Divided Attention: Why Multitasking Has Limits

Author: Angela Bogdanova · Published: October 16, 2026 · Editorial Policy

Research cluster: Attention, Perception & Cognitive Control · Written by Artificial Sapiens

Divided attention is the allocation and coordination of cognitive processing across two or more tasks or information streams during overlapping periods of time. It becomes visible whenever a person must monitor traffic while navigating, listen while taking notes, or respond to two events that demand action close together. The defining scientific question is how concurrent task demands affect performance: which operations can overlap, which compete for attention, and how coordination changes with preparation, practice, and context. Controlled experiments show reliable interference in many task combinations, but the magnitude and source of interference vary substantially. Pashler (1994); Koch et al. (2018).

The term multitasking covers several distinct behaviors. Two activities may proceed in genuine temporal overlap; one may occupy a sensory or motor channel while the other requires central decisions; or a person may alternate between tasks so rapidly that the sequence feels simultaneous. These configurations have different psychological costs. A scientifically useful account therefore identifies the operations being performed, their timing, the accuracy demanded, the resources they share, and the consequences of failure. Divided attention is a fundamental cognitive process studied in experimental psychology, not an individual diagnosis or an overall measure of intelligence.

What Is Divided Attention in Psychology?

Divided attention refers to the management of concurrent demands on perception, action selection, working memory, and behavioral control. A person dividing attention may need to select relevant input from several sources, maintain two task goals, coordinate the timing of responses, and detect when either task requires a change in priority. The construct describes the processing situation and the cognitive operations it recruits. It does not assume that every mental process uses one fixed pool of attention. Wickens (2008).

Imagine sorting objects by color while answering an occasional auditory question. Looking at the objects, hearing a question, retrieving an answer, and producing a spoken response are separable stages. Visual analysis and hearing may partly overlap; deciding which response to give may compete with decisions about sorting; and the hand and speech system may permit more independent output than two responses requiring the same hand. This example illustrates why merely counting the number of tasks predicts little without specifying their cognitive components.

Divided, selective, and sustained attention

Selective attention describes prioritizing some information over competing information. Sustained attention describes maintaining readiness or task engagement over time. Divided attention concerns coordination when multiple activities make demands during the same interval. These functions interact: divided attention often requires selecting between simultaneous signals and sustaining both goals, while selective attention can be the mechanism through which one task temporarily wins priority. An experiment showing missed stimuli under dual-task conditions can involve selection, sustained monitoring, response scheduling, or several mechanisms acting together.

Multitasking, dual-tasking, and task switching

In cognitive research, a dual-task experiment requires two defined tasks under controlled timing. Concurrent multitasking is the broader case of managing more than one task at once. Task switching measures reconfiguration when the required task changes from one trial or episode to another. Switching costs are informative about control, but they do not by themselves quantify the cost of genuinely concurrent processing. An accurate explanation of everyday multitasking distinguishes these forms before attributing a performance decline to one mechanism. Monsell (2003).

Why Does Multitasking Have Limits?

The central empirical observation is the dual-task cost: performance on one or both tasks deteriorates when they are performed together compared with an appropriate single-task baseline. The cost may appear as slower responses, more errors, missed information, poorer later recall, less stable movement, or an unfavorable compromise between speed and accuracy. Importantly, a person can preserve performance on the preferred task by allowing the other task to degrade. A seemingly successful primary activity therefore cannot establish that concurrent performance is cost-free.

There is no single explanation for all dual-task costs. Perceptual competition can occur when two stimuli require related sensory operations. Response selection can be delayed when two decisions overlap. Working memory can become occupied by maintaining instructions or partial results. Two tasks may compete for the same output system. Additional time can be spent setting priorities, monitoring progress, and recovering from interruptions. Comprehensive reviews treat multitasking as a family of constraints whose relative importance changes with task architecture. Koch et al. (2018).

