Weapon focus refers to the decreased ability to give an accurate description of the perpetrator of a crime by an eyewitness because of attention to a weapon present during that crime. In the first experiment, subjects viewed a mock crime scene in which a weapon was either highly visible or mostly hidden from view. Subjects in the highly visible weapon group recalled significantly less feature information. Overall, memory accuracy scores were negatively correlated with self-reported arousal. The second series of experiments tested the weapon focus effect in a nonemotional situation in which the "time in view" of both the weapon and the individual's face were manipulated. A series of six slides were used in which either the weapon or the face was not in view for specific intervals within the sequence. The weapon focus effect was found to occur within a nonarousing, environmentally stark setting and was dependent on the percentage of time the weapon was visible.

Forty students served as subjects in two groups of witnesses to which a surprise filmed crime was shown. All subjects gave free recall statements about the crime. Following either a directed thinking interval or a diversionary task to block rehearsal, witnesses gave two additional statements. Hypermnesia for hits and memory intrusions was observed, but d' scores did not show hypermnesia. The witnesses fell into two groups: "good" witnesses who had ascertained the correct schema and "bad" witnesses who were obviously on the wrong track. While most witnesses showed short-term hypermnesia for correct responses (hits), the good witnesses showed a significant increase in memory intrusions (false alarms) during the third recall trial (p < .01). Only with correct-schema witnesses did d' scores above the guessing level occur. Witnesses who had ascertained the correct schema but who were interrogated too often tended to make things up to suit the demands for detail, with a resultant increase in memory intrusions and lower d'.

No abstract