Memory Versus Media: Creating False Memories with Virtual Reality


At the Stanford University Virtual Human Interaction Lab, 6-year-old Sam listened intently to an experimenter describing a fictitious scenario where Sam was swimming with two friendly orca whales in beautiful blue water. When Sam was questioned if he could recall ever swimming with the orca whales in real life, he vehemently shook his head.

Later on, Sam’s classmate Adam came to the lab. Adam donned a virtual-reality helmet that immersed him into a world completely different from the lab’s physical surroundings. Within a few seconds, Adam could see himself underwater frolicking with two orca whales. The three-dimensional virtual version of Adam was a near mirror reflection of his physical body — with short brown hair, alert brown eyes and the corners of his lips turning up in a natural smile. He appeared to really enjoy swimming with the orca whales.

After a few minutes the underwater world faded, and the experimenter helped the boy lift the virtual-reality helmet off of his head. The experimenter asked Adam, “Do you remember swimming with these whales in real life?” Unlike Sam, Adam wasn’t so quick to respond, and he bit his lip a little as a concentrated look came across his face. “I think … yes, I think I remember now,” he replied.

Human memory is complex. Humans encode, store and retrieve bits of information constantly; and while these processes are frequently accurate, the human mind is fallible. Human memory errors range from failure to retrieve stored events to what scientists call false memories
— “remembering” events that did not occur, as in Adam’s case. Most of such memory errors do not garner national attention, but false memories of sexual abuse — supposedly planted and elicited by therapists — became the subject of several prominent lawsuits in the 1990s. These lawsuits drew the attention of the general public and many scholars and motivated research on false memories.

In order to understand how Adam developed a false memory of swimming with the orca whales, we must first be familiar with how the human brain stores and retrieves bits of information and identify where this process is vulnerable to error. Numerous studies on false memory creation have revealed that repeated interviewing, leading questions and biased interviewers, along with other causes, can lead to memory errors in everyone, and with children in particular.


Memories come from a variety of sources, such as television, imagination, books and first-hand physical experiences, but according to the Source Monitoring Framework (SMF) developed by Marcia Johnson, a professor at Yale University, memories are rarely stored with tags that identify their sources. This means that each time a memory is retrieved, the brain attempts to determine the source of the memory. During this rapid and mostly automatic source-monitoring process, memories are evaluated and attributed to particular sources.

Each type of source produces memories with unique characteristics. For example, physical-world memories — memories of events that actually happened to the rememberer — are extremely rich in perceptual detail. Memories developed from a book, however, usually do not contain as much detail. One of the indicators the human brain uses to determine a memory’s source is the amount of detail stored in a memory. Another indicator of a memory’s source is the amount of cognitive effort the brain uses to organize and store the memory. Little cognitive effort is needed to store and organize the memory details for a physical-world event. On the other hand, it takes substantially more cognitive effort to organize and store the details of an event that was merely read about in a newspaper. The brain is able to differentiate memories that require high effort to organize and store from memories that require low effort, and the brain uses this information to draw useful conclusions about the sources of our memories.

The problem is that today’s digital media, such as video games and websites, as well as the virtual-reality helmet that Adam donned, are so rich with detail — from three-dimensional graphics and personalized avatars to high-quality sound and more — that it is easy for the human brain to organize and store memories derived from detail-rich experiences with digital media. Thus, it becomes difficult for the brain to distinguish between memories of physical-world experiences and memories of experiences with digital media — both of which tend to be rich in detail and relatively easy to organize and store. In a recent study published in the journal Media Psychology, we discovered that source-monitoring errors occur in direct proportion to media richness.


Previous research on human memory has shown that false memories can develop in response to a combination of self-relevant information and different types of media. For example, in one study conducted by Maryanne Garry and her colleagues, adult participants viewed a manipulated childhood photograph falsely depicting each participant riding in a hot-air balloon as a young child. When shown the manipulated photographs, half of the adult participants came to erroneously recall riding in a hot-air balloon as a child. Other studies have explored the brain’s ability to distinguish memories derived from written narratives, mental imagery or virtual-reality simulations from memories of physical-world experiences.

To further this line of research and test the hypothesis that source-monitoring errors occur in direct proportion to media richness, at Stanford University’s Virtual Human Interaction Lab we tested a new form of media for studying false memory formation: Immersive Virtual Environments (IVEs). IVEs are created using a combination of hardware, software and interactive devices which make the user feel surrounded by a virtual, computer-generated environment. Immersive Virtual Environment Technology (IVET) continually tracks the user’s position and updates the images that the user sees in the head-mounted display (see Fig. 1, Panel C). In addition, using photographs of people and places, IVEs can be personalized to include virtual humans or environments that resemble specific people or locations in the real world. A three-dimensional model of child’s head (see Fig. 1, Panel A) can be attached to a virtual body and included in a virtual-reality simulation. Furthermore, virtual humans can be animated with a high degree of behavioral realism. Experiences in virtual environments, unlike experiences with less rich media, may be encoded into human memory with such detail that they become very difficult to distinguish from memories of physical-world experiences.

In the Stanford University study, 55 preschool and elementary children listened to the experimenter verbally describe a false childhood scenario where the children either swam with friendly orca whales or shrunk to dance with a stuffed mouse. The children were then randomly assigned to one of four memory-prompt conditions that varied according to media richness.

One-quarter of the children were instructed to sit quietly for several minutes; one-quarter were asked to sit and imagine the childhood scenario for several minutes; another quarter were immersed in a virtual-environment simulation where they saw another virtual child either swim with orcas or shrink to dance with a stuffed mouse; and the remaining children were immersed in a virtual-environment simulation where they saw a virtual child designed to look like the participant swimming with orca whales (see Fig. 1, Panel B) or shrinking and dancing with a stuffed mouse.

Next, all participants were asked if they could remember swimming with the orcas whales or dancing with the stuffed mouse in the real world. Results revealed that while preschool children were equally likely to develop false memories across all memory-prompt conditions, elementary children recalled significantly more false memories in the mental imagery and IVET condition where they saw a virtual child that looked like themselves than in the condition where they were asked to sit quietly after listening to the false scenario.


The results reveal that mental imagery (which the participant must actively engage) and IVET self-relevant simulations (which can be completely controlled by a third party and only passively perceived by the participant) are both powerful in eliciting false memories in children. This study suggests that third parties may be able to elicit false memories without the consent or mental effort of an individual. With this in mind, scholars and professionals alike must critically examine the use of detail-rich digital media in places such as courtrooms and counseling centers. There are many benefits and precautions to consider before implementing digital-media technology; different forms of media may manipulate the human experience in dramatic and predictable ways.

Be the first to comment

Leave a Reply

Your email address will not be published.