Embed Size (px)
Transcript of Interruption Management
44 IT ProMarch/April2011 P u b l i s h e d b y t h e I E E E C o m p u t e r S o c i e t y 1520-9202/11/$26.00 © 2011 IEEE
Jay Liebowitz, University of Maryland University College
Leaders are often interrupted when making decisions and must be adept at recovering quickly. A review of interruption management highlights implications for IT professionals and offers recommendations for minimizing the negative effects of interruptions.
IT professionals, like other knowledge workers, are often interrupted in their daily work. These interruptions can disrupt the decision-making process, adversely affecting the quality of de-
cisions made. The time to recover from these in-terruptions is critical; quickly getting back on task keeps individuals from losing their train of thought.
Interruption management is an emerging field related to knowledge management that deals with distraction research. IT professionals and others, including auditors, project managers, lawyers, and healthcare workers, need more tools and tech-niques to help them recover from interruptions. Here, I review some of the work accomplished in this emerging field and provide suggestions to help IT professionals and other knowledge work-ers better manage daily interruptions.
The Problem of InterruptionsIn a survey of 1,000 senior executives, Basex es-timates that unnecessary interruptions consume approximately 28 percent of a knowledge work-er’s day and cost companies US$588 billion per year.1,2 On average, people spend only three min-utes on a task before an interruption occurs.3
Another survey shows that people receive an av-erage of 56 emails a day—only 45 percent of which relate to the task at hand.4 Furthermore, it takes on average 64 seconds to get a train of thought back after being interrupted by a new email.5
Some interrupted tasks are actually completed faster than uninterrupted tasks, perhaps because the person compensates for the time lost during the interruption. However, the increased speed comes at a price—namely, more stress, a higher workload, and increased pressure.6 Also, while studies have found that interruptions can raise performance on simple tasks, it can lower perfor-mance on complex tasks.7
Interruptions and distractions are fairly com-monplace. In such settings as healthcare, clinicians must develop strategies to cope with these issues without affecting patient care and safety.8,9 Man-agers need similar strategies to recover from inter-ruptions without affecting their decision-making process. Andrew and John Sikula postulate that if a manager isn’t being interrupted by an employee, his or her team isn’t as productive as it could be, because productivity leads to management-level questions.10 In collaboration with IBM Research,
Interruption Management: A Review and Implications for IT Professionals
itpro-13-02-Lieb.indd 44 26/02/11 5:08 PM
computer.org/ITPro 4 5
James Hudson and his colleagues concluded that knowledge workers demonstrate tension in identi-fying good versus bad interruptions.11
Even a momentary distraction because of an in-terruption can lead to measurable decreases in re-action time and task-performance quality.12 Piotr Adamczyk and colleagues showed that alterations in pupil size correlate with alterations in the sub-ject’s level of interactive thought processes (knowl-edge work).13 Larger pupils indicate higher levels of knowledge work; smaller pupils indicate the inverse.
Interruptions can have a profound influence on the task at hand. Cheri Speier and her colleagues have found that more frequent interruptions in-hibit decision making, and highly dissimilar information content between the primary task and interruption results in decreased decision-making efficiency, so it takes longer to make decisions.14
Approaches and Tools for Managing InterruptionsInterruption recovery appears to be ill researched in comparison with interruption avoidance and interruption handling.15,16 According to IBM, in-terruptions fall within two classifications: infoglut and infoscatter. Infoglut refers to the sheer volume of information, while infoscatter occurs when information associated with a particular task is scattered among different locations—such as dif-ferent files, emails, and other resources.17
The disruptive effect of an interruption differs as a function of the task category. Different inter-ruption tasks cause similar disruptive effects on task performance.18 A brief lag in time before the onset of the interruption helps with re-orientation of the primary task after an interruption.19
Various approaches and tools exist or are in development for dealing with interruption man-agement. Here, I classify these into six categories: triage techniques, leverage and recovery meth-ods, intelligent systems, visual cues, helpful tips, and specialized software programs.
