Von Restorff Effect: Why the Odd One Out Is the One You Recall

An item is remembered better when it breaks the pattern of the items around it. The word in the accent color in the demo above is not easier to read, shorter or more familiar than the other fourteen. What it has is a background: fourteen words that are alike in the one respect in which it differs. Hedwig von Restorff's insight, which Reed Hunt restated in 1995 after going back to her 1933 paper, is that the effect belongs to the relation between the item and its set rather than to anything about the item on its own. Put the same word in a list where every word is a different color and the advantage disappears.

Two processes have been offered, and the evidence points to both. At encoding, an item that violates the pattern draws extra attention and extra processing, which is why a word in a different font, size or color is studied longer than its neighbors even when nobody is told to study it. At retrieval, the isolate is in a category of its own, so almost any cue reaches it, while the fourteen similar words compete with one another. The retrieval account is what explains the strangest part of the finding: von Restorff herself put the odd item early in the list, where perceptual surprise has the least room to work, and the advantage still appeared.

The effect is not unconditional, and two findings mark its edges. Stephen and Constance Schmidt reported in 2017 that under moderately distracting conditions they found isolation effects for items in the middle of a list but not for items near the beginning, and that early isolation by category actually hurt recall of the distinct items compared with a list where everything matched. Attention at study and a retrieval environment that favors the isolate both have to be in place. Siri-Maria Kamp and colleagues reported in 2026 that in their first three experiments the isolation advantage was absent on recognition tests unless recall was tested first, and that three further experiments, which took more observations per person and controlled the order of testing, found isolation effects in recognition and in source memory. Free recall is where the effect is most robust; whether recognition shows it depends on how the test is run.

What this adds up to for a page, a slide or a set of notes is narrower than the usual advice. One thing picked out of many similar things is likely to be the thing recalled. Pick out five, and there is no longer a background for any of them to stand against.

Hedwig von Restorff published "Über die Wirkung von Bereichsbildungen im Spurenfeld" in Psychologische Forschung in 1933, working in Wolfgang Köhler's laboratory in Berlin; she and Köhler published a second paper on the same memory problem in 1937. The 1933 paper was never published in English, which is one reason the effect is often described loosely, and most English-language accounts of it rest on Hunt's 1995 review, which worked from a translation by Andrea Dorsch.

Her method was a comparison between lists rather than a comparison within one list. In a homogeneous list, every item came from the same class. In an isolated list, one item came from a different class, for example a number among syllables, while the rest stayed alike. The control for the isolated item was the same item in the same serial position in the homogeneous list, which is the comparison that makes the effect a fact about isolation rather than about the item. Later reviews report that she placed the odd item early in the list, at the second or third position, which keeps the advantage from being explained by surprise at the moment it appears.

Hunt's reading, in "The subtlety of distinctiveness: What von Restorff really did," is that she was not demonstrating that striking things are memorable. She was demonstrating that memory is organized by similarity and difference, so that an item defined against a uniform background is retrieved more easily than an item that shares its description with every other item on the list. The practical rule that follows is about the background, not the item.

The modern version of her experiment is the demo above. Paolo Vitali and colleagues ran it in 2006 with 25 lists of 10 words each, one word per list printed in double-size font, and free recall after every list. Young volunteers recalled 60 percent of the isolated words against 45.7 percent of the ordinary ones, and elderly volunteers showed the same gap at a lower level, 31.2 percent against 20.2 percent. Tamra Bireta, Aimee Surprenant and Ian Neath reported a comparable age pattern in 2008 with an immediate memory task.

The effect is used deliberately far more often than it is suffered. The rows below are illustrations, not findings from the studies cited on this page.

The isolation effect is a benefit when you control what stands out and a cost when something else does, so the practical rules are about who is choosing the odd item.

• Pick one thing. The advantage comes from a uniform background, so a page with one highlighted term works and a page with ten does not.
• Make the isolate the thing worth recalling. A number in a different color is remembered whether or not it matters, so the choice of what to isolate is the whole decision.
• Expect the rest to stay ordinary. The effect is an advantage for the odd item measured against the same item in a uniform list; it is not a method for remembering everything.
• When you are the audience, ask what the uniform field was. If a presentation left you with one vivid figure, that is the figure that was isolated, not necessarily the most important one.
• For learning, do not rely on it alone. Spacing and retrieval practice do more for a whole list than distinctiveness does for one item, and the two can be combined.

The von Restorff effect is one of several reasons an item is recalled out of proportion to the rest of a list.

• It is not the serial-position effect, the U-shaped curve in which the first and last items are recalled best. That is an advantage of position; isolation is an advantage of difference, and in von Restorff's own experiments it appeared at an early position where primacy was already helping.
• It is not the bizarreness effect, the claim that strange or absurd material is remembered better than ordinary material. Isolation requires only that one item differ from its neighbors, and the odd item can be perfectly mundane.
• It is not the humor effect, the better recall of funny items, which is one particular way for an item to differ from a set.