Effect of self-interruption and external interruption on error detection

Knowledge workers are frequently bombarded with interruption and are required to constant multitask. Previous observational studies found that frequent interrupted activities cause more errors and induce feelings of stress and frustration. Therefore, the aim of the current research is to investigate...

Full description

Bibliographic Details
Main Author: YAU, Sze Yuen
Format: Others
Language:en
Published: Digital Commons @ Lingnan University 2015
Subjects:
Online Access:https://commons.ln.edu.hk/psy_etd/2
https://commons.ln.edu.hk/cgi/viewcontent.cgi?article=1003&context=psy_etd
Description
Summary:Knowledge workers are frequently bombarded with interruption and are required to constant multitask. Previous observational studies found that frequent interrupted activities cause more errors and induce feelings of stress and frustration. Therefore, the aim of the current research is to investigate how interruption affects error detection performance. Current error detection research focused on the effectiveness of different checking methods. In this thesis, we concentrate on the psychological mechanism of error detection. A series of experiments was carried out to examine the effects of self-interruption (i.e. the pilot study and Study I) and external interruption (Study II) on error detection performance respectively. The pilot study and Study I focus on the effects of working memory (WM) load and capacity. The pilot study employed a think-aloud technique to verify the predictions on WM and self-interruption. The results suggest that low-capacity individuals (LWMC) rehearsed more frequent than high-capacity individuals (HWMC). In other words, LWMC have more self-initiated interruptions during the primary error detection task. Study I was carried out to test the generated predictions from the pilot study. A reliable interaction effect WM load × capacity was found: LWMC performed significantly worse in higher WM load conditions; however, HWMC’s performances were unaffected by higher WM load. Study II focuses on the effect of interruption task types and position. There was no difference between the different interruption task types proposed. However, a significant main effect was found in interruption position: participants performed significantly worse in terms of both error detection and resumption when they were interrupted just before the actual field is displayed (i.e. between-fields interruption) compared to when they can see what is in the field (i.e. within-field interruption). The results are explained in terms of Salvucci and Taategen’s (2008) threaded cognition. The concurrent execution of error detection and WM tasks in Study I is interpreted as concurrent multitasking performance; whereas the sequential execution of error detection and interrupting tasks is interpreted as sequential multitasking. The current study contributed to the understanding of error detection performance by examining the roles of both self-interruption and external interruption and extends the application boundary of threaded cognition to interpret the effect of interruptions.