Case Study 2: Attention Residue — Why You Can't Focus After Checking Email

The Problem You Have Already Felt

You are in the middle of something that requires concentration — a paper, a project, a difficult email that requires care. You pause for a moment, not quite stuck but not quite in the flow, and you check your messages. You find something: not a crisis, but not nothing. A question that needs an answer, an update that changes something you thought you knew, a thread you didn't know had started without you. You spend ninety seconds on it — maybe two minutes — and then you return to what you were doing.

Except you don't, quite. You are sitting in front of the same task, but you are not in the same cognitive relationship to it as you were. Something is different. The flow doesn't come. You write a sentence and delete it. Your mind is partly elsewhere. If you were asked to describe the experience, you might say you feel scattered, or that you have lost your train of thought, or just that something is off.

What you are experiencing is attention residue. And the person who named and studied it, Sophie Leroy, has spent the better part of two decades documenting exactly what it is, why it happens, and what it costs.

Sophie Leroy and the 2009 Paper

Sophie Leroy published the foundational attention residue paper in the journal Organizational Behavior and Human Decision Processes in 2009 under the title "Why Is It So Hard to Do My Work? The Challenge of Attention Residue When Switching Between Work Tasks." The paper introduced the concept, provided the first experimental evidence for it, and sketched the theoretical framework that subsequent work has elaborated.

Leroy's starting point was an observation that should be familiar to anyone who works in a knowledge economy: we switch between tasks constantly, and the standard assumption has been that the cost of switching is primarily the time lost to the switch itself — the few seconds or minutes spent reorienting to the new task's demands. But Leroy suspected that the cost was longer and deeper than this — that something of the first task stayed with the person who had left it.

Her theoretical framework drew on the Zeigarnik effect — the well-documented finding, first described by Lithuanian-Soviet psychologist Bluma Zeigarnik in 1927, that incomplete tasks are better remembered than completed ones. Zeigarnik found that waiters who had not yet delivered orders remembered those orders well; once the order was delivered and the task complete, their memory for it rapidly faded. The incomplete task maintained a kind of representational priority, staying active in memory in a way that completed tasks did not.

Leroy extended this insight to the context of cognitive performance. If incomplete tasks remain representationally active, she reasoned, they should also be competing for attentional resources — drawing on working memory capacity that the person has now formally committed to a different task. The residue is not merely a memory phenomenon. It is an attentional one. And it should have measurable performance consequences.

Study 1: The Interruption Paradigm

In her first study, Leroy created an experimental task-switching scenario. Participants were assigned to complete a word search puzzle before turning to a second, separate task involving reading and assessing job applicant resumes. The critical manipulation was whether they completed the puzzle or were interrupted before completion.

Participants in the interrupted condition were told, mid-puzzle, that they needed to move on to the resume task. Participants in the completed condition finished the puzzle before transitioning.

Before starting the resume task, all participants completed a lexical decision task — a measure in which they saw strings of letters and had to rapidly indicate whether each was a real word or a nonword. Leroy embedded words related to the first task (the puzzle) among the distractors to measure whether participants' attention was still partially directed toward it. Interrupted participants responded faster to task-1-related words than completed participants, indicating that their attention was still partially oriented toward the incomplete puzzle even though they had officially moved on.

Critically, the performance measure on the resume task — the new task — showed that interrupted participants performed significantly worse. They processed fewer resumes, made less accurate evaluations, and showed lower performance quality across multiple indicators. The incomplete first task was actively impeding performance on the second task.

Study 2: Time Pressure and the Completion Drive

Leroy's second study introduced a wrinkle that made the findings more nuanced and more practically relevant. She examined what happens when people are under time pressure as they complete or leave a first task.

The finding was counterintuitive: when participants worked under high time pressure on the first task, the subsequent attention residue on the second task was reduced compared to those working under low time pressure. The mechanism, as Leroy understood it, involved the degree to which attention had been genuinely disengaged from the first task before the switch.

Under time pressure, participants appeared to mentally close out the first task more completely before transitioning — the time constraint served as a forcing function for task disengagement. Under low pressure, the more relaxed transition allowed attention to linger, maintaining the active residue of the incomplete mental model.

