Retroactive Interference Psychology Definition

Short answer

• Retroactive interference is a memory effect in which new learning interferes with recall of earlier learning. In plain terms, later information gets in the way of older information. Retroactive means backward. The newer material acts backward in time on the older material. A common example is learning a new phone number and then struggling to recall the old number that was once automatic. The old number may still be stored, but the new number competes with it when you try to retrieve it. Retroactive interference is most likely when the old and new material are similar. It is one explanation within interference theory, which holds that forgetting can happen because memories compete, not only because memories fade with time. It does not mean the older information is permanently erased, and it does not by itself mean a person has a memory disorder. It describes a normal pattern of competition between memories.

What Retroactive Interference Means in Psychology

In psychology, retroactive interference is defined by order and direction. First, a person learns one set of material. Later, the person learns a second, related set. When tested on the first set, recall is poorer than it would have been without the second set, or the second set intrudes into the answer. The key test is whether the newer material changes performance on the older material.

University teaching material from the College of the Canyons, published through Social Sci LibreTexts, puts it this way: retroactive interference refers to new activities during the retention interval that interfere with retrieving a specific, older memory. The retention interval is the time between learning and testing. During that interval, memories can consolidate, which aids retention, but new experiences can also undermine the memory. That combination matters. The same interval that allows a memory to stabilize also allows new material to arrive and compete with it.

Encoding means taking information in and forming a memory. Storage means keeping it over time. Retrieval means bringing it back when needed. Retroactive interference is expressed at retrieval: the moment of recall is when the competition shows up, even if the interfering learning happened earlier. A learner may have encoded the old item well and stored it, yet still fail to retrieve it cleanly because a newer, similar item answers the cue first.

This distinction matters for study. If old information was never learned well, extra study of the old material may help. If it was learned but is now blocked by newer material, a better cue, a clearer distinction between the old and new items, or practice in telling them apart may help more than simply rereading the old notes. In practice both problems can occur together, and a student who keeps producing the new answer is not necessarily failing to study. The newer response may simply be fresher, more recently rehearsed, and easier to reach.

Retroactive interference also differs from forgetting through lack of use. A faded memory may be weak on its own. A memory affected by retroactive interference may be blocked by a rival. That is why the same person can recall the old item in one setting and fail in another. When the cue is distinctive, the old item may win. When the cue is shared with the new item, the new item may win.

Retroactive Interference and Proactive Interference in Brief

Proactive interference is the mirror image of retroactive interference. It occurs when earlier learning impairs later learning or recall. The older material acts forward on the newer material. For example, trying to remember a new password while the old password keeps coming to mind first is proactive interference.

The two effects are measured differently. Retroactive interference is measured by performance on the earlier material. Proactive interference is measured by performance on the later material. Retroactive points backward, from new to old. Proactive points forward, from old to new. To tell them apart, ask which material is being tested. If the test is on old material and new material interferes, it is retroactive interference. If the test is on new material and old material interferes, it is proactive interference.

For a full treatment of the forward direction, including its classic laboratory studies and the release from proactive interference finding, see Proactive Interference Psychology Definition. This article keeps proactive interference brief on purpose, because it is a separate topic with its own article. The two articles are companions: together they cover both directions of interference, and neither should be read as covering both.

The Classic Test: Learn A, Then Learn B, Then Recall A

The standard laboratory test of retroactive interference uses a simple sequence. Participants first learn an original set of material, often called List 1. They then learn an intervening set of material, called List 2. Finally, they are tested on List 1. A control group learns List 1 and then rests, or does an unrelated task, instead of learning List 2. If the group that learned List 2 recalls List 1 worse than the control group, the difference is attributed to retroactive interference from List 2.

The most famous version uses paired associates. Participants learn pairs such as cat and tree, jelly and moss, book and tractor. This is the A-B list: a stimulus word A paired with a response word B. The experimental group then learns a second list in which the same stimulus words are paired with new responses, such as cat and glass, jelly and time, book and revolver. This is the A-C list: the same A, a new C. When tested on the first list, the experimental group recalls the original responses less accurately than the control group, which never learned the second list. Learning the A-C pairs interfered with recall of the A-B pairs.

