Open any job description for a role in Singapore's financial district and “ability to multitask in a fast-paced environment” appears alongside Excel proficiency as a basic professional requirement. The assumption is clear: handling several things simultaneously makes a worker more effective. It is one of the most thoroughly disproved ideas in cognitive science, and the workplaces that most need the research — high-speed finance, technology, and healthcare — are also the ones least likely to have acted on it.
Singapore's working culture compounds the problem. The Ministry of Manpower (MOM) records among the longest average working weeks in developed Asia, and knowledge-sector roles in the CBD and in tech corridors like one-north are defined by near-constant interruption: instant messages, meeting pings, open-plan offices, and the expectation of immediate response. The Health Promotion Board (HPB) has flagged cognitive fatigue as a rising workplace concern, yet the structural fix — reducing task-switching — is rarely the first one organisations reach for.
Here is what the research shows, and what to do instead.
What Research Shows About Task-Switching Costs and Attention Residue
The term “multitasking” implies parallel processing — two operations running simultaneously. For most cognitive work, that is not what is happening. The brain's prefrontal cortex — the region responsible for executive function, goal-directed attention, and working memory — operates as a largely serial processor. When you switch between tasks, you are not running two processes in parallel; you are alternating between them rapidly, and every switch carries a measurable cost.
The research confirms this with precision. Rubinstein, Meyer, and Evans (2001), published in the Journal of Experimental Psychology: Human Perception and Performance, identified two distinct cognitive operations triggered by every task switch: goal shifting (updating what you are trying to accomplish) and rule activation (loading the relevant mental rules for the new task). Both operations take time. For simple, familiar tasks the delay is small. For complex, rule-governed work — financial modelling, code review, clinical documentation — the cost is substantial. Across their switching experiments, the researchers found that this “mental blocking” added up to a 40 per cent increase in the total time required to complete complex tasks compared with completing them sequentially. Not a marginal inefficiency: close to half of potential cognitive throughput is lost.
Working memory limits compound the problem. Nelson Cowan (2001) in Behavioral and Brain Sciences revised the classic seven-item short-term memory estimate down to approximately four items held simultaneously. That ceiling does not expand under pressure. Loading two cognitively demanding tasks does not double output — it divides a fixed and small resource between them.
The research suggests a second mechanism operating beyond the moment of the switch itself: attention residue. Sophie Leroy (2009), in Organizational Behavior and Human Decision Processes, showed that when people move to a new task before completing the previous one, part of their cognitive resources remains allocated to the unfinished work. This “attention residue” competes directly with attention on the current task. Participants who switched mid-task performed measurably worse on the second task than those who completed task one first — and the performance gap persisted even after they reported having mentally moved on.
Research from Gloria Mark at the University of California, Irvine, published in ACM CHI proceedings, documented that after a workplace interruption, it takes an average of over 20 minutes to return to the original task with full engagement — and that the average knowledge worker is interrupted more than once every 11 minutes during a typical workday. Full cognitive recovery is arithmetically impossible at that switching rate.
READ ALSO: Working Memory and Executive Function: What Your Bloodwork Does Not Show
How Multitasking Reduces Productivity in High-Pressure Finance, Tech, and Healthcare Roles
For roles where precision matters — financial modelling, compliance review, software engineering, clinical decision-making — the task-switching penalty translates into concrete errors and missed details, not merely slower throughput.
The research confirms that frequent multitasking degrades the specific cognitive control capacities these roles depend on. Ophir, Nass, and Wagner (2009), published in the Proceedings of the National Academy of Sciences, tested heavy and light media multitaskers on a battery of cognitive control tasks. Heavy multitaskers — the people most practised at managing multiple information streams simultaneously — performed significantly worse at filtering irrelevant information, maintaining task goals in working memory, and switching efficiently between task sets than light multitaskers. The cognitive trait that high-interrupt environments select and reward corresponds to lower measured performance on precisely the tasks those roles require.
Singapore's Monetary Authority of Singapore (MAS) risk management frameworks for financial institutions formally identify human cognitive error as a category of operational risk. The structural sources of that error — task fragmentation, attention residue, sustained interruption — are rarely part of the same conversation as the compliance frameworks designed to catch the mistakes they produce.