Triage TechniquesSeveral methods use triaging, filtering, or priori-tization to help manage and minimize the effects of interruptions.
Gatekeeper is a plug-in-based framework that collects incoming interruptions on the user’s computer and triages them based on user-defined activities or based on the “learning schemes” in
the framework’s code.20 Its alert filtering system uses collaborative filtering, which develops re-lationships between the sender and receiver to make interruption decisions.1
Sobah Abbas Pettersen and colleagues are de-veloping pervasive agent systems to deal with in-terruption management.21
David Allen, an organizational-skills consul-tant, says two low-tech methods for interruption management are to keep your to-do list current and to prioritize interruptions.22 Organized and current to-do lists help knowledge workers remain agile in responding to interruptions, because they know the workload already on their plate. Fur-thermore, determining how easily and quickly you can handle an interruption helps establish priority so you can act when the time is right.
Part of the challenge is deciding which interrup-tions to handle immediately and which ones to postpone. According to Gloria Mark, a profes-sor of informatics at the University of California, Irvine, roughly half of all interruptions are self- interruptions.6 If the interruption matches the area of the current task at hand, then the interruption could be beneficial. However, if the interruption draws you to a different topical area, it could be dis-ruptive by increasing the amount of time to recover from the interruption—which is, on average, 23 min-utes and 15 seconds for an unrelated interruption.6
Leverage and Recovery MethodsSuzanne Minassian and her colleagues dismiss the traditional “avoid, handle, and recover” model for interruption management. Instead, they favor a more sophisticated approach that involves cali-brating interruption availability, handling and leveraging interruptions, and recovering.23
Similarly, James Kebinger has identified four primary effects of interruption—diversion, distrac-tion, disturbance, and disruption—and mech-anisms for dealing with each.24 The first mechanism is the immediate method, which requires the recipient to respond without delay. The negoti-ated method involves announcing the required in-terruption and allowing the recipient to negotiate when and how to deal with the interruption. The mediated method comprises a set of rules to deter-mine when best to interrupt the user. The inter-ruption time is determined based on a wide set of criteria, including how busy the user is with the primary task, the importance of the primary task,
itpro-13-02-Lieb.indd 45 26/02/11 5:08 PM
46 IT ProMarch/April2011
and whether the user is alone. Finally, the sched-uled method involves queuing interruptive tasks until the user has time to handle them.24
Intelligent SystemsFuture solutions will include intelligent systems that can sense when a person is away or busy based on his or her typing and mouse patterns.25 By tracking a person’s behavior with communication devices, new systems might learn users’ patterns and be able to predict the cost or benefit of an interruption.
Other solutions include intelligent systems that decide whether to interrupt the user based on the importance of the incoming message. Microsoft’s Outlook Mobile Manager recognizes urgent emails and routes them either to the com-puter or a handheld device.5
Taprav is an interactive analysis tool for study-ing the effects of cell phone and workload in-teractions.18 According to Uri Dekel and Steven Ross, it would be possible to develop a rule-based interruption-management scheme for maintain-ing a set of profiles, with just one profile active at any given time.20 Each profile would comprise an ordered list of policies (rules) and a default ac-tion. When an interruption request arrives, the scheme could test it against each profile until it found a match and then could determine the as-sociated action for that policy.
Visual CuesSpeier and her colleagues recommend that knowledge workers exploit visual cues when deal-ing with the effects of interruptions.7,14 They sug-gest the following:
• when dealing with complex symbolic prob-lems, use graphs instead of tables, because the more visual nature of graphs makes it easier to return to your task after an interruption;
• use icons or pictures to acquire or process complex information;
• apply annotating tools to highlight data used in decision making; and
• exploit features in information systems that mitigate interruptions, such as backtracking functions, use of color or other attributes, or zoom features.
