Using Interruption Models to Test Interruption Studies
June 20, 2008 at 1:37 pm | In Attention Management, User experience, interruption science | 1 CommentYesterday I posted up a set of interruption models. I mentioned in that post that I’d write another entry on how they can be used to test interruption study methodologies. I know that sounds pretty arcane – mostly of interest to people doing interruption studies or interpreting their findings. That may not sound like too many of you, but one survey in particular, from Basex, has gotten into a lot of popular press for its easy-to-digest dollar amount for “unnecessary” interruptions in the U.S. ($650,000,000,000). It’s used by pop press journalists whenever they write about a fuzzy info-stress topic, but want to show this is really important and add a drop of academic-sounding data. Any of them wanting to delve deeper can select from hundreds of academic papers on interruption, attention, and human-computer interface (interruptions.net has a great list), but none of those have a big dollar figure to quote.
My attempts to determine the methodology of the Basex study have been unsuccessful so far. The way I would evaluate its legitimacy is the same way I’d evaluate any interruption study’s legitimacy – by lining it up against the models I’ve presented to see how accurately it would count them. Clearly not all interruptions are “bad” or “unnecessary” – many of the interruption models I listed have a positive net closed-loop benefit. A seemingly valid methodology that simply asks people how often they were interrupted (or observes them and records interruptions) and how much time they lost can provide a very inaccurate conclusion. Each model I list (except maybe the jerk model and blast model) could be easily miscounted by a poor survey methodology.
For example, I believe the Help-me model to be a large proportion of interruptions. This is where one person needs a little bit of someone’s time to provide a good deal of benefit to them. A study that just counts interruptions and their cost would only count the costs and not the benefits to the interrupter which is often many multiple higher than the cost. Only net closed-loop benefit analysis would hunt down the person that interrupted them and determine the value to them and add it back in. That’s difficult to do in a survey, but essential for an accurate estimate. Alternately a survey could ask how often you interrupted other people and how much benefit you got.
As another example, the Help-you model is common as well. This is where someone is interrupted to be told they should stop or modify what they’re doing, perhaps due to new information that’s just come in. But a methodology that only asks about the cost in time of each interruption in negative terms may miss the positive value the interruptee places on the interruption.
One more example: The Interaction model would throw any survey off if it doesn’t properly define “interruption” versus the simple act of collaboration. I defined interactions as interruptions that take place within the task the person is currently working on. Many people wouldn’t even consider this really an interruption. Survey takers may randomly include interactions fitting this model as interruptions, possibly incorrectly counting each positive benefit as a negative.
Interruption Models
June 19, 2008 at 2:59 pm | In Attention Management, Information Work, interruption science | 3 CommentsWell, we’ve gone quickly through the cycle of seasons here in Chicago, passing from winter to spring to construction. When working in my home office I’m now faced with a random barrage of interruptions from beeping trucks, pile drivers, and loud workmen that can’t afford walkie-talkies. Living in a part of Chicago that was fully built 50 years ago, many feel the need to tear down perfectly good houses and erect new ones to match the current style (the “large brick block covering every allowable inch in 3 dimensions” school of architecture). I think this inspired me to develop a list of interruption models that I posted over at the Collaboration and Content Strategies blog. I figure I should post them here as well for greater input. These are still open for debate – so your comments and feedback are welcome.
Each has an example of how it would apply, followed with a sample numerical calculation based on the dollars gained or lost by the organization based on the interruption (assume this is $ based on time x fully loaded pay rate).
- Help-me model: Bill needs a moment of Stu’s time to proceed with his work
- Value to interrupter (80) + value to interruptee (-20) = Net closed-loop benefit (60)
- Help-you model: Bill takes the time to let Stu know he needs to change his task approach
- Value to interrupter (-10) + value to interruptee (50) = Net closed-loop benefit (40)
- Jerk model: Mick is an jerk that likes bugging other people about fantasy football, hurting both their productivity
- Value to interrupter (-20) + value to interruptee (-30) = Net closed-loop benefit (-50)
- Machine interrupt model: Stu’s PC crashes. This distrubs Stu and has no benefit to the PC
- Value to interrupter (0) + value to interruptee (-50) = Net closed-loop benefit (-50)
- Break model: Bill’s thinking has been getting less effective and he finds himself spinning on a simple task, so he interrupts himself and decides he needs a mental break. He returns to work more refreshed and effective
- Value to interrupter & interruptee (5) = Net closed-loop benefit (5)
- Interaction model: Stu and Bill are working on a task together, expecting each other’s input, and neither would really consider this an “interruption”
- Value to interrupter (5) + value to interruptee (5) = Net closed-loop benefit (10)
- Alert model: A fire alarm goes off while Stu is working, interrupting him and saving his life
- Value to interrupter (0) + value to interruptee (100) = Net closed-loop benefit (100)
- Scheduled interruption model: Stu is working hard on a task that requires concentration, but has to stop at 10:00 for a scheduled meeting, which interrupts his train of thought and will require recovery time upon resuming. For this example, it is assumed the meeting is a project update for another project that Stu doesn’t get much out of but is obligated to attend
- Value to interrupter (0) + value to interruptee (-10) = Net closed-loop benefit (-10)
- Lazy model: Mick could figure out his task alone if he applied some time and effort, but it just seems easier to ask his smarter colleague Stu. Too bad Mick will never learn to help himself and will keep bothering Stu
- Value to interrupter (5) + value to interruptee (-7) = Net closed-loop benefit (-2)
- Training model: Bill is stuck in his task and needs to ask his smarter colleague Stu for information. Bill learns a valuable lesson that can be immediately applied and Bill is now that much better at his job
- Value to interrupter (10) + value to interruptee (-7) = Net closed-loop benefit (3)
- Blast model: Mick shouts out to the room to see if anyone wants to go to lunch. No one wants to because Mick is a jerk, so they are annoyed
- Value to interrupter (1) + value to interruptees (-50) = Net closed-loop benefit (-49)
- Social interruption model: Stu stops by his co-worker Bill’s desk and interrupts him to find out how his daughter is feeling after she got out of the hospital
- Value to interrupter (?) + value to interruptees (?) = Net closed-loop benefit (positive?)