This design is useful because it isolates the direction of the effect. The second list comes after the first, so any disadvantage for the first list cannot be explained by the first list interfering with the second. It also isolates similarity: the stimulus word is identical in both lists, so the same cue now points to two different answers. The learner must select the older answer while the newer answer competes for the same cue.

The design also has a clear limit. Word lists in a laboratory are not a semester of study, a workplace procedure, or a lifetime of phone numbers. The lists are short, the material is arbitrary, and the delay is often brief. The design shows that retroactive interference can happen under controlled conditions. It does not by itself show how large the effect is in everyday life, where material is meaningful, spaced over time, and supported by many cues. Everyday examples in this article are illustrations of the pattern, not replications of the laboratory result.

Why Similarity Matters Most

Similarity is the strongest general condition for retroactive interference in classic research. When two sets of material share meaning, sound, structure, or context, they are harder to keep apart. A new list of words from the same category as an old list interferes more than a new list from a different category. A new procedure that shares steps with an old procedure is easier to confuse with that procedure than one that looks and feels different.

The classic demonstration comes from McGeoch and McDonald in 1931. Participants learned a word list to perfect recall. They then spent ten minutes resting, learning numbers, learning nonsense syllables, or learning words related to the original list, such as adjectives, antonyms, or synonyms. Recall of the original list fell as the intervening task grew more similar to it, from about four items after rest to about one item after synonyms. Because the time interval was identical across conditions, time alone could not explain the difference. The similarity of the intervening material did the work.

Similarity can take several forms. Semantic similarity means the items mean similar things, such as two sets of vocabulary from related languages. Sound based similarity means the items sound alike. Structural similarity means the items share a pattern, such as two passwords with the same format or two procedures with the same steps in a different order. Contextual similarity means the items were learned in the same place, at the same time, or for the same purpose. Each form of overlap gives the newer memory another route into the retrieval attempt for the older memory.

Distinctiveness works in the opposite direction. When older material has a feature the newer material does not have, that feature can act as a tag. The tag might be a vivid example, a different category, a different location, or a clear contrast stated in words. It does not need to be dramatic. It needs to separate the old item from the new item at recall. A small but consistent difference, such as always writing the old term with its defining contrast beside it, can be more useful than a large but irrelevant difference.

Recency also matters. Newer memories are often more available because they were rehearsed more recently. Availability is not accuracy, and under time pressure the most available answer may slip out before it is checked. That is one reason an old password can feel lost the week after a change and then return later, when the new password is no longer quite so fresh and the old one has had a chance to be cued distinctly.

Response Competition, Unlearning, and Where Researchers Disagree

Researchers do not fully agree on what happens inside the memory when retroactive interference occurs, and the article should not pretend otherwise. Two explanations have dominated, and modern accounts often combine them.

The first explanation is response competition at retrieval. Both the old and the new memories remain stored, but they share a cue, so they compete when the cue is presented. The newer response is often stronger or fresher, so it wins, and the older response is blocked. On this account, the old memory is not gone. It is outcompeted. A distinctive cue, more time, or a changed context can sometimes let the old memory win again, which fits the common experience of an old number returning after the new one stops dominating.

The second explanation is unlearning, sometimes called response unlearning. On this account, learning the new response actively weakens the association between the cue and the old response. The old association is not merely blocked at the moment of recall. It is degraded by the new learning itself. Early interference theorists took this view seriously, and it remains part of the debate.

The evidence does not settle cleanly for one account in all cases. The A-B, A-C paradigm shows competition clearly, because the same cue points to two answers. Other findings, such as partial recovery of the old response with distinctive cues or after a delay, fit competition better than permanent unlearning. Still other findings fit a weakening of the old association. The honest summary is that both processes may contribute, and their balance may vary by material, delay, and test format. What the classic studies establish firmly is the effect itself: new, similar learning impairs recall of older learning. The exact internal mechanism remains a live question, not a settled fact.

A related modern idea is retrieval induced forgetting, in which retrieving one memory makes related, competing memories harder to recall later. This is not identical to retroactive interference, because the interfering event is retrieval rather than new learning, but it points in the same direction: competition between similar memories is an active process, not a passive fading. Brain imaging research has also shown that repeatedly retrieving one memory can actively suppress the brain patterns of its competitors, with the amount of suppression predicting how much those competitors are later forgotten. That finding reframes interference as partly adaptive: a system that suppresses competitors can retrieve the currently wanted answer faster, at the cost of the older answer.