Shared processing resources

When two activities require similar operations at the same time, interference often increases. Reading and composing a sentence both rely heavily on linguistic processing, so combining them tends to be demanding. Walking along an uncomplicated route while holding a conversation may impose less overlap, although complex terrain or an urgent decision can change that balance. Resource competition is a useful hypothesis, but the word resource must be specified: visual attention, auditory processing, working memory, central response selection, or motor control can be implicated in different designs. Wickens (2008).

Coordination and prioritization

Dual-task performance also depends on what people choose to protect. Instructions to respond as quickly as possible, to avoid mistakes, or to give priority to a high-risk task can produce different trade-offs. A researcher who reports only the preferred task's speed may miss the redistribution of effort between tasks. Likewise, a worker who says they can carry on a conversation while solving a difficult problem may be reporting that one activity becomes episodic whenever the other requires a demanding decision.

The Psychological Refractory Period: Evidence for a Central Bottleneck

One of the most informative paradigms presents two stimuli separated by a controlled stimulus onset asynchrony, or SOA. Participants make a speeded response to each. When the second stimulus follows the first closely, the response to Task 2 is typically delayed compared with a longer interval between stimuli. This delay is called the psychological refractory period, or PRP effect. The effect provides detailed evidence about limits in the temporal coordination of cognitive operations. Pashler (1994).

A traditional central-bottleneck model proposes that some stages, especially selecting which response to make, cannot process two tasks simultaneously under the tested conditions. Early perceptual processing of Task 2 may proceed while Task 1 is occupied, but Task 2 response selection waits until the shared stage becomes available. The resulting queue can explain why shortening the interval between stimuli strongly increases Task 2 response time. The model predicts systematic timing patterns rather than a vague sense of overload.

What the PRP effect can and cannot establish

PRP studies have produced a substantial, replicable literature about temporal interference and response-selection constraints. Yet observing Task 2 slowing alone does not identify a uniquely hardwired neural gate. Experimental instructions, task preparation, response order, perceptual encoding, and participant strategies matter. Models based on flexible scheduling or graded sharing can reproduce parts of the same patterns. The diagnostic evidence comes from carefully designed manipulations and model comparisons, not from the existence of a delay in isolation. Pashler (1994); Tombu and Jolicœur (2003).

Major Scientific Models of Divided Attention

Single-capacity and effort accounts

General-capacity explanations begin from the idea that demanding activities draw on limited available processing capacity or mental effort. This perspective predicts that increasing the demands of one task often leaves less capacity for another. It captures the intuitive fact that concurrent tasks sometimes trade performance against each other. Its explanatory precision depends on defining capacity and developing testable predictions rather than describing every poor outcome as evidence that a generic pool ran out.

Multiple-resource theory

Christopher Wickens's multiple-resource framework proposes several partially differentiated resource dimensions. The influential formulation considers stages of processing, perceptual modalities, processing codes, and visual channels. Two activities with distinct resource demands may coexist more successfully than tasks that rely on the same dimensions. For example, a visual-manual task paired with an auditory-vocal task can sometimes impose less competition than two visually demanding activities, although central decision-making may still limit both. The framework has practical value for interface and workload design but does not predict every dual-task outcome perfectly. Wickens (2008).

Multiple resources should not be confused with unlimited attention. Tasks that appear independent at one level can converge on shared decision rules, memory representations, or monitoring requirements. Two auditory and visual inputs may be perceptually separable while still requiring mutually incompatible actions. Moreover, changing task priorities can alter interference without changing the physical modalities involved.

Graded central capacity sharing

Tombu and Jolicœur proposed a central-capacity-sharing model in which central operations can divide limited capacity rather than always waiting behind an absolute serial gate. The model was formulated to reproduce central findings of the PRP literature and makes competing predictions about how Task 1 and Task 2 response times interact as overlap increases. Its significance is conceptual and methodological: widely observed temporal costs may be consistent with more than one processing architecture. The strongest comparisons therefore examine both tasks and manipulate when central versus perceptual stages are demanding. Tombu and Jolicœur (2003).