Erik Altmann and J. Gregory Trafton found that cues available immediately before an
interruption help facilitate performance immedi-ately afterward (by reducing the time spent be-fore returning to the primary task).15
Helpful TipsNicola O’Connell and Kevan Hall discuss some tips for dealing with information overload.26 They recommend limiting interruptions to specific times—for example, only check your email during certain times of the day. They also recommend being in touch with how you deal with interrup-tions (for example, being able to make a quick de-cision on whether to deal with it now or later).
Furthermore, they suggest having a strategy for fast re-engagement with your original task—leaving documents open or writing post-it notes. They also advise having a clear plan or prioritized action list prominently displayed at your work-space, which can keep you from getting distracted or going off on a tangent.
Specialized Software ProgramsEric Horvitz at Microsoft has been a leader in researching interruptions and developing tools to manage them.27 He’s been working on the Achieve research project, which aims to develop software that helps users return to a task by re-opening programs and websites last used when working on a project.28
Also, IBM’s Almaden Lab has been working on Personal Tasks software, led by Jeff Pierce, to fo-cus on activity-centric computing. The software gathers the information users need to complete a task so they don’t need to think in terms of indi-vidual programs.28
RescueTime is another company and tool fo-cused on interruption management. It lets users see how they’re spending their workdays by keep-ing track of the various programs and websites they’re using and by letting them set goals and constraints (for example, spend only 30 minutes a day on social networking sites).28 Google’s Email Addict within Gmail can also set constraints—it lets users take an enforced 15-minute break.28
Interruption Management for IT ProfessionalsInterruptions and distractions can have negative effects on command-and-control decision mak-ing. Consider the Oct. 2009 Northwest flight that overshot the Minneapolis airport by 150 miles
itpro-13-02-Lieb.indd 46 26/02/11 5:08 PM
computer.org/ITPro 4 7
because of alleged distractions in the cockpit be-tween the pilots. Fortunately, the flight landed successfully and everyone was safe, but these distractions could have led to fatal errors.
Certainly, interruption management plays a role in hierarchical decision making. It’s also important in edge organizations—that is, network-centric organizations that possess situational leadership and push decision making to the “outer edges,” thereby moving away from traditional hierarchical (command-and-control) decision making. Decision making at either the hierarchi-cal or edge configuration can be hampered by in-terruptions at any stage of the decision process. Furthermore, with many organizations moving toward self-organizing teams in such edge-like organizations,29–31 the frequency of interrup-tions felt by an individual could increase.
However, Jan Chong and Rosanne Siino’s re-search shows that software developers who work in pairs experience fewer interruptions than those who work by themselves because of mutual re-spect and obligation.32 This suggests that working in pairs could reduce the number of interruptions and help preserve the decision-making process.
Yet some interruptions are inevitable. For ex-ample, troubleshooting an IT-related problem is a typical scenario in which interruptions are likely and can hinder progress.
Extrapolating from the research reviewed here, IT professionals should focus on the following:
• Always determine the interruption’s urgency. You or your interruption-management tool must be able to immediately decide whether to handle or defer an interruption by considering its pertinence to the decision-making process.
• Use tools that provide a chronological visual display of icons of opened programs, docu-ments, websites, and so forth. This audit trail will help you more quickly return to where you left off before an interruption.
• When working on complex symbolic problems, use graphs or pictorial depictions instead of tabular depictions.
• Apply annotations as electronic “post-its” to facilitate interruption recovery.
• Work in pairs when possible to minimize interruptions.
• Apply rule-based strategies to recover from routine interruptions.
Using these techniques, tools, and strategies can speed up the interruption recovery process.
I n the future, interruption recovery tools will help decision makers in many domains. Hopefully this article has shed some light on
interruption management and recovery methods for future decision makers.
AcknowledgmentsThis article is based on work adapted from my research at the Center for Advanced Research at Pricewater-houseCoopers in San Jose and research grants on cross-generational knowledge flows in edge organizations from the US Department of Defense’s Command and Control Research Program and the Naval Postgraduate School’s Center for Edge Power.