Recall, Recognition, and Why the Test Format Changes the Result

Retroactive interference shows up most clearly on recall tests, where the learner must produce the answer from a cue alone. It shows up less clearly, and sometimes barely at all, on recognition tests, where the learner must pick the correct answer from presented options. This difference matters for interpreting both laboratory results and everyday experience.

On a recall test, the cue must select one answer from competing candidates. If the new answer is fresher or stronger, it is selected instead of the old one, and the old one looks forgotten. On a recognition test, the correct old answer is presented again, and that presentation can itself act as a strong cue that bypasses the competition. The learner may recognize the old password in a list even while failing to produce it from memory alone.

This has a practical consequence. Failing a recall test does not prove the old information is gone, and passing a recognition test does not prove it can be recalled when needed. A student who recognizes old terms on a multiple choice test may still struggle to produce them on an essay question after learning similar new terms. Testing both formats, or testing recall with a distinctive cue, gives a fairer picture of what is actually retained.

The same logic applies to relearning. An apparently forgotten item is often relearned faster than a genuinely new item, a pattern sometimes called savings. Savings suggests the old memory left a trace even when recall failed. That trace is evidence against a pure erasure account of retroactive interference, and in favor of competition or partial weakening. The article does not claim savings occurs in every case, but where it occurs, it is a useful check against the claim that new learning simply deletes old learning.

The Misinformation Effect: A Real World Relative

One of the most discussed real world relatives of retroactive interference is the misinformation effect in eyewitness memory. University teaching material describes it this way: after witnessing an event, a person may hear other people describe it from their own perspective, and that new information may interfere with or disrupt the person’s own recollection. In some cases the person may even come to remember the event happening exactly as others described it. The teaching material presents this as a type of retroactive interference occurring during the retention interval.

The connection is real but needs care. In the laboratory A-B, A-C paradigm, the interfering material is new learning of a similar list. In the misinformation effect, the interfering material is misleading information supplied after the event, often in questions or conversation. The direction is the same, new information acting backward on an older memory, but the mechanism is debated. Some researchers emphasize that the original memory is altered or overwritten. Others emphasize that the original memory remains but is outcompeted at retrieval by the newer, misleading version, or that the person reports the misleading version for social reasons rather than because memory itself changed. The article does not settle that debate. It notes the parallel and the disagreement, because a reader who meets both ideas should know they are related, not identical.

The practical lesson is narrower and safer. Memory for an event is more vulnerable to later information than everyday intuition suggests, especially when the later information is similar, plausible, and delivered confidently. That vulnerability is a reason to record important details promptly and to be cautious about details learned from other accounts after the fact. It is not a reason to distrust all memory, and it is not a diagnosis of any person’s memory.

Consolidation, Sleep, and the Retention Interval

The retention interval is not empty time. During it, a memory consolidates, meaning it becomes more stable in long term storage. University teaching material notes that memories can consolidate during the retention interval, aiding retention, while new experiences during the same interval can undermine the memory. Retroactive interference therefore competes with consolidation: the same window that stabilizes a memory also admits new material that can interfere with it.

Sleep is often discussed in this context, because sleep supports consolidation in general. That statement needs qualification. Sleep research supports a broad role for sleep in memory consolidation, but it does not show that a single night of sleep makes any particular memory immune to retroactive interference, and this article makes no such claim. What can be said safely is that a memory tested immediately after learning, before consolidation and before much interfering material has arrived, will often look stronger than the same memory tested later after a busy interval. The interval matters because of what fills it, not only because of how long it is.

This also explains why cramming similar topics back to back can be risky. A student who studies one set of similar terms and then immediately studies a second set has given the first set almost no protected consolidation time, and has filled the retention interval with the most interfering material possible: highly similar new learning. Spacing the two sets, or separating them with a clearly different topic, reduces that risk. It does not eliminate interference, and it does not guarantee retention, but it changes the conditions in the direction the classic research predicts.

For readers planning study time, a related guide on this site covers How to Create a Study Plan, which discusses spacing and sequencing in practical terms. The study plan guide and this article answer different questions: the guide organizes time, while this article explains one mechanism that makes sequencing matter.