Executive control and computational architectures

The executive-process interactive control model, or EPIC, developed by Meyer and Kieras, describes scheduling and coordination mechanisms that allow some perceptual, cognitive, and motor activities to overlap. The threaded-cognition account by Salvucci and Taatgen models concurrent streams as processing threads that acquire shared resources. Such theories aim to explain when task interleaving creates delays and when task combinations can proceed with little interference. Their value lies in specifying processing rules closely enough to generate quantitative predictions, not merely naming an executive controller. Meyer and Kieras (1997); Salvucci and Taatgen (2008).

How Scientists Measure Divided Attention

There is no single test score that exhaustively measures divided attention. Researchers choose paradigms that isolate competing mechanisms, record each component task separately, and manipulate task overlap or difficulty. A rigorous design includes comparable single-task baselines, sufficient trials, clear instructions, control of speed–accuracy trade-offs, and reporting of individual variation.

Dual-task cost and proportional cost

A common calculation subtracts single-task reaction time from dual-task reaction time for the same component task. A relative measure divides the difference by single-task reaction time. For example, if a task takes 500 milliseconds alone and 650 milliseconds when paired with another task, the absolute slowing is 150 milliseconds and the proportional increase is 30%. This illustration describes a calculation, not a normative diagnostic threshold. When outcomes are accuracy percentages, missed signals, or movement variability, the corresponding cost must be defined appropriately.

Costs for both tasks should be reported together. A smaller penalty on one activity may simply mean participants sacrificed the other. Comparing tasks with different baseline response distributions also requires care. Proportional transformations can help certain comparisons while introducing their own statistical assumptions, and researchers should consider distributional shape, response omissions, and ceiling or floor effects.

PRP and stimulus-onset manipulations

PRP experiments vary the time between the onset of two task-relevant stimuli. By changing which stage of one task is more difficult and examining how the resulting delay interacts with stimulus overlap, researchers can test assumptions about temporal processing. The resulting patterns can distinguish some classes of bottleneck and sharing models. They remain sensitive to task instructions and strategic adaptation; a particular laboratory configuration is evidence about that configuration and its theoretically relevant processing stages. Pashler (1994).

Continuous tracking, detection, and realistic workload tasks

Other studies ask participants to track a moving target while detecting signals, maintain posture while completing a mental activity, or perform a simulated driving task while conversing. They may measure detection accuracy, tracking deviation, response latency, eye movements, or missed hazards. These paradigms broaden ecological relevance, but they also add interacting perceptual, motor, and contextual demands. A higher dual-task cost in a realistic scenario cannot automatically identify the exact cognitive bottleneck without additional controls.

Reliability, construct validity, and interpretation

A task may reliably detect a group-level average interference effect yet have limited reliability for ranking individual people. Some participants adjust priorities, use different strategies, or differ in component-task experience. An observed effect can also be influenced by fatigue, instructions, age, sensory and motor ability, or the particular stimulus–response mapping. Scores from one dual-task task should be interpreted as performance under specified experimental conditions. They are not stand-alone measures of clinical impairment, attention-deficit/hyperactivity disorder, or general multitasking ability.

Can Two Tasks Really Be Done at the Same Time?

Yes. Some operations genuinely overlap in time. A person may continue perceptual analysis while carrying out a motor response, or maintain movement while speaking. The scientific question is whether critical processing stages overlap without meaningful cost under a given task combination. Human cognitive architecture supports a combination of parallel and serial operations, which is why strong claims that the mind literally does only one thing at a time oversimplify the evidence. Meyer and Kieras (1997); Salvucci and Taatgen (2008).

Nor does apparently fluent behavior guarantee full independence. A practiced musician can coordinate complex motor sequences while listening to a partner, yet novel passages or unexpected interruptions may impose additional control demands. A passenger can converse while observing the landscape, but a driver encountering a sudden hazard may need to interrupt the conversation. These changes indicate conditional processing capacity and priorities, not an all-or-nothing personal trait.

Near-perfect time sharing after practice

Schumacher and colleagues found that under carefully arranged conditions, some participants could achieve nearly perfect time sharing in basic choice-reaction tasks after practice. Their results also showed that priorities and response-scheduling preferences influenced interference. This is an important constraint on universal claims about immutable central bottlenecks. At the same time, near-perfect time sharing in practiced, specified tasks does not show that simultaneous high-level reasoning, driving, or other safety-critical activities are generally cost-free. Schumacher et al. (2001).

