Memory

Sensory Register
Coding: sensory (smell, touch, sight, sound)
Duration: short
Capacity: very large

Short-term memory
Coding: Mainly acoustically, also visually
Duration: 30 seconds
Capacity: 7+-2 (5-9) Improved by chunking e.g SOSABCFBI

Long-term memory
Coding: Semantically (meaning)
Duration: Lifetine
Capacity: Unlimited

• Info transfers from SR to STM through re-coding (attention)
• Info is lost through decay
• Info transferred from STM to LTM through rehearsal

+ Research evidence for STM and LTM coding. BADDELEY. 75 ppts. List A- acoustically similar, B- acoustically dissimilar, C-semantically similar, D-semantically dissimilar. Found for STM list A they did worst, For LTM list C did worst, as that's how they're encoded.

+ Research evidence for LTM. BAHRICK. 700 ppts between 17-74. Last 15 years recalled 90% ex-school friend names and faces. Last 48 years: 80% names and 70% faces. So very long.

+ Case study evidence.  Clive Wearing caught virus which affected turning STM into LTM. But could still play piano, suggesting seperate stores.

+ Serial postion effect, primacy (LTM- words rehearsed) and recency (STM). Stop seperate stores. GLANZER AND CUNITZ

- Model is oversimplified. Suggests single types of STM and LTM, however research suggests several types.

- Focuses too much on structure (3 stores) and not enough on processes.

1) Central Executive (CE)
Filters information and directs to the seperate slave systems. It proccesses all sensory info. 

2) Slave Systems:
Phonoligical Loop (auditory info)
'Inner ear'- stores words recently heard
'Inner voice'- sub-vocal repetition and speech production

Visuo-spatial sketchpad (visual info)
'Inner scribe' - spatial information
'Inner cashe'- form and colour

Episodic Buffer (added in 2000)
Temporary store of info from CE, VSS and PL. 

+ Research evidence. BADDELEY. Participants asked to track moving light and classify angles in an 'F' and found ppts found it difficult because both use VSS (Inner scribe- spatial info) and (Inner cashe- light)

+ PET scan evidence. Have shown different areas of the brain activated during different tasks, suggesting seperate stores.

- Lack of evidence for Central Executive. It can be explained as controlling attention rather than a memory store. 

Explicit (Concious)

EPISODIC (episodes)- memories experienced e.g 18th Birthday. Emotions play a part in remembering. 
HIPPOCAMPUS

SEMANTIC (meaning)- general knowledge about world e.g How many legs cats have. Linked to Epsisodic memory as knowledge learnt from experience. 
FRONTAL AND TEMPORAL LOBES

Implicit (Unconcious)

PRODEDURAL (procedure)- performance of actions e.g how to use a knife and fork. 
NEOCORTEX

EPISODIC AND SEMANTIC

+ Research evidence. Tulving. 6 volunteers injected with liquid gold and scanned to detect it. Closed eyes and asked to think about Episodic and Semantic memories. Golden injected 60 secs after memorising. Found differences in blood flow patterns suggetsing different parts of the brain.

- Difference between episodic and semantic systems is unclear. They overlap. 

PROCEDURAL

+ Case study evidence. H.M Epilepsy meant destruction of hippocampus and temporal lobes. Unable to form explicit (concious) memories he still could form new procedural memories.

- Case studies with brain damage to procedural memory are very rare and cases are specific

Information in LTM getting confused with other information

Proactive Interference
Old information disrupting new information e.g recalling old phone number not new

Retroactive Interference
New information disrupting old information e.g calling an old friend Judy (new friend) instead of Julie.

+ Research evidence. Schmidt
Asked 700 former students between 11-79 to recall street names. Found those who had moved more confused more street names as Retroactive interference (new information affecting old). 
- extraneous variables e.g playing on the streets affects memory

- Research tends to be laboratory based. Lacks mundane realism. 

- Interference Theory only explains forgetting when there's two sets of info. 

