Retrieval-enhanced suggestibility (RES) is a memory effect in which taking an initial test about a witnessed event can increase a person’s later susceptibility to misleading information presented after that test. Chan, Thomas and Bulevich reported the reversed testing effect in 2009. It is conditional, not a reason to abandon learning quizzes: retrieval can still strengthen accurate original details while also affecting learning of certain later false details.
Did you know that the order of learning activities can change how later information is remembered? In the RES paradigm, a person first witnesses an event, retrieves it during an interim test, then encounters a misleading narrative and finally answers a new memory test. Vocabulary becomes a technology of accuracy when we track what was originally observed, what was recalled and what arrived from a later source.
This guide explains the 2009 experiments, Chan and Langley’s 2011 finding that accuracy benefits and suggestibility costs can coexist, Gordon and Thomas’s 2014 tests of learning mechanisms, and later research on misinformation formats, retention intervals and warnings. It includes original examples for Primary and Secondary English, Mathematics, Science, History, General Paper, translation and professional source records.
Editorial note: No classroom scenario in this guide is a real incident, a measured retrieval-enhanced suggestibility experiment or a protocol for influencing witnesses. The research is explained with conditional results and practical source-awareness safeguards.
Contents: why an early memory test can change the effect of later information
Open 54 chapters and research checks
- 1. Testing can sometimes have an unexpected side effect
- 2. The order of information is essential
- 3. Chan, Thomas and Bulevich’s 2009 study
- 4. Retrieval practice usually helps retention
- 5. The initial event supplies the ground truth
- 6. The interim test reactivates event details
- 7. Misinformation arrives after the initial test
- 8. The final test measures later memory
- 9. Why the result initially seemed paradoxical
- 10. The post-event narrative has a separate status
- 11. A fictitious classroom example
- 12. Memory tests can focus attention on certain features
- 13. Attention to misinformation can increase learning
- 14. Reconsolidation was one proposed mechanism
- 15. A dual-mechanism account was investigated
- 16. The 2011 study found benefits and costs together
- 17. A correct detail and a misleading detail are not equivalent
- 18. The first test may provide practice with concepts
- 19. The 2014 Gordon–Thomas experiments
- 20. Original information remained accessible in one test
- 21. The modified–modified free-recall test has a purpose
- 22. Testing effects can depend on what gets retested
- 23. Repeated interim tests have been investigated
- 24. The delay is a major boundary condition
- 25. A 2017 study compared short and longer retention
- 26. Format of misinformation is another moderator
- 27. A question can introduce an assumption
- 28. A narrative can feel coherent even when wrong
- 29. A 2020 study observed protective testing conditions
- 30. Initial-test errors can persist independently
- 31. Retrieval-enhanced suggestibility is not retrieval-induced forgetting
- 32. It is not the general testing effect
- 33. It differs from post-event misinformation alone
- 34. It differs from continued influence after correction
- 35. Source monitoring contributes to accuracy
- 36. Warnings have been studied experimentally
- 37. Warning timing matters in later research
- 38. The 2025 warning study added a test
- 39. A warning cannot change original ground truth
- 40. A quiz should not be blamed for all errors
- 41. High-stakes interviews need professional methods
- 42. Testing before accurate feedback is a different situation
- 43. A school vocabulary case separates materials
- 44. Secondary English can compare event and report
- 45. Science requires clear observation records
- 46. Mathematics distinguishes measured and inferred values
- 47. History requires dated source chains
- 48. General Paper should separate claims and later corrections
- 49. Translation must preserve temporal source labels
- 50. Professional documents need an audit trail
- 51. A five-stage experimental timeline
- 52. A two-outcome result table is essential
- 53. An eight-stage classroom evidence audit
- 54. The final lesson is retrieval with trustworthy context
Quick Test: RES requires a particular sequence
| Stage | What the study manipulates or measures | Common error to avoid |
|---|---|---|
| Original event | Controlled film, image or narrative | Treating later wording as original observation |
| Interim retrieval | Test versus no-test comparison | Calling all quizzes harmful |
| Post-event material | Accurate, misleading or control details | Assuming source information is direct perception |
| Final memory test | Recall or recognition scored against original event | Assuming a wrong answer proves the original trace was erased |
| Separate outcomes | Original detail accuracy and misinformation endorsement | Reporting only the cost while hiding retrieval benefits |
The timeline is the scientific core of RES. Without testing before misleading post-event information and a suitable comparison, the concept cannot be inferred from one wrong answer.
