The Zeigarnik effect: what Bluma Zeigarnik found, and why it rarely replicates

The Zeigarnik effect is the claim that people remember interrupted, unfinished tasks better than tasks they completed. Bluma Zeigarnik reported it in 1927, after experiments in Kurt Lewin’s laboratory in Berlin. Her participants named interrupted tasks about twice as often as completed ones. Most later replications failed, and a 2025 meta-analysis of 38 publications found no overall memory advantage. What does hold up is a related tendency: people go back to an interrupted task when they get the chance.

• Zeigarnik calculated a ratio for each participant and then took the average. In her main experiment the result was 1.9, which she summarised as “approximately twice as well”.
• In her own data, the effect disappeared in several conditions: tired participants, recall on the next day, advance knowledge of all tasks, and a sudden change of situation before recall.
• The famous waiter story is Kurt Lewin’s anecdote, and it exists in two versions. It is not part of Zeigarnik’s data.
• An exact replication in the 1960s found the reverse pattern in most participants. The 2025 meta-analysis put the usual measure at 0.99, where 1.0 means no difference.
• The Ovsiankina effect, the urge to resume an interrupted task, appeared in about 67% of cases across 21 publications. That part of the theory has much better support.

What the Zeigarnik effect is

The APA Dictionary of Psychology defines the Zeigarnik effect as “the tendency for interrupted, uncompleted tasks to be better remembered than completed tasks”. The entry adds that some theorists relate it to gestalt principles of organisation, applied to memory instead of perception.

The idea came from Kurt Lewin’s field theory. In Lewin’s view, an intention to do a task creates a “quasi-need”, a state of tension that pushes towards completion. Finishing the task releases the tension. An interrupted task leaves the tension in place. Zeigarnik, Lewin’s doctoral student, asked whether that leftover tension changes memory. In the English version of her paper, the question reads: “What is the relation between the status in memory of an activity which has been interrupted before it could be completed and of one which has not been interrupted?”

Zeigarnik was born in 1901 in Prienai, Lithuania, and enrolled at the Friedrich Wilhelm University of Berlin in 1922. Her dissertation appeared in 1927 in the journal Psychologische Forschung. The details below come from the abridged English translation of 1938. They also come from a 2020 history of the effect by the psychologist Colin MacLeod.

Where the waiter story comes from

Most explanations of the Zeigarnik effect start with a waiter. The story belongs to Lewin, not to Zeigarnik, and it is told in two different ways. One of them comes from a person who was there. MacLeod quotes both.

The best-known version comes from the historian of psychology E. G. Boring, in 1957. Lewin and his friends sat in a Berlin restaurant for a long time. When Lewin asked what he owed, the waiter knew the amount at once. Some time after Lewin paid, he asked the waiter again how much he had paid. “The waiter no longer knew.”

The second version comes from Donald MacKinnon, who was at the table. In it, the waiter knew exactly what each person had ordered, without notes. About half an hour after payment, Lewin asked him to write the bill again. The waiter was indignant. “I don’t know any longer what you ordered,” he said. “You paid your bill.”

In one version the waiter remembers the total, and in the other he remembers each order. MacLeod notes that this difference fits Bartlett’s view, from the same period, that memory is reconstructive. Both versions make the same point: the waiter remembered the details while the bill was still unpaid. The anecdote gave Lewin the idea. Zeigarnik’s experiments tested it.

What Zeigarnik’s 1927 experiments found

Zeigarnik tested 164 participants one at a time: students, teachers and children. She also ran two group tests, with 47 adults and 45 children. Each participant did 18 to 22 short tasks, each meant to take 3 to 5 minutes. The tasks were manual or mental: making a box from cardboard, modelling clay figures, puzzles and arithmetic.

The instruction was to complete the tasks “as rapidly and correctly as possible”. Half of the tasks were interrupted. Zeigarnik chose the moment when the participant seemed most engrossed, put the next task on the table and said, “Now do this, please.” After the last task, she cleared the table and asked the participant to name the tasks. She counted only the tasks named before the first pause in recall.

Her measure was the number of interrupted tasks recalled (IR) divided by the number of completed tasks recalled (CR). She calculated this ratio for each participant and then averaged it. A value of 1.0 means equal recall. A value of 2.0 means twice as many interrupted tasks.

