neur.ro

Zeigarnik effect

type: effect · origin: study · evidence: failed to replicate · themes: learning and memory

Claim assessed: Interrupted tasks are remembered better than finished ones.

Published · how we work

In short

The Zeigarnik effect is the idea that you remember interrupted tasks better than finished ones. The classic explanation says an intention creates a tension that stays active until you finish. The idea is well known and often invoked, but a recent meta-analysis did not find the memory advantage. It did find a related effect, named after Maria Ovsiankina: people tend to resume interrupted tasks on their own.

What it says

The original studies. Building on Kurt Lewin’s ideas, Bluma Zeigarnik (1927) gave participants a series of 18–22 short tasks and interrupted half of them. She then asked them which tasks they remembered. They named interrupted tasks more often: for every 10 finished tasks recalled, about 16 interrupted ones. With a similar method, Maria Ovsiankina (1928) observed that participants resumed interrupted tasks on their own when given the chance. We have not read the original studies; we describe them from the meta-analysis below.

The meta-analysis. Ghibellini and Meier (2025) collected the studies that used the same method, from 1927 on: 38 publications on the Zeigarnik effect, 20 on the Ovsiankina effect and one on both. In the 37 publications after Zeigarnik, interrupted tasks were recalled as well as finished ones: the ratio between them was 0.99. Measured another way, interrupted tasks made up 49% of all recalled tasks, so no more than half. In the 8 publications with enough data, the computed advantage was small. The authors note that the measure Zeigarnik used inflates the result, because a few participants with extreme values sway it heavily.

Top: for each finished task recalled, Zeigarnik's participants recalled 1.61 interrupted tasks; in the 37 later publications, 0.99, the same number. Bottom: Ovsiankina's participants resumed 91% of interrupted tasks; in the 20 later publications, 67%. What is left of the two effects Ghibellini and Meier, 2025 Interrupted tasks recalled per finished one 1 means the same number Zeigarnik, 1927 1.61 37 later publications 0.99 Interrupted tasks resumed when participants had the chance Ovsiankina, 1928 91% 20 later publications 67% The memory advantage did not hold up; the tendency to resume tasks did.
The later values are the meta-analysis averages, weighted by number of participants and computed without the original studies. The 1927 and 1928 values are the ones the authors computed from the original studies.

The role of the situation. The authors grouped the experimental conditions by atmosphere. When the tasks were presented neutrally, the ratio was 0.96. When they were presented as an intelligence test, it was 0.88, so interrupted tasks were recalled somewhat less often. Only in a relaxed atmosphere, with the tasks presented as a mere test of the materials, did the ratio rise slightly above 1, to 1.07. Participants’ achievement motivation did not change the result much. The authors’ conclusion: the Zeigarnik effect is not general, but depends, at most, on the situation and the person. MacLeod (2020), in a review they cite, says the same: at best, the effect hinges on certain individual characteristics; at worst, it simply does not replicate.

The Ovsiankina effect. In the 20 later publications, interrupted tasks were resumed 67% of the time on average, when participants were left alone with them or were given a new task in between. The tendency appeared in adults and in children. Studies described by the authors show that it dropped when participants knew they were being paid for taking part and when the new task resembled the interrupted one.

Example

A student deliberately stops halfway through an exercise, convinced this will help them remember the material. The evidence does not support that they will remember it better. At most, it supports that they are somewhat more likely to come back to the exercise. The example is an illustration, not a study cited here.

How to apply it

The steps below are a practical approach we propose, based on the studies cited.

  1. Don’t rely on interruption to remember something. On average, interrupted tasks were not recalled better than finished ones.
  2. Write down what you left unfinished. Since recall isn’t guaranteed, a list is safer than the feeling that the task “stays on your mind”.
  3. If you want to come back to something, leaving it started may help. People resumed interrupted tasks about two thirds of the time, but the figure is indicative, not a guarantee.
  4. Be skeptical when the Zeigarnik effect is used as an explanation. The meta-analysis authors note that it is often cited as settled fact, although it does not replicate consistently.

Limits and nuances

Sources

See also: Clickbait (curiosity gap), Serial position effect, Spacing and retrieval practice