1. Testing can sometimes have an unexpected side effect
Retrieval-enhanced suggestibility, abbreviated RES, occurs when an initial memory test makes people more susceptible to misleading information presented later about the same event, under certain experimental conditions.
This is surprising because retrieval practice often improves later memory for information that was actually studied. RES does not overturn the educational value of testing. It identifies a particular situation: a person retrieves an original event, then encounters misleading post-event material and later reports memory that may incorporate that material.
2. The order of information is essential
The sequence in the classic RES paradigm begins with a witnessed event, continues with an initial recall test, then presents post-event information that can contain misinformation, and ends with a final memory test.
Changing that order changes the research question. Misinformation introduced before the first test is not the same manipulation. A pupil explaining RES should name the four stages rather than simply say that questioning people causes false memories. The effect depends on retrieval occurring before a later misleading source.
3. Chan, Thomas and Bulevich’s 2009 study
Jason C. K. Chan, Ayanna K. Thomas and John Bulevich published Recalling a Witnessed Event Increases Eyewitness Suggestibility: The Reversed Testing Effect in Psychological Science in 2009.
Their research found that an immediate cued-recall test could increase later misinformation susceptibility among younger and older adult samples under the studied procedure. The result was termed a reversed testing effect because retrieval usually benefits memory. It did not prove that all eyewitness interviews necessarily make recall less accurate.
4. Retrieval practice usually helps retention
A well-supported finding in memory research is that attempting to recall learned material can improve later retention compared with only restudying it, under many conditions. This is often called the testing effect or retrieval-practice effect.
RES raises a narrower issue: after retrieval, exposure to a later source containing misleading details may change final reports. The ordinary benefits of retrieval for correct original details can coexist with susceptibility to the later misinformation. A responsible explanation must retain both outcomes.
5. The initial event supplies the ground truth
In a controlled laboratory paradigm, participants may watch a film or other stimulus with details known to the researchers. The original material provides a reference for judging whether later reported details were actually present.
Without that ground truth, a researcher cannot easily distinguish a misinformation-induced answer from an accurate memory of the event. A classroom activity should similarly retain the original fictional text or image, but should not claim to replicate the experimental effect without suitable controls.
6. The interim test reactivates event details
After the original experience, some participants answer questions about particular features. That is the retrieval manipulation. A comparison group may undertake a filler task rather than recall the event at that point.
The later comparison concerns whether this initial testing changes the influence of subsequently encountered material. It is not simply a comparison of people with good and bad memory. Random assignment and consistent materials help isolate what the interim test contributes.
7. Misinformation arrives after the initial test
Participants then encounter a post-event narrative or another source describing the original event. Some details match what happened, while selected critical details may be incorrect.
The misleading details are not part of the original event. They are later communication. This distinction is essential to the meaning of RES. A pupil must identify which source supplied each item, not treat the original film and subsequent narrative as interchangeable evidence.
8. The final test measures later memory
At the end of the sequence, participants answer questions about what they originally witnessed. If they report a later false detail, the researcher can compare that response with the original event and the experimental conditions.
One incorrect response is not sufficient to establish the size of RES. A proper study compares tested and untested groups, misleading and control details, and appropriate outcomes. The reported effect is a systematic contrast, not a diagnosis of a particular witness.
9. Why the result initially seemed paradoxical
If retrieval practice strengthens an original memory, researchers might expect the tested group to resist misinformation better than those who did not retrieve the event. The 2009 findings contradicted this simple prediction under the studied conditions.
This does not imply that memory testing itself is intrinsically harmful. Instead, it suggests retrieval can alter how subsequent information is processed. A scientific explanation must identify both the benefit to original recall and the possible cost when misleading material follows.
10. The post-event narrative has a separate status
A later summary can include accurate details, omissions, interpretations and fabricated or erroneous features. Even when it discusses the same incident, it is not a direct substitute for the witnessed event.
The educational skill is source separation. A student can remember the story in the later narrative without having directly observed every detail it contains. Vocabulary such as witnessed, reported and suggested distinguishes those information channels.
11. A fictitious classroom example
Imagine an original school illustration showing a blue notebook on a desk. Students recall what they saw. Later they read a clearly labelled fictional paragraph that mistakenly calls the notebook green.
