Interrupting yourself can be more disruptive than being interrupted by someone else

At work or study, whenever you choose to break away from your main task to do something else – such as leaving an email you're in the middle of writing to go check Facebook instead – you are effectively interrupting yourself. It's obvious that self-interruptions risk hurting your focus, but you might not realise just how much. A new study in Computers in Human Behaviour shows that in certain contexts interrupting yourself can be even more disruptive than an external interruption, and that it has to do with the brain power that's needed whenever you make a decision to pause your main task to do something different.

Ioanna Katidioti and her colleagues at the University of Groningen had 28 young adult participants complete a computer task that was designed to simulate a typical work situation in which the primary objective was to answer client emails, while a secondary objective was to participate in a live text chat with a friend. Answering the client emails involved remembering the product the client was enquiring about, looking up the price in a browser, then emailing back the answer. The live chat with a friend involved answering questions about favourite things, like books and restaurants.

To complete each stage of the experiment, the participants were told that they had to answer 10 client emails and exchange 15 chat messages with their friend. The most important detail is that for some stages, the participants chose when to switch from answering emails to chat with their friend (i.e. self-interruption), whereas in other stages, the live chat was forced on them (i.e. external interruption – the chat window came to the foreground of the computer screen and the email window went out of focus).

One other thing about the task set-up – the researchers programmed the simulation so that the external interruptions actually occurred at the same moments in the email answering and price-look-up process as the self-interruptions – this was to rule out the effect of interruption timings on the results (I'll explain more about this later).

It seems counter-intuitive, but the participants actually took longer to answer client emails in the stages of the experiment that involving self-interruption as opposed to external interruption – 24.83 seconds to answer 10 emails, on average, compared with 23.36 seconds (a statistically significant difference).

Past research has shown that work interruptions are often particularly disruptive because of the time it takes us to get back into our primary task. However, in the current study, the slower performance during self-interruption stages of the experiment was not due to participants taking longer to return to emailing as compared with after external interruptions.

This prompted Katidioti and her colleagues to reason that perhaps the greater cost of self-interruptions is due to effects that happen prior to the interruption. To test this, they repeated the first experiment with more participants and this time recorded their pupil dilation – a well-established marker of mental effort. They found that pupil dilation occurred one second earlier prior to self-interruptions, as compared with external interruptions, indicating that the decision of whether and when to self-interrupt takes a  degree of mental effort, which has an adverse effect on subsequent performance on the primary email task.

This results suggest that when you're working on your computer and you have social media apps open in the background, there's going to be a significant cost to your performance, even if you have all the alerts turned off. This is because you've now got the mental burden of deciding whether and when to stop what you're doing, to go check your social apps. By contrast, when you have the app alerts switched on, or the phone rings, or there's someone at the door, this decision is effectively made for you.

There is a major caveat here related to the timings of interruptions. One reason that real-life external interruptions can be so disruptive, compared with self-interruptions, is that they can occur when you're in a particularly tricky part of your primary task, whereas when you self-interrupt,  you can make sure you only switch tasks at more convenient moments. The current study removed this part of the equation by ensuring the external interruptions were timed similarly to self-interruptions.

What does this mean for real-life working? The researchers said their results suggest the ideal situation would be to have external interruptions – such as app alerts – that are programmed to occur only at low-demand moments of the primary task. For example, perhaps one day Facebook or other alerts could be designed to monitor your work behaviour to only appear on-screen when you're doing something that you've identified as relatively easy. Such a set-up would save you the mental effort that the current study suggests is involved in self-interruptions, while also removing the usual main disadvantage of external interruptions – their inconsiderate timing.

Further reading

—Interrupt me: External interruptions are less disruptive than self-interruptions