I talked this over with Mike Gotta, who brought up the point of reciprocity. One enters into an implicit social contract that they will be gracious about interruptions in exchange for getting to interrupt others when needed. The Help-me model should be encouraged as it has a net benefit for the organization, but it can also have a net benefit for Stu if he gets some of Bill’s time the next time he needs it. He also pointed out that interruptions tied to communities can be worthwhile as people search for expert opinions and information.
For individuals feeling stressed and overloaded this list of models could help guide some introspection about the degree to which interruptions are causing the stress and which models need to be reduced.
For the owner of an attention management project, surveying information workers for the types of interruptions they are experiencing can help optimize the communication flows and interruptions.
For anyone presented with an interruption study (particularly those showing extremely high negative impact by interruptions) it provides a firetest of the study’s assumptions. These models can be run through the methodology of the study to see how accurately it would count the net closed-loop benefit. I’ll post more on this later.
Google Lands Crushing Blow to Email Addiction With New Feature
June 8, 2008 at 11:44 am | In Attention Management, User experience | 7 CommentsWell, that headline is what I’d like to write anyways. But, of course, solving email addiction is beyond the capabilities of a mere software behemoth. Still, Google took a humorously kitschy attempt in some new lab features for Gmail just released.
By going into Gmail settings (the “Labs” tab) and enabling the “Email Addict” feature, you get a “Take a break” link added to your email:
Then, whenever you click on it, your screen blanks out and you get the following message:
At least until you reload the page and get back to your email.
Cute. Even though it’s just for fun, it does acknowledge that email addiction is on people’s minds. Maybe not those of Google or the programmers themselves, as they may have meant this as a satiric swipe at their users who think this is a problem. After all – why would they want its users to reduce their usage of email and IM when they seem to thrive on more and more personal information from users being stored on their servers? Google needs bytes to live. <zombie voice> “More bytes …” </zombie voice>.
Well, in any case, it’s a nice email addiction / information overload / attention management joke. And it plays off the idea that people who are addicted to something have little ability to help themselves anymore and need external help.
If Google really wanted to help these users I think there are some real features they could have added:
- Mail arrival schedules (hourly, morning/noon/evening, morning/night, daily): Remember waiting by the (real) mail box for the postman to arrive? Unless you are expecting something to act on today, why not do that again and break the unconscious habit the rest of the day of checking for it? You could set it for the frequency (for example, every hour on the hour) and create a whitelist of certain people or messages that get “express” delivery without waiting.
- Measurement capabilities. Like many behavioral changes, measurement is often a key starting point and more. This feature would provide measurements on the number of times email is checked and useful stats on frequency (per day, per hour) and graph when checking was done over time. Granted this is a bit difficult when it’s just left open, so maybe this feature would have to be enabled and would turn off automatic refreshes. Once people really see how much they check email reflexively they will be surprised and may do more to curb it if they think this is a problem.
- Slow delivery. I find myself checking mail more often when I’ve just sent a bunch of emails because I am now waiting on the responses. This creates an echo effect then where, for example, 20 emails sent out prompt 12 emails back (some quick, some slow, like clapping in a large cathedral). I then respond to 8 of those, 5 people then respond back, etc until the echo dies away. If the emails aren’t very urgent, using slow delivery (they go out in a bundle the next morning for example) would take a burden off for response checking and possibly enable some reflection that would have you change or rescind the messages before they are sent. The “slow design” movement and slowmail have been advocating an approach like this for some time. I think you’d turn this feature on as a default and then only flag messages individually if they need instant delivery.
- Tokens. What if you only had a certain number of tokens per day or week to spend on checking email? Maybe you start with 10 tokens in the morning and it costs you one each time you check email. If administrators are having trouble with load, they could raise the cost to 2 tokens first thing in the morning or right after lunch. You’d start noticing how often you’re really checking (see measurement above) and start planning out your checking better throughout the day. I would recommend that extra tokens can carry over to the next day so you’re not encouraged to do a bunch of frantic checking at the end of the day. Similar attempts to putting a price on email activity have been made for sending email (see Serios from Seriosity).
- A free e-book on Zen. OK, this one is a bit out there. Maybe it’s just me, but while email is ostensibly about communication and human connection, so often it seems to be all about one person and controlling. Someone checks because they want to see if someone found the joke they sent out was funny, if they got someone else to finally admit they were right or agree to do what they said, if everyone else in the group agreed to their restaurant choice. What does it mean about me if people don’t respond to me, listen to me, include me? If my email/IM/message board posting/blog posting falls in the internet forest and no one responds, am I silent and irrelevant? Like sound, does my message only matter if it causes something to resonate in someone’s head? A reminder now and then to “be the water, not the rock” and “let things be and take what comes” may be all that some people need.
Blog at WordPress.com. | Theme: Pool by Borja Fernandez.
Entries and comments feeds.