Baddeley (1966): had to recall 4 different groups of words, straight after (STM) and then 20 mins after (LTM) to which they did worse with semantically similar words - coded semantically


Miller (1956): can hold 7 plus or minus 2 items, 5 letters as well as 5 words with chunking

Cowan (2001): capacity of STM is 4 chunks


Bahrick (1959): 392 students from Ohio to recognise names of people from a yearbook, and then accoustically the names. Within 15 years (90%), 48 years (70%) for photos and 60% for 15 and 30% for names: high external validity

Peterson and Peterson (1959) found that STM is 18 to 30 seconds but goes down the longer the retention interval: stimulus material is artificial and lacks external validity (constant syllable letters)

1 of 16

Multi-Store Model of Memory

Sensory Register: duration of material is half a second, only goes through with attention

STM: can only hold 5 items instead of 9. Maintenance rehearsal if we repeat it to ourselves

LTM: Prolonged rehearsal leads to this. 

Negatives: no evidence for sensory register, Craik and Watkins (1973) found that maintenance rehearsal keeps it in STM but elaborative is needed for LTM, Shallice and Warrington (1970) studied KF and found that he could recall semantically but not accoustically, MORE THAN ONE STM STORE

2 of 16

PET Scans | HM | CW

PET Scans

  • Inject patient with radioactive tracer that imitates glucose
  • Body processes it like so and tracer is found in areas of more activity
  • PET detects radiation, shows a map of activity
  • Combined with MRI and CT scans

Henry Molaison

  • Had procedural and semantic memory
  • No episodic memory (hippocampus removed for epilepsy)
  • Rehearsal allowed him to keep info in the STM, but never to the LTM

Clive Wearing

  • Exactly the same, had amnesia after a virus in 1985
3 of 16

Types of LTM

Episodic: to recall events, memories are "time stamped", requires physical effort. Semantic: to recall meaning, requires physical effort .Procedural: to do skills, requires no physical effort

Clinical Evidence: HM/CW had impaired episodic memory, semantic ones unaffected. Procedural memories intact (CW could still play a piano) and HM could remember what a dog was but couldn't remember stroking one

Neuroimaging evidence: Tulving (1984) made people do memory tasks while PET scanning them, and found episodic/semantic were from the  prefrontal coretex. Left one was semantic, right was episodic

Real life applications: Belleville (2006) found episodic memories in old people could be improved 

Cohen and Squire (1980): episodic and semantic memories are stored as one - declarative and non-declarative memories

Lack of control with all sorts of variables

4 of 16

The Working-Memory Model

Central Executive: attentional processes, monitors incoming data and allocates slave systems to tasks - very limited processing capacity

Phonological Loop: slave system that deals with auditory information and preserves order

  • Phonological store: stores words you hear
  • Articulatory process: allows maintenance rehearsal (repeating sounds/words in a loop to keep it in the working memory when needed) - capacity is 2 seconds of what you can say

Visuo-Spatial Sketchpad: stores visual or spatial information and has limited capacity of 3 or 4 objects (Baddeley 2003)

  • Visual cache: stores visual data
  • Inner scribe: records arrangement/position of data

- Logie (1995)

Episodic Buffer: time sequencing of events, storage component of CE and limited capacity of 4 chunks (Baddeley 2012, added this in 2000), limked to WMM and perception

5 of 16

Working Memory Model Evaluation

Shallice and Warrington (1970): studied KF and found he could process visual info but had difficulty with sounds and yet could recall letters and digits. Suggests his PL had been damaged. Supports the existence of a separate visual and accoustic store (evidence from brain damaged patients might not be reliable as they are unique cases) - clinical evidence

Baddeley et al (1975) got people doing 2 visal tasks, tracking a light andn describing the letter F. This is increased becauswe they compete for the same slave system. Supports VSS that processes visual input - dual task performance

Central executive is not fully explained (Baddeley 2003) and there has to be more than just "attention" - WMM not fully explained

Baddeley et al (1975) found word length affect (harder to remember longer words) because in the articulatory process there is 2 secs worth of what you can say. This disappears if you give a articulatory suppression task. 

Braver et al (1997) got people to do tasks involving the CE whilst doing a brain scan; found lots of activity in the prefrontal coretex (demands on the CE increase, harder work to fulfil function)

6 of 16

Declarative/Non-Declarative Memory

Semantic and Episodic Memories (Declarative) and Procedural Memories (Non-Declarative)

Established by Cohen and Squire in 1980

7 of 16


  • Proactive: older memory interferes with a new one
  • Retroactive: newer memory interferes with an older one

Forgetting happens when one memory blocks another, causing one/both of them to be distorted/forgotten. Interference - makes it harder for memories to be located

McGeoch and McDonald (1931): "studied RI - similarity between two sets of materials" - participants had to remember them with 100% accuracy

  • Synonyms
  • Antonyms
  • Words unrelated to the originals
  • Nonsense syllables
  • 3 digit numbers
  • No new list - they rested

Performance depended on the nature of the second list. The most similar produced the worst recall - interference is the strongest when the memories are similar

8 of 16

Interference - Evaluation

Positives: 1000s of experiments carried out into forgetting, shows RI and PI are likely for interference (lab studies control effects of irrelevant influences), Baddeley and Hitch (1977): "rugby players to remember names of teams they'd played - accurate dependent on no of games played in meantime, not the time scale)

Negatives: stimulus material was artificial (consonant syllables) so it makes interference more likely because it isn't real, majority of lasb experiments maximised for interference and may not reflect the real world

Tulving and Psotska (1971)

Participants had 5 lists of 24 words - organised into 6 categories. Categories presumed to be obvious. 70% recall but this fell when they were given another list. Cued recall test at the end, given names of categories - 70%

9 of 16

Retrieval Failure

Encoding Specificity Principle: Tulving (1983): "if a cue is to help us, it has to be present at coding and at retrieval. If there's different cues, they'll forget. 

