The brain dump method is only half a method

Eighty adults in the fifth of E. J. Masicampo and Roy Baumeister's experiments on unfulfilled goals were asked to think about one thing they still had to get done in the next few days. Some of them listed every way they could imagine getting it done, without committing to any of them. Others wrote one plan: how, when, and where. Then everybody read the first 600 words of a detective novel, and afterwards said how much the task had been on their mind while they read. The list-makers were still thinking about it, at 3.37 on a seven-point scale. The planners were at 2.14, and the people who had written about an errand they had already finished sat between the two at 2.60.

A list of options left the nagging worse than thinking about finished business. Emptying your head is the easy part. The work happens in the ten minutes afterwards, when two or three items get a specific time and place, because that is the only step in the whole procedure anyone has actually run an experiment on.

A brain dump is one sitting in which you write down every commitment, worry and half-formed idea you're carrying, without sorting or judging any of it. Nobody disputes that part.

The pages ranking for the method agree with each other about almost everything and cite almost nothing. InstantOwl asks for five to ten minutes and three buckets. Quik.md asks for fifteen minutes and three passes. Vanulos asks for twenty minutes, then a ten-minute walk, then five categories. All three explain why the method works by naming the Zeigarnik effect, and a meta-analysis published last year put that effect at a recall ratio of 0.99 across 38 publications, which is the number you get when there's nothing there.

Where the open loop came from, and what it originally asked of you

The phrase belongs to David Allen. His company has been publishing the same definition since at least 2011: "Every commitment unfinished is an 'open loop'; and when it is tracked in your psyche, instead of your system, it will require energy and attention to track and maintain." The capture phase that later became the brain dump is the first of that framework's five, ahead of Process, Organize, Review and Do, and the same page is careful about what emptying means. "By the way, the word 'empty' doesn't mean complete the work, just define it and organize it."

The same company put it more bluntly in March 2026. "You don't need to finish it. You need to define it." Their three steps are to clarify what a successful outcome would be, decide the very next action, and put both somewhere you trust. None of that is writing a long list quickly.

The popular version kept the fun half. It dropped the half that carried the claim, turning the dump into the technique and the sorting into an optional appendix, which left the justification to come from somewhere else.

The Zeigarnik effect is not holding that argument up

Romain Ghibellini and Beat Meier pooled every study they could find on Bluma Zeigarnik's 1927 result, and published the numbers in Humanities and Social Sciences Communications. On the most commonly used measure, the ratio of average recall for interrupted tasks to average recall for finished ones, they got 0.99 across 38 publications. Removing Zeigarnik's own data leaves it at 0.99. Interrupted tasks made up 49.16 percent of everything people recalled, which is a hair under half, and the eight publications that reported enough detail to compute an effect size averaged d = 0.15. Their conclusion is one sentence: "the Zeigarnik effect lacks universal validity."

Something in the old story does survive, and it's the other half nobody quotes. Maria Ovsiankina's finding, that people go back to an interrupted task when given the chance, came out at a weighted resumption rate of 67.00 percent across 21 publications. You do return to unfinished things. You just don't remember them better than the things you finished, which is the claim the brain dump guides need and don't have. What actually predicts whether you get back to something looks more like the quiet middle of a long project, where measured effort drops off.

Vanulos states the opposite outright, saying the effect has "been replicated so consistently across cognitive science that it carries her name." The same page reaches for George Miller's 1956 paper to establish that working memory holds seven items, give or take two. Miller meant it loosely. Nelson Cowan's review of that literature quotes the intent directly: the number "was meant more as a rough estimate and a rhetorical device than as a real capacity limit," and the figure the evidence supports is three to five chunks. It's a measure of what you can hold in mind for a few seconds anyway, not a count of the commitments you've been carrying around since Tuesday.

The brain dump method, sorted by what has been measured

Here is the standard procedure with its evidence attached, and the empty rows marked empty.

The ten minutes after the dump

It's worth reading those experiments closely, because the effect they found is narrower than "write it down." In the first of them, 73 undergraduates wrote about two important errands due in the next few days, and were told to pick tasks where they didn't yet know when, where, or how. Sixty-nine made it into the analysis. One group then wrote out, in detail, when and where and how they would do each one.

Everyone read the first 3,200 words of The Case of the Velvet Claws and answered eight questions about the plot. The unfulfilled group reported being distracted by their errands at 3.00 and got 6.13 questions right. The planners reported 1.77, which was no different from people who had written about finished errands, and got 6.94. Their expectations of actually completing the tasks were identical to the other group's, 6.28 either way, so this is not confidence doing the work.

