Weapon focus is a factor affecting the reliability of eyewitness testimony. Weapon focus signifies a witness to a crime diverting his or her attention to the weapon the perpetrator is holding, thus leaving less attention for other details in the scene and leading to memory impairments later for those other details. Elizabeth Loftus, Yuille and Burns, have all been associated with studies showing the existence of a weapon focus effect. According to a 2001 survey of eyewitness experts, 87% found the effect sufficiently reliable to form the basis of expert testimony in criminal trials.
In the field of forensic psychology, researchers have validated the weapon focus effect and shown that a witness will remember less about a crime, or the perpetrator of a crime, when a weapon is present, as opposed to if the weapon is not present at an identical crime. As for the reason the phenomenon occurs, the two leading explanations attribute it to the cognitive arousal of the victim, or to the overall unusualness of the situation.
In one of the earliest known investigations of weapon focus, Johnson and Scott (1976) had two groups of participants come in to what they thought was a laboratory study of human memory. In actuality they were to take part in a simulated interaction intended to determine whether the presence of a weapon would influence eyewitness memory for an event. Participants in the control condition sat in a waiting room where they overheard a conversation between two people following which a man exited with greasy hands and a grease pen. In the weapon condition participants sat in the same waiting room, but instead they heard a violent argument - including furniture being thrown around - following which a man came out holding a blood-stained knife. During a photo line-up, the control participants were more likely to accurately identify the man they say in the waiting room relative to participants in the weapon condition (49% versus 33% correct identifications).
The study conducted by Johnson and Scott (1976) represents one of the few simulation studies available, likely due to the ethical issues surrounding the exposure of research participants to a putatively threatening scenario. For this reason much of the research conducted on the weapon focus effect has made use of videos or slide shows. In one of the first such experiments, Loftus, Loftus and Messo (1987) had participants watch a video in which a young man approached the counter of a fast food restaurant, presented an object to the cashier, accepted money and left. In the control condition the man presented a cheque to the cashier whereas in the weapon condition the man presented a gun. Specialized equipment tracked the participant's gaze as they viewed the video to determine with what frequency (and for how long) they fixated upon the item of interest (the cheque or the gun). Relative to the control condition, participants in the weapon condition looked at the item the man was holding more frequently and for greater duration. Further, when tested for the details of the event, performance was better for the control condition relative to the weapon condition - with the exception that participants in the weapon condition were more likely to recall what object the man was holding (a gun).
Since the initial research conducted by Johnson and Scott (1976) and Loftus et al. (1987) others have demonstrated a similar effect using not weapons but unusual objects. For example, Pickel (1998) demonstrated an effect comparable to weapon focus using a video in which a man approached a cashier and presented a whole raw chicken or miniature Pillsbury Dough Boy instead of an expected item such as a wallet. From her finding, Pickel (1998) argued that the weapon focus arose from the unusual nature of the object in the relation to the context in which it was presented. The relative contributions of arousal and unusualness remains one of the primary theoretical issues in this literature, with some authors arguing for a contribution of both.
Another significant challenge to the weapon focus effect has been its ecological validity. Specifically, many theorists have argued that the effect is limited to the laboratory. These claims have been supported by the relative absence of applied evidence supporting the effect. Several reports have been published looking for evidence of a weapon focus effect using records of actual criminal events. According to the laboratory findings summarized above, the prediction had been that eyewitness memory would be worse for weapon crimes compared to non-weapon crimes. Many of the primary studies failed to support this prediction. Even so, a recent meta-analysis conducted by Fawcett et al. (2011) has demonstrated that when the data for all of the applied studies are combined, there is a small but reliable effect suggesting that weapon presence impairs actual eyewitness memory. This finding supports the laboratory studies conducted on this topic.