Making Memory Research Mirror Real Life

Targeting attentional rhythms • Segmenting our daily lives • Remembering fictional information • A bit better

Although memory has been extensively studied since the 19^th century and the beginning of psychological research, gaps in understanding remain. The key mechanisms that impact remembering or forgetting in any given moment and how people separate factual memories from fictional memories remain poorly understood. As a result, how best to intervene when memory fails—in old age, clinical populations, or everyday life—is yet unknown. 

Psychological scientists are trying to address these gaps by designing studies that better represent how people actually remember—or forget—experiences in their daily lives. They are moving away from outdated, simplistic methodologies—such as memorizing lists of words or responding with yes/no answers—and creating updated models that better cover the full spectrum of different types of memory. With new tools and technologies, researchers are able to capture how memory works in more authentic ways.

Shawn Schwartz, a doctoral candidate in APS Fellow Anthony Wagner’s Memory Lab at Stanford University, wanted to study memory in ways more reflective of remembering in the real world.

“A lot of the way memory is discussed is that it’s like a binary decision or some set of decision processes where you either remember it or you don’t,” he said in an interview with the Observer. “Sometimes it feels like memory is more continuous. It could be measured in a more continuous way than some yes or no type of sensitivity.” 

By asking questions with answers on a continuum—for example, identifying an item’s color on a color wheel instead of stating whether it was blue or red—more subtle differences can be detected, an important consideration for studying memory in both healthy aging and Alzheimer’s disease.

In order to catch any decline, “You want to get any subtle hint as early as possible, [so] you definitely need a sensitive paradigm,” said Haopei Yang, a postdoc in Wagner’s lab. 

Asking for answers on a continuum allows researchers to plot the distribution of responses in a quantitative way. The distribution of responses is much closer to the true color for younger adults than for older adults, who have a wider distribution. The wider range reveals the differences between older people who may be “super agers,” people who follow the norm for healthy aging, and those beginning to show signs of dementia. 

“It is quite fascinating how much information you can get when you use a continuous type of response,” said Schwartz. 

Targeting attentional rhythms

In a 2025 Current Directions in Psychological Science paper, Schwartz and Yang highlighted recent advances and tools that can help get at these subtle differences and understand both the mechanisms behind episodic memory and when best to intervene.

When attention slips, episodic memory also suffers. To witness this in the lab, researchers typically distract subjects—with flashing lights or multiple tasks that compete for attention, for example. But again, Schwartz and Yang wanted to get closer to real-life experiences and so chose to use closed-loop biofeedback approaches, such as measuring pupil diameter or posterior alpha power via scalp electroencephalogram (EEG), to monitor attention in real time and provide an intervention when attention slips.

“We have found initial evidence that we can, at least through the mechanism of the pupil, causally detect, intervene, and improve or recover episodic memory performance back to baseline,” said Schwartz. “All by way of the participant not even knowing that the experiment was being manipulated by their own attention.” 

The rhythmic structure of attention, which fluctuates in the theta and alpha frequency ranges, also provides a target to improve memory. For example, some research indicates troughs in hippocampal theta oscillations may be optimal for encoding memories. It would follow that information presented in a suboptimal phase—such as a peak in theta—may not be remembered as well.

“When you are in the optimal phases, you are more likely to perform the test better versus when you’re in the suboptimal phases,” said Yang. Using high-temporal-resolution recording methods like EEG or even something as simple as pupil size can tell you, “‘Now this participant is going to be in the optimal phase and let’s present this thing that they’re going to remember now … [or] let’s wait until the optimal phase comes online,’” said Yang.

“It’s quite fascinating that such a simple and one-dimensional signal that’s the conjunction of all these multidimensional processes that are going on in the brain could be used to actually improve someone’s performance,” said Schwartz.

Segmenting our daily lives

Like Schwartz and Yang, Maverick Smith, an assistant professor at Truman State University, wanted to learn more about how people actually remember experiences, adopting a different approach than how it’s typically tested in the lab.

Although memory studies classically ask subjects to memorize words or pictures, “Most of us don’t go around trying to remember a list of words,” said Smith. “The types of information that people really need to remember and try to remember are information that is much more dynamic: that conversation that I had with my significant other last night, or did I take my medication this morning?”

In other words, rather than encountering information as discrete items, people are usually trying to remember things from a continuous flow of activity. To do that, they break experiences into meaningful chunks. For example, when making breakfast, people might parse cooking eggs, pouring juice, and preparing the toast into three separate segments. Segmentation may guide attention, help coordinate when information is moved from working memory to long-term memory, and reduce interference at retrieval. 

