2a. One theory of forgetting in STM is displacement. Forgetting in terms of capacity (7+/- 2: Miller, 1956). Once STM is full, any new information pushes out or ‘displaces’ old information. Old information is therefore forgotten.
2b. A theory of forgetting in LTM is Interference. According to this explanation, similar memories interfere, disrupt and confuse each other. The more similar the information is the more likely interference will occur. There are two types of interference: Proactive (forward) interference: when old memories affect new memories e.g. calling a new boyfriend by an old boyfriend’s name. And Retroactive (backward interference) new memories affect old memories e.g. after driving a bus making mistakes when driving the car.
3. People are often able to recall events accurately and in detail, but eyewitness testimony (EWT) can be unreliable. Given the potentially serious consequences of inaccurate testimony, psychologists have carried out extensive research into the reliability of EWT, why people remember inaccurately and to find ways of improving their recall.
There are many studies which suggest that EWT is unreliable. Loftus has shown that Eyewitness testimony can be affected by the wording of questions and people’s willingness to make the information fit their schemas, which has serious implications for the judicial system since juries are often convinced solely on EWT. Loftus and Palmer (1974) showed the effect of the wording of questions by changing a single verb about the speed of a crashing car. Using the verb ‘smashed into’ led to a faster speed estimate (41mph) than use of the verb contacted (32mph). Loftus also shows evidence that the style of questioning can dramatically change witness testimonies. Eye witnesses asked about a car passing a barn were later more likely to say they saw a barn, when in fact the barn didn’t exist. Loftus believes that in the light of misleading information, the original memory is deleted and replaced by a new false memory.
Loftus (1979) also reported a study where participants either saw a man emerge from the lab holding a pen or another holding a paper knife covered in blood. The Participants were then asked to identify the person- those who had witnessed the man holding the pen identified the person 49% of the time, whereas those who witnessed the man with the bloodstained paper knife were only successful 33% of the time. Loftus concluded that the fear and anxiety caused by the sight of the weapon narrowed the focus of attention so the Participants could give very detailed accounts of the central details, but not to peripheral details. This finding had become known as ‘weapon focus’.
Criticisms of these studies is that they are very low in ecological validity in that they do not take place in real life situations and that the emotional effects associated with seeing incidents are likely to affect memory recall, but this cannot be replicated in a lab setting. Loftus has also been criticised for her method of testing recall. People are much more accurate if they are asked questions in a logical order and if they are not forced to answer if they are unsure. The cognitive interview method is based on four instructions: to recreate the context of the original incident, to report every detail, to recall the event in different orders and to change perspectives. People recalled many more items in the cognitive interview than the standard (41.15% to 29.4%) but the errors made were very slightly higher. Many studies have investigated the cognitive interview technique and have found that it provides more accurate information. So perhaps Loftus simply wasn’t asking the right questions.
Baddeley questions that EWT research is clear cut. He argues that research doesn’t prove the deleting of a memory, but interferes with its retrieval. This suggests that reconstructive questioning of witnesses is very important. Yuille & Cutshall (1986) used witnesses of real life crimes, in contrast to Loftus’ artificial experiments, and found memories to be accurate, detailed and not easily distorted, so EWT can be reliable. Foster et al (1994) did a study where one group was led to believe a robbery was a real event, and there testimonies would effect the trial, and the other group knew it was just a simulation. Recall was much higher for the first group, showing that knowing there will be consequences creates more accurate memories. In conclusion, EWT can be inaccurate if misleading information is given which is believable and not incredibly significant (e.g. saying a gun instead of a knife) and if witnesses have little reason to distrust it or question it (as in lab experiments).