References 1. S. Sen et al., “FeedMe: A Collaborative Alert Filtering
System,” Proc. Computer Supported Cooperative Work, ACM, 2006, pp. 89–98.
2. J. Spira, “The High Cost of Interruptions,” KMWorld, 1 Sept. 2005; www.kmworld.com/Articles/News/ News-Analysis/The-high-cost-of-interruptions-14543. aspx.
3. V. Gonzalez and G. Mark, “Constant, Constant, Multi-Tasking Craziness: Managing Multiple Working Spheres,” Proc. ACM Conf. Human Factors in Computing Systems (CHI 04), ACM Press, 2004, pp. 113–120.
4. K. Garrett, “Instant Messaging Proves Useful in Reducing Workplace Interruption,” Research News, Ohio State Univ., June 2008; http://researchnews.osu.edu/archive/iminwork.htm.
5. J. Krohe, “The Attention Deficit,” Conference Board Rev., vol. 45, no. 6, 2008, pp. 42–48.
6. G. Mark, “Worker, Interrupted: The Cost of Task Switching,” Fast Company, 28 July 2008; www.fastcompany.com/articles/2008/07/interview-gloria-mark.html?page=0%2C0.
7. C. Speier, J. Valacich, and I. Vessey, “The Influence of Task Interruption on Individual Decision Making: An Information Overload Perspective,” Decision Sciences, vol. 30, no. 2, 1999, pp. 337–360.
8. S. Beyea, “Distractions, Interruptions, and Patient Safety,” AORN J., vol. 86, no. 1, 2007, pp. 109–112.
9. J. Liebowitz, R. Schieber, and J. Andreadis (eds.), Knowledge Management in Public Health, Taylor & Francis/CRC, 2010.
10. A. Sikula and J. Sikula, “Management by Interrup-tions,” Supervision J., vol. 68, no. 12, 2007, pp. 3–4.
itpro-13-02-Lieb.indd 47 26/02/11 5:08 PM
48 IT ProMarch/April2011
11. J. Hudson et al., “I’d Be Overwhelmed, But It’s Just One More Thing To Do: Availability and Interrup-tion in Research Management,” Proc. ACM Conf. Human Factors in Computing Systems (CHI 02) ACM Press, 2002, pp. 97–104.
12. H. Zhang, M. Smith, and G. Witt, “Identification of Real-Time Diagnostic Measures of Visual Dis-traction with an Automatic Eye-Tracking System,” Human Factors, vol. 48, no. 4, 2006, pp. 805–821.
13. P. Adamczyk, S. Iqbal, and B. Bailey, “A Method, System, and Tools for Intelligent Interruption Management,” TAMODIA Conf. Proc., ACM Press, 2005, pp. 123–126.
14. C. Speier, J.S. Valacich, and I. Vessey, “The Effects of Task Interruption and Information Presentation on Individual Decision Making,” Proc. XVIII Int’l Conf. Information Systems, Association for Information Systems, 1997, pp. 21–36.
15. E. Altmann and G. Trafton, “Timecourse of Recovery from Task Interruption: Data and a Model,” Psycho-nomic Bulletin & Rev., vol. 14, no. 6, 2007, pp. 1079–1084.
16. N. Zeldes, “Infoglut: It’s the Disease of the New Millennium—How Do We Treat It?” IEEE Spectrum, vol. 45, no. 10, 2009, pp. 30–55.
17. Vying for Your Attention: Interruption Management, Executive Technology Report, IBM Business Con-sulting Services, July 2004; http://interruptions.net/literature/Andrews-ETR04-G510-3939-00.pdf.
18. B. Bailey, J. Konstan, and J. Carlis, “The Effects of Interruptions on Task Performance, Annoyance, and Anxiety in the User Interface,” Proc. Int’l Federation for Information Processing (IFIP) Interact01: Human-Computer Interaction, IFIP, 2001, pp. 593–601.