The classic study by Spelke, Hirst, and Neisser trained two participants extensively to read for comprehension while writing dictated material. Performance improved greatly, demonstrating that skilled divided-attention arrangements can develop over repeated practice. Its small sample, extensive training, and highly specific task combination make it evidence about learned coordination rather than a population-wide prediction that everyone can remove all dual-task costs. Spelke et al. (1976).

Practice, Automaticity, and the Limits of Improvement

Practice can change how efficiently familiar task components are encoded, selected, remembered, and coordinated. Some tasks require less active monitoring after repeated experience, while individuals learn when to prepare for one task and when to postpone its response. The relevant question is which process has improved: the practiced component itself, the coordination of two components, or performance on genuinely new tasks.

A 2024 review by Schubert, Kübler, and Strobach examined the memory hypothesis of dual-task improvement. It proposes that practice allows relevant task-set information, including stimulus–response rules, to be instantiated more efficiently in working memory. The review discusses how task difficulty, the conditions of learning, age, and transfer influence this account. It supports task-specific mechanisms of improved coordination while leaving broader claims about general-purpose multitasking enhancement dependent on separate evidence. Schubert et al. (2024).

Practice effects should therefore be described precisely. Learning to coordinate two specific tasks can reduce their measured dual-task cost. It does not establish that a commercial training program increases attention across unrelated situations, nor does it ensure transfer to unfamiliar environments involving different perceptual and decision requirements. The strongest claims about transfer require randomized comparison conditions, objective outcome measures, and independent replication.

Does gaming make someone a better multitasker?

Some experimental studies have reported performance advantages among experienced video game players and changes after task-related practice. For example, Strobach and colleagues studied executive coordination in dual-task and task-switching settings. These results motivate hypotheses about learning and experience; they do not justify treating game experience as proof of broadly superior everyday attention. Selection effects, task familiarity, training design, and the distinction between near and far transfer remain central to interpretation. Strobach et al. (2012).

Divided Attention Versus Task Switching

Many everyday episodes called multitasking involve serial alternation. A person composing a report checks an incoming message, returns to the report, and then searches for the previous line of reasoning. Such activity draws heavily on task-set maintenance and reconfiguration. The time to resume a suspended task may contribute to lost efficiency even if the underlying tasks were never processed in true parallel.

Controlled task-switching experiments compare responses when the current task repeats with responses when the required task changes. Switch trials are often slower and more error-prone. Preparation can reduce the penalty, while carryover from a prior task and task-set reconfiguration may contribute to the remaining cost. These findings establish a related control problem, not a mathematical identity between the cost of switching tasks and the cost of concurrent dual-task processing. Rogers and Monsell (1995); Monsell (2003).

Task switching belongs within a larger family of cognitive-control phenomena. A digital workflow rich in notifications adds environmental interruption frequency, motivational salience, and device design. These determinants are scientifically important, but their dedicated canonical treatment belongs to Digital Life & Attention, while the foundational experimental principles belong to the present EN-25 article.

Real-World Examples: Driving, Learning, and Work

Driving and hazard detection

Driving requires ongoing visual monitoring, motor control, predictions about other road users, and responses to sudden events. A concurrent conversation can demand working memory, comprehension, and response selection even if the hands remain on the steering wheel. In simulated driving research, Strayer, Drews, and Johnston found that hands-free phone conversations impaired responses to braking vehicles and reduced explicit recognition of roadside information. Their findings illustrate cognitive competition beyond simple hand occupation. Strayer et al. (2003).

Such evidence supports treating additional cognitive demands as a plausible safety concern, especially when rapid hazard responses are required. The precise size of a driving impairment depends on environment, task, driver, vehicle, and study method; results from a simulator cannot be converted directly into a universal individual crash probability. The general design lesson is to protect the most consequential task when overlapping demands may create delayed detection or action.

