Hall II · Psychological Phenomena, Tasks & Thought Experiments
Deese-Roediger-McDermott false memory list
Try it
Thread: Memory & learning
Thread: Reasoning & bias (secondary)
content note: deception
Station · Try it
A short memory demonstration. You will read twelve words one at a time, spend twenty seconds on a small counting task, and then say which of nine words you saw. Part of what this station is doing is deliberately not explained yet. Starting it registers an anonymous "debrief owed" marker in this browser, and the page will settle that debt — a full explanation of what really happened — before you leave the demonstration. Nothing you read, click or answer here is stored or sent anywhere.
The interactive station needs JavaScript, which is off — nothing is hidden from you because of that. The full explanation below covers everything the station demonstrates.
A debrief is owed the moment you start
A short memory demonstration: you will read a list of words, do twenty seconds of something else, then be asked which words you saw. Part of what this station does is deliberately not explained up front. Starting it registers one anonymous “debrief owed” marker in this browser — no answer or timing is stored — and the station settles that debt before the demonstration ends.
Honest limitation: if you abandon this page permanently and clear your browser storage, no website can force a debrief afterwards. The debt survives completion, refresh, back navigation and same-browser return.
The experience
The station above shows you twelve words, one at a time, sends you off to count
backwards for twenty seconds, then offers nine words and asks which you saw. It says in
advance that part of what it does is not yet explained, and settles that debt before you
leave. Run it first — the next paragraph gives the game away.
In the experiment this station restages, thirty-six
students heard six lists of twelve words, each list built entirely from the leading
associates of one word that was never read out, and then wrote down everything they
could remember. The missing word came back on 40 per cent of the lists — about as often
as words that genuinely had been read to them from the middle of the list.Convergent / replicated Nothing
was wrong with those students' memories. The lists were simply very good.
What is happening
The trick is older than the paper that made it famous.
In 1959 James Deese built thirty-six lists, each one the twelve primary associates of a
word he never read aloud, and found that the chance of a particular word intruding into
recall was predicted by how strongly the list pointed at it — a correlation of .87
between intrusion probability and associative strength. His report then sat almost
unread: Roediger and McDermott counted fourteen citations in total, and only one since
1983, before they picked it up thirty-six years later.Historical / descriptive
A list like this is a set of arrows aimed at a word that is not there. Reading table,
sit, legs and sofa stirs the neighbourhood those words live in, and its hub gets
stirred along with them — but nothing in the trace is stamped with where it came from.
Asked later whether you saw the hub, you find something to consult, and it feels like the
rest.
On a recognition test the illusion is
sharper than in recall. In Roediger and McDermott's first experiment the false-alarm
rate for the missing word was .84, against a hit rate of .86 for words that had actually
been studied, while unrelated words drew false alarms only 2 per cent of the time. Asked
to rate their confidence on a four-point scale, subjects picked the top point — sure it
was old — for the missing word 58 per cent of the time.Convergent / replicated
What this shows
What lifts this above a party trick is the quality of the experience, not the frequency
of the error. In a second experiment with
twenty-four fifteen-word lists, false recall rose to 55 per cent, and the experimenters
added a question borrowed from Tulving: for every word you call old, say whether you
actually remember its presentation or merely know it was there. Of the never-presented
words that people called old after studying and then recalling a list, 72 per cent were
judged remembered rather than known.Supported, with limits They were not guessing under pressure or being agreeable: they were reporting
the felt texture of recollection about an event that did not occur — which is why so
small a procedure ended up in arguments far beyond word lists.
What this does not show
The tempting jump is from "I recognised a word that was never shown" to "so my memories
of my own life may equally be invented". That jump is what is in dispute.
Whether this illusion indexes anything about
false memory outside the word list is contested. Ost and colleagues gave 120 people a
standard misinformation experiment — CCTV footage followed by misleading information
about it — and a DRM test as an apparently unrelated filler; both effects came out
robustly, and the correlations between them averaged r = -.01, with reliability and
statistical power ruled out as explanations. Reviewers of the paradigm point the other
way, linking individual differences in this illusion to false autobiographical memories
and even to fantastic beliefs, and argue that the question can only be settled by
isolating the several distinct processes that produce even the simplest false
memory.Mixed / contested And the station measures nothing about you: one list, no norms, no score,
nothing kept.
