What Is Multitasking? Task-Switching Research and Practical Ways to Handle Interruptions

September 14, 2026Shusaku Yosa

Multitasking means handling several activities at once or moving between them. Carrying several projects is different from repeatedly redirecting attention while working. To understand why work stalls, distinguish what is running in parallel, why you interrupt it, and how you return. The examples below connect task-switching research with practical choices about concentration and responsiveness.

Distinguish multitasking from task switching

When you leave a proposal to read email and return to check figures, the information you consciously handle changes. By contrast, sorting documents while a file processes uses waiting time; it does not necessarily mean making two complex judgments simultaneously. Treating both situations as identical can lead to rules that block useful work.

Competing tasks that require attention can create demands when you reset goals and rules. That does not justify claiming that humans cannot perform any two processes at the same time. Familiarity, information demands, task type, and urgency all matter.

Separate projects in flight from the action you are performing now. Five projects might include four awaiting replies. A single project can still involve repeated switches between notifications and documents. Project count alone is an incomplete measure of the problem.

What task-switching studies establish and what they do not

Rubinstein and colleagues compared repeated and alternating geometric-classification and arithmetic tasks in four experiments. They examined switching-time costs and differences associated with rule complexity and task cues. Read the original task-switching paper.

The findings help explain why switching demands depend on conditions. They do not establish a fixed percentage of productivity lost during a working day, an ideal number of projects, or a universally effective focus-block length. Laboratory response time and a day containing meetings, production, and waiting are different measures.

Leroy’s research addresses attention remaining with an unfinished task when someone moves to another. It provides context for that experience, without establishing an identical duration of interference for every person or activity. See the attention-residue study overview.

At work, you can observe time spent finding documents, recalling a decision, and preparing to resume. A practical preparation log is useful, but it is not a direct measurement of the psychological mechanisms tested in those studies.

Separate focus work from productive use of waiting time

Analysis, writing structure, and planning often benefit from keeping several assumptions in view. A short gap can still suit a bounded action, such as checking one source or correcting a chart unit. A 30-minute gap between meetings is not automatically unusable.

If a task is blocked, record the condition that will let it move again before switching. A reply deadline or review notification can replace repeated checks. Emergencies, safety issues, and incidents may appropriately interrupt a focus period; the decision depends on your responsibilities.

Situation: illustrative	Reason to switch	Response	Restart condition or next action
Editing ad copy	Routine chat	Read at the agreed 11:00 check-in	Verify evidence for paragraph 3
Editing the same copy	Urgent measurement incident	Follow the incident procedure	After the incident owner clears it, resume paragraph 3
Review requested	Waiting on another person	Move to available work	When review is complete, address 3 required changes

The time and number of changes are conditions of this example, not recommended thresholds. Agree which messages can wait and what acknowledgment the sender needs.

Use a worked interruption and restart log

A restart note locates the next action; it does not need to be a long activity report. “Continue proposal” leaves you to reconstruct the work. Record the file and position, the decision already made, the unresolved question, and the first action to take.

Interrupted	Location	Established	Unresolved	Restart action	Preparation
10:20	Proposal p.4	Audience draft	Scope of evidence data	Check the source aggregation scope	4 minutes
14:10	Same proposal p.4	Source data checked	Chart-axis unit	Correct the unit	2 minutes

These are fictional observations. Four plus two equals six minutes of recorded restart preparation for these two interruptions. It is neither total daily loss nor a research average. Define preparation consistently rather than treating essential quality checks as wasted time.

On returning, check whether assumptions changed while you were away. If feedback changes the direction, update the next action instead of automatically continuing from an outdated position.

Agree on interruption rules and review work in progress

Silencing notifications will not solve an operating model that demands immediate replies through several other channels. Agree on the normal request location, the urgent route, acknowledgment expectations, and cover during absence. Accepting a request and starting it immediately are separate commitments.

Record routine requests with their purpose, due date, and dependencies.
Interrupt through the agreed route when the urgency criteria apply.
Before starting new work, leave a restart note and identify any displaced plan.
When work accumulates, discuss changes to scope, timing, or ownership.

If you limit work in progress, test a limit against capacity, task size, waiting, and quality checks. Three is not a universal optimum. A lower limit could also reduce the available response to urgent work. State how waiting items are counted and review more than completion count.

Evaluate time, quality, and responsiveness together

Choose the problem you want to investigate. Long returns to analysis call for a preparation log; missed responses call for response-time checks. Over several days, record task type, switch reason, restart preparation, corrections, and missed response commitments. Then try a small change, such as a restart note or a protected work period.

Fewer switches are not an improvement if necessary messages are missed or errors rise. Shorten the focus period, provide cover, or allow selected urgent notifications when the evidence calls for it. The aim is useful progress with acceptable quality and responsiveness.

Different workloads, deadlines, or staffing can explain a before-and-after difference. Repeat observations on broadly comparable work before deciding what to keep. Start with one interruption: record what is established and the next action, then check whether returning becomes easier.