All articles
6 min readfocus / productivity

A 3-second interruption doubles your errors: evidence and the fix

In my AI trainings, the most common complaint has nothing to do with AI. People tell me they cannot think anymore. They blame the volume of work. The research points at fragmentation instead, and the bill arrives as errors rather than lost minutes.

Six studies below, then the protocol I use.

A 3 second interruption is not free.

Erik Altmann, Greg Trafton and David Hambrick ran a controlled experiment where participants performed a task made of ordered steps, the kind of work where losing your place is the failure mode. They interrupted people for a few seconds and measured what happened next. Interruptions averaging 2.8 seconds, roughly the time to perform one step of the task, doubled the rate of sequence errors. Interruptions averaging 4.4 seconds tripled them 1.

The doses here are seconds, and the damage lands as wrong steps rather than slowness. Non-sequence errors did not move, which points at the mechanism: an interruption costs you your position in the work, and little else.

Anything with ordered steps inherits this: a deployment checklist, a financial model, a migration, a code review, a client proposal that builds an argument.

You finish on time and pay in stress.

Gloria Mark, Daniela Gudith and Ulrich Klocke tested 48 participants in a controlled setting, interrupting them while they worked. The counterintuitive result: people completed interrupted tasks in less time, with no difference in quality. They compensated by working faster 2.

Interruptions feel survivable because the work still ships. The same study measured the price of that compensation: more stress, higher frustration, more time pressure, more effort 2. The cost lands in your nervous system rather than your calendar.

Put the two studies together and you get the honest picture. Interrupted work gets delivered on schedule. It carries sequence errors you find later, and it leaves you with an afternoon where you feel wrecked without being able to say why.

Part of your attention stays behind.

Sophie Leroy named the third mechanism: attention residue. In her experiments, participants who switched from one task to another kept thinking about the first one, and their performance on the second suffered. The residue appears when you switch while the first task feels unfinished 3.

The useful part is what came next. Leroy and Thomas Glomb tested a fix across four studies: a ready to resume plan. Before turning to the interrupting task, you take a moment to note where you are and what you will do when you come back. Attention residue dropped, and performance on the interrupting task no longer suffered 4.

It costs about 30 seconds. It is the highest return habit in this whole article.

The batching advice is half right.

Batching your messages is the standard advice, and the evidence is more mixed than the advice suggests.

The good news first. Kostadin Kushlev and Elizabeth Dunn asked 124 adults to limit email checking to 3 times a day for one week, and to check freely the other week. The limited week produced less daily stress, and people felt less distracted by email 5. Nicholas Fitz and colleagues ran a randomized field experiment with 237 participants on notification delivery. Batching notifications 3 times a day made people more attentive, more productive and in a better mood. Batching hourly did almost nothing. Turning notifications off entirely backfired: more anxiety and more fear of missing out 6.

Now the caveat. Mark and colleagues tracked 40 information workers in their real workplace for 12 workdays with computer logging, biosensors and daily surveys. They found no evidence that batching email led to lower stress. The variable that predicted stress and low perceived productivity was the total daily time spent on email 7.

Three times a day beats hourly and beats zero. But if you batch and still spend 3 hours in the inbox, you have changed the schedule and kept the problem.

This field also carries a famous overclaim. The 2017 brain drain study reported that the mere presence of your own phone on the desk reduced cognitive capacity 8. Productivity talks still quote it. A pre-registered replication in 2022 found no such effect 9, and a 2023 meta-analysis of 22 studies found a significant pooled effect for memory only, small at that (g = -0.23), with attention and general cognitive performance not significant 10. Put your phone away if it helps you. Do not sell it as settled science.

The protocol.

Run this for one week. Each step is falsifiable.

  1. Write the resume note before each switch. Two lines, the same shape each time: where I am, what I do next. Before you open Slack, before the meeting, before you answer the call. This is the Leroy and Glomb intervention 4, and it is the step to keep if you keep only one.
  2. Pick 3 fixed message windows. For example 09:30, 13:30, 17:00. Hourly batching did almost nothing in the field experiment, and cutting notifications to zero made people anxious 6. Three windows is the dose that worked.
  3. Cap total inbox time, on top of the windows. Set a timer, 60 minutes a day, all windows combined. Total duration is the variable that tracked with stress in the in situ study 7.
  4. Give sequential work a hard block. Two hours, notifications off at the operating system level, for anything with ordered steps: deploys, models, reviews, migrations. This is where a 3 second ping buys you a doubled error rate 1.
  5. Keep exactly one interrupt channel open. A phone call, nothing else. Tell your team that is the emergency path. You are not unreachable, you are addressable through one door.
  6. Count your self-interruptions for 3 days. A tally mark each time you switch with no one asking you to. The number runs higher than you expect, and it gives you a baseline for the other five steps.

Working interrupted but faster remains available as a trade, and the studies show people take it, paying in stress and frustration 2. Protect instead the specific hours where a lost place in a sequence costs real money.

Sources.

  1. Altmann, E. M., Trafton, J. G., & Hambrick, D. Z. (2014). Momentary interruptions can derail the train of thought. Journal of Experimental Psychology: General, 143(1), 215-226. doi.org/10.1037/a0030986
  2. Mark, G., Gudith, D., & Klocke, U. (2008). The cost of interrupted work: more speed and stress. CHI '08, Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. doi.org/10.1145/1357054.1357072
  3. Leroy, S. (2009). Why is it so hard to do my work? The challenge of attention residue when switching between work tasks. Organizational Behavior and Human Decision Processes, 109(2), 168-181. sciencedirect.com
  4. Leroy, S., & Glomb, T. M. (2018). Tasks interrupted: how anticipating time pressure on resumption of an interrupted task causes attention residue and low performance on interrupting tasks, and how a ready-to-resume plan mitigates the effects. Organization Science, 29(3), 380-397. doi.org/10.1287/orsc.2017.1184
  5. Kushlev, K., & Dunn, E. W. (2015). Checking email less frequently reduces stress. Computers in Human Behavior. doi.org/10.1016/j.chb.2014.11.005
  6. Fitz, N., Kushlev, K., Jagannathan, R., Lewis, T., Paliwal, D., & Ariely, D. (2019). Batching smartphone notifications can improve well-being. Computers in Human Behavior, 84-94. doi.org/10.1016/j.chb.2019.07.016
  7. Mark, G., Iqbal, S. T., Czerwinski, M., Johns, P., Sano, A., & Lutchyn, Y. (2016). Email duration, batching and self-interruption: patterns of email use on productivity and stress. CHI '16, 1717-1728. doi.org/10.1145/2858036.2858262
  8. Ward, A. F., Duke, K., Gneezy, A., & Bos, M. W. (2017). Brain drain: the mere presence of one's own smartphone reduces available cognitive capacity. Journal of the Association for Consumer Research, 2(2), 140-154. doi.org/10.1086/691462
  9. Ruiz Pardo, A. C., & Minda, J. P. (2022). Reexamining the "brain drain" effect: a replication of Ward et al. (2017). Acta Psychologica, 230. pubmed.ncbi.nlm.nih.gov/36007374
  10. Böttger, T., Poschik, M., & Zierer, K. (2023). Does the brain drain effect really exist? A meta-analysis. Behavioral Sciences, 13(9), 751. doi.org/10.3390/bs13090751

Keep reading