t a l l a p e
The Science of Interruptions

Posted on Monday 17 October 2005

The artice “Meet the Life Hackers”, in yesterday’s New York Times Magazine, has been on my mind all day.

Briefly, the article (which you should go read, if you haven’t already) is about the pace of work, and the relative lack of uninterrupted time in a given work day. It’s about the difficulty that workers have in focusing on one task for any appreciable amount of time, and how that undermines productivity. It also discusses the research that has been done on the problem of “interruption time,” and some of the ways that various companies are trying to minimize distractions that pull people off of one task and onto another.

I read this as someone who is constantly interrupted; it’s an essential element of my job description, actually, that long-term projects are constantly interrupted by short-term crises. In a good day I can spend half of my productive time working on making things better, or planning, or preparing. The rest of my work day is spent dancing from project to project, interruption to interruption, without much focus.

Two conclusions of the article jumped out at me.

The first is that interruptions are, if presented properly, a fundamental and beneficial part of the work day. People actually get more done when they get interrupted “well”. Often interruptions add a crucial piece of information, or a different viewpoint, to a task that’s already underway, and as a result make that task (or those tasks!) easier.

The second, and this is one that’s near and dear to my heart, is that desktop (as in computer desktop) real estate matters. Mary Czerwinski, a researcher at Microsoft, did a study in which she asked subjects to complete a number of tasks as quickly as possible. Some of the tasks were computer-based (find a particular web page, for example, along with cutting and pasting and organizing a list) and some were “real world” (memorizing a phone number). She had some of the research subjects go through the list of things while working on a 15-inch monitor. Others did the same list while working on a 42-inch monitor. The results: those working on the larger monitor worked 10% faster minimum, and some worked almost half again as fast as those on the smaller monitor.

Now, I’m feeling pretty good about myself after reading this, because I’ve been making this point for years. When I bought a new computer back in 2003, I made a point of buying the largest screen available, and I love it. At work, I use two monitors, because I’ve found that I can get more done when I don’t have to page back and forth. If I could use 4, or 6, instead, I would. That would be wonderful.

I’m not sure where I’m going with this; I keep getting interrupted, and I lost my train of thought.


3 Responses to “The Science of Interruptions”

  1. pgw Says:

    hmm. not so sure about that last bit. there’s an IT guy in the basement of the law school who uses six or eight monitors, including one for his DVD of the day. it’s kind of creepy.

  2. Jeff Says:

    Ooooh… shiny. *poing*

    It’s all about the alt-tab. alt-tab VNC connection to customer site, alt-tab Eclipse development environment, alt-tab check my email, alt-tab back to VNC.

  3. Ian Says:

    Well, I have spent the better part of the last forty years “dancing from project to project, interruption to interruption” and, yes, I did manage to get a phenomenal number of things, including raising you, done. So on one level, my life suggests that the inverse correlation posited between quantity of interruption and quantity of output is in error. That said, I have just had the experience of six weeks that included long stretches of uninterrupted time to do my work and am prepared to conclude that THAT is the “real estate” that matters, not desktop video screens. Making oneself more efficient and making oneself truly productive are not the same thing, I don’t think. I will concede, however, that some kinds of interruptions enhance creativity, particularly when they come as islands in reflective time sufficient to process the input.

Leave a Reply