“Segmentation seems to be this encoding process that people are engaging in that is influencing how it is that they later remember that experience,” said Smith. “People that chunk the experience in the same way as others or as the normative of the group tend to have better memory for that experience later.” 

In a 2025 Current Directions in Psychological Science paper, Smith described how interventions that improve segmentation may also improve memory.

Participants who watched videos and pushed a button at what they perceived as event boundaries—the end of pouring the juice, for example—remembered the videos better up to one month later than participants who were told only to try to remember the video or those told to push buttons at regular intervals. However, these results depended on age: Older adults’ memories did not improve with this type of segmentation task.

Instead, Smith thinks aging adults need more support to remind them when to segment. To investigate this, he edited the videos to include loud tones and arrows on the screen when an event boundary occurred. With this help, memory improved in cognitively healthy older adults.

Smith suspects that aging adults may lose some of the ability to initiate segmenting themselves.

“Maybe it’s the case that in healthy aging and in Alzheimer’s disease, there’s just some sort of decline or even deterioration that’s happening in the ability to maintain goals and segment that experience,” he said. “But if the environment is constructed in such a way that can skew the systems that are already in place and to scaffold them, then you do see the benefit.” 

Myth: Eyewitness Testimony is the Best Kind of Evidence

Smith hopes that future work can address how to help people become better at segmenting. “Maybe one day your YouTube video or your Netflix has an option for movie editing that will make your memory for this video better,” he said.

Remembering fictional information

It’s crucial that the memory of that YouTube or Netflix video be categorized as fiction or fact. Memories of fictional information—from videos, novels, games, and pretend play—can affect people’s behaviors, beliefs, and attitudes, and can be a potent source of misinformation if not separated from factual memories. But past models of memory didn’t account for memories of fictional information and didn’t theorize how fact and fiction are processed differently. 

“I looked around for different memory theories that could account for this distinction that people can still recall events from novels or films, and they know that they are not real,” said Pierre Gander, an associate professor of cognitive science at the University of Gothenburg. “There didn’t seem to be any theory that could account for this.” 

To rectify this, in a 2025 paper in Perspectives on Psychological Science, Gander expanded David C. Rubin’s 2022 model of memory to include another dimension for fictional memory. But rather than call it the definitive model, it is really just a proof of concept that memories of fictional information can and should be incorporated into memory models.

“We see it as one example of how a memory model could accommodate this,” he said. “It’s just one possibility.” 

Gander also discussed three cognitive mechanisms that may help separate fact from fiction: content-based reasoning (does the memory seem fantastical, impossible, or incompatible with reality?), source monitoring (where did you learn the information from?), and associative links to the concept of fiction (such as knowing unicorns and Santa Claus are make-believe).

Although source monitoring may be the main mechanism, Gander sees them all working together, providing input from different perceptual details and content. To help prevent misclassifying information, understanding the conditions where confusions occur is key.

Stress or split attention while reading or watching something fictional can lead to mistakes during encoding, leading to it being misremembered later as fact, said Gander.

A bit better

Memory mistakes can not only lead to confusion. They can also be detrimental and life-threatening, from someone forgetting vital medication to an older person wandering around lost. Memory mistakes on the witness stand can lead to wrongful convictions, making understanding those memory mishaps crucial (see sidebar). 

Related content: Careers Up Close: Ryan Fitzgerald Eyes Eyewitness Identification

Along with new technologies, tools, and models, these types of studies are leading to discoveries about how to prevent these mistakes from occurring. They are also leading to a better understanding of memory by moving the field beyond the older methods of memorizing lists or asking yes/no questions. 

“That’s definitely the foundation of memory and understanding how it is that memory works,” said Smith. “But I think we can do a little bit better.”

Back to top

Feedback on this article? Email [email protected] or login to comment.

Related content we think you’ll enjoy

References 

Gander, P., Szita, K., Falck, A., & Lowe, R. (2025). Memory of fictional information: A theoretical framework. Perspectives on Psychological Science, 20(2), 308–324. 

Schwartz, S. T., Yang, H., Xue, A. M., & Wagner, A. D. (2025). Attending to remember: Recent advances in methods and theory. Current Directions in Psychological Science.

Seale-Carlisle, T. M., Quigley-McBride, A., Teitcher, J. E. F., Crozier, W. E., Dodson, C. S., & Garrett, B. L. (2024). New insights on expert opinion about eyewitness memory research. Perspectives on Psychological Science, 20(5), 903–924.

Smith, M. E., & Zacks, J. M. (2025). Event segmentation interventions improve memory for naturalistic events. Current Directions in Psychological Science.