Multitasking is often treated as a simple matter of doing more than one thing at once. In practice, the harder problem is deciding what deserves attention now, what can wait, and when those priorities need to change. A new randomized pilot trial suggests that explicitly practising that kind of flexible attention allocation may help cognitively healthy older adults improve on untrained tasks that demand similar executive-control skills.
Researchers tested a web-based intervention called the Breakfast Game, a simulated everyday environment in which participants had to prepare several foods while simultaneously setting tables. All participants practised the same basic game, but half were repeatedly instructed to shift which component deserved most of their attention. The other half were told to treat all components as equally important.
The distinction mattered. The group trained to change priorities showed greater improvement on one of the game’s timing measures and on all three prespecified near-transfer cognitive outcomes. Those advantages survived correction for multiple comparisons. However, the trial involved only 38 participants, and it found no corresponding improvement in more distant outcomes such as mood, self-efficacy or engagement in cognitive activities in daily life. The results therefore offer an encouraging proof of concept rather than evidence that the programme prevents cognitive decline or improves everyday functioning.
Training the skill behind the task
The study focused on executive control, the collection of mental processes used to coordinate behaviour, update information, switch between demands and inhibit inappropriate responses. These abilities matter whenever several goals compete for limited attention.
A persistent challenge in cognitive training research is transfer. People often become better at the task they repeatedly practise, but that does not necessarily mean the improvement extends to other tasks or to everyday life. Researchers commonly distinguish near transfer, where improvement appears on an untrained task that relies on similar processes, from distant transfer, where benefits appear in substantially different domains.
The team tested whether a strategy called Emphasis Change could strengthen transfer. Instead of simply rehearsing the Breakfast Game, participants in the experimental condition were told to vary their priorities across rounds. In some rounds, 75% of their score depended on table setting. In others, the emphasis shifted to cooking, accurate cooking times or synchronising when foods finished. The active-control group played the same game but maintained a fixed strategy in which all elements were treated as equally important.
A small randomized trial conducted entirely online
The researchers enrolled 38 cognitively unimpaired adults aged 60 to 74 years and randomly assigned 19 to the Emphasis Change group and 19 to the active control group. Mean age was 65.8 years, with a standard deviation of 3.6 years, and participants had an average of 16.4 years of education. Women made up 81.6% of the sample. About 78.9% identified as White, 18.4% as Black or African American and 2.7% as Asian, while 15.7% identified as Latinx or Hispanic.
The sample was also relatively comfortable with computers. Its mean computer-literacy score was 26.92, above the comparison value of 24 cited for older adults. This is important because a remote cognitive intervention may work differently for people with less access to technology or less confidence using it.
Participants completed 12 training sessions, generally three times per week. The programme was fully remote, although researchers supervised selected sessions by videoconference and monitored activity through the platform. Analyses followed the intention-to-treat principle. Linear mixed-effects models tested changes over time, group differences and, most importantly, time-by-group interactions. Baseline cognitive performance was included as a covariate, and the researchers used Benjamini-Hochberg false-discovery-rate correction across the six primary outcomes.
Retention was high for a pilot intervention. Thirty-four of 38 participants completed all sessions within two months, giving adherence and retention rates of 89.4%. Those completers took an average of 36.8 days to finish the 12 sessions. The mean interval between sessions across the full sample was 3.4 days.
Practice helped both groups, but changing priorities added something
The active-control group improved through practice. Across the 12 sessions, the number of correctly completed tables rose significantly, with a model coefficient of 0.11 and p below 0.001. Their cooking stop-time difference also improved, with a coefficient of -0.53 and p = 0.02. There was no significant practice effect for cooking discrepancy, where the coefficient was 0.05 and p = 0.53.
The experimental group also improved with repetition, but across all three game measures. Correct tables increased with a coefficient of 0.12 and p below 0.001. Cooking stop-time difference improved with a coefficient of -1.57 and p below 0.001, while cooking discrepancy improved with a coefficient of -0.15 and p = 0.01.
The more revealing comparison was whether improvement differed between the two groups. For cooking stop-time difference, the time-by-group interaction was -1.52, with p = 0.003 and a false-discovery-rate-adjusted q value of 0.018. The effect size was small, Cohen’s d = 0.13. This indicates that explicitly changing attention priorities produced a greater improvement in synchronising the end times of multiple cooking tasks than practice alone.
The second cooking measure moved in the same direction but did not survive the researchers’ corrected threshold. Cooking discrepancy had a time-by-group coefficient of -0.15, p = 0.05 and q = 0.06. The correct-table measure likewise did not show a statistically significant interaction after correction.
The stronger signal appeared on tasks participants had not trained
The most interesting result concerned near transfer. Participants completed an Alphanumeric Task in which they had to divide attention while prioritising one component, as well as a broader executive-control composite derived from measures of working memory, cognitive flexibility, divided attention and inhibition.
