How do false memories form?
How a rebuilt memory takes in later information, related ideas and imagined scenes, and what pooled studies say about how often it happens.
Answer to the question
How do false memories form?
False memories form when a memory is rebuilt at recall and the rebuild takes in something that was not part of the event: later information, related ideas or an imagined scene. Roediger and McDermott found a word never shown was recalled 40 percent of the time, and a pooled analysis classed 30.4 percent of implantation cases as false memories.
Key takeaways
- Loftus, Miller and Burns (1978) showed 1,242 participants a slide series of an accident; misleading information given afterwards lowered recognition accuracy, most when introduced just before the final test.
- In the Roediger and McDermott word-list experiments (1995), a word never shown was recalled 40 percent of the time in the first experiment and 55 percent in the second.
- Imagining a childhood event inflated confidence that it occurred in Garry and colleagues' 1996 study.
- The source monitoring framework (Johnson, Hashtroudi and Lindsay, 1993) holds that judging where a memory came from rests on qualities of the experience, and that those judgements are open to error.
- A pooled analysis of eight implantation studies (423 participants) classed 30.4 percent of cases as false memories and a further 23 percent as accepting the event to some degree (Scoboria and colleagues).
Most false memories are assembled from parts that were real, and the false detail slots into the frame so smoothly that it carries the same confidence as everything around it. The person did visit the place, did know the people and did feel the feeling.
That pattern runs through the laboratory work on how false memories form: the error is rarely a blank filled with nonsense, and more often a plausible addition made while the memory is being rebuilt.
Every recall is a rebuild
Frederic Bartlett argued in 1932 that remembering is reconstructive, with each recall composed afresh from stored fragments and from what the person expects to have happened. Charan Ranganath makes the same case in current terms in Why We Remember, where recall uses the machinery that also builds imagined scenes, which is why a vivid scene can be put together without the event ever having occurred.
A memory that is rebuilt on demand has an opening in it, because anything near the right fragments can be pulled into the rebuild. The same process that lets a person recall a day nobody recorded also lets a detail from another day, another telling or another person's account sit inside the scene without any sense of having been added. The trace itself is set out in What is an engram How memories are stored, and the pull toward a warmer, tidier past in Your Nostalgia Is Lying to You comes from the same assembly. False memories are what assembly produces when one of the pieces on the bench did not come from the event.
How false memories form from added information
Information that arrives after an event can end up inside the memory of it. Elizabeth Loftus, David Miller and Helen Burns tested this in 1978 across five experiments and a pilot with 1,242 participants, who watched slides of a single accident and were later given consistent, misleading or irrelevant information about it. Misleading information lowered accuracy on both a yes-no and a two-alternative recognition test, and it had a larger effect when it was introduced just before the final test and not straight after the event. A simple demand-characteristics explanation could not account for the results, and the authors concluded that what a witness hears after an event, consistent or misleading, is integrated into the memory of that event.
Outside suggestion is one route. Association is another, and it needs no outside voice at all. In the Roediger and McDermott experiments of 1995, modelled on earlier work by James Deese, participants studied lists of twelve words such as bed, rest and awake, each list built around one word that was never shown, here sleep. On immediate free recall the missing word was produced 40 percent of the time and was later recognised with high confidence. A second experiment with an expanded set of lists reached 55 percent, and on a later recognition test the false alarms to these words came at a rate comparable to the hits for words that had really appeared.
In both routes the false detail is reported with confidence, and nothing in these studies required the real event to have been forgotten first. Most false memories of this kind sit beside true ones, built from the same material.
Why a pictured scene feels remembered
A false detail still has to feel like a memory, and imagination supplies much of that feeling. This is the step that turns a suggestion into one of the false memories people report as their own. Maryanne Garry, Charles Manning, Elizabeth Loftus and Steven Sherman reported in 1996 that people who first rated their confidence in a set of childhood events, then imagined some of them, rated those events as more likely to have happened, for every target item. A scene that has been pictured carries sensory detail and a sense of place, and those are the qualities people usually take as signs of a real experience.