Everyday Examples

Everyday examples illustrate the pattern but are not proof. A laboratory effect and an everyday slip can look similar for different reasons. The examples below are best read as illustrations to check against the definition, not as evidence on their own.

A new phone number is the most familiar example. A person who has used an old number for years may find, after memorizing a new number, that the old number no longer comes to mind cleanly. The new number answers the cue first. Months later, when the new number is itself familiar and no longer novel, the old number may return in some contexts. That return fits competition better than permanent erasure.

Passwords and codes follow the same logic. After a password change, a person may type the new password when an old account still requires the old one, or may blank on the old code for a rarely used system after learning a new code for a daily system. The error shows what the new learning was, and shows that the cue for the old task overlaps with the cue for the new task. It does not show that the old code was never learned.

Language learning gives a richer example. A student who learned Spanish vocabulary first and then studies Portuguese, which shares many similar words with different meanings, may find Spanish words harder to retrieve cleanly, or may produce Portuguese forms when aiming for Spanish. The teaching material notes the reverse direction too, with a previously learned language making a new similar language harder in some respects. Both directions occur. The retroactive case is the one where the newer language disturbs the older one.

Workplace routines and motor skills follow the same logic. An employee who learns a new software system with similar menu labels may struggle to recall the old system’s exact steps for a task that still runs on the old system. A musician who learns a new fingering for a passage may find the old fingering harder to reproduce cleanly on demand. A driver who learns a new route may find the old route’s turn sequence harder to recall. In each case the error is most likely when the old and new versions share cues and when the new version has been practiced more recently. These examples do not show permanent loss. The older version often returns when the person slows down, uses a distinctive cue, or checks the source.

Retroactive Interference in Study

Study is where retroactive interference causes the most practical trouble, because study deliberately fills the retention interval with new, similar material. A student revising for an exam may study one topic, then study a closely related topic, then be tested on the first topic and find the second topic’s terms intruding. The student has not necessarily studied badly. The sequencing itself created competition.

Several common study patterns invite retroactive interference. Studying several similar chapters in a row without a break or a change in format maximizes overlap. Rereading the older material immediately after the newer material can feel fluent, because recognition is easy, while hiding the recall problem that will appear on the test. Learning new terminology that reuses familiar words with new technical meanings creates a special risk: the familiar word cues the everyday meaning or the older technical meaning first.

The confusion may not appear while rereading, because rereading feels fluent. It may appear on a test, when the student must select the right term from similar candidates. A practice test forces that choice and reveals which items compete. That diagnostic value is one reason retrieval practice is widely recommended in educational research, though this article does not claim it eliminates retroactive interference. It reveals it, which is the first step to addressing it.

For practical guidance on organizing study sessions, see How to Create a Study Guide and How to Focus While Studying. The study guide article covers organizing material so related topics are distinguishable, and the focus article covers managing attention during study sessions. Both are practical companions to this explainer: they address how study is organized, while this article explains why similar material studied in sequence can interfere.

Strategies Studied for Protecting Older Learning

No single strategy removes retroactive interference in all settings, and this article makes no promise that any strategy will work for a particular person. The strategies below are drawn from laboratory and educational research, framed as general information only, not as treatments and not as advice for any health condition.

Test the old material before learning the new material. A short recall test on the older topic, taken before the newer topic is studied, both reveals what is actually retained and strengthens the older memory’s retrieval path. Testing after the newer topic is also useful, but testing before gives the older material a recent, successful retrieval that can help it compete later.

Make the old and new material distinctive. Add a distinguishing feature that separates the old item from the new item, present at encoding and available at retrieval. A student might pair each term with a unique example, a contrasting definition, or a different visual layout in notes. The principle is that distinction gives retrieval something to select on besides general familiarity. A comparison that lists only similarities strengthens overlap. A comparison that lists the defining difference for each item strengthens discrimination.

Separate highly similar topics when possible, and compare them deliberately when they must be studied together. Learning similar items back to back with the same cues maximizes competition. Spacing them, or learning one with different examples, can reduce overlap. When comparison is needed, focus on defining differences, not only on shared features. This is the study equivalent of the similarity finding: reduce shared cues, and competition falls.

