By Emma Young
Where were you at 8am two Tuesdays ago? If it’s a little tricky to recall, what if I presented you with a map with four location flags to choose from, each about 3-4 km apart, with one marking your actual location on that time and date? Are you confident that you’d pick the right one?
If you are confident, the good news from a new paper in Psychological Science is that you’re more likely to be right than if you’re not too sure. The bad news is that when a group of students in Melbourne, Australia was tested in this way, they picked the wrong location 36% of the time. The study shows that this type of memory is pretty fallible — and yet it’s the type, of course, that’s needed for a criminal alibi. Failures in recalling where you were have no doubt contributed to serious miscarriages of justice, write Elizabeth Laliberte at the University of Melbourne and her colleagues.
The 51 students all downloaded onto their phones an app that had been developed by two of the researchers. Over a four week period, this app continuously tracked their movement and orientation, and also took GPS location and audio snapshots every ten minutes. Software stripped any speech from these audio files, but preserved background noise, such as birdsong or building work. (The app was designed to keep raw participant data confidential from even the researchers, and the participants were able to pause data collection for periods.)
After a week’s delay, the participants were tested, again via the app. For each of 72 trials, they had to pick where they were on a specific time and date from four markers on a Google map. (They were able to zoom in and out and pan as they pleased). One of these markers represented the correct location, and the other three represented places they’d been on other occasions.
The team’s analysis revealed that, as a group, the students were correct only about two-thirds of the time. Accuracy did increase with confidence, a finding that supports the idea that (under certain circumstances) greater confidence is linked to better memory. But the far from perfect performance of these university students, with presumably fairly regular schedules, doesn’t bode well for the rest of us — especially people without regular office hours.
The team highlights the case of American man Ronald Cotton, who in 1985 was convicted on charges of rape and burglary and sentenced to life plus 50 years in prison. Ten years after the conviction, DNA evidence exonerated him. Cotton’s initial defence had not been helped when his alibi had been shown to be false. “Rather than report where he’d been at the time of the crime, Cotton mistakenly recalled where he had been the week before. The error probably contributed significantly to his wrongful conviction,” the researchers write.
The team’s analysis suggests that this error is relatively common: choosing a “right-day, wrong-week” location accounted for 19% of their participants’ errors. Participants also frequently chose a location that matched the stipulated hour but got the day wrong — this error accounted for 8% of the mistakes. (They also sometimes picked an incorrect location in which they had made similar patterns of movement or where the background sounds were similar to the correct location — though these errors were less common). “Both of these figures are well above chance rates and suggest that the error that Ronald Cotton made might be quite common,” the team writes.
Clearly, the findings could have practical, legal applications. Knowing just how fallible this type of alibi memory is — and understanding the most common errors that people make — could help lawyers to ask the right questions of defendants and witnesses, to try to get closer to the truth of where someone really was at a certain time.
But there are more general implications for understanding memory, and for the way that memory research is done. Experiments are often lab-based, raising questions about real-world applicability. Passive, unobtrusive, privacy-conscious monitoring technologies like this new app could be used to gather and analyse all kinds of real-world data. As the researchers write, “We are now able to capture the ground truth of people’s lives and to test their ability to retain those events.” And this, they hope, could change how a good deal of memory research is done.