Harvard Health Publishing notes that the subjective experience of multitasking can feel highly productive because each task switch carries a small novelty response. That felt sense of engagement does not correlate with output data — which is part of why the productivity myth has proven so durable in workplace culture.
READ ALSO: Cortisol and Cognitive Performance: What Chronic Stress Does to Your Prefrontal Cortex
What Serial Focus Does Instead: The Research Evidence
The research confirms that sequential task execution — completing a defined unit of one task before beginning the next — produces higher accuracy and shorter total completion time for complex cognitive work. The same Rubinstein et al. (2001) data that demonstrated a 40 per cent time loss under switching conditions also showed near-elimination of that loss under serial conditions. The gain does not come from working faster; it comes from not incurring the cognitive re-loading cost that switching requires.
Circadian physiology interacts with this directly. Executive function and working memory follow a daily cycle tied to the sleep-wake rhythm, with a well-documented decline in the post-lunch window. This dip in prefrontal availability coincides with the period when most Singapore open-plan offices generate their highest volume of messages, requests, and collaborative interruptions. Protecting a focused work block in the first part of the working day — before the interrupt load accumulates — is a direct application of the task-switching evidence, not a personal preference.
The same principle applies to meeting sequencing. Back-to-back meetings followed immediately by solo analytical work require rapid cognitive mode shifts between social processing and generative construction. A brief transition window — not a check of messages, but a genuine pause — reduces the carry-over from the social cognitive mode into the analytical one. The task-switching literature gives a mechanistic reason why that transition time is not wasted.
READ ALSO: Sleep and Cognitive Recovery: Why Your Brain Cannot Optimise on Six Hours
How to Shift to Sequential Task Execution in a Multitasking Environment
The structural changes that reduce task-switching are largely within individual control, even in organisations that treat multitasking as the default operating mode.
- Reserve the first 90 minutes for one task, before messages. Before Slack, before email. In Singapore's financial and technology sectors, the window before regional markets fully open and meeting load ramps up carries the lowest interrupt load and, for most people, the highest prefrontal availability. Use it for the most cognitively demanding work first.
- Complete a defined unit before switching. Not an entire project — a defined unit: finish the section, close the analytical loop, write the summary note. Leroy's attention residue data shows the performance cost is driven by incompleteness, not by duration. Completing a unit before switching eliminates the residue; abandoning it mid-way sustains it into the next task.
- Batch notifications rather than monitoring them continuously. Cleveland Clinic's guidance on cognitive load and attention management notes that the average knowledge worker checks messaging applications many times per hour. Each check is a task switch with a residue cost. Three fixed check-in windows per day — morning, midday, late afternoon — reduce that overhead without meaningfully affecting responsiveness in most roles.
- Write a one-sentence save state before switching. Before moving to a new task, write down where you are and what the next step would be. This externalises the working memory load the previous task was occupying and reduces attention residue by giving the unfinished work a clear, retrievable state. It takes under 30 seconds and consistently reduces re-entry time after the switch.
None of these changes require organisational approval to implement. They are sequencing decisions — which is precisely what the cognitive science literature identifies as the primary lever for reducing task-switching costs.
READ ALSO: Circadian Rhythm and Cognitive Performance: When Your Brain Is at Peak Capacity
Now You Know What the Standard Approach Misses
The professional premium on multitasking is a misread of what high performance in knowledge work actually requires. The research does not show that focused workers are less responsive or less adaptable — it shows they produce more accurate work, in less total time, with lower cognitive cost. The trait selected for in “must be able to multitask” job descriptions is tolerance for interruption. That is a different capacity from cognitive throughput, and the difference is measurable in the research record.
For Singapore's finance, technology, and healthcare professionals, the shift to serial task execution is one of the higher-return adjustments available — without a clinician visit, a supplement, or a budget. The mechanism is established, the data is replicable, and the tools are already in your calendar.
Now you know what the standard approach misses. The guide shows you what to do instead, starting tomorrow.
