We’ve reported on flashbulb memory before, with the Talarico and Rubin study and the MacKay and Ahmetzanov study. First observed in 1977 by Brown and Kulik, flashbulb memories—memories about shocking events—were supposed to be more vivid and long-lasting than normal memories. Jennifer Talarico and David Rubin seemed to have put a damper on the whole concept of flashbulb memory with their finding that while flashbulb memories are more intense and people are more confident about them, they are no more accurate than normal memories. Donald MacKay and Marat Ahmetzanov, using an experimental test, found that when the memory is directly related to the shocking event itself, it is more likely to be remembered than a normal memory.
In a report published in the journal Memory roughly simultaneously with Talarico and Rubin’s Psychological Science article, Susan Hornstein, Alan Brown (a different Brown from the Brown of Brown and Kulik fame), and Neil Mulligan seem to have come to the opposite conclusion from Talarico and Rubin. They quizzed psychology students at Southern Methodist University one week after the death of Princess Diana. When the same students were retested 3 and 18 months later, their memories were just as accurate. Further, more confident students were more accurate in their responses. What’s going on here? Can these divergent results be reconciled?
Let’s take a closer look at Hornstein, Brown, and Mulligan’s study. Like Talarico and Rubin, they tested participants nearly immediately after the shocking event occurred and retested several months later. Unlike Talarico and Rubin, they did not simultaneously ask respondents about a normal memory for comparison.
What they did find is something that Talarico and Rubin did not: a significant correlation between emotion and memory accuracy. Take a look at this chart of accuracy ratings.
Students were asked whether how intense their emotions were on learning of Princess Diana’s death. The students with high or medium emotional intensity had significantly more accurate memories of the event than those with low emotional intensity. So why didn’t Talarico and Rubin find such a correlation? One possibility is a ceiling effect: surely it’s plausible that none (or at least very few) of the respondents at Duke felt a “low” level of emotional intensity when they were surveyed the day after the attacks.
But what about Talarico and Rubin’s comparison case? Didn’t they also ask respondents about an ordinary event occurring a few days before September 11? Why was no emotional correlation found there? The answer might lie in MacKay and Ahmetzanov’s experimental study. They found that emotion increased accuracy only when the details of the shocking memory were directly related to the item being recalled. Perhaps the fact that students took a survey on the day after September 11 gave those “neutral” memories preceding September 11 a significant emotional valence. Remember, only a medium level of emotional activity was necessary for Hornstein et al.’s participants to have more accurate memories. If that was the case, then both the shocking and neutral memories would have been coupled with strong emotion, leading to more accurate memories. Talarico and Rubin would be able to discern no difference between the two.
But it’s also possible that something else was going on—the array of emotions and thoughts following a shocking event can be staggeringly comples. These complexities are among the difficulties of studying flashbulb memory. Each “flashbulb” event is different. A UK study of flashbulb memory for Margaret Thatcher’s resignation found over 60 percent memory retention, but a similar study for the same event with U.S. participants found only 28 percent retention. It does seem fair to say that under certain circumstances, we are more likely to accurately remember events than in others, but learning exactly what those circumstances are will require a great deal of additional research.
Hornstein, S.L., Brown, A.S., & Mulligan, N.W., (2003). Long-term flashbulb memory for learning of Princess Diana’s death. Memory, 11(3), 293-306.