upa - home page JUS - Journal of usability studies
An international peer-reviewed journal

Switching Between Tools in Complex Applications

Will Schroeder

Journal of Usability Studies, Volume 3, Issue 4, August 2008, pp. 173-188

Article Contents

Recommendations (What's Next)

Making tool switching painless and quick may not be enough, and perhaps not even the highest design priority. Experienced users employ techniques-and choose tools-that appear rapid and efficient because using them has come to seem easy. Some are effective and some are not. For example, every user has his or her way to access help and support information. These patterns can't all be optimal, even if users think they are. Some users get better results than others; their tool choices and switching strategies will be the ones to reinforce in design.

Development of new designs to support tool switching highlights some areas of further investigation that are likely targets for work carried out with the techniques that are discussed in the following sections.

User self-monitoring

The notion of best practices and much of the content of training hinge at least in part on the assumption that users can monitor the quality of their work. In terms of this study, users discover through evaluation of work progress that they need to change tools. Our more narrow interest would be whether design could support this monitoring activity or not.

Studies of pair programming shed interesting light on this area: Experimental work pairing experts with novices shows experts' awareness of task barriers (contradictions) when observing novices encountering them (Raeithel & Velichovsky, 1996). It is not clear, however, that these experts monitor their own activity with the same care. Pair programming, in which one user codes while a teammate takes a supervisory view and (theoretically) makes good switching decisions, should model any system that notifies users of contradictions in their ongoing work (Domino, Collins, Hevner, & Cohen, 2003). An ethnographic study of pair programming in practice suggests that this is not so easy, asking at one point "how it would even be possible for two people working at different levels of abstraction to successfully sustain a conversation at all" (Chong & Hurlbutt, 2007, p. 2). Also, the authors repeatedly observed that the best-functioning pairs are also peers at skill level and experience (Chong & Hurlbutt, 2007). Advice, however good, administered injudiciously, can be counterproductive. Experiments harnessing the joint attention of two users (Raeithel & Velichovsky, 1996) appear to be more encouraging, suggesting that appropriately designed software agents might support the monitoring task better than peers hampered by their social baggage.

Task specificity of tools

Nardi (1996) suggests classification of tools as more or less task specific according to the variety of tasks that a task is suited for. As an example, the task of comparing two files can be carried out in most editors, but awkwardly; the same task can be handled much more quickly and efficiently in a difference editor. It's plausible that users resist switching to a task-specific difference editor simply because the comparison is only a small part of the overall task, and the difference editor is not hospitable to most of what the user is trying to accomplish. The effect of task specificity in tool choice appears in frequency of switching in tasks where use of the tool is appropriate, and in delay (inefficient attempts to complete the work) before moving to the tool, and then moving back.

Previous | Next