Practice source monitoring. When an error occurs, ask which source the answer came from, rather than only marking it wrong. Did this answer come from the old chapter or the new chapter? Writing that source beside the error turns an intrusion into diagnostic information. Over time the learner can see which new items repeat as intruders and target those pairs for distinctive practice.

Use distinctive cues at retrieval. Try to recall the old item from its own cue, not from a cue shared with the new item. If the new item comes first, the cue is shared and needs a distinguishing feature or a second cue. A location, a date, a specific example, or a category label attached to the old item can serve as that second cue. Laboratory benefits depend on material, delay, test format, and prior knowledge, so check results with actual recall tests rather than assuming a method change has solved the problem.

Protect consolidation time where the schedule allows. A short break, a change to a clearly different topic, or a night of sleep between studying similar topics gives the older material a less crowded retention interval. This is not a guarantee, and busy schedules do not always allow it, but the classic research predicts that a crowded interval filled with similar material is the worst case for the older memory.

What Retroactive Interference Is Not

Retroactive interference is not decay. Decay proposes that memory traces weaken over time when not used. Retroactive interference proposes that a memory is blocked or displaced by a competing newer memory. Both can contribute to forgetting, but they are different explanations. The McGeoch and McDonald finding is the classic reason to keep them apart: when the time interval is held constant and only the similarity of the intervening material changes, recall still changes, which time alone cannot explain.

It is not encoding failure, in which information was never stored in usable form because of distraction or misunderstanding. It is not retrieval failure in general, which covers any case where stored information cannot be accessed because a cue is weak or mismatched. Retroactive interference is a specific retrieval problem in which the blocking item is an identifiable newer, similar memory.

It is not amnesia and it is not a diagnosis. Amnesia refers to a more severe, clinically defined loss of memory, often linked to injury or disease, and is outside this explainer. Retroactive interference in laboratory tasks occurs in healthy participants learning ordinary material, and everyday slips do not by themselves indicate a disorder. This article provides general information only. It does not diagnose, treat, or advise on any mental health condition. A person with persistent or worrying memory problems should seek advice from a qualified professional through appropriate channels.

It is also not the same as negative transfer in general, although the ideas overlap. Negative transfer is the broader term for prior learning hindering new learning or performance, and it is usually discussed in the forward direction. Retroactive interference is the memory focused version in the backward direction, centered on recall of earlier material after newer learning.

How to Spot It and Limits of the Evidence

To check for retroactive interference in study, identify the intruding item, check similarity, check order, and test the cue. Is the wrong answer a newer item? Are old and new items similar in meaning, sound, format, or context? Did errors rise after the newer material was learned? Does the old item return when you use a distinctive cue, or on a recognition test? Together these checks help choose a remedy. Weak encoding needs better initial learning. A mismatched cue needs a better cue. Competition needs distinctiveness and discrimination.

Laboratory research is strong on control and weaker on everyday realism. Classic participants learned word lists or paired associates in short sessions, often university students. Tasks were designed to isolate interference, not to mimic a semester of study or a workplace changeover. Everyday examples are therefore illustrations, not replications. A password slip may involve interference, habit, motor automaticity, and time pressure together. A second limit is generalization. A finding with word lists does not guarantee the same size of effect for meaningful material, longer delays, or different test formats. A third misreading is to treat new learning as harmful. New learning is usually valuable, and the goal is to manage competition, not to avoid learning new things that are still needed.

A final limit concerns measurement. Because retroactive interference is defined by comparison with a control group that did not learn the intervening material, everyday cases can rarely prove it conclusively. A student who forgets old terms after a new chapter may have experienced interference, decay, weak encoding, or all three. The definition gives a useful way to test the hypothesis, by checking for a newer intruder, similarity, and recovery with a distinctive cue, but it does not license certainty about any single slip.

Frequently Asked Questions

What is retroactive interference in simple terms?

Retroactive interference is when new learning makes it harder to remember older, similar information. The new memory acts backward in time and competes with the old memory at retrieval. Learning a new phone number and then struggling to recall the old number is a simple illustration.

What is the difference between retroactive and proactive interference?

Retroactive interference is new material interfering with old material. Proactive interference is old material interfering with new material. Ask which material is being tested. If the test is on old material and new material interferes, it is retroactive interference. If the test is on new material and old material interferes, it is proactive interference. The companion article on Proactive Interference Psychology Definition covers the forward direction in full.

