Key Studies: “Weapon focus” and its effects on eye-witness memories (Loftus, 1987)

Travis DixonCognitive Psychology, Criminology, Key Studies, Studies and Theories

Weapons focus could explain poor eye-witness memories because our attention is drawn to the weapon and we don't pay attention to other important details.

From decades of research we know that memory is not a passive cognitive process, but it is an active reconstructive one. As Elizabeth Loftus says, memory is not like a tape recorder that records things accurately and plays it back for us, but it’s more like a wikipedia page that anyone can go in and change. Loftus should know as she’s been studying memory since the 1970s and is the most prominent psychologist working in this field. 

Loftus’s studies and others like them have had an enormous influence in the American legal system (and in other countries, too, I’m sure) whereby if someone is identified by an eye-witness as the guilty person they are not automatically assumed to be guilty. This is because hundreds of accused criminals (mostly men) have been released from jail because DNA evidence shows they did not commit the crime they were in jail for. In many cases they were imprisoned because an eye-witness got it wrong.


These people started “The Innocence Project” – a project with the mission of releasing innocent people from prison, many of whom are there because of faulty eye-witness testimony. To date, over 300 people have been proven innocent and released. (wikicommons)

So how can eye-witnesses be wrong so often? This is something that researchers have been studying for years. One possible explanation is “weapon focus” – the cognitive phenomenon whereby if a weapon is present our attention is drawn to the weapon and away from other details of the scene, which means our memory of other details (like the face of the person holding the weapon) is reduced.

Key Study #1: The original weapon focus study (Johnson and Scott, 1976¹)

One critique of Loftus’ original research involving car crashes was that she was testing memory in situations where there was little emotional impact or stress involved. Studies into the effect that the presence of a weapon has on memory, addressed this limitation.

The participants in this study were asked to wait in a waiting room. There were two conditions in this experiment:

  • Weapon condition
  • No weapon condition

No weapon condition: Participants overheard a conversation about some equipment not working and then a man walked through the room holding a “grease pen”*.

Weapon condition: While participants were waiting, they heard a loud argument, crashing chairs and breaking glass next door. A man then walked through the room holding a paper-cutter knife with bloody hands.


This first study showed that the ability to correctly identify the criminal was reduced if they were holding a paper cutter knife like this one (

Afterwards, the participants were asked to identify the man from 50 photographs. In Condition A, 49% of the participants correctly identified the man. In Condition B, 33% of the participants could identify the man.

It seems from this study that our memory of a person is impaired if they are holding a weapon. However, Loftus and others noted several problems with the design of this experiment and so conducted further research into the phenomenon of weapon focus.

Can you spot the problems with Johnson and Scott’s experiment? What is the IV in this experiment and  was it controlled? i.e. were both conditions the same except for that one change in the IV?

Key Study #2: A controlled experiment on weapon focus (Loftus, Loftus and Messo, 1987) 

The aim of this study was to investigate the phenomenon of weapon focus in more tightly controlled experimental designs than were originally carried out.

In one experiment, 36 students experienced one of two different experimental conditions: weapon condition or control condition. The participants viewed a scenario by watching a slideshow (a series of 18 images shown for 1.5seconds each). The scenario was the same in both conditions – it’s a fast food taco restaurant and customers are standing in line to order. The second man in line then approaches the cashier. What he does next is the only difference in the two conditions:

  • No weapon (control) condition: He hands her a cheque and she hands him some money
  • Weapons condition: He pulls a gun on her and she hands him some money

The weapon/check appeared in four of the slides. They then tested their memory of the event. They also measured weapon focus by seeing where the participants’ eyes were focused while watching the scene (using video taped footage of the participant viewing the scene).

The Man Hid The Gun Behind Their Backs, Robbery, Crime, Kidnappi

The presence of a gun in the slide series was the only difference in the two conditions and it had a negative effect on memory, possibly because of weapon focus.

The results:

General memory: there was not much difference in general memory of the scene (50% in the control compared with 45% in the weapon condition)

Identifying the man: participants in the control condition accurately identified the man in a line-up of possible people 39% of the time. This is compared to the much lower rate of 11% in the weapon condition.

-Weapon Focus: Those in the weapon condition focused on the item (the gun) more than those in the no weapon condition (check). The rate was 3.7 times on average they looked at the gun, compared to 2.4 times looking at the check.

A second experiment using 80 participants gathered similar results – when watching the same slide series the control condition could accurately identify the man from a line-up of 12 people 35% of the time, compared with only 15% of the time in the weapon condition.

These results provide more solid evidence for the existence of a weapon focus when witnessing a crime – people are more likely to focus on a weapon than they are on a harmless object. This could also have negative effects on their reliability of memory as they are not focusing on the person but rather on the object.

Critical Thinking Considerations

  • Why might people be less able to correctly identify the man if he was holding a weapon? (Analysis)
  • What do these studies suggest about the reliability of memory? (Application)
  • Do these studies suggest that memory is reconstructive?  (Analysis)
  • Are these results generalizable? For example, is population validity an issue – both of Lotus et al’s 1987 experiments used students as participants (18-31years old). Would we expect the same results for everyone? Could we expect some people to be less affected by weapons focus who might have a high chance of witnessing a crime?
  • These results are also based on viewing a series of still images. Does this reflect what happens in real-life? Why might it make a difference to the validity of the results?
¹As cited in Loftus et al. 1987.
*This is probably a grease gun which is used for adding grease to a machine, or a grease pencil (which just looks like an ordinary pencil which is why I think it was probably a grease gun as it makes sense for them to use this as it’s an unusual item).


Loftus, Elizabeth & Loftus, Geoffrey & Messo, Jane. (1987). Some facts about “weapon focus.”. Law and Human Behavior. 11. 55. 10.1007/BF01044839.

(Full study)