Research and limitations
Two caveats sit close to this page. The first: the effect is a property of the list at
least as much as of the reader.
When Stadler, Roediger and McDermott collected
normative data for thirty-six of these lists, the spread was enormous: 60 per cent or
more of subjects falsely recalled window and sleep after the matching lists, with false
recognition above 80 per cent, while the list built from king produced 10 per cent false
recall and 27 per cent false recognition.Convergent / replicated Our station uses one published list and
reports one session, so a shrug from you says something about that pairing, not about
your memory.
The second concerns whether knowing helps. Being
warned in advance does not simply switch the illusion off, and how much it helps depends
on the person and on practice. Given an explicit warning before study and five
study-test trials on the same list, young adults virtually eliminated false recall by the
second trial; older adults used the warning to cut false recall on the first trial and
then could not reduce it any further across the remaining four, which the authors read as
a breakdown in spontaneous, self-initiated source monitoring rather than an inability to
follow instructions.Supported, with limits That is why this station owes a debrief rather than a warning: a
warning would have blunted the demonstration without making anyone immune, and told you
less.
Research drawer — claims, sources and limits
Every highlighted claim above is recorded here with its evidence grade and sources. Grades are per claim; this site never computes an overall truth score.
↩ In the experiment this station restages, thirty-six students heard six lists of twelve words, each list built entirely from the leading associates of one word that was never read out, and then wrote down everything they could remember. The missing word came back on 40 per cent of the lists — about as often as words that genuinely had been read to them from the middle of the list.
Roediger, H. L., III, & McDermott, K. B. (1995). Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21(4), 803-814. doi:10.1037/0278-7393.21.4.803 https://psych.hanover.edu/classes/cognition/Papers/RoedigerMcDermott%201996%20DRM%20False.pdf · accessed 2026-08-24 Support note: Experiment 1, pp. 804-805: 36 Rice University undergraduates; six lists taken from Deese (1959) — chair, mountain, needle, rough, sleep and sweet — each the first 12 associates from the Russell and Jenkins (1954) norms, read aloud at about 1 word per 1.5 s. The paper also records that "in a few instances, we replaced 1 of the first 12 associates with a word that seemed, in our judgment, more likely to elicit the critical word" (p. 805), so "built entirely from the leading associates" holds exactly for the chair list printed on that page and approximately for the other five. Figure 1 caption: "probability of recall of the studied words was .65, and probability of recall of the critical nonpresented item was .40", and the text states the critical item "was recalled with a probability of .40, or with about the same probability as items that had been presented in the middle of the list".
Gallo, D. A. (2010). False memories and fantastic beliefs: 15 years of the DRM illusion. Memory & Cognition, 38(7), 833-848. https://doi.org/10.3758/MC.38.7.833 · accessed 2026-08-24 Support note: Abstract characterises the DRM effect as an established "associative memory illusion, whereby people falsely remember words that were not presented" that has "broadly influenced basic theories of memory in cognitive psychology and neuroscience" over fifteen years of research — supporting the "replicated, not a one-off" half of the grade rather than the specific 40 per cent figure.
↩ The trick is older than the paper that made it famous. In 1959 James Deese built thirty-six lists, each one the twelve primary associates of a word he never read aloud, and found that the chance of a particular word intruding into recall was predicted by how strongly the list pointed at it — a correlation of .87 between intrusion probability and associative strength. His report then sat almost unread: Roediger and McDermott counted fourteen citations in total, and only one since 1983, before they picked it up thirty-six years later.