This finding has implications for how people exit tasks: the quality of disengagement — the degree to which a natural or deliberate stopping point is reached — affects how cleanly attention transfers to the new task. A task left hanging in the middle of a thought process generates more residue than a task left at a natural pause point.

Extension: Leroy and Glomb (2018) — The "Being Present" Study

In 2018, Leroy and Theresa Glomb published an important extension examining the connection between attention residue and psychological presence — the degree to which a person's mind is fully engaged with what they are currently doing.

The key finding was that individuals who were more psychologically present in their current task showed lower performance decrements from task-switching, not because they had better working memory capacity, but because they had developed better habits of cognitive disengagement — they were more practiced at mentally closing out a previous task before fully transitioning.

This finding introduced an important practical dimension: attention residue is not entirely a fixed property of the switching situation. The habits of mind people bring to transitions affect how much residue accumulates. People who regularly practice completing tasks to natural stopping points before switching, or who develop deliberate "closing rituals" for tasks they must leave incomplete, appear to carry less residue into subsequent work.

The 2018 paper also examined the role of worrying about incomplete work. Residue is not only a function of leaving tasks unfinished — it is also generated by the ongoing monitoring process through which the brain continues attending to those tasks. People who are high in worry show more attention residue from incomplete tasks than people lower in worry, consistent with the idea that residue is partly driven by the active maintenance of the incomplete task's representation.

Leroy's original research focused on workplace task-switching — moving between projects, assignments, and responsibilities in an organizational context. But the connection to social media checking behavior is direct and important.

Consider what happens when a social media notification arrives during a focused work or study session and the user responds to it. The notification opens a new task: a social situation that has developed, a message that requires response, a conversation that has taken a new turn. Even a fifteen-second engagement with the notification introduces a task that is almost certainly incomplete at the moment of disengagement — a social exchange does not resolve in fifteen seconds, and the brain knows it.

The user returns to their original work carrying the residue of that social exchange: monitoring its status, planning potential responses, replaying its content, maintaining awareness of its emotional valence. These monitoring activities occupy working memory and compete for attentional resources with the primary task. The user is cognitively in two places simultaneously, and their performance in neither place is as good as it would be in an undivided state.

The problem compounds with multiple notifications. Three or four brief social media checks during a two-hour study session do not simply create three or four brief disruptions. They create three or four simultaneous attention residues that stack and interact. The student who surfaces from a study session feeling scattered and unproductive may have spent most of that session nominally engaged with their work but actually carrying the accumulated residue of multiple social media interruptions.

Open-Plan Offices and the Structural Residue Problem

While this book is primarily concerned with social media, the attention residue literature emerged from workplace research and has significant implications for how knowledge work is structured more broadly. Open-plan offices — in which workers are co-located without physical barriers and share ambient sound and visual space — are essentially structural residue generators.

Research on open-plan office effects consistently finds that workers in open-plan environments are interrupted more frequently than those in private offices, and that they report lower ability to concentrate, higher stress, and lower satisfaction. The attention residue framework provides a specific mechanism for these effects: frequent environmental interruptions — a colleague's question, an overheard conversation, a visual event in the peripheral field — create constant low-level task-switching with its associated residue costs, degrading cognitive performance throughout the workday even when individual interruptions are brief.

The similarity to heavy social media notification environments is worth noting. Both open-plan offices and social media notification streams are environments characterized by high-frequency, low-predictability interruption. Both generate structural attention residue. And both are environments that workers and users often describe as acceptable or even enjoyable in the moment — the social presence that makes open-plan offices feel energetic is the same social signal density that makes social media feeds feel engaging — while producing measurable costs to deep cognitive work.

The Notification Design Implication

If every notification that arrives during a focused task creates an incomplete task and therefore generates attention residue, the question of notification design becomes a question with direct cognitive consequences.

Notification systems for social media platforms are calibrated to maximize the probability of response. They are sent at times when users are likely to be engaged with other activities (because those are the times when users check their phones most eagerly when they do respond). They are framed emotionally — "Liked your photo," "Commented on your post," "Mentioned you in a story" — in ways that create immediate social relevance and therefore strong salience responses.

A notification that produces a thirty-second phone check during study creates, at minimum, several minutes of attention residue. If the check reveals something emotionally significant — a social conflict, a disappointing response, unexpected news — the residue may extend for much longer. A student who checks their phone twice during a two-hour study session and encounters ordinary social media content will likely carry ten to twenty minutes of cumulative attention residue from those checks, fragmenting what might otherwise have been coherent focused work.

