A study conducted by the U.S. Navy and Yale University found that eyewitness testimony is often largely inaccurate, even when victims get a long look at violent criminals, and police and juries may be giving this evidence too much credibility. After a unique study of 509 Navy and Marine officers undergoing high-level and low-level stress during elite hostage survival training at Fort Bragg, N.C., researchers discovered that few of the participating top officers could accurately identify their “interrogators” and “guards” during a line-up conducted 24 hours after the grueling sessions. When asked to select their captors from a line-up, only 30% of the high-stress group made correct identifications and only 62% of the low-stress group made accurate identifications. When sequential photos were used, the accuracy rate of the high-stress group rose to 49% and the low-stress rate improved to 76%. Researchers found that nearly 7 out of 10 high-stress participants made mistaken identifications, and there was no relationship between the participant’s confidence level and the accuracy of their memory. In fact, officers who were absolutely positive that they had selected the right person were no more likely to be correct than officers who expressed some doubt. “Memory in healthy people is not inherently terribly accurate. There’s a substantial amount of error. Maybe we should demand more evidence,” said Charles A. Morgan III, a Yale psychiatrist and lead author of the study. One explanation for these identification errors may be the high levels of hormones such as cortisol and adrenaline that result from stress and may degrade spatial memory. (New Hampshire Register, Science Section, June 21, 2004). See Resources.