On the Alphanumeric Task condition prioritising equations, the time-by-group coefficient was 0.10, with p = 0.005, q = 0.015 and Cohen’s d = 0.48. For the condition prioritising visual detection, the coefficient was 0.04, p = 0.019 and q = 0.028. Its effect size was d = 0.77, the largest of the reported near-transfer effects.
The executive-control composite also favoured the Emphasis Change group. Its time-by-group coefficient was 1.33, with p = 0.019, q = 0.038 and d = 0.40. All three near-transfer interactions therefore remained significant after false-discovery-rate correction.
That pattern is more informative than simply showing that participants learned the Breakfast Game. The transfer tasks used different content, stimuli and response formats. The common ingredient was the need to coordinate or flexibly allocate attention. The findings are therefore consistent with the possibility that participants were learning a strategy rather than merely memorising a game routine.
The benefits did not spread everywhere
The study also provides an important boundary around that interpretation. No significant time-by-group interactions appeared for the distant-transfer measures. These assessed general self-efficacy, engagement in cognitive activities, depressive symptoms and anxiety.
This means the experiment does not show that several weeks of attention-priority training broadly improved participants’ psychological well-being or everyday cognitive engagement. The authors note that participants began with low mood symptoms and relatively high self-efficacy and leisure engagement, which may have left limited room for improvement. Even so, a null result remains a null result, and broader benefits require direct evidence.
Participants generally found the intervention acceptable. About 82.3% rated their overall experience at eight or higher on a ten-point scale. Around 79.4% rated enjoyment at seven or higher, and 42% of participants reported becoming more aware of how they allocated attention, managed time or divided focus.
Why flexible attention may matter
Many everyday tasks are not difficult because each component is complicated. They are difficult because several simple demands arrive together. Cooking while answering a message, keeping track of appointments while planning transport, or switching between online tasks all require a person to decide where limited attention should go.
Emphasis Change makes that decision explicit. Rather than encouraging participants to spread attention evenly, it repeatedly forces them to identify the current priority and then change it. The researchers suggest that this may engage metacognitive processes involved in understanding and controlling attention allocation.
The near-transfer results fit that explanation, but they do not establish the underlying neural mechanism. No neuroimaging was performed in this trial. The authors point to earlier work on variable-priority training as a reason to investigate whether overlapping executive-control networks help explain transfer in future studies.
Promising evidence, with substantial limits
The randomized active-control design is a strength because both groups received the same game and similar contact, allowing the researchers to isolate the added contribution of the changing-priority instructions more cleanly than a no-training comparison would.
But this was explicitly a pilot proof-of-concept trial, not a definitive efficacy trial. Thirty-eight participants provide limited statistical power and make effect estimates less stable than those from a large randomized study. The sample was highly educated, predominantly female and predominantly White. Participants were also cognitively unimpaired and relatively computer literate. Results should not automatically be extended to people with mild cognitive impairment, dementia, lower educational attainment or limited digital access.
The trial was short and did not establish whether the gains persist after training stops. It also did not demonstrate reduced dementia risk. Cognitive training and dementia prevention are related research areas, but improvement on executive-control tests over several weeks is not evidence that a programme prevents neurodegenerative disease.
Finally, near transfer is not the same as everyday functional improvement. The absence of significant distant-transfer effects makes that distinction particularly important. A larger trial would need longer follow-up, a more diverse sample and outcomes tied more directly to daily functioning before stronger clinical claims would be justified.
What the pilot adds
The study’s contribution is therefore narrower, but useful. It suggests that how people practise may matter as much as what they practise. Repeatedly changing attention priorities produced measurable advantages on several untrained tasks that shared the underlying demand for flexible executive control.
For cognitive-training research, that offers a testable direction: interventions may transfer more effectively when they train a general strategy that can be redeployed across contexts, rather than simply asking people to repeat one task until they become proficient at it.
For now, the findings support further testing rather than a consumer prescription. The next question is whether the same pattern survives in a larger, more diverse trial and whether near-transfer gains translate into durable changes in the complicated multitasking demands of everyday life.
Source Information
Study: Bartolotta, K. K., Gal, M., Hernandez, M. et al. Training executive control and multitasking abilities in older adults: training gains and transfer effects of an online pilot randomized controlled trial.
Journal: Scientific Reports, volume 16, article 30006 (2026).
Publication date: 28 September 2026.
DOI: 10.1038/s41598-026-56974-9.
Study design: Fully remote, assessor-blinded pilot randomized controlled trial with an active control condition.
Sample: 38 cognitively unimpaired adults aged 60 to 74 years, randomized equally between experimental and active-control groups.
Intervention: Twelve online Breakfast Game sessions. The experimental group used Emphasis Change instructions to vary attention priorities, while the control group practised the same game without changing priorities.
Key caution: This was a small proof-of-concept pilot. Near-transfer findings are preliminary, distant-transfer outcomes were not significant, and the study did not test dementia prevention or long-term clinical benefit.