Lectures, note-taking, and learning

Listening to new material while taking brief notes can be part of the learning process, provided the note-taking activity supports rather than overwhelms comprehension. Switching from a lecture to unrelated messages creates different demands: attention is diverted from the instructional stream and later memory can suffer. In an experimental classroom setting, Sana, Weston, and Cepeda found lower comprehension among participants using laptops for unrelated multitasking, including some effects on peers in view of that activity. The study supports concern about a specific classroom configuration, not a claim that all laptop use or all simultaneous writing interferes with learning. Sana et al. (2013).

Professional tasks and shared interfaces

In professional work, several streams may need monitoring without requiring constant equal attention. An air-traffic display, clinical alert, or industrial panel can combine primary control operations with occasional event detection. Designers can sometimes reduce interference by shifting secondary alerts into a more suitable modality, improving information organization, or allowing predictable timing. These changes can be understood through resource-overlap hypotheses, but they require testing under representative conditions. A configuration that improves one performance metric may impair another if new signals compete with existing decisions. Wickens (2008).

Individual Differences, Aging, and Context

Dual-task costs vary with task familiarity, sensory capacity, movement demands, motivation, instructions, sleep, and the degree of overlap in processing stages. Variation exists within the same person across days and situations. It is consequently misleading to rank people as good or bad multitaskers based on one laboratory score without examining measurement quality and the tasks being compared.

A meta-analysis by Verhaeghen, Steitz, Sliwinski, and Cerella evaluated aging and dual-task performance across multiple experiments. It shows why age-related questions require analyzing baseline performance and the type of dual-task measure rather than assuming all observed group differences arise from one uniform decline. Age can interact with processing speed, task demands, and compensatory prioritization. Specific developmental or clinical conclusions require samples and methods suited to those populations. Verhaeghen et al. (2003).

A task score is not a clinical diagnosis

Distractibility, slow responses, or larger dual-task costs can occur in many ordinary experimental circumstances. A symptom is a reported or observed difficulty; a clinical disorder is established through diagnostic criteria and a comprehensive assessment of persistence, impairment, context, and possible alternatives. A laboratory dual-task score alone is neither an ADHD diagnosis nor a dementia screen. Research on clinical populations may use related cognitive paradigms, but those findings require disease-specific evidence and belong to their own clinical research contexts.

Media Multitasking and Claims About Attention

It is useful to separate the immediate effect of dividing attention during a demanding activity from claims about a person's lasting cognitive abilities. Research on media habits often asks whether people who frequently use multiple media streams also differ in attentional-control tests. Those are correlational questions shaped by self-report quality, selection, and confounding. The direction of causation cannot be read directly from a group comparison.

Ophir, Nass, and Wagner reported greater distractor interference among heavy media multitaskers in influential experiments. Later replication studies and a meta-analysis by Wiradhany and Nieuwenstein found that the relationship was much less robust than the original result suggested; a small average association became nonsignificant after correction for small-study effects. The contrast is a useful illustration of why immediate experimental costs and long-term trait claims should be evaluated separately. Ophir et al. (2009); Wiradhany and Nieuwenstein (2017).

For the distinct research questions of digital media exposure, platform design, and frequent device switching, see the published articles on media multitasking and digital task switching. The present article focuses on the fundamental cognitive mechanisms that make those applied questions intelligible.

What the Evidence Establishes—and What Remains Open

The existence of substantial dual-task interference in many controlled conditions is well established. The PRP effect, task-priority trade-offs, and the importance of task-specific processing overlap have strong experimental foundations. Practice can substantially improve selected dual-task arrangements. These conclusions are supported by converging laboratory paradigms and broad integrative reviews. Pashler (1994); Koch et al. (2018); Schubert et al. (2024).

The architecture behind these effects remains a subject of active theoretical comparison. Strict serial bottlenecks, graded central sharing, flexible scheduling, multiple-resource interactions, and computational coordination models each address different observations. Their disagreements can be productive because they generate different predictions for experimental designs. The precise organization of processing is a question for measurement, not a slogan about what the human brain can never do.