A final question asks what colour appeared in the original picture. The activity can demonstrate which source contained the inaccurate colour, but it is not evidence of RES unless the design includes suitable test and no-test comparisons. The aim is source awareness, not inducing real autobiographical errors.
12. Memory tests can focus attention on certain features
An interim question asking about a notebook may make the notebook a particularly relevant object in the participant’s representation. When a later narrative refers to that object, readers may give the description more attention.
This is one possible mechanism for enhanced learning of post-event misinformation. The learner is not necessarily trying to absorb falsehood. The initial retrieval task can change which details are salient in subsequent reading, so later information about those details may be encoded strongly.
13. Attention to misinformation can increase learning
The RES literature investigates whether testing potentiates new learning, including information that happens to be wrong. If the initial test makes a detail especially relevant, the person may process its later description more deeply.
This principle is not a condemnation of all new learning after a quiz. Accurate feedback and explanations are valuable. The specific risk arises when the later source is misleading. A responsible educational setting should provide correct, clearly identified feedback rather than unsupported suggestions.
14. Reconsolidation was one proposed mechanism
The 2009 authors discussed the possibility that retrieval can make an episodic memory more susceptible to later interference, a mechanism sometimes framed in terms of reconsolidation.
That interpretation is a hypothesis, not proof that every memory becomes biologically unstable whenever recalled. Later experiments found RES under conditions not requiring a reconsolidation-blocking account. A public explanation should present reconsolidation cautiously and avoid turning it into an automatic memory-editing rule.
15. A dual-mechanism account was investigated
A dual-mechanism account considered whether the initial test both strengthens later learning of misinformation and creates susceptibility through interference with the original memory.
Chan and Langley’s 2011 paper tested this account across additional conditions. The study found that enhanced learning of misinformation remained an important explanation and that blocking reconsolidation was not necessary for RES in their experiments. That narrows claims about one specific biological mechanism.
16. The 2011 study found benefits and costs together
Chan and Moses Langley published Paradoxical Effects of Testing: Retrieval Enhances Both Accurate Recall and Suggestibility in Eyewitnesses in 2011.
They found RES after a one-week delay while also observing a robust testing benefit for original event details that were not contradicted by later misinformation. This is central to the scientific meaning of the phenomenon: retrieval can support accurate memory and enhance later misinformation learning under different item conditions. One overall label of good or bad is inadequate.
17. A correct detail and a misleading detail are not equivalent
A post-event narrative may accurately repeat some features and alter others. A tested participant might later remember genuine features more accurately while also reporting certain false narrative details more often.
Therefore, a research table should separate control details, originally witnessed details and contradicted features. An average accuracy number may conceal opposing effects. A strong educational account explicitly says which memory component improved and which became more vulnerable.
18. The first test may provide practice with concepts
Questions about a particular object can make related nouns and descriptions more accessible when the later narrative mentions them. The person may recognise the topic more readily, regardless of whether the later sentence is accurate.
That mechanism concerns learning opportunities, not a magical transformation of truth. A later source that supplies valid correction can also benefit from attention. In education, testing should be coupled with reliable content, clear provenance and feedback that does not introduce unintended inaccuracies.
19. The 2014 Gordon–Thomas experiments
Leamarie Gordon and Ayanna Thomas published Testing Potentiates New Learning in the Misinformation Paradigm in Memory & Cognition in 2014.
Their experiments compared conditions with identical interim and final questions, related but different questions and a standard misinformation procedure. They examined whether retrieval of exact critical details was needed, whether attention allocation changed and whether original information became inaccessible. The study refined the explanation beyond a generic claim that testing erased memory.
20. Original information remained accessible in one test
Gordon and Thomas reported that RES did not coincide with reduced accessibility of original information in their modified–modified free-recall measure. This finding weighs against interpreting every false response as proof that the original memory was destroyed.
Instead, the later misleading detail may have been learned strongly and compete at the final test. Source confusion or response demands may also matter. A responsible guide distinguishes observed test behaviour from the content of an unobservable memory trace.
21. The modified–modified free-recall test has a purpose
This experimental method is used to probe whether information that was not produced in an initial answer can still be recovered under a revised retrieval instruction. It can help distinguish inaccessible content from effects of the normal response rule.