The last row tested a simple explanation: perhaps the shock of the interruption made the task memorable. If so, a task that was interrupted and then finished should be remembered just as well. It was not. Zeigarnik also told some participants that an interrupted task would be resumed later, and others that it would not. The ratios were 1.7 and 1.8, so the expectation of a return did not explain the result.

Two other findings fit her theory. Interrupted tasks were named first three times as often as completed ones. Of her 22 tasks, 17 were recalled better when interrupted. In her summary, unfinished tasks were remembered “approximately twice as well as completed ones”, because at the time of recall “there still exists an unsatisfied quasi-need”.

When the effect disappeared in Zeigarnik’s own data

Most summaries stop at “twice as well”. The same paper reports many conditions in which the advantage shrank, vanished or reversed. Zeigarnik saw these results as support for the tension theory: anything that weakens the separate tension of each task should remove the effect.

The time of interruption mattered too. Tasks interrupted near the middle or end had a 90% advantage. Tasks interrupted soon after the start had 65%. Tasks with no clear end, such as marking crosses on a sheet, showed no marked difference. Participants whom Zeigarnik called “ambitious” showed an advantage of 175%. Children averaged 150%, against 90% for adults.

These details matter for anyone who wants to use the effect. Even in the original study, the advantage depended on immediate recall, on engrossed participants and on a calm transition to the recall question. With tiredness during the tasks, or with recall one day later, Zeigarnik’s own data show a much smaller advantage, none, or the reverse.

What later replications found

MacLeod’s review traces the next 40 years. The first follow-up, by Schlote in 1930, did not support Zeigarnik. A 1938 study by Marrow did: 30 participants did 20 paper-and-pencil tasks, and the ratio was 1.77. According to MacLeod, the name “Zeigarnik effect” first appeared in J. W. Atkinson’s 1953 dissertation. Before that, researchers spoke of the “interrupted task paradigm”.

Atkinson also showed that the effect depends on the person. With skill-oriented instructions, people high in the need for achievement recalled more interrupted tasks. People low in that need showed the reverse. Other studies found the reverse in anxious situations, when a failure could threaten self-esteem. The explanation offered at the time was that an unfinished task is a failure, and people may avoid recalling it.

By 1964, a review by E. C. Butterfield concluded that “the Zeigarnik effect is far from being the invariable result”. It added: “Frequently, more completed than incompleted tasks are recalled.” Butterfield also pointed to a design problem. Participants spend more time on completed tasks, so completed tasks get more chance to be learned.

The strongest test came from A. van Bergen, who began an exact replication in 1961. She tested 34 participants, mostly university students. The ratio was 0.88. Nine participants recalled more interrupted tasks, five recalled equal numbers, and 20 recalled more completed tasks. In her 1968 dissertation, she found no effect in any of her seven experiments. By MacLeod’s count, fewer than a third of the 44 replication papers she reviewed reported the effect. She called it a “non-problem”. MacLeod’s own verdict: “At best, it would appear to hinge on certain individual difference characteristics; at worst, it is simply not replicable.”

Is the Zeigarnik effect real? What the 2025 meta-analysis found

In 2025, Romain Ghibellini and Beat Meier of the University of Bern published the first meta-analysis of the Zeigarnik and Ovsiankina effects. Their database search found 1,455 records. After screening, 59 publications remained: 38 on the Zeigarnik effect, 20 on the Ovsiankina effect and 1 on both.

The authors calculated the effect in three ways, because studies reported different measures. Zeigarnik’s own measure, the average of each person’s ratio, gives too much weight to extreme values. Their example shows the problem with two people. Person A recalls 8 interrupted and 4 completed tasks, a ratio of 2. Person B recalls 4 interrupted and 8 completed tasks, a ratio of 0.5. The average ratio is 1.25, which looks like an effect. But both people recalled 12 tasks, 6 of each kind on average, so the ratio of the averages is 1.

With the most common measure, interrupted and completed tasks were recalled equally. Only 8 publications gave enough data for an effect size. Their average was dz = 0.15, which the authors call a small effect. The experimental setting changed the result a little. Across 16 publications, the ratio was 1.07 in relaxed conditions and 0.96 in neutral conditions. It was 0.88 when the tasks were framed as a test of ability.

