LONDON (UNITED KINGDOM)
Scientific American [New York NY]
July 1, 2025
By Bernice Andrews and Chris R. Brewin
New studies underscore the difficulty of implanting entirely fictional events in a person’s recollection
[Print publication in the July/August 2025 issue. Online publication April 8, 2025]
How much can we trust our memories? We know that our mind keeps an imperfect record of the past. We can forget or misremember details, with frustrating consequences. Our attention can be diverted in ways that make it all too easy to miss key events.
But a particularly disturbing idea is that we readily form false memories—that is, that we can become convinced we experienced something that never actually occurred. The suggestion that it is easy to create false memories of entire events is often used to cast doubt on the reliability of a plaintiff’s testimony in a court case. For example, lawyers representing movie executive Harvey Weinstein cited this idea to raise questions about several women’s allegations against him.
Recently we had the opportunity to take a closer look at this concept by analyzing data from a study designed to replicate one of the most iconic experiments on false memories to date.
This experiment, by American psychologists Elizabeth Loftus and Jacqueline Pickrell, was published in 1995. Loftus had demonstrated decades earlier that one can manipulate people’s memories of visual details by posing questions that contain misinformation. She then wanted to learn whether it was possible to implant an entire false memory for a childhood event that had never happened. To that end, in the 1995 study, she and Pickrell misled participants into believing that, according to their parents or older sibling, when they were about five years old they had been lost in a shopping mall and then found by an older woman.
Over the course of two sessions the researchers strongly encouraged 24 participants to remember and describe all that they could about this experience (which the parents or older sibling denied had actually happened). The experimenters evaluated the participants’ responses and concluded that one quarter of them had been led to remember the suggested fake event either partially or fully. Loftus had previously claimed that some therapists could implant false memories of childhood sexual abuse in their clients. This “lost in the mall” experiment offered evidence that such a thing might indeed be possible. Over the years other scientists have established false memories of events in study participants, such as knocking over a punch bowl at a wedding, traveling in a hot air balloon or putting slime in a teacher’s desk.
In a 2017 paper, we identified two big questions that have been hanging over these studies. The first is: How confident can we be in the experimenters’ false-memory judgments? For example, would the participants agree that they not only believed in the false event on their relative’s say-so but had an actual memory of it? And second, what exactly was it that the participants remembered? Could some of those recollections have been true memories? What does a “partial” false memory consist of? Our analysis digs into these questions and suggests that the body of research on false-memory induction must be treated with caution; it is probably much more difficult to convince someone of a false memory than past work has suggested.
In 2023 Irish psychologist Gillian Murphy and her colleagues closely repeated the “lost in the mall” study, following the original methods. They used a larger sample of 123 people and reported that 35 percent of participants had a false memory, 10 percent more than in the original study. When asked, however, less than half as many participants (14 percent) said they had a memory of the fake event.
The data that were gathered by Murphy’s team and transcriptions of what participants had actually said were made freely available to other researchers, reflecting a move toward greater transparency in psychological research. We were impressed by this open approach to science, which is the only way to establish whether the claims made for memory implantation stand up to the scrutiny of independent researchers. For the first time, it was possible to examine what was really going on.
Before reanalyzing the data, we broke the “lost in the mall” story down into its six core elements: the person was around age five, was lost for an extended period, had cried, was found by an older woman and was reunited with their family, and this event occurred in the specifically suggested shopping location.
Our findings raise serious questions about claims made in court that it is easy to impart false memories.
To our surprise, none of the participants in the study remembered all six elements. Those rated as having a full false memory recalled fewer than three of the details on average, and those described as having a partial false memory recalled about one detail. Even more strikingly, 20 percent of those with a “full” memory and 60 percent of those with a “partial” memory did not explicitly remember the defining detail of being lost.
We also found that half of those judged to have a false memory had actually been lost before or experienced an analogous situation but not in a way suggested by the experimenters. In all cases, these participants described real events that they clearly distinguished from the suggested fake event. One participant said, “My memory is completely different than the other [suggested] memory.” Another said, “I don’t really remember that one…. But me getting lost in the shop was a regular occurrence.” Others were so uncertain about the suggested details in the fake story that their testimony would have little value in court. One participant commented, “I don’t even know if I ever did get lost in the shop before, so I’m not sure if it’s completely constructed or whether it’s the right memory.”
Taking everything into account, we estimated that only five participants could reasonably be claimed to have a false memory, rather than the 43 that were originally claimed. The participants were clearly very engaged by the study and approached the task of weighing up what, if anything, they remembered about the suggested event in a sophisticated way. Their comments revealed, for example, that they compared the scenario with other episodes of being lost, thinking about who would have been present and considering whether the mall was as suggested. Labeling their musings as a false memory does not capture these important aspects of their experience.
Our findings raise serious questions about claims made in court that it is easy to implant entire false memories in others. The great majority of these so-called false memories were much more limited, and held with much less conviction, than reports about this type of experiment led us to expect. As long as these questions remain open, psychologists should be very cautious about how they present findings on memory implantation to others. It is easy to overstate the relevance or generalizability of scientific evidence.
Nonpsychologists can take comfort in these findings. Although memory is limited and sometimes wrong, completely false memories are not easy to implant. Most of the time memory does a good enough job. And although it is valuable to apply critical distance and skepticism when considering the reliability of memory—particularly in legal contexts—we should not be too quick to throw out a person’s testimony simply because it could be imperfect.
Are you a scientist who specializes in neuroscience, cognitive science or psychology? And have you read a recent peer-reviewed paper that you would like to write about for Mind Matters? Please send suggestions to Scientific American’s Mind Matters editor Daisy Yuhas at dyuhas@sciam.com.Rights & Permissions
Bernice Andrews is an emeritus professor of psychology at Royal Holloway, University of London. She studies memory, adverse life experiences, post-traumatic stress disorder and research methodology.More by Bernice Andrews
Chris R. Brewin is an emeritus professor of clinical psychology at University College London. He studies post-traumatic stress disorder, trauma and memory.More by Chris R. Brewin