A later study in the same paper shows what happens when the plan is not sincere. Ninety-seven students were given a brainstorming task to do at the end of a packet, and a suggested method for doing it well. One group simply committed to trying hard. Another committed to the method in so many words: when I get to the final task, I will write down the letters of the alphabet and list sea creatures for each one. Before either group got near the brainstorm, everyone spent five minutes on 25 anagrams. The group carrying an unplanned goal solved 6.55, the control group 8.39, and the planners 9.55. Seventy percent of the planners went on to use the method they had committed to, against 13 percent of the goal group and 6 percent of the controls, and the anagram advantage was fully mediated by whether people actually used it.

The mediation is the whole story. A plan you don't intend to keep doesn't buy the attention back, which is why it has to name a moment you actually mean. Peter Gollwitzer and Paschal Sheeran got to the same place by pooling 94 tests of naming the when, the where and the how, putting the average effect on goal attainment at d = .65. We used that result when arguing that a reading habit is built on a fixed time and place.

So the last five minutes of a dump look like this. Read the pile once, pick the two or three items that are loudest, and write each one as a sentence with a time and a place in it: after I get back from lunch tomorrow, I call the dentist from the kitchen. Leave the rest of the pile alone.

Does a brain dump actually reduce stress

Relief is what the whole genre sells, and Vanulos puts mental clarity in the title. Half of that promise is supported. The other half was tested twice in the same paper and came back empty both times.

Masicampo and Baumeister measured four emotions after people made their plans. Anxiety ran at 3.15 in the control group, 4.11 among people carrying an unfulfilled task, and 4.06 among the planners. Measured again after a delay and a reading task, it ran 1.95, 3.63 and 3.32. The planners stayed as anxious as the people with no plan at all, while reporting significantly fewer intrusive thoughts than them, and the authors say so plainly: "plans did not alleviate anxiety or restore contentment."

That is not nothing. Getting your attention back while still feeling unsettled is a real trade, and it's the trade on offer. What has been tested on the worry itself is a different exercise. Gerardo Ramirez and Sian Beilock had ninth graders spend ten minutes writing out their worries about the exam they were about to sit, the first final of their high school careers. Among the students who habitually got anxious about tests, the writers beat the controls by 6 percent, t(52) = 2.08, d = 0.57, a B-plus against a B-minus, and ended up performing like their calmer classmates. Students who were not especially test-anxious showed no difference at all. That was prose about one imminent thing, not a list of forty.

The strongest evidence for a dump is at bedtime, and it says write more

Michael Scullin's team brought 57 healthy adults aged 18 to 30 into a sleep lab, gave them five minutes with pen and paper before lights out at 10:30 p.m., and recorded the night with polysomnography. Half wrote a to-do list for the coming few days. Half wrote about what they had finished over the previous few days. The list-writers fell asleep in 15.82 minutes against 25.09, t(55) = 2.32, p = .02, d = .63.

Then the advice runs into trouble. Within the to-do list group, the more items somebody wrote, the faster they fell asleep, r(26) = -.39. In the completed-activity group the trend ran the other way. A longer, more specific list was better. InstantOwl tells you to stop at the timer because "the goal is an empty head, not a complete inventory," and that instruction is guessing against the only measurement anybody has taken.

Read the limits before you rearrange your evening around it. This was one night, in a laboratory, with young adults screened for sleep disorders, using the first night in a lab as a stand-in for a bad night at home. The authors took no measure of arousal or worry, so the reason the writing helped is still open. When people wrote fewer than ten items, the gap between the two conditions narrowed.

What to do with the pile

Keep it, if you will reopen it. Quik.md gets this one right without citing anybody, warning that a dump into a notebook you never reopen "teaches your brain that externalizing does not work." There's an experiment behind that. Benjamin Storm and Sean Stone found that saving a file before studying a new one improved memory for the new one, and that the benefit disappeared when the saving process was described as unreliable. Trust in the container is doing the work, not the writing.

Don't throw items out on the day you write them. A dump made at 11 p.m. is a poor judge of what matters, and the sorting pass is where people quit anyway, which quik.md concedes in its own numbers: "forty items times thirty seconds of filing is twenty minutes of drag." Carry three commitments out of it. Leave the rest for the plan you write on Sunday, which is a better moment to reread a pile than the night you wrote it.

Then do the dump again next week. The point of a second one is that it'll be shorter, and the items that come back twice are the ones you've been avoiding rather than forgetting.