The authors concluded that the Zeigarnik effect “lacks universal validity”. They also suggested a reason for the original result. In the 1920s, a professor’s laboratory had great authority, and participants worked with high involvement. Today, they argue, people are used to constant interruptions. This explanation is a hypothesis from the authors. The meta-analysis did not test it.

The Ovsiankina effect: the part that holds up

In 1928, Maria Ovsiankina, who also worked in Lewin’s laboratory, studied what people do after an interruption. When she gave them a period with few demands, they tended to go back to the interrupted task without being asked. This is the Ovsiankina effect. It follows from the same tension theory, but it measures action, not memory.

In the 2025 meta-analysis, people resumed interrupted tasks in 67.00% of cases across 21 publications. Without Ovsiankina’s own study, the figure was 66.79%. The authors call this “well above the chance rate of 50%”. They also warn that the studies measured resumption in different ways, so the percentage is an indication, not an exact value.

Two studies in the meta-analysis show how this tendency behaves. In a 1982 study by McGraw and Fiala, participants who knew that they would be paid resumed interrupted tasks noticeably less often. The authors read this as an extrinsic goal replacing the wish to finish. Our article on intrinsic and extrinsic motivation explains why a reward can change behaviour this way. In a 1969 study by Malerstein, some patients with Korsakoff’s syndrome, a severe memory disorder, resumed a puzzle that they could not recall. The meta-analysis authors note that boredom during a 20-minute wait may also explain this, so the result is uncertain.

Why unfinished tasks stay on your mind, and how a plan stops it

Many people know the feeling that an unfinished task keeps coming back to mind. That experience is not the same claim as the Zeigarnik effect. Zeigarnik measured which tasks people could name at the end of a session. The feeling concerns thoughts that intrude while you do something else.

The second claim has more direct modern evidence. In 2011, E. J. Masicampo and Roy Baumeister published a set of studies on unfulfilled goals. According to their abstract, an unfinished goal caused intrusive thoughts during an unrelated reading task. Words related to the goal were also more accessible in the mind, and performance on an unrelated anagram task fell.

The same studies found a simple way to stop this. When participants made a specific plan for the unfinished goal, the intrusions and the interference disappeared. The participants who later carried out their plans were the ones who had no more intrusions. The authors suggest that a committed plan lets the mind put the goal on hold until the time that the plan sets. This is one paper with several studies. We did not find a published replication.

Does the Zeigarnik effect help you study?

Many study tips say: stop in the middle of a chapter, and the unfinished topic will stay in your memory. The evidence above does not support that tip. Also, Zeigarnik measured memory for the tasks themselves, not for the content that people learned while they did them. Across studies, interrupted tasks were not recalled better. In Zeigarnik’s own data, the advantage was small or absent after one day and in tired participants. Exams almost always come after more than one day.

Two parts of the research are more useful for studying:

1. 1.If you must stop, write down the next specific step and when you will do it. In the 2011 studies, a specific plan stopped intrusive thoughts about the unfinished goal.
2. 2.If an interrupted task is easier to restart, use that. The Ovsiankina studies suggest that people often return to unfinished work, but they do not show that this improves learning.

For remembering the material, other methods have much stronger evidence. Testing yourself and spreading study sessions over days both improve long-term memory. Our article on spaced repetition explains how spacing works and why it beats cramming.

One more detail from the meta-analysis is worth knowing. In two older studies, a demanding activity between the tasks and the recall made people recall more completed tasks. The authors attribute this reversal to retroactive inhibition, also called retroactive interference: new mental work disrupts the memory of earlier material.

Keeping the terms apart

The Zeigarnik effect, the Ovsiankina effect, quasi-need, the interrupted task paradigm and the need for achievement come from the same research tradition. They are easy to mix up on an exam. In Psychology World you can save terms like these to one list. Then you can drill the list as flashcards, as multiple-choice questions once it has four or more terms, or by typing the answer. What each study mode does is set out on the home page. Our guide to memorising psychology terminology explains which study methods have the strongest evidence.