Eyewitness Memory and the Flashbulb Debate

The misleading information effect

Loftus and Palmer (1974) showed participants a film of a car crash and asked how fast the cars were going when they hit / collided / bumped / contacted / smashed into each other. The mean estimate rose from about 31.8 mph for contacted to about 40.8 mph for smashed. In a second experiment participants returned a week later and were asked whether they had seen broken glass; there was none in the film, but those given smashed reported it more than twice as often as those given hit. The verb had not merely biased an answer on the day — it altered what was reported a week later. Loftus and Zanni (1975) made the same point with a single article: asking about the broken headlight produced far more false sightings than asking about a broken headlight.

Two rival readings of these results survive, and IB rewards candidates who state both. The substitution account says post-event information overwrites the original trace. The response-bias account says the trace is intact and the wording merely shifts what the witness is willing to say. Loftus and Zanni's article manipulation is difficult for a pure substitution account, because a single word carries no new content about the scene.

Weapon focus, and what happens outside the laboratory

Loftus, Loftus and Messo (1987) tracked eye movements and found that participants fixated longer on a weapon than on a neutral object, and were worse at identifying the person holding it — the weapon focus effect. But Yuille and Cutshall (1986) interviewed real witnesses to a fatal shooting in Vancouver several months afterwards and found recall largely accurate and highly resistant to the misleading questions deliberately inserted by the researchers. The contrast is the heart of the ecological-validity argument in this topic: witnesses to a filmed crash have nothing at stake, while witnesses to a real crime encode under genuine consequence.

Flashbulb memory

Brown and Kulik (1977) proposed flashbulb memories: vivid, long-lasting, detailed memories of the circumstances in which shocking news was received, supposedly produced by a special biological mechanism. Neisser and Harsch (1992) tested the claim by collecting accounts the morning after the Challenger explosion and again about two and a half years later; accuracy was low and some participants scored zero, yet confidence remained high, and several rejected their own handwritten original account. Talarico and Rubin (2003) added the missing control: they compared memories of 11 September 2001 with everyday memories from the same period and found consistency declined at the same rate for both, while only the flashbulb memories kept their vividness and their felt accuracy. Sharot et al. (2007) found that participants who had been closer to the World Trade Center showed selective amygdala engagement when recalling that day, suggesting personal proximity, not the newsworthiness of the event, is what matters.