A learner need not memorise the technical method to understand its contribution. The research question was whether original details were still available even when the participant selected misinformation in another test. The outcome challenges an overly simple memory-overwriting narrative.
22. Testing effects can depend on what gets retested
An interim test may ask exactly the detail later targeted by misinformation, or it may ask a related feature. These designs can produce different salience and response conditions.
A researcher should not assume every question has the same effect on all later information. The 2014 study explicitly manipulated the relationship between interim and final tests. For teaching, it means feedback should match the item and correct the specific misconception rather than merely revisit a broad topic.
23. Repeated interim tests have been investigated
Chan and LaPaglia published research in 2011 examining how repeated retrieval attempts influenced later suggestibility to misleading post-event information. They manipulated the number of initial tests and delays.
Their experiments found that more retrieval attempts could increase misinformation susceptibility in some conditions, but the effect was qualified by factors such as retention interval and whether testing was manipulated within or between participants. One should not generalise that every additional quiz always makes memory less accurate.
24. The delay is a major boundary condition
The interval between the initial test, later misleading material and final memory assessment changes what information remains accessible. Retrieval’s long-term benefits can emerge more clearly after some delays.
A study showing higher suggestibility immediately may differ from one testing after several days. The research needs to report each interval. A school article should not announce a single fixed time after which memory becomes vulnerable. Timing interacts with method and source quality.
25. A 2017 study compared short and longer retention
Research reported in 2017 found that the effect of testing on misinformation susceptibility could increase or decrease depending on the retention interval. Under its studied conditions, an immediate design showed more susceptibility while a delayed design showed protective effects of initial testing.
This directly challenges any universal rule that testing always worsens eyewitness memory. The phenomenon is real under some conditions, but its direction depends on how the experiment is organised. Good teaching acknowledges the contradiction and its methodological explanation.
26. Format of misinformation is another moderator
LaPaglia and Chan’s 2013 research compared misinformation embedded in narratives with misleading information presented through questions. Their study found that interim testing could increase susceptibility for narrative-based misinformation while reducing it for question-based misinformation under the tested conditions.
This is important because a paragraph describing an event and a question about the event have different processing demands. A responsible summary cannot treat all later information formats as equivalent. The research tells us to examine how misinformation was delivered.
27. A question can introduce an assumption
A post-event question such as ‘What colour was the missing bag?’ can presuppose that a bag existed. A paragraph can instead assert that the person carried a particular bag.
Both can introduce misleading details, but the recipient processes them through different tasks. The educational habit is to identify the actual information source and whether the detail was in the original event. This helps preserve evidence without deliberately designing suggestive questioning.
28. A narrative can feel coherent even when wrong
A later account may organise the event into a smooth story, including plausible causes or objects. A reader who completed an interim test may find the described details especially relevant and absorb them more readily.
The narrative’s coherence does not establish its accuracy. Students should compare the later account with the original record where available. A vivid explanation can be false even when its parts connect naturally. Vocabulary should distinguish stated cause, observed cause and suggested cause.
29. A 2020 study observed protective testing conditions
Pereverseff, Bodner and Huff studied household-scene materials and later suggested objects in a 2020 publication. They found a protective effect of testing on several final measures under their design while also noting that initial-test intrusions could persist.
The result is important precisely because it complicates the simple RES story. Initial retrieval can sometimes aid discrimination of original details. When students learn the phenomenon, they should know that protective effects have also been reported and that the format and delay are central.
30. Initial-test errors can persist independently
If a participant produces an unpresented object during the first test, that intrusion may later remain available. Such an effect differs from adopting a false detail first encountered in the post-event narrative.
Researchers should classify where the misinformation originated. A person might generate an error themselves before reading the misleading account or learn it from the later source. A final wrong response alone cannot reveal the full path. Source tracking is essential.
31. Retrieval-enhanced suggestibility is not retrieval-induced forgetting
Retrieval-induced forgetting concerns difficulty recalling some related unpractised material after selectively retrieving other items under particular experimental conditions. RES concerns increased susceptibility to post-event misinformation after initial testing.
The two can interact with retrieval processes, but they have different designs and outcome measures. A student should not use forgetting as a catch-all for any false memory response. Some RES experiments find original information remains accessible despite later errors.
32. It is not the general testing effect
The ordinary testing effect describes improved later retention due to retrieval practice. RES concerns a potential cost when misleading content is encountered after that retrieval.