Platform designers are aware of interruption patterns at the aggregate level — they can measure the relationship between notification cadence and user behavior across millions of users. Whether notification timing is explicitly calibrated to maximize interruption of competing tasks is difficult to determine from outside the organizations. What is clear is that notification systems are not designed around user cognitive welfare, and that the costs they impose — in terms of attention residue and associated performance decrements — are real, measurable, and largely invisible to the users who experience them.

What Reduces Attention Residue?

The research literature suggests several approaches that reduce attention residue, some at the level of individual behavior and some at the level of environmental and system design.

Natural stopping points: Leroy's research consistently shows that leaving tasks at natural completion points — even within an ongoing project — reduces the residue carried into the next activity. The implication is that before any transition (including checking a social media notification), taking a moment to reach a natural pause point in the current task reduces the cognitive cost of the interruption.

Deliberate disengagement rituals: The Leroy and Glomb (2018) finding that psychological presence moderates residue suggests that developing deliberate habits around task transitions helps. Brief rituals that signal task closure — writing down where you are, noting what the next step will be, explicitly telling yourself "I am pausing this task until X" — appear to reduce the brain's ongoing maintenance of the incomplete task's representation.

Notification batching: Checking notifications on a schedule (once per hour, once per study session) rather than responding to each as it arrives converts multiple interruptions into a single planned transition. The residue from a single deliberate notification check is substantially lower than the cumulative residue from multiple reactive checks.

Task closure before switching: Evidence from Leroy's time pressure study suggests that switching under conditions that allow genuine cognitive closure produces less residue. Structuring work in defined units with explicit stopping points — rather than leaving everything perpetually open — reduces the baseline residue burden.

The Broader Picture: Why This Research Matters

Attention residue research matters because it makes visible a category of cognitive cost that is otherwise invisible. Most people intuitively understand that using their phone during work or study is somewhat costly. They understand it at the level of the specific minutes of distraction involved. What they do not readily perceive is the residue — the continuing cost that extends beyond the interruption itself, the way that each check degrades the quality of the work that follows it.

The invisibility of residue makes it difficult to motivate protective behavior. If you check your phone for ninety seconds during study, you experience ninety seconds of obvious disruption and then return to work. You do not experience the subsequent fifteen minutes of degraded performance as a consequence of the check — you experience it as just being a little scattered, or a little tired, or not quite in the zone. The causal chain is not obvious.

Leroy's research makes the causal chain visible. It provides a name and a mechanism for the experience that many people have described but not been able to locate. And it offers a path forward: not simply "use your phone less" but rather "understand the specific ways in which incomplete tasks generate ongoing cognitive costs, and design your transitions accordingly."

For Maya, sitting down to study after fifty minutes of TikTok, the attention residue model adds a layer to the explanation of her difficulty. She is not only dealing with extraneous load accumulation and attentional transition costs. She is carrying the residue of multiple incomplete social processing tasks — the conversation she half-followed, the emotional content she didn't fully process, the social situation she observed without resolution. Her cognitive workspace is occupied. The calculus problem set has competition it did not ask for.

Key Takeaways from This Case Study

Attention residue is the cognitive phenomenon whereby incomplete tasks retain representation in working memory and compete for attentional resources even after the person has formally switched to a new task.
Sophie Leroy's foundational 2009 research demonstrated the effect experimentally: participants interrupted mid-task showed measurably worse performance on a subsequent task compared to those who completed a natural stopping point.
The mechanism is the ongoing maintenance of the incomplete task's representation — the brain continues monitoring and processing it, occupying executive resources that are formally committed to the new task.
Social media checking during focused work creates classic attention residue conditions: each check opens an incomplete social processing task that remains active after the phone is put down.
Multiple interruptions create stacking residues that compound, potentially degrading sustained performance far beyond what the actual interruption time would suggest.
Strategies that reduce residue include: reaching natural stopping points before transitions, developing deliberate disengagement rituals, batching notifications rather than responding reactively, and building work into defined units with explicit closure.
Notification systems designed around engagement maximization are, by their logic, designed to interrupt users during competing tasks — the residue cost of these interruptions is real and substantial but largely invisible to users.