Roediger, H. L., III, & McDermott, K. B. (1995). Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21(4), 803-814. doi:10.1037/0278-7393.21.4.803 https://psych.hanover.edu/classes/cognition/Papers/RoedigerMcDermott%201996%20DRM%20False.pdf · accessed 2026-08-24 Support note: p. 804: "he developed 36 lists, 12 words per list. Each list was composed of the 12 primary associates of a critical (nonpresented) word"; "Deese obtained a correlation of .87 between the probability of an intrusion in recall (from one group of subjects) and the probability of occurrence of the word as an associate to members of the list (from a different group)"; and "since the Social Science Citation Index began publication in 1969, the article has been cited only 14 times, and only once since 1983". The 36-year gap is 1959 to 1995.
Deese, J. (1959). On the prediction of occurrence of particular verbal intrusions in immediate recall. Journal of Experimental Psychology, 58(1), 17-22. Journal of Experimental Psychology, volume 58, issue 1, pages 17-22 (1959); PubMed record PMID 13664879 confirms author, title, journal, volume and pagination. · citation only Support note: The primary publication being described. Not consulted directly: the PubMed record confirms author, title, journal, volume and pagination but carries no abstract, and no full text was reachable, so every detail above is taken from Roediger and McDermott's account of it rather than from Deese himself.
↩ On a recognition test the illusion is sharper than in recall. In Roediger and McDermott's first experiment the false-alarm rate for the missing word was .84, against a hit rate of .86 for words that had actually been studied, while unrelated words drew false alarms only 2 per cent of the time. Asked to rate their confidence on a four-point scale, subjects picked the top point — sure it was old — for the missing word 58 per cent of the time.
Roediger, H. L., III, & McDermott, K. B. (1995). Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21(4), 803-814. doi:10.1037/0278-7393.21.4.803 https://psych.hanover.edu/classes/cognition/Papers/RoedigerMcDermott%201996%20DRM%20False.pdf · accessed 2026-08-24 Support note: Experiment 1 Recognition, p. 806 and Table 1: "the hit rate was 86% and the false-alarm rate for the standard type of unrelated lures was only 2%"; "the false-alarm rate for the critical nonstudied lures (.84) approached the hit rate (.86), t(35) < 1". Table 1 gives the proportion of items given a rating of 4 (sure old): studied .75, critical lure .58, weakly related lure .04, unrelated lure .00.
Ost, J., Blank, H., Davies, J., Jones, G., Lambert, K., & Salmon, K. (2013). False memory is not equal to false memory: DRM errors are unrelated to the misinformation effect. PLoS ONE, 8(4), e57939. https://doi.org/10.1371/journal.pone.0057939 · accessed 2026-08-24 Support note: Independent replication of the recognition effect in a different laboratory and country: 120 University of Portsmouth undergraduates showed 68 per cent false recognition of critical lures against 3 per cent for new items, F(2,238) = 621.13, p < .001 — a smaller lure rate than Roediger and McDermott's, which is why the claim quotes their figures as theirs rather than as a universal constant.
↩ In a second experiment with twenty-four fifteen-word lists, false recall rose to 55 per cent, and the experimenters added a question borrowed from Tulving: for every word you call old, say whether you actually remember its presentation or merely know it was there. Of the never-presented words that people called old after studying and then recalling a list, 72 per cent were judged remembered rather than known.
Roediger, H. L., III, & McDermott, K. B. (1995). Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21(4), 803-814. doi:10.1037/0278-7393.21.4.803 https://psych.hanover.edu/classes/cognition/Papers/RoedigerMcDermott%201996%20DRM%20False.pdf · accessed 2026-08-24 Support note: Experiment 2, pp. 807-808: 30 subjects, 24 lists of 15 associates from Russell and Jenkins (1954) presented by tape at a 1.5-s rate. Results: "Subjects recalled the critical nonpresented word on 55% of the lists". Table 2 and text: in the study + recall condition the critical lure was called old .81 of the time, of which remember judgments were .58 — "with remember judgments being made 72% of the time (i.e., .58 divided by .81 times 100) for words that had never been presented". The remember-know procedure is attributed in the paper to Tulving (1985). Graded supported_with_limits rather than convergent_or_replicated: unlike the recall and recognition claims above, no independent replication of the remember/know split was opened for this entry, so the warrant is one experiment (n = 30) in the constituting paper.
