Why a mildly engaging break may beat both idleness and distraction
The intuition behind a low-demand filler task is a Goldilocks balance. A task with no demands at all may let you keep consciously chewing on the problem, which is the opposite of stepping away. A high-demand task, by contrast, may consume the very mental resources the background restructuring needs. The low-demand condition in the direct comparison required some attentional focus yet still left resources available for what the authors call an unconscious analytical restructuring process—a covert, goal-directed search that reorganizes the problem without your deliberate effort [1].
That framing comes from the hypothesis of Unconscious Analytic Thought, which holds that solving an insight problem depends on this hidden analytic work rather than on a passive fading of the problem from mind [1]. On this view, the filler task's job is to occupy attention just enough to keep conscious fixation at bay without monopolizing the cognitive machinery the background work relies on. It is worth noting this is a proposed mechanism consistent with the results, not a directly observed process.
The break itself helps; the demand level fine-tunes it
Stepping away from a creative problem and doing something else can help in its own right. In one study, taking a break from the problem and engaging in another task was helpful for creative problem solving [3]. That study varied the kind of incubation activity and whether people could choose it, but it did not manipulate how demanding the filler task was, so it speaks to the value of a break rather than to the ideal demand level [3].
Where demand was compared directly, the picture sharpens: low demand outperformed both non-demanding and high-demanding tasks for insight problems [1]. The advantage showed up as a high percentage of successes shortly after the filler task ended, which fits the idea that the restructuring happens during the break and surfaces once the filler task stops [1]. The effect held across visual, numerical, and verbal filler tasks, suggesting the demand level mattered more than the filler task's content in that study [1].
Does the same pattern hold for spatial problems?
Not every creative task has been tested the same way. A planned experiment is set to compare low- and high-demand visual-spatial filler tasks on a visual and a spatial creative problem, with no moderate condition [2]. It is designed to test whether a low-demand filler task leads to the lowest number of problems solved—a result seen previously with a visual problem—or whether that pattern instead appears for the spatial problem [2]. Because this is a planned study, its results are not yet in hand, so the low-demand advantage for insight problems should not be assumed to transfer automatically to spatial creative problems.
The practical takeaway for now: if you want a break to help you crack a stubborn problem, choose something mildly engaging—enough to hold your attention, not enough to drain it. The strongest direct evidence favors low demand over both none and high demand for insight problems [1], while the boundary conditions for other kinds of creative problems remain an open question [2].
Sources used in this answer
Different incubation tasks in insight problem solving: evidence for unconscious analytic thought
Among non-demanding, low-demanding, and high-demanding incubation tasks, the low-demanding task produced the highest insight-problem success rate regardless of task type, supporting the Unconscious Analytic Thought account.
Problem Solving Task Performance: Spatial Incubation Task
A planned experiment will compare low- and high-demand visual-spatial incubation tasks on visual and spatial creative problems, testing whether the low-demand disadvantage seen for a visual problem replicates or shifts to the spatial problem.
The effect of creative task types, incubation activities during incubation period, and choice in tasks on the incubation effect
Taking a break and doing another task helped creative problem solving, and incubation effects for a convergent thinking task differed by activity and task choice, though filler-task demand level was not manipulated.
A demonstration of incubation in anagram problem solving
The study examined incubation in anagram problem solving and whether longer incubation periods aid resolution, without addressing filler-task demand levels.
Forgetting in Creative Problem Solving for College Students
Filler tasks were reported to promote the incubation effect, with forgetting during the incubation period examined, but the activities were not classified by demand level.