What are some examples of retroactive interference?

Common examples include forgetting an old phone number after memorizing a new one, blanking on an old password after learning a new one, producing words from a newly learned similar language when aiming for an older one, and struggling to recall the steps of an old procedure after learning a new similar procedure. In each case the newer, similar item intrudes when the older item is wanted.

Why does retroactive interference happen?

It happens because memories that share meaning, sound, structure, or context compete for the same retrieval path. Newer memories are often more recently rehearsed and more available, so they may be retrieved first even when wrong for the current task. Similarity increases competition, and a shared cue makes it harder to select the correct source. Researchers still debate whether the old memory is mainly blocked at retrieval, partly unlearned, or both.

Does retroactive interference mean memory loss?

Not necessarily. It describes competition at retrieval, not necessarily permanent loss. The older information can often be recalled with a distinctive cue, on a recognition test, or after relearning faster than a new item, which suggests a trace remains. Persistent, severe, or worrying memory problems should be discussed with a qualified professional, not diagnosed from an explainer article.

How can retroactive interference be reduced?

Research points to general approaches, with no guarantee for any individual. Test old material before studying new similar material, make old and new items distinctive, separate highly similar topics when possible, compare them on defining differences when they must be studied together, practice source monitoring, and use distinctive retrieval cues. These approaches aim to reduce competition and improve source discrimination, not to erase the new memory.

Is retroactive interference the same as forgetting?

No. Forgetting is a broad outcome with many causes, including decay, encoding failure, retrieval failure, and interference. Retroactive interference is one specific cause within interference theory. It applies when an identifiable newer, similar memory interferes with recall of earlier material. Other forgetting may have different causes and may need different remedies.

What is retroactive inhibition?

Retroactive inhibition is an older term for the same phenomenon, used widely in early twentieth century research. McGeoch and McDonald in 1931 wrote of retroactive inhibition, while later literature generally uses retroactive interference. The terms refer to the same backward acting effect of new learning on older recall.

Sources & references

• Andrade, Mehgan, and Neil Walker. Cognitive Psychology, section 5.4, Decay vs. Interference. Social Sci LibreTexts, College of the Canyons. https://socialsci.libretexts.org/Bookshelves/Psychology/Cognitive_Psychology/Cognitive_Psychology_(Andrade_and_Walker)/05:_Working_Memory/5.04:_Decay_vs._Interference
• McGeoch, John A., and William T. McDonald. Meaningful relation and retroactive inhibition. American Journal of Psychology, 1931, volume 43, issue 4, pages 579 to 588. https://doi.org/10.2307/1415159
• McLeod, Saul. Proactive and Retroactive Interference: Definition and Examples. Simply Psychology. https://www.simplypsychology.org/proactive-and-retroactive-interference.html
• Underwood, Benton J., and Leo Postman. Extraexperimental sources of interference in forgetting. Psychological Review, 1960, volume 67, issue 2, pages 73 to 95. https://doi.org/10.1037/h0041865
• Underwood, Benton J. Interference and forgetting. Psychological Review, 1957, volume 64, issue 1, pages 49 to 60. https://doi.org/10.1037/h0044616
• Chandler, Carole C. Specific retroactive interference in modified recognition tests: Evidence for an unknown cause of interference. Journal of Experimental Psychology: Learning, Memory, and Cognition, 1989, volume 15, issue 2, pages 256 to 265. https://doi.org/10.1037/0278-7393.15.2.256
• Murayama, Kou, Miyatsu, Toshiya, Buchli, Dorothy, and Storm, Benjamin C. Forgetting as a consequence of retrieval: A meta-analytic review of retrieval-induced forgetting. Psychological Bulletin, 2014, volume 140, issue 5, pages 1383 to 1409. https://doi.org/10.1037/a0037505
• Wimber, Maria, Alink, Arjen, Charest, Ian, Kriegeskorte, Nikolaus, and Anderson, Michael C. Retrieval induces adaptive forgetting of competing memories via cortical pattern suppression. Nature Neuroscience, 2015, volume 18, issue 4, pages 582 to 589. https://doi.org/10.1038/nn.3973