Dismal Interrupted - The effects of unplanned and unplanned interruptions on the learning and performance of a complex procedural task
Open Access
- Author:
- Oury, Jacob D
- Graduate Program:
- Information Sciences and Technology
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- July 28, 2022
- Committee Members:
- Jeffrey Bardzell, Program Head/Chair
Frank Ritter, Chair & Dissertation Advisor
Xiaolong Zhang, Major Field Member
Ling Rothrock, Outside Unit & Field Member
Aiping Xiong, Major Field Member - Keywords:
- Learning
cognitive modeling
ACT-R
Interruptions
cognitive load - Abstract:
- Interruptions are already ubiquitous throughout society, and the attention-driven economy may be training us to constantly switch tasks and refocus our attention without ever lingering on one activity. Previous studies of interruptions during work have found many negative outcomes (e.g., more errors, higher workload, slower task time) and some positive outcomes (e.g., reducing mental clutter, better problem solving). Research towards understanding the effects of interruptions on work has seen significant progress towards a unified theory of interruptions and cognition, however, there are significant gaps in our understanding of interruptions with regards to learning. Seemingly paradoxical results are common in this area. For example, a visual learning task with periodic interruptions finds more disruption from interruptions during the easier trials. Previous studies on interruptions and learning do not address the procedural memory consolidation and learning, and I have developed an experimental study and analysis to assess how interruptions affect learning and retention for a complex procedural skill over multiple days. I hypothesized that interruptions would disrupt task performance (e.g., slower task time) and that interruption warning signals would mitigate the disruption. I conducted a between-subjects study (N = 47) in three groups with interruption presence (Yes, No) and interruption warning (Yes, No) as factors. I found that interruptions have a complex effect pattern. The interruption warning appeared to increase workload, leading to higher forgetting rates between days. Task time was generally similar across conditions, but interruptions reduced responsiveness to error feedback. Finally, I analyzed these results and showed that workload may be a stronger predictor of disruption than interruptions, and I proposed methods for accounting for cognitive load in a cognitive model for the experiment task.