Memory (AQA AS-Level Psychology)

Describe the Multi-Store Model

Atkinson and Shiffrin 1968 MSM.

Sensory register (modality-specific coding, very high capacity but untestable, duration less than ½ a second) attention moves information from SR to STM.

Short term memory (acoustic coding Baddeley 1966, capacity 7 +-2 Miller, Cowen chunks, Simon size of chunks, 18-30 seconds duration Peterson and Peterson). With rehearsal information can move from STM to LTM.

Long term memory (semantic coding Baddeley 1966, potentially unlimited capacity, duration a lifetime, or at least 48 years Bahrick 1975).

Evaluate the Multi-Store Model

- MSM is machine reductionistic, using a computer model oversimplifies the complexity of the human brain. Ignores other factors affecting memory e.g IQ or brain damage.

– All research is concluded through inferences, memory is not observable, correlational data (which does not mean causation).

+ Lots of supporting evidence e.g Peterson & Peterson (trigrams) or Bahrick (STM acoustic, LTM semantic, word lists). High control lab studies BUT
– Supporting research has low mundane realism and is subject to demand characteristics

Describe the Working Memory Model

Baddeley and Hitch 1974 WMW.

Central executive focuses/monitors/coordinates the working memory, allocating to slave systems.
Phonological loop processes auditory information – phonological store is the inner-ear and articulatory process is the inner verbal rehearsal.
The visuo-spatial sketchpad stores visual information – visual cache stores visual data and the inner scribe handles spatial data.
Episodic buffer sequences information from both the PL and VSS.

Evaluate the Working Memory Model

- Case study EVF provides opposing evidence for the CE, brain damage, could reason but not make decisions, suggesting the CE is more complicated than previously thought alluding to sub-components, causing us to question the accuracy of the rest of the WMM.

+ Baddeley’s dual task studies 1976, when both tasks use the same slave system, their ability to multitask is impaired. However if dual tasks are using different slave systems, they can complete tasks simultaneously. Provides evidence for multiple components of WMM.

– All research into WMM is in a lab setting so high control, but uses artificial tasks, so have low mundane realism and don’t reflect behaviours of everyday lives, so low ecological validity.

Describe interference as an explanation for forgetting

Interference is when two pieces of information cause confusion due to the similarity between them. Retroactive interference is when new information inhibits the recall of old information. Proactive interference is when old information inhibits recall new information.

Evaluate interference as an explanation for forgetting

+ McGeoch and McDonald, participants learned a word list to 100% accuracy, then learned another word list. They found that the more similar the second list was to the first (synonyms), the worse participants did in recall.

- Tulving and Psotka found retrieval failure to be a better explanation for forgetting because when given cues, participants recalled a word list to 70% accuracy regardless of how many lists they had learned in between.

- Many of the studies involve learning word lists, artificial task, lack ecological validity.

Describe retrieval failure as an explanation for forgetting

Retrieval failure is when you cannot access information due to the lack of cues to trigger the memory. The encoding-specificity principle was proposed by Tulving, it stated cues present at encoding must also be present at retrieval. These cues can either be context-dependant or state-dependant.

Evaluate retrieval failure as an explanation for forgetting

+ Godden and Baddeley 1975 used divers to investigate whether memory was better for words learned and recalled in the same or different environments (on land or under water). They found better accuracy if recalling in the same environment as learned = context cues.

+ Goodwin et al. 1969 taught male participants a set of words either drunk or sober. They were asked to recall the words 24 hours later when either drunk or sober again. They found participants showed better recall when in the same state as learned the words = state cues.

- Interference offers an alternate explanation for forgetting.

- Many of the studies involve learning word lists, artificial task, lack ecological validity

Describe how post-even discussion affects the accuracy of eyewitness testimony

Post-event discussion involves eyewitnesses discussing the event after it occurs.

Gabbert 2003 – pairs of participants each watched a film clip of the same crime but from different perspectives. They were allowed to discuss together before answering a questionnaire on their recall of the crime.
71% of Ps who had post-event discussed, claimed to have recalled information they did not see in their perspective of video.
60% said the girl in the video was guilty despite them not seeing her commit the crime.

Evaluate research into post-event discussion

- Gabbert 2003 lacks ecological validity, the participants know they were taking part in an experiment, meaning they would have been paying close attention to the crime, which is not usually the case with real life crimes.

Describe how leading questions affect the accuracy of eyewitness testimony

Leading questions are those that suggest a certain answer in the way they are worded, leading to the eyewitness to recall events incorrectly.

Loftus and Palmer 1974 – 45 student participants were shown videos of car accidents and given a questionnaire to complete. One of the questions used different verbs, it was a leading question: “How fast was the car travelling when it ___ the other car?”. The verb options were hit, contacted, smashed, collided, bumped.
“Contacted” had a mean speed of 31.8mph.
“Smashed” gave a mean speed of 40.8mph.

Evaluate research into leading questions

+ Loftus and Palmer’s research has led to real life application, ensuring innocent people are not wrongly convicted of crimes due to poor recall of an eyewitness.

+ Led to the development of the cognitive interview (Geiselman 1985) to improve witness’ memory retrieval.

Describe how anxiety affects the accuracy of eyewitness testimony

Johnson and Scott in 1976 investigated weapon focus. Participants were told they were taking part in a lab study and were sat in a waiting area beforehand, this participant either witnessed a man walk out of a room holding either a greasy pen or a knife covered in blood. Participants were later asked to ID him from a line up of photos.
Low-anxiety group 49% accurate.
High-anxiety group 33% accurate.
Researchers concluded anxiety/weapon focus steals attention from other details of the event and therefore decreases accuracy of EWT.

Christianson and Hubinette 1993 asked people who had witnessed a real bank robbery questions about the crime. They found people closer to the action (victims) had a more accurate recollection of the event than bystanders. Suggesting high levels of anxiety increase EWT.

Yerkes-Dodson Law explains the contradictory findings of researchers, it argues anxiety increases accuracy of EWT up to a point, but when the anxiety gets too great, performance decreases.

Evaluate research into how anxiety affects eyewitness testimony

- Johnson and Scott’s research has ethical issues, causing participants high levels of anxiety.

+ Christianson and Hubinette used a natural study, so it has high ecological validity.

Describe the cognitive interview

The Cognitive Interview was proposed by Geiselman in 1985 to improve the effectiveness of police interviews.

Composed of four components: Context reinstatement (to provide contextual and emotional cues to discourage retrieval failure), report everything (seemingly irrelevant details may act as a cue for a more important memory), reverse order (disrupts schemas) and changed perspective (also disrupts schemas).

Evaluate the cognitive interview

+ Kohnken 1999 conducted a meta-analysis of 53 studies comparing CI to standard interview. Showed a 34% increase in correct information for the CI BUT
– found an 81% increase in incorrect information.

- Kebbell and Wagstaff 1966 report the CI is impractical and takes too long. It also requires specialist training which is not a good use of police funding. So police use some aspects but not all. (Thames Valley Police)

+ May aid elderly witnesses more. Mello and Fisher 1996 compared younger and elderly participants responses to the CI verses the standard interview and found it was more effective for older people.