AR-Cues Change Users’ Strategy for Dealing with Deferrable Interruptions

Paper Title

AR-Cues Change Users’ Strategy for Dealing with Deferrable Interruptions

Publication Info

• Topic area: Human-Computer Interaction (HCI) and interruption management.
• Keywords: Augmented reality, task resumption, interruptions, coarse breakpoints, deferrable tasks, stress reduction, cognitive load, AR-cues, multitasking, interruption strategies.

Background and Problem

• Problem / challenge: Interruptions in work environments often disrupt task performance, increase cognitive load, and lead to errors or stress. Strategies like reaching coarse breakpoints help mitigate these effects but are time-consuming and effort-intensive. Existing solutions, such as environmental cues, have shown limited effectiveness.
• Significance: Managing interruptions effectively is critical in high-demand environments like healthcare, aviation, and emergency services, where errors and delays can have severe consequences.
• Motivation and related work: Previous research highlighted the benefits of coarse breakpoints and AR-cues for task resumption. However, the interaction between AR-cues and deferrable interruptions remains underexplored, particularly regarding their impact on cognitive strategies, task performance, and stress levels.

Solution

• Proposed approach: The study investigates how AR-cues influence interruption management strategies, task resumption performance, and stress levels in deferrable interruptions using a pill-sorting task paradigm.
• Novelty:
  1. Demonstrates that AR-cues reduce reliance on coarse breakpoints without eliminating the strategy entirely.
  2. Shows AR-cues significantly reduce resumption errors compared to coarse breakpoints.
  3. Provides evidence that AR-cues lower stress levels associated with frequent interruptions.
  4. Explores how AR-cues affect strategy development over time.
• Procedure and key techniques:
  • Participants performed a pill-sorting task with interruptions triggered at varying distances from coarse breakpoints.
  • AR-cues (red arrows) indicated the next correct step after interruptions.
  • Metrics included resumption lag, resumption errors, and subjective stress levels (SUDS).
  • Experimental design: 2 (cue vs. no cue) × 4 (distance to coarse breakpoint) mixed factorial design with repeated measures.

Results

• Concrete findings:
  • AR-cues reduced the percentage of interruptions accepted at coarse breakpoints (e.g., 41.7% vs. 6.0% at distance 3).
  • Resumption errors were nearly eliminated in the AR-cue group (0.7%) compared to the no-cue group (12.0%).
  • Resumption lag was shorter in the AR-cue group (e.g., 2967ms vs. 4466ms at distance 3).
  • Stress levels increased significantly in the no-cue group (mean SUDS: pre=20.5, post=31.0) but remained stable in the AR-cue group (pre=23.0, post=22.6).
• Advantage over baselines:
  • AR-cues outperformed coarse breakpoints in reducing errors and stress while maintaining efficiency.
  • Enabled faster responses to interruptions without sacrificing accuracy.
• Experiments / evaluation:
  • Participants: 50 university students (25 per cue group).
  • Metrics: Resumption lag, resumption errors, SUDS stress levels.
  • Statistical analyses: ANOVA, post-hoc tests, and exploratory analyses.
• Limitations and future work:
  • Task simplicity limits generalizability to complex work environments.
  • Predictable interruption timing may not reflect real-world dynamics.
  • High interruption frequency may have influenced strategies.
  • Future research should explore nested interruptions, varying urgency, and bio-physiological stress measures.

Summary

This study demonstrates that AR-cues significantly reduce reliance on coarse breakpoints, resumption errors, and stress levels in deferrable interruptions. While coarse breakpoints remain a useful strategy, AR-cues offer a more efficient and robust solution for task resumption, especially in time-critical environments. Findings suggest AR-cues could enhance workflows in interruption-prone settings, such as healthcare and emergency services, and support the adoption of AI-driven technologies. Future research should investigate AR-cues in more complex scenarios and varying interruption conditions.