CDF= info is in LTM but cannot be accessed. 

Context Dependent Failure
external environment is different from how it was coded e.g learning info in the classroom and recalling it in the exam hall

GODDEN AND BADDELEY- asked diver ppts to learn info out of water and recall underwater. Found recall was worse when conditions were different. Supporting content dependent failure.

State Dependent Failure
internal environment is different from how it was coded e.g learning info while sober and having to recall when drunk

OVERTON- asked ppts to learn information when drunk and recall when sober. Found recall was worse when internal conditions were different. Supporting state-dependent failure

+ Highly controlled studies making replicable but low ecological validity. 

+ Real world applications. Witnesses to crimes taken back to where it occurred.

EWT= evidence given to police by someone who has witnessed a crime. 

Schemas- affect EWT as thy produce biases and 'filling in the gaps' leading to false comments.

3 Stages:
1) Acquisition (encoding)
2) Storage (retention)
3) Retrieval (retreval)

Effect of misleading information:
Misleading questions- 'What colour were his eyes?' 
Post-event discussion- misleading info added after an event

LOFTUS AND PALMER
HSBCC (hit, smashed, bumped, collided, contacted)
45 university students shown clips of car crashes and asked how fast they were going with verb. Found Smashed= 40mph and contacted= 31mph. showing misleading questions affect memory.
- artificial task (watching videos) lacks mundane realism, unethical

EWT research is very lab based with little emotion, real life crimes produce anxiety.

1) Weapon focus- attention paid to weapon rather than peripheral details
It may divert attention away, so worse EWT

LOFTUS- ppts asked to sit in waiting room. 1) man emerged after discussion with a pen and greasy hand. 2) argument, man emerged holding bloody paperknife. Found only 33% successfully identified man KNIFE and 49% identified man PEN. So weapon focus.

CHRISTIANSON & HUBINETTE- 110 witnesses to 22 bank robberies. Onlookers and workers threatened. Those who had been threatened had more accurate recall. AGAINST.

2) Inverted U-Hypothesis- relationship between anxiety and recall. curved graph. There is an optimal point.

YUILLE & CUTSHALL- witnesses to shooting in Canada, those with high anxiety had worse recall. But very high = accurate. Against IUH. But later found they were closer to event.

+ Studies on real life witnesses e.g Yuille and Cutshall have high ecological validity. BUT ethical isses. Psychological harm.

- Incosistent findings. Lab studies e.g Loftus, found anxiety (weapon focus) reduces accurate recall. But real life research e.g Yuille and Cutshall & Christianson and Hubinette found accuracy increased with anxiety. 

- Individual differences may affect EWT e.g age, gender

4 techniques:
Change of narrative order- recall the event in a different chronological order e.g end to beginning. Reduces prior knowledge of schemas.
Mental reinstatement of context- recall environmental and emotional aspects of the event. e.g the weather was sunny. 
Change of perspective- report from a different point of view e.g the criminal
Report everything- every detail of the crime is reported, even if little evidence. May trigger further details.
CI- long time to recall, interviewers do not interrupt, open questions
Standard Police Interview- brief, closed and direct questions

+ Reserach suggests effective. Geiselman found the CI produced more accurate memories than the SPI. Showed violent crime to ppts after 48 hrs interviewed. CI= 41.5 correct answers and SPI= 29.4 correct items

- CI is time consuming 

- Hard to work out which techniques are most successful. as so many 

Trust building between interviewer and witness

Often used with vulnerable people e.g children/ mental health issues

• Interviewer must not distract witness
• ask open-ended questions
• get witnesses to speak slowly
• encourage witnesses not to guess, so no false info
• reduce anxiety

Can't use CI as children wouldn't be able to fulfill 'change of narrative order' or 'change of perspective'

+ Found to be more accurate. BUT also more incorrect info

- Requires a trained professional and is time consuming

- Produces vast amounts of info which is hard to filter