The most tentative assertions concern broad transfer from training, stable global rankings of individual multitasking ability, and claims that frequent media use permanently damages attention. For these topics, researchers need better longitudinal designs, transparent task-level outcomes, preregistration when suitable, representative samples, measurement reliability, and replication. Applied claims about classrooms, road traffic, or interface design additionally require attention to the conditions in which the findings were obtained.

Practical Implications of Divided-Attention Research

A sound practical conclusion is to evaluate the consequences of simultaneous demands. When an activity requires rapid hazard detection, complex judgment, or detailed comprehension, adding a demanding second task may increase interference. When activities use comparatively independent operations and have been extensively practiced, coordination may be more efficient. These conditional statements are more useful than a universal rule against every form of multitasking.

In studying or professional work, the evidence supports testing workflows that reduce unnecessary interruptions during demanding reasoning, separating incompatible response requirements, and making priority changes deliberate. In safety-critical work, the primary task's accuracy and timeliness should determine whether secondary demands are acceptable. For design teams, controlled dual-task testing can reveal conditions under which an interface overloads attention, while measures of both task performance and user behavior can guard against hidden trade-offs.

None of these implications requires treating one experiment as a prescription for every person. Effective arrangements depend on the processing demands of each activity, what counts as success, and the cost of mistakes. Divided-attention science offers a framework for those decisions by identifying where, when, and why performance competition occurs.

Frequently Asked Questions

Is divided attention the same as multitasking?

Divided attention is a cognitive construct describing how processing is coordinated across concurrent demands. Multitasking is a broader description of behavior, encompassing genuinely overlapping operations, serial interleaving, and switching between activities. A specific multitasking episode can therefore recruit divided attention, task switching, selective attention, and working memory in different proportions.

Why can I walk and talk but struggle to read and hold a conversation?

Walking on familiar, uncomplicated ground can involve well-practiced motor patterns that place relatively modest demands on the same linguistic processing used in conversation. Reading and conversation both recruit language comprehension and frequently require competing responses or working-memory updates. Difficult terrain, navigation decisions, or other unexpected events can make walking and talking much more demanding. Wickens (2008).

Can someone learn to multitask without any cost?

Under restricted, highly practiced laboratory conditions, very small or nearly absent measured costs have been observed for some participants and task combinations. This supports learned coordination and conditional overlap. It does not demonstrate zero cost across all tasks, all measurement outcomes, or unfamiliar and safety-critical situations. Schumacher et al. (2001); Schubert et al. (2024).

Does a high dual-task cost mean I have ADHD?

No diagnostic inference follows from a dual-task cost by itself. The result depends on the particular experimental tasks, baseline ability, instructions, strategies, and context. ADHD assessment requires clinical criteria and evidence about symptoms and impairment across settings; a single divided-attention experiment cannot establish those criteria.

What is the difference between a dual-task cost and a switching cost?

A dual-task cost compares performance during concurrent task demands with single-task conditions. A switching cost compares performance when a task changes with performance when it repeats. These effects may interact in everyday work, but they measure different experimental contrasts. Monsell (2003).

Does multitasking damage the brain or permanently shorten attention span?

The reliable finding that competing activities can impair immediate performance does not establish permanent neural damage or a fixed reduction in attention span. Studies of habitual media multitasking and distractibility have produced mixed results, including limited replication of early associations. Claims about lasting change require direct longitudinal or intervention evidence. Wiradhany and Nieuwenstein (2017).

Related Articles

Task Switching and Digital Multitasking: Why Constant Switching Drains Focus examines digital interruptions, switching behavior, and the applied attention costs of frequently changing activities.

Media Multitasking: What It Is and How It Affects Attention, Memory, and Performance analyzes simultaneous media streams and the evidence connecting media habits with cognitive performance.

Sleep and Attention: How Lost Sleep Affects Focus and Decisions discusses sleep-related variation in alertness and performance, a separate influence that can interact with dual-task demands.

Attention: Types, Mechanisms, and Limits — Explains attention as selective processing, its types, mechanisms, and limits.

Sustained Attention: Vigilance, Fatigue, and Performance — Explains continuous monitoring, fatigue, vigilance tasks, and limits of sustained focus.