↩ Whether this illusion indexes anything about false memory outside the word list is contested. Ost and colleagues gave 120 people a standard misinformation experiment — CCTV footage followed by misleading information about it — and a DRM test as an apparently unrelated filler; both effects came out robustly, and the correlations between them averaged r = -.01, with reliability and statistical power ruled out as explanations. Reviewers of the paradigm point the other way, linking individual differences in this illusion to false autobiographical memories and even to fantastic beliefs, and argue that the question can only be settled by isolating the several distinct processes that produce even the simplest false memory.
Ost, J., Blank, H., Davies, J., Jones, G., Lambert, K., & Salmon, K. (2013). False memory is not equal to false memory: DRM errors are unrelated to the misinformation effect. PLoS ONE, 8(4), e57939. https://doi.org/10.1371/journal.pone.0057939 · accessed 2026-08-24 Support note: Abstract and results: 120 participants watched CCTV footage, received misleading post-event information, completed free recall and recognition tests, and did a DRM test "as an ostensibly unrelated filler task"; "Despite obtaining robust misinformation and DRM effects, there were no correlations between a broad range of misinformation and DRM effect measures (mean r = -.01). This was not due to reliability issues with our measures or a lack of power."
Gallo, D. A. (2010). False memories and fantastic beliefs: 15 years of the DRM illusion. Memory & Cognition, 38(7), 833-848. https://doi.org/10.3758/MC.38.7.833 · accessed 2026-08-24 Support note: Abstract supplies the opposing side: "Some applicability is evident from research linking individual differences in the DRM illusion to false autobiographical memories (e.g., misremembering public events) and fantastic autobiographical beliefs (e.g., memories from past lives). But which aspects generalize?" — and argues that "multiple and often opposing psychological processes cause even the most basic false memories", so isolating component processes is required to answer the question.
↩ When Stadler, Roediger and McDermott collected normative data for thirty-six of these lists, the spread was enormous: 60 per cent or more of subjects falsely recalled window and sleep after the matching lists, with false recognition above 80 per cent, while the list built from king produced 10 per cent false recall and 27 per cent false recognition.
Stadler, M. A., Roediger, H. L., III, & McDermott, K. B. (1999). Norms for word lists that create false memories. Memory & Cognition, 27(3), 494-500. https://doi.org/10.3758/BF03211543 · accessed 2026-08-24 Support note: Abstract: "We tested false recall and false recognition for critical targets from 36 lists. Despite the fact that all lists were constructed to produce false remembering, the diversity in their effectiveness was large--60% or more of subjects falsely recalled window and sleep following the appropriate lists, and false recognition for these items was greater than 80%. However, the list generated from king led to 10% false recall and 27% false recognition."
↩ Being warned in advance does not simply switch the illusion off, and how much it helps depends on the person and on practice. Given an explicit warning before study and five study-test trials on the same list, young adults virtually eliminated false recall by the second trial; older adults used the warning to cut false recall on the first trial and then could not reduce it any further across the remaining four, which the authors read as a breakdown in spontaneous, self-initiated source monitoring rather than an inability to follow instructions.
Watson, J. M., McDermott, K. B., & Balota, D. A. (2004). Attempting to avoid false memories in the Deese/Roediger-McDermott paradigm: Assessing the combined influence of practice and warnings in young and old adults. Memory & Cognition, 32(1), 135-141. https://doi.org/10.3758/BF03195826 · accessed 2026-08-24 Support note: Abstract: half of each age group "were given an explicit warning about the DRM paradigm prior to encoding and were asked to attempt to avoid recalling any associated but nonpresented words"; "When a warning about false recall was provided, young adults virtually eliminated false recall by the second trial. Even though old adults also used warnings to reduce false recall on Trial 1, they were still unable to decrease false memories across the remaining four study-test trials"; the authors conclude that old adults "have a breakdown in spontaneous, self-initiated source monitoring". Graded supported_with_limits because this is a single study of one repeated list, and the wider warning literature is itself mixed.