Multitasking Bar: Prototype and Evaluation of Introducing the Task Concept into a Browser

Introduction

Currently with the ubiquity of the Internet, we are doing more and more of our jobs online (UCLA, 2003; USC, 2009; GVU, 1998), with the most important workplace for these jobs being the browser. On the other hand, humans are naturally multitasking beings, often either doing several tasks simultaneously and alternatively, or executing a single task through several working sessions. For instance, there are several reports in the literature that a great many managerial tasks can be characterized by their brevity, variety, and fragmentation (González & Mark, 2004; Mintzberg, 1970, 1980). This leads to simultaneously executed activities, interruptions and resumption of tasks both on and off the Web (Mayer, 2007, p. 173-278). Meanwhile, a diary study of knowledge workers has identified an average of 50 task shifts per person during a typical working week (Czerwinski, Horvitz, & Wilhite, 2004, p. 175-182), and a log study has identified an average of 3.74 e-mail or IM-driven task shifts per hour, taking between 8 and 10 minutes on average for the person to return to the interrupted task (Iqbal & Horvitz, 2007, p. 677-686).

Cognitive psychologists have studied many aspects of multitasking or task switching, providing several definitions of multitasking (Carlson & Sohn, 2000; Miyata & Norman, 1986, p. 265-284). Just et al. (2001) and Rubinstein et al. (2001) stated that multitasking is the ability of humans to handle simultaneously the demands of multiple tasks through task switching, and that it allows people to cope with ever increasing complex environments by handling multiple tasks through task switching (Burgess, 2000, p. 465-472; Lee & Taatgen, 2002, p. 572-577). We follow these definitions in this chapter and concentrate especially on tasks that depend on Web resources, typically Web pages. When users multitask while working on the Web, they switch among several tasks running simultaneously, often suspending and then resuming their task(s). In doing so, they tend to follow one of two common patterns, which we define as Multi-Tasks (MT) and Multi-Session-Task (MST). In this chapter, Multi-Tasks (MT) is defined as the pattern of executing a set of tasks simultaneously, where the execution often depends on bundles of Web pages. MST on the other hand, is the pattern of executing a single task spanning multiple Web sessions (as in MacKay’s [2008] definition). For example, whilst shopping on an e-commerce web site, and doing other task(s) on the Web at the same time, i.e., planning the itinerary of a vacation, constitutes MT. Furthermore, the shopping task is classed as an MST if it spans more than one Web session; in other words, if at some point the browser is closed and the shopping task temporarily suspended, and then later resumed in a new instance of the browser.

These patterns of MT and MST in Web use are very common, and have frequently been reported in the literature. Spink et al. (2002) found that multitasking information seeking and searching is common human behavior as many IR system users conduct information seeking and searching on related or unrelated topics. The study by Sellen et al. (2002) of knowledge workers’ Web use reported that 40% of the “information gathering” activities they observed were not completed in a single sitting, often due to external interruptions. MacKay and Watters (2008) found that users with tabbed browsing typically worked on several tasks during a single web session. Morris et al. (2008) presented evidence that users often conduct multi-session tasks, such as Web investigations, and they also found that “such tasks are not adequately supported by current tools.”