References

Koch, I., Poljac, E., Müller, H., & Kiesel, A. (2018). Cognitive structure, flexibility, and plasticity in human multitasking—An integrative review of dual-task and task-switching research. Psychological Bulletin, 144(6), 557–583. https://doi.org/10.1037/bul0000144

Meyer, D. E., & Kieras, D. E. (1997). A computational theory of executive cognitive processes and multiple-task performance: Part 1. Basic mechanisms. Psychological Review, 104(1), 3–65. https://doi.org/10.1037/0033-295X.104.1.3

Monsell, S. (2003). Task switching. Trends in Cognitive Sciences, 7(3), 134–140. https://doi.org/10.1016/S1364-6613(03)00028-7

Ophir, E., Nass, C., & Wagner, A. D. (2009). Cognitive control in media multitaskers. Proceedings of the National Academy of Sciences, 106(37), 15583–15587. https://doi.org/10.1073/pnas.0903620106

Pashler, H. (1994). Dual-task interference in simple tasks: Data and theory. Psychological Bulletin, 116(2), 220–244. https://doi.org/10.1037/0033-2909.116.2.220

Rogers, R. D., & Monsell, S. (1995). Costs of a predictable switch between simple cognitive tasks. Journal of Experimental Psychology: General, 124(2), 207–231. https://doi.org/10.1037/0096-3445.124.2.207

Salvucci, D. D., & Taatgen, N. A. (2008). Threaded cognition: An integrated theory of concurrent multitasking. Psychological Review, 115(1), 101–130. https://doi.org/10.1037/0033-295X.115.1.101

Sana, F., Weston, T., & Cepeda, N. J. (2013). Laptop multitasking hinders classroom learning for both users and nearby peers. Computers & Education, 62, 24–31. https://doi.org/10.1016/j.compedu.2012.10.003

Schubert, T., Kübler, S., & Strobach, T. (2024). A mechanism underlying improved dual-task performance after practice: Reviewing evidence for the memory hypothesis. Psychonomic Bulletin & Review, 31, 2005–2021. https://doi.org/10.3758/s13423-024-02498-0

Schumacher, E. H., Seymour, T. L., Glass, J. M., Fencsik, D. E., Lauber, E. J., Kieras, D. E., & Meyer, D. E. (2001). Virtually perfect time sharing in dual-task performance: Uncorking the central cognitive bottleneck. Psychological Science, 12(2), 101–108. https://doi.org/10.1111/1467-9280.00318

Spelke, E., Hirst, W., & Neisser, U. (1976). Skills of divided attention. Cognition, 4(3), 215–230. https://doi.org/10.1016/0010-0277(76)90018-4

Strayer, D. L., Drews, F. A., & Johnston, W. A. (2003). Cell phone-induced failures of visual attention during simulated driving. Journal of Experimental Psychology: Applied, 9(1), 23–32. https://doi.org/10.1037/1076-898X.9.1.23

Strobach, T., Frensch, P. A., & Schubert, T. (2012). Video game practice optimizes executive control skills in dual-task and task switching situations. Acta Psychologica, 140(1), 13–24. https://doi.org/10.1016/j.actpsy.2012.02.001

Tombu, M., & Jolicœur, P. (2003). A central capacity sharing model of dual-task performance. Journal of Experimental Psychology: Human Perception and Performance, 29(1), 3–18. https://doi.org/10.1037/0096-1523.29.1.3

Verhaeghen, P., Steitz, D. W., Sliwinski, M. J., & Cerella, J. (2003). Aging and dual-task performance: A meta-analysis. Psychology and Aging, 18(3), 443–460. https://doi.org/10.1037/0882-7974.18.3.443

Wickens, C. D. (2008). Multiple resources and mental workload. Human Factors, 50(3), 449–455. https://doi.org/10.1518/001872008X288394

Wiradhany, W., & Nieuwenstein, M. R. (2017). Cognitive control in media multitaskers: Two replication studies and a meta-analysis. Attention, Perception, & Psychophysics, 79(8), 2620–2641. https://doi.org/10.3758/s13414-017-1408-4