Delays and user performance in human-computer-network interaction tasks

This article describes a series of studies conducted to examine factors affecting user perceptions, responses, and tolerance for network-based computer delays affecting distributed human-computer-network interaction (HCNI) tasks.

HCNI tasks, even with increasing computing and network bandwidth capabilities, are still affected by human perceptions of delay and appropriate waiting times for information flow latencies.

Conducted were 6 laboratory studies with university participants in China (Preliminary Experiments 1 through 3) and the United States (Experiments 4 through 6) to examine users' perceptions of elapsed time, effect of perceived network task performance partners on delay tolerance, and expectations of appropriate delays based on task, situation, and network conditions.

Results across the six experiments indicate that users' delay tolerance and estimated delay were affected by multiple task and expectation factors, including task complexity and importance, situation urgency and time availability, file size, and network bandwidth capacity. Results also suggest a range of user strategies for incorporating delay tolerance in task planning and performance.

HCNI user experience is influenced by combinations of task requirements, constraints, and understandings of system performance; tolerance is a nonlinear function of time constraint ratios or decay.

Appropriate user interface tools providing delay feedback information can help modify user expectations and delay tolerance. These tools are especially valuable when delay conditions exceed a few seconds or when task constraints and system demands are high. Interface designs for HCNI tasks should consider assistant-style presentations of delay feedback, information freshness, and network characteristics. Assistants should also gather awareness of user time constraints.