Marcia Johnson, Shahin Hashtroudi and Stephen Lindsay built the framework that explains why this matters. In their 1993 review of source monitoring, the decision about where a memory came from, whether seen, heard, told or imagined, rests on qualities of the experience such as perceptual detail and the thoughts that went with it. The judgement uses flexible criteria and is open to error, and the review covers cases in which fiction is folded into fact and familiarity is credited to the wrong source.
A rebuilt scene with strong detail and no label on its origin passes the check. That is the point at which a suggestion, an association or a pictured scene is taken for an experience, with nothing to mark it as borrowed.
Four questions help weigh a vivid memory before it is trusted:
- Where did this detail come from: something seen, something told, or something pictured?
- Is the feeling of certainty being counted as support for the detail?
- Is there a record, such as a dated message or a photograph, that settles the small specific?
- Has the scene been retold or imagined many times, adding routes for new material?
How often it happens in studies
Estimates of how often false memories form vary between studies, partly because the studies defined a false memory differently and used different ways of inducing one. The best single estimate comes from work that fixed the definition first. Alan Scoboria and colleagues pooled reports from eight published implantation studies, 423 participants in all, and had independent raters code transcripts against seven criteria: accepting the suggestion, elaboration beyond it, imagery, coherence, emotion, memory statements, and not rejecting the suggestion. The authors described this as the most valid estimate of false memory formation in the implantation literature to date.
By that scheme, 30.4 percent of cases were classified as false memories and another 23 percent as having accepted the event to some degree. The rate of false memories was higher under particular conditions. When the suggestion included self-relevant information and an imagination procedure, and was not accompanied by a photo depicting the event, the memory formation rate across those cases was 46.1 percent, which is well above the 30.4 percent found for the full set of eight studies, whose transcripts were all coded by independent raters against the same seven criteria.
Common questions
How do false memories form?
They form when a memory is rebuilt at recall and the rebuild takes in something that was not part of the event: information given afterwards, ideas related by meaning, or a scene that was imagined. Studies show each route producing confident reports of material that was not experienced.
Does someone have to push a false memory on you?
No. In the Roediger and McDermott word-list experiments, a word that was never shown was recalled 40 percent of the time in the first experiment, with no outside suggestion beyond the related words themselves.
Can a false memory feel as sure as a true one?
In the same experiments the missing words were recognised with high confidence, and in a second experiment false alarms to them were comparable in rate to hits on words that had appeared. Garry and colleagues found that imagining a childhood event inflated confidence that it occurred.
How common are false memories in implantation studies?
A pooled analysis of eight studies with 423 participants classified 30.4 percent of cases as false memories and a further 23 percent as accepting the event to some degree. The rate was 46.1 percent under one combination of conditions.
Sources
Ranganath, C. Why We Remember. Seed book; vault hub Books/Why We Remember.
Loftus, E. F., Miller, D. G., & Burns, H. J. (1978). Semantic integration of verbal information into a visual memory. Journal of Experimental Psychology: Human Learning and Memory, 4(1), 19-31. doi:10.1037/0278-7393.4.1.19 https://doi.org/10.1037/0278-7393.4.1.19
Roediger, H. L., & 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://doi.org/10.1037/0278-7393.21.4.803
Garry, M., Manning, C. G., Loftus, E. F., & Sherman, S. J. (1996). Imagination inflation: Imagining a childhood event inflates confidence that it occurred. Psychonomic Bulletin & Review, 3(2), 208-214. doi:10.3758/BF03212420 https://doi.org/10.3758/BF03212420
Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin, 114(1), 3-28. doi:10.1037/0033-2909.114.1.3 https://doi.org/10.1037/0033-2909.114.1.3
Scoboria, A., Wade, K. A., Lindsay, D. S., Azad, T., Strange, D., Ost, J., & Hyman, I. E. (2017). A mega-analysis of memory reports from eight peer-reviewed false memory implantation studies. Memory, 25(2), 146-163. doi:10.1080/09658211.2016.1260747 https://doi.org/10.1080/09658211.2016.1260747