A meta-analysis of incubation research found that stepping away from a problem helps most when the break is filled with something undemanding — for word problems, a boring task beat resting outright

Getting stuck on a problem costs more than the minutes spent staring at it.

Push through anyway and the same three approaches tend to circle back, none of them any better than the last. Walk away entirely, and there’s a real chance of never coming back to it, or coming back an hour later no closer to an answer than before.

Both routes waste the same thing: a window where a fresh angle was actually available, closed by handling the break badly.

The common advice is to just step away and let it “incubate.” What that advice usually leaves out is what to do during the step away part, and it turns out that detail changes whether the break works at all.

What a meta-analysis of incubation research actually did

The idea that a break can help solve a problem has been tested in psychology labs for decades, with individual studies reaching different conclusions about how much it helps, or whether it helps at all.

A meta-analysis takes a large batch of those separate studies and combines their results into one clearer picture, which is what psychologists Ut Na Sio and Thomas Ormerod did with more than 100 tests of the incubation effect.

Ormerod is a professor of psychology at the University of Sussex who has spent years researching how people solve problems and where they get stuck along the way.

Their review, published in Psychological Bulletin in 2009, confirmed that stepping away from a problem does help on average. That much matched the common advice. What it also found was that the size of the benefit depended heavily on one detail nobody usually thinks to mention: what a person actually does during the break.

The finding that complicates the standard advice

Sio and Ormerod reported that “low cognitive demand tasks yielded a stronger incubation effect than did rest during an incubation period when solving linguistic insight problems.” In plainer terms, for word-based problems, filling the break with something simple and undemanding worked better than doing nothing at all. Full rest, the version of a break most people would assume is the most restorative, came out behind a boring task.

That result runs against the instinct to treat a break as an empty space, something you clear entirely so the mind can settle. The finding suggests the opposite ordering. An empty, restful break did less for these problems than one with a small, easy task sitting inside it.

It’s worth being precise about how solid that particular ordering is. The studies feeding into this part of the review were controlled experiments, with the type of break assigned by the researchers rather than chosen by participants, so the comparison reflects a deliberately manipulated variable and not just two groups that happened to behave differently on their own.

That’s stronger footing than a single observational study would offer. It’s still a summary built from more than 100 separate experiments run by different teams over different years, with results that vary from one to the next, so treating the finding as a settled law rather than a well-supported pattern would go further than the evidence does.

This pattern shows up most clearly for word problems, not everything

The meta-analysis covered more than one kind of problem, including visual and spatial puzzles alongside the word-based ones, and a break of some form helped across the board.

The specific ordering, an undemanding task beating full rest, is documented for linguistic insight problems, the kind where the answer usually hinges on rethinking a familiar word or phrase.

Whether that exact ordering holds for a visual puzzle, a design choice, or a strategic decision at work isn’t something this part of the review pins down.

The safer takeaway is narrower than the headline version: for word-based sticking points specifically, an occupied break has the edge over an empty one.

Why doing something might beat doing nothing

The meta-analysis establishes that this pattern shows up reliably. It doesn’t hand over a settled explanation for why, and it would be overreaching to claim otherwise.

What seems to hold up as a reasonable read of it is that true rest gives the mind too much room, room that tends to get filled with worry about the very problem you were trying to step away from, or with nothing at all. A task that’s simple enough not to demand real attention, but active enough to occupy the part of the mind that would otherwise circle back to the problem, may be what actually creates distance from it.

That’s a plausible account, not a proven mechanism, and it’s worth holding it that loosely. What the data supports more directly is the pattern itself: for this type of problem, a light task beat sitting still.

What this changes about how to take a break

The practical version of this isn’t complicated, though it does cut against the instinct to either power through or fully check out. The studies behind this finding used simple filler tasks, not deep rest and not an open-ended scroll through a phone. A reasonable translation of that for an ordinary afternoon might look like folding laundry, washing a small stack of dishes, or taking a short walk without a podcast running in one ear. Enough activity to occupy the hands and a sliver of attention, without asking for any real thought.

A feed that keeps refreshing with new information probably isn’t the same thing. It asks for attention in a different, more demanding way than folding a shirt does, so lumping the two together as equally undemanding may be part of why so many breaks that involve a phone don’t end up helping the stuck problem underneath them.

None of this means every stuck moment calls for a laundry basket. It means the space between working on a problem and solving it might be better spent lightly occupied than left empty, at least for the kind of problem where the answer depends on a fresh angle rather than more grinding.