The cost of interrupted work is stress, not speed
Interrupted participants finished faster with similar quality — and reported more stress, frustration and effort. A time log alone will not show that cost.
The short version
In a lab study of email work, people who were interrupted finished their tasks faster than when they were left alone, with no loss of quality. But they paid for it: after only about 20 minutes of interrupted work they reported more stress, frustration, time pressure, effort and mental workload. The hidden cost of interruptions is strain, not lost minutes.
What the researchers did
Forty-eight participants, most of them German university students, played the role of a human resources manager answering a full inbox — 12 emails per condition — using a fact sheet. Each person worked under three conditions in counterbalanced order:
- No interruptions.
- Same-context interruptions: a "supervisor" asked HR questions related to the email task.
- Different-context interruptions: the supervisor asked about unrelated topics, such as how many hot dogs to order for a company party.
Interruptions arrived about every two minutes, by phone or instant message. The researchers measured time spent on the task (excluding time handling interruptions), errors, politeness and email length, and participants rated their workload, stress, frustration, time pressure and effort after each block.
What they found
- Participants spent less time on the task when interrupted — about 20 minutes versus about 23 without interruptions — and wrote somewhat shorter emails.
- There was no difference in errors or politeness.
- Whether the interruption was related to the task made no difference to the time cost, and phone versus instant message made no difference either.
- Interrupted work came with significantly higher stress, frustration, time pressure and effort, and mental workload was highest for unrelated interruptions.
The authors' interpretation: people compensate for expected interruptions by working faster and writing less, at the price of a more stressful working state.
Limits and caveats
This is a single lab experiment with 48 people and short sessions, so it cannot say whether people would adapt over time or whether the strain would keep building. Quality was measured narrowly (errors and politeness). The finding that interrupted people were faster applies to this kind of routine email task; it should not be read as "interruptions don't cost time" for complex work. Relatedly, the often-quoted claim that task switching wastes "up to 40%" of productive time is not a result from this study.
What it means for how you work
A time log can make interrupted days look efficient, because the cost does not show up in minutes. If you track your time, it is worth also noting how many times you switched and how the day felt — a quick end-of-day rating captures what the clock misses. To reduce the interruptions themselves, see the evidence on batching notifications and blocking distracting sites.