Why would a shower beat a hard task — or lose to one?
There are two rival explanations for why stepping away from a problem sometimes helps. One, often called the unconscious work view, proposes that automatic processes keep activating problem-related information while you are away, and that a low-effort task is best because it keeps your conscious mind from returning to the problem [1]. The other, the return-to-act view, proposes that the break lets false assumptions fade so you can re-encode the problem in a new way, and that a demanding task is best because it provides the maximum distraction [1]. These two theories make opposite predictions about effort, which is exactly why the question is testable.
The practical upshot is that "showering" is not a magic ingredient. What matters in these accounts is how much the break occupies your attention and whether it keeps you from consciously chewing on the problem. A shower is a candidate low-demand activity; a demanding task is the high-demand alternative. The theories disagree about which one should win.
The results so far point in two directions at once
In the earlier experiment from one lab, the low-demand incubation condition led to completing the creative problems fastest, measured as total time working on them. The high-demand condition, by contrast, showed a trend toward the most problems solved correctly [1]. That is a split result: low effort looked better on speed, high effort looked better on accuracy. The authors themselves flag that some problems had very different solution rates and that the conditions were between-subjects, which limits how firmly the comparison can be drawn [1].
A separate study tested incubation activities more directly across two kinds of creative thinking. For convergent thinking — problems with a single correct answer, like the remote association task — incubation effects differed according to the activity and whether participants could choose their task, and the group with no choice showed significant improvement over the choice group. For divergent thinking — open-ended tasks like listing alternative uses — there was no difference in incubation effects according to activity or choice [2]. Notably, that study could not find a difference in incubation effect between the incubation activities themselves [2].
So the honest summary is that low effort has some support on speed, high effort has a weak signal on accuracy, and at least one study found the choice of activity made no measurable difference. The evidence does not establish that showering-type breaks beat demanding tasks.
Why the answer is still unsettled
Part of the problem is that "incubation activity" is not one thing. A follow-up experiment is now testing low- and high-demand visual-spatial incubation tasks on visual and spatial creative problems, precisely to see whether performance changes when the break engages cognitive resources similar to the problem [3]. That matters because a break that uses the same mental resources as the task may behave differently from one that does not. The same follow-up also notes it will check whether the earlier low-demand result reappears for the visual problem or instead shows up for the spatial one [3].
The broader creativity literature treats incubation as a real phase in artistic and scientific work, but does not settle the effort comparison. One account describes a long incubation phase in science and poetry without comparing low- versus high-effort breaks [5], and a management-oriented discussion mentions pausing for incubation without testing effort levels [4]. Those sources support the idea that stepping away can matter; they do not tell us whether a shower beats a hard task.
What this means if you are choosing how to take a break
If your goal is to get through creative problems quickly, the low-demand result is the one piece of direct evidence in its favor [1]. If your goal is to get more problems right, the high-demand trend is suggestive but was only a trend, and the same study's own limitations caution against leaning on it [1]. And because another study found no difference between activities [2], you should not assume the choice of break is decisive.
A reasonable reading is that the type of break is probably not the main lever, and that the effect may depend on the kind of problem and on how closely the break's mental demands overlap with the task [3]. The specific image of a shower as an ideal incubation chamber is not something these studies test. What they test is effort level, and on that question the evidence is mixed rather than settled.
Sources used in this answer
Problem Solving Task Performance 2
A previous experiment found a low-demand incubation task led to the fastest completion of creative problems, while a high-demand task showed a trend toward the most problems solved correctly, with noted limitations of differing solution rates and between-subjects conditions.
The effect of creative task types, incubation activities during incubation period, and choice in tasks on the incubation effect
Incubation effects on convergent thinking differed by incubation activity and task choice, with the no-choice group improving significantly over the choice group, while divergent thinking showed no differences by activity or choice and no difference was found between incubation activities.
Problem Solving Task Performance: Spatial Incubation Task
A planned experiment will test whether low- and high-demand visual-spatial incubation tasks affect visual and spatial creative problem solving, including whether the earlier low-demand result replicates for the visual or the spatial problem.
When sparks fly: Igniting creativity in groups
A management-context discussion mentions pausing for incubation in creative problem solving but does not compare low-effort and demanding incubation tasks.
Creativity: An elixir of life
A discussion of creativity in artistic and scientific contexts describes long incubation phases but does not compare low-effort and demanding incubation activities.