide inadequate control feedback. Instances of suboptimal performance can be alleviated by better feedback de- sign that takes into account the temporal dynamics of the human-system interaction.

2007-08, Payne, Stephen J., Duggan, Geoffrey B., Neth, Hansjörg

When participants allocated time across 2 tasks (in which they generated as many words as possible from a fixed set of letters), they made frequent switches. This allowed them to allocate more time to the more productive task (i.e., the set of letters from which more words could be generated) even though times between the last word and the switch decision ("giving-up times") were higher in the less productive task. These findings were reliable across 2 experiments using Scrabble tasks and 1 experiment using word-search puzzles. Switch decisions appeared relatively unaffected by the ease of the competing task or by explicit information about tasks’ potential gain. The authors propose that switch decisions reflected a dual orientation to the experimental tasks. First, there was a sensitivity to continuous rate of retur - an information-foraging orientation that produced a tendency to switch in keeping with R. F. Green’s (1984) rule and a tendency to stay longer in more rewarding tasks. Second, there was a tendency to switch tasks after subgoal completion. A model combining these tendencies predicted all the reliable effects in the experimental data.

2007, Gray, Wayne D., Neth, Hansjörg, Schoelles, Michael J.

2006, Neth, Hansjörg, Khemlani, Sangeet S., Oppermann, Brittney, Gray, Wayne D.

Tardast (Shakeri 2003; Shakeri & Funk, in press) is a new and intriguing paradigm to investigate human multitasking behavior, complex system management, and supervisory control. We present a replication and extension of the original Tardast study that assesses operators’ learning curve and explains gains in performance in terms of increased sensitivity to task parameters and a superior ability of better operators to prioritize tasks. We then compare human performance profiles to various artificial software agents that provide benchmarks of optimal and baseline performance and illustrate the outcomes of simple heuristic strategies. Whereas it is not surprising that human operators fail to achieve an ideal criterion of performance, we demonstrate that humans also fall short of a principally achievable standard. Despite significant improvements with practice, Tardast operators exhibit stable sub-optimal performance in their time-to-task allocations.

2008, Neth, Hansjörg, Müller, Thomas

Taking a lead from existing typologies of actions in the philosophical and cognitive science literatures, we present a novel taxonomy of actions. To promote a notion of epistemic agency we distinguish theoretical (mental state-directed) from practical (world-directed) actions. Our basic structural unit is that of a teleological frame, which spans one specific goal of an agent. Relative to a given teleological frame, actions can be classified as focal (directed towards the end) or ancillary (directed towards a means). The framework is applied to further illuminate previous attempts to distinguish between pragmatic and epistemic actions (Kirsh & Maglio, 1994). Physical actions that substitute or support mental processes are reclassified as practical ancillary actions that are strategically contingent alternatives to theoretical actions.

2007, Carlson, Richard A., Gray, Wayne D., Kirlik, Alex, Kirsh, David, Payne, Stephen J., Neth, Hansjörg

2007, Neth, Hansjörg, Myers, Christopher W.

2008, Schlimm, Dirk, Neth, Hansjörg

To analyze the task of mental arithmetic with external representations in different number systems we model algorithms for addition and multiplication with Arabic and Roman numerals. This demonstrates that Roman numerals are not only informationally equivalent to Arabic ones but also computationally similar - a claim that is widely disputed. An analysis of our models' elementary processing steps reveals intricate trade-offs between problem representation, algorithm, and interactive resources. Our simulations allow for a more nuanced view of the received wisdom on Roman numerals. While symbolic computation with Roman numerals requires fewer internal resources than with Arabic ones, the large number of needed symbols inflates the number of external processing steps.

2007, Myers, Christopher W., Neth, Hansjörg

2007, Neth, Hansjörg, Sims, Chris R.