The same initial test can produce useful benefits for correct original details and a risk for certain later false details. These are not logical contradictions because they concern different items and measures. A useful explanation states each outcome precisely.
33. It differs from post-event misinformation alone
A standard post-event misinformation experiment can show distortion after misleading material without any interim retrieval test. RES adds the critical manipulation of testing before misinformation exposure.
The research question is whether that earlier testing changes susceptibility relative to a no-test control. Without that comparison, a later error demonstrates possible misinformation influence, not necessarily retrieval enhancement. The term should be reserved for the appropriate design.
34. It differs from continued influence after correction
Continued influence concerns false information shaping reasoning after a correction or retraction. RES does not require a retraction at all; the later narrative can introduce misinformation before the final memory test.
The practical concern in both is accurate source use, but the experimental timelines differ. A teacher should identify whether the task studies initial retrieval, later suggestion or corrected causal reasoning. Clear vocabulary helps prevent scientific concepts from merging.
35. Source monitoring contributes to accuracy
A reader should distinguish what appeared in the original event from what was encountered in the later narrative. Such provenance can help prevent treating a post-event statement as direct observation.
But source monitoring alone is not necessarily sufficient under every RES condition, and an individual may remember both versions while choosing the wrong one. The Source-Monitoring Framework explains a relevant aspect of attribution. The actual study must test how source instructions affect performance.
36. Warnings have been studied experimentally
Thomas, Bulevich and Chan’s 2010 research found that warning participants about inaccuracies in the post-event narrative could promote eyewitness accuracy and reduce the RES pattern in its experimental conditions.
That does not mean a vague warning will work regardless of timing or wording. The warning must be relevant and processed. Researchers have subsequently investigated when it appears and whether its benefit remains after a delay.
37. Warning timing matters in later research
Chan, O’Donnell and Manley reported in 2022 that warnings issued soon after misinformation exposure reduced RES under the tested procedures, including when the final test was delayed. Warnings delivered forty-eight hours after misinformation were less effective.
This supports attention to the order and timing of credible corrections. It does not provide a universal forty-eight-hour cutoff for every real incident. A public message should clearly identify what may be inaccurate as soon as responsible evidence permits, while acknowledging uncertainty.
38. The 2025 warning study added a test
Later research published in 2025 investigated warnings delivered before final retrieval and how their timing affected memory errors in repeated-testing misinformation procedures. It reported effective warnings near exposure in some conditions, with reduced benefit when delayed.
The study adds to evidence that warning design matters, but it should be read with its experimental interval and sample. It does not show that all witnesses can be repaired by one script or that every late correction is useless. Outcomes remain conditional.
39. A warning cannot change original ground truth
If the witnessed film contained a blue notebook, a later warning about possible inaccuracies does not alter the film. It changes the information participants have when making later memory decisions.
This is a basic but important distinction. Psychological interventions can influence report accuracy without changing past events. Vocabulary words such as originally observed, later suggested and retracted preserve the timeline needed for responsible evidence discussion.
40. A quiz should not be blamed for all errors
A student’s incorrect answer after reading two sources could arise from an ambiguous question, poor baseline learning, distraction or a mistaken inference. RES is a group-level experimental effect requiring an appropriate sequence and control comparison.
An educator should examine the actual learning materials and feedback rather than assign a cognitive mechanism from one wrong answer. Good teaching asks whether the pupil understood the original content and whether later information was authoritative. Those are actionable questions.
41. High-stakes interviews need professional methods
Real eyewitness interviews involve legal, ethical and procedural constraints. The RES literature can inform research and caution about later misinformation, but classroom activities should not be presented as forensic protocols.
Investigators rely on validated evidence-gathering practices, records and independent corroboration. A witness’s memory should not be manipulated to demonstrate an effect. The safest public lesson is that questions and subsequent accounts must be handled carefully and accurately.
42. Testing before accurate feedback is a different situation
In school learning, an initial retrieval quiz is commonly followed by corrective feedback from a teacher or trusted resource. The later information is intended to be accurate rather than misleading.
The RES findings do not justify abandoning quizzes in that context. Retrieval practice remains a useful learning tool when designed appropriately. Accurate feedback, explanation and transfer tasks help pupils strengthen correct knowledge. The misinformation paradigm addresses a narrower risk.
