Human memory might be even more unreliable than currently thought. In a new study, scientists found that it’s possible for people to form false memories of an event within seconds of it occurring. This almost-immediate misremembering seems to be shaped by our expectations of what should happen, the team says.
What Was Your First Experience with Super Mario Bros? | io9 Interview
Over the past few decades, it’s become clear that our process of remembering is deeply flawed. Studies time and again have shown that our memory is routinely inaccurate, while others have found that you can easily convince people to falsely believe that past events have happened in their lives, from getting lost in a mall to being viciously attacked by an animal. Tragically, people have even been accused or convicted of horrific crimes on the basis of someone’s memories purportedly “recovered” later in life (in at least some of these cases, the convictions were eventually overturned or the accusers themselves recanted their memories).
Much of the research on misremembering has focused on long-term memory, with an emphasis on childhood events from long ago. But scientists in the Netherlands, UK, and Canada wanted to take a closer look at the potential inaccuracy of short-term memory.
“This study is unique in two ways, in our opinion. First, it explores memory for events that basically just happened, between 0.3 and 3 seconds ago. Intuitively, we would think that these memories are pretty reliable,” lead author Marte Otten, a neuroscientist at the University of Amsterdam, told Gizmodo in an email. “As a second unique feature, we explicitly asked people whether they thought their memories are reliable—so how confident are they about their response?”
G/O Media may get a commission
To do this, they recruited hundreds of volunteers over a series of four experiments to complete a task: They would look at certain letters and then be asked to recall one highlighted letter right after. However, the scientists used letters that were sometimes reversed in orientation, so the volunteers had to remember whether their selection was mirrored or not (for example, correctly identifying whether they saw c vs ↄ). They also focused on the volunteers who were highly confident about their choices during the task.
Overall, the participants regularly misremembered the letters, but in a specific way. People were generally good at remembering when a typical letter was shown, with their inaccuracy rates hovering around 10%. But they were substantially worse at remembering a mirrored letter, with inaccuracy rates up to 40% in some experiments. And, interestingly enough, their memory got worse the longer they had to wait before recalling it. When they were asked to recall what they saw a half second later, for instance, they were wrong less than 20% of the time, but when they were asked three seconds later, the rate rose as high as 30%.
According to Otten, the findings—published Wednesday in PLOS One—indicate that our memory starts being shaped almost immediately by our preconceptions. People expect to see a regular letter, and don’t get easily fooled into misremembering a mirrored letter. But when the unexpected happens, we might often still default to our missed prediction. This bias doesn’t seem to kick in instantaneously, though, since people’s short-term memory was better when they had to be especially quick on their feet.
“It is only when memory becomes less reliable through the passage of a tiny bit of time, or the addition of extra visual information, that internal expectations about the world start playing a role,” Otten said.
Of course, this is only one study, involving a piece of knowledge that’s strongly reinforced throughout our lives (letters). So Otten and her team hope to keep testing the fallibility of short-term memory in other ways.
“I am personally very interested in finding a way to test the effects of social knowledge, such as prejudice or stereotypes and individual beliefs, on short-term memory. Do the expectations that we have about people based on, for example, their gender, almost immediately start shaping what we remember about, say, their voice or facial expression? Or do I after only a few seconds start slightly misremembering certain data-representations, because it does not fit my beliefs about, say, climate change?” she said. “This is obviously a bit more complicated to explore than just asking people to look at a display of 6 letterlike items, but I look forward to exploring this further.”