Context-Dependent Forgetting: Godden and Baddeley (1975): "Asked divers to learn/recall visa versa on land and underwater (there were 4 trials). When contexts of learning/recall matched, it was 40% higher than non-matching conditions because of external cues

State-Dependent Forgetting: Carter and Cassaday (1998): "Gave anti-histamine drugs to participants which made them drowsy (internal physiological state) where they had to learn/recall on the drug and off the drug - mismatch between internal state at learning/recall so when cues are absent, there is more forgetting

Positives: Eynseck (2010): "supporting evidence means better validity"

Negatives: Baddeley (1997): "context effects are not actually that strong because environments are different enough in reality". (1980): "might relate to the kind of memory being tested - replicated it with recognition but found there was no CDE effect"

10 of 16

Leading Questions

A question that is worded in such a way that it insinuates an answer

Loftus and Palmer (1974): "participants to watcch films of a car accident and then asked them..

"How fast was the the car's going when they ___ into each other" - contacted/collided/smashed/bumped/hit

Contacted = 31.8mph

Smashed = 40.5mph

  • Response-bias explanation: the wording of the question has an effect on how they answer
  • Substitution explanation: the wording of the question alters the memory (some reported broken glass) - LOFTUS AND PALMER

Research has big se in the real world, Loftus (1975) - they can lead to inaccurate EWT

Tasks are artificial, lack the stress of a real incident (some evidence emotions affect memory) and could tell us very little about its effect on EWT

11 of 16

Post-Event Discussion

When co-witnesses dicuss a crime with each other

Gabbert (2003): "got particvipants (in pairs) to watch a video of the same crime from different POVs (one person could see the title of a book being carried by a young woman) then they had to discuss before recall

71% of participants mistaken recalled aspects of the crime they didn't see

In a control group, mistaken recall was 0%

Witnesses go along with each other for social approval/they believe they are wrong - memory conformity

Explains inaccuracy of EWT

Tasks are artificial and lack stress of real incident (emotion affect)

Foster et al (1994): "EWT in real life has more important consequences, lab studies don't"

Zaragosa and McCloskey (1989): "Demand characteristics"

12 of 16

Fight Or Flight


Yuille and Cutshall (1986)

"Real-life shooting in a gun-shop in Vancouver. Shop owner shot a thief dead. Interviews 4-5 months after, compared to original statements. Asked to rate how stressed they were using a 7 point scale and if they were suffering emotionally e.g. drowsiness"

Very accurate, little change. 88% of highest stress levels were accurate  compared to 75% for less stressed

Field studies lack control over extraneous variables that may affect recall and its impossible to assess

Ethical issues in real life studies might cause harm

13 of 16

Weapons Focus

Johnson and Scott (1976)

"Participants think they were taking part of a lab study. Heard argument in next room. In low anxiety, man walked through with a pen and grease on hands. Same argument in high anxiety, but there was a paper knife covered in blood

Picked out the man from a set of 50 photos. 49% identified the man with the pen, 33% with the knife. Tunnel theory of memory says that witnesses memory goes to the weapon because of anxiety

Pickel (1998): hairdressing salon video - scissors, pistol, wallet, and a chicken. EWT poorer in high unusualness. Weapon focus due to unusualness rather than anxiety

Demand characteristics

14 of 16

Cognitive Interview

REPORT EVERYTHING: encouraged to report everything regardless of significance, might trigger other memories

REINSTATE THE CONTEXT: try to imagine the crime scene, context-dependent forgetting triggering other memories

REVERSE THE ORDER: prevents people from reporting their expectations of how it happened and prevents dishonesty

CHANGE PERSPECTIVE: recalling from other people's views, disrupts scheme on recall. Scheme for particular scenes generates expectations rather than what did

ENHANCED COGNITIVE INTERVIEW: Fisher et al (1987): "establish eye contact, relinquish it, reduce eyewitness anxiety and open ended questions"

YERKES-DODSON LAW: Deffenbacher (1983): "applied to EWT and found optimal recall is midpoint of low and high anxiety" - too simplistic and only assumes poor performance is due to physiological arousal

15 of 16

Cognitive Interview Evaluation

Kebbell and Wagstaff (1986): "many police forces have not been able to train people in CI techniques because it takes a lot of time"

He found that there was a 81% of correct information but also 61% of incorrect information

Milne and Bull (2002): "combination of report everything and context reinstatement produced better recall - 2 techniques should be used even if the others aren't"

Kohnken et al (1999): "meta-analysis of 50 studies, found ECI provided more correct information which shows there are real practical benefits"

Variations of CI are used and the same is true in real life

16 of 16


No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Memory resources »