43. A school vocabulary case separates materials
A Primary pupil sees a picture of a blue book and later answers a question about its colour. The tutor then presents a clearly labelled fictional account mistakenly calling the book green and asks the child to identify the discrepancy.
The exercise supports source awareness but should not aim to make the child falsely remember the image. It is not an RES experiment without a no-test control and measured final outcomes. The educational goal is to compare the original picture with later language accurately.
44. Secondary English can compare event and report
A Secondary class reads a fictional eyewitness account and a later newspaper-style summary containing a disputed detail. The student records which source introduced that detail and whether it was established in the first account.
Then the learner writes a source-attributed summary, using reported and observed correctly. This helps prevent blending separate narratives. The activity does not prove retrieval enhanced suggestibility but develops the underlying evidence vocabulary.
45. Science requires clear observation records
A Science pupil may perform a measurement, write an initial note and later encounter a classmate’s contradictory interpretation. The original instrument record should be preserved separately from the later explanation.
A retrieval quiz can help learn the mechanism, but the measured result must be verified from the actual observation. Do not turn an uncertain explanation into data. RES research motivates careful handling of post-event accounts, not a recommendation against reviewing experiments.
46. Mathematics distinguishes measured and inferred values
An event description may contain a measured distance and a later estimated distance. Repeatedly answering questions about the event can make both numbers memorable, but only one may belong to the original instrument record.
Students should label the measurement source, unit and uncertainty and use the correct value in calculations. A remembered number is not automatically an observed one. This is an evidence-handling analogy, not a direct mathematics replication of RES.
47. History requires dated source chains
A historical event may have a contemporary document and later retellings that add dramatic details. A student who repeatedly practises the first account can still encounter and potentially remember a false later addition.
Use a document timeline and attribute each detail to the correct source. Do not invent independent eyewitness corroboration simply because several books repeat the same claim. Historical truth requires original evidence and careful interpretation, irrespective of psychological test effects.
48. General Paper should separate claims and later corrections
An argumentative text may cite an early statistic, while a subsequent report revises it. Students should know which value was available at each stage and which is currently supported.
A strong GP essay uses the updated reliable source and preserves the initial claim only as historical context where relevant. The research on RES offers a caution about learning later material, but it cannot decide the actual statistic’s validity. Source methods and data must settle that question.
49. Translation must preserve temporal source labels
A multilingual incident report may distinguish an original observation from a later witness statement. If a translation removes subsequently alleged or replaces it with was seen, the target text changes evidential status.
The translator should preserve who supplied each detail and when. A fluent translation is not faithful if it converts suggestion into direct perception. Accurate vocabulary and source chronology are necessary for reliable communication.
50. Professional documents need an audit trail
An operational incident record may contain an initial event description, a staff debrief and an updated explanatory note. The organisation should keep those stages traceable rather than silently rewriting the original observations.
The purpose is transparency, not to induce a particular memory outcome. A well-maintained record distinguishes measured data from later interpretations. RES research underscores why information encountered after initial recall must be clearly sourced, especially when accuracy matters.
51. A five-stage experimental timeline
Construct a conceptual table with original event, initial retrieval test or control, later narrative containing verified or misleading details, final test and scoring against original ground truth.
The table makes it easy to see whether a reported experiment actually manipulated interim testing before misinformation. Do not add invented effect sizes. A scientific claim should name the conditions, participant sample, delays and outcome before explaining why the result was surprising.
52. A two-outcome result table is essential
For each condition, researchers can separately consider correct recall of original noncontradicted details and erroneous endorsement of later misinformation. Reporting only one of these outcomes can hide the paradox that testing sometimes helps one while harming the other.
A pupil should learn that memory is not one single score. A study can show both benefits and costs under different items. Precise vocabulary keeps the claims transparent and prevents sensational headlines from replacing actual results.
53. An eight-stage classroom evidence audit
Record what was originally shown, what questions were asked first, what later account was presented and which details were new. Identify the actual source of each disputed item.
Next check whether the final interpretation preserves the original evidence, determine if any warning was supplied and use authoritative records rather than subjective familiarity for consequential decisions. This is a source-literacy routine, not an attempt to induce false memories in pupils.
54. The final lesson is retrieval with trustworthy context
Retrieval-enhanced suggestibility shows that initial testing can alter how later misleading information affects memory, sometimes increasing errors even while strengthening correct original details. The literature also documents protective effects and important moderators involving delay, format and warnings.