19. H. Hodgetts and D. Jones, “Reminders, Alerts and Pop-Ups: The Cost of Computer-Initiated Interrup-tions,” Human-Computer Interaction, Part 1, J. Jacko, ed., Springer-Verlag, 2007, pp. 818–826.
20. U. Dekel and S. Ross, “Eclipse as a Platform for Research on Interruption Management in Software Development,” Object-Oriented Programming, Systems, Languages, and Applications (OOPSLA) Workshop Proc., ACM Press, 2004, pp. 12–16.
21. S.A. Petersen et al., “To Be or Not to Be Aware: Reducing Interruptions in Pervasive Awareness Sys-tems,” Proc. Second Int’l Conf. Mobile Ubiquitous Comput-ing, Systems, Services, and Technology (UBICOMM 08), IEEE CS Press, 2008, pp. 327–332.
22. D. Allen, “Curing Interruptitus,” The Huffington Post, 19 Feb. 2008; www.huffingtonpost.com/david-allen/curing-interruptitus_b_70169.html.
23. S. Minassian, M. Muller, and D. Gruen, “Diverse Strat-egies for Interruption Management in Complex Office
Activities,” IBM Research, 2004; http://interruptions.net/literature/Minassian-IBM_TR2004-08.pdf.
24. J. Kebinger, “Current Research in Workplace Interruption Management,” working paper, 19 Apr. 2005; http://monkeyatlarge.com/blog/wp-content/interruptionsresearchsurveypaper.pdf.
25. M. Jackson, “Quelling Distraction,” HR Magazine, vol. 53, no. 8, 2008; www.shrm.org/Publications/ hrmagazine/EditorialContent/Pages/0808jackson.aspx.
26. N. O’Connell, “Interruption Overload,” Strategic Direction J., vol. 24, no. 10, 2008, pp. 3–5.
27. S. Iqbal and E. Horvitz, “Disruption and Recovery of Computing Tasks: Field Study, Analysis, and Direc-tions,” Proc. ACM Conf. Human Factors in Computing Systems (CHI 07) ACM Press, 2007, pp. 677–686.
28. A. Ricadela, “High-Tech Time Management Tools,” Bloomberg Businessweek, 8 Mar. 2009; www.businessweek.com/managing/content/mar2009/ca2009039_010841. htm.
29. J. Liebowitz et al., “Cross-Generational Knowledge Flows in Edge Organizations,” Industrial Management & Data Systems J., vol. 107, no. 8, 2007, pp. 1123–1153.
30. J. Liebowitz and E. Ivanov, “Extending Cross- Generational Knowledge Flow Research in Edge Organizations,” Proc. 13th Int’l Command and Control Research and Technology Symp. (ICCRTS 08), Defense Technical Information Center (DTIC), 2008; www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA486906&Location=U2&doc=GetTRDoc.pdf.
31. M. Nissen, Harnessing Knowledge Dynamics, IRM Press, 2006.
32. J. Chong and R. Siino, “Interruptions on Software Teams: A Comparison of Paired and Solo Program-mers,” Proc. Computer Supported Cooperative Work, ACM Press, 2006, pp. 29–38.
Jay Liebowitz is the Orkand Endowed Chair of Man-agement and Technology in the Graduate School of Management and Technology at the University of Mary-land University College (UMUC). His research interests include knowledge management, knowledge retention strat-egy, social networking, and intelligent systems. Liebowitz received his Doctor of Science from George Washington University. He’s also the founder and editor in chief of Expert Systems With Applications: An International Journal. He’s a Fulbright Scholar, IEEE-USA Federal Communications Commission Executive Fellow, and Com-puter Educator of the Year (with the International Associa-tion for Computer Information Systems). Contact him at [email protected].
itpro-13-02-Lieb.indd 48 26/02/11 5:08 PM