For vocabulary education, the lesson is not to stop using retrieval practice. It is to pair useful testing with accurate feedback and clear source boundaries, so that later language cannot be mistaken for an original observation.
Full worked case: two descriptions of one fictional notebook
A fictional classroom image shows a blue notebook on a table. A pupil is later asked to recall the object’s colour. A subsequent story text mistakenly says green, and a final question asks what appeared in the original image. The class can compare the event, interim question and later description to identify exactly where the wrong colour entered.
This is a source-audit illustration rather than a replication of RES. A real experiment would need test and no-test groups, controls, timing and formal outcome measures. For school learning, the relevant skill is distinguishing an original observed detail from a later unverified statement.
Eight-stage source-aware retrieval workbook
- Preserve the exact original event or source before any later discussion.
- Mark whether an initial recall test occurred and which details it asked about.
- Identify all post-event descriptions and their separate source origins.
- Underline any details that contradict the controlled original record.
- Distinguish the later narrative’s wording from actual observation.
- Check what the final test measures and how responses are scored.
- Record both original-detail retention and misinformation errors separately.
- Use correct feedback and clear provenance rather than unsupported suggestions in actual teaching.
Frequently asked questions
1. What does RES mean?
Retrieval-enhanced suggestibility, where interim testing can increase susceptibility to later misleading information under certain conditions.
2. Who reported the influential 2009 effect?
Chan, Thomas and Bulevich in Psychological Science.
3. Is this the same as the ordinary testing effect?
No. The usual testing effect benefits retention; RES concerns later misinformation interference.
4. Does an early test always increase error?
No. Later studies have reported protective effects and moderators involving delay and information format.
5. Can retrieval help correct memory and increase later errors?
Yes. Some experiments found both outcomes for different details or measures.
6. Does RES mean the original memory was destroyed?
No. Some experiments found original information remained accessible despite later misinformation reports.
7. Do warnings matter?
Warnings sometimes reduced RES, and later work found timing of warnings important under the tested conditions.
8. How is RES different from post-event misinformation alone?
RES requires an earlier retrieval-test manipulation before the misleading post-event information.
9. Should schools stop giving retrieval quizzes?
No. Retrieval practice remains useful, especially with correct feedback and appropriate materials.
10. What is the practical lesson?
Keep original event evidence and later source claims separate when interpreting memory.
Primary experiments, qualifications and connected vocabulary reading
- Chan, Thomas and Bulevich (2009): Recalling a Witnessed Event Increases Eyewitness Suggestibility — original reversed testing effect.
- Chan and Langley (2011): Paradoxical Effects of Testing — simultaneous accurate recall benefits and suggestibility costs.
- Gordon and Thomas (2014): Testing Potentiates New Learning in the Misinformation Paradigm — attention and preserved original accessibility.
- LaPaglia and Chan (2013): Testing Increases Suggestibility for Narrative-Based Misinformation but Reduces It for Question-Based Misinformation — format moderation.
- Chan and colleagues (2011): The Dark Side of Testing Memory — multiple tests and retention intervals.
- The Effect of Testing Can Increase or Decrease Misinformation Susceptibility Depending on the Retention Interval (2017) — temporal boundary condition.
- Thomas, Bulevich and Chan (2010): Testing Promotes Eyewitness Accuracy With a Warning — warning conditions.
- Chan, O’Donnell and Manley (2022): Warning Weakens Retrieval-Enhanced Suggestibility Only When It Is Given Shortly After Misinformation — warning timing.
- Timing of Pre-Retrieval Warnings Matters in Reducing Memory Errors (2025) — later warning experiments.
- Pereverseff, Bodner and Huff (2020): Protective Effects of Testing Across Misinformation Formats — conditions with retrieval protection.
- Vocabulary | What Is the Post-Event Misinformation Effect? — baseline paradigm.
- Vocabulary | What Is the Source-Monitoring Framework? — distinguishing original and later source memories.
- Vocabulary Learning Hub — memory research and accurate language.
Retrieval-enhanced suggestibility shows why learning outcomes depend on the full sequence: original event, retrieval, later information and final test. Some studies find stronger misinformation effects after interim testing, while others reveal protection under different timing and formats. The useful educational response is accurate source tracking and trustworthy feedback, not abandoning retrieval practice.