memory Flashcards
(22 cards)
SHORT TERM MEMORY
Ao1
CAPACITY
Jacobs measured digit span by asking participants to repeat in increasing numbers of digits until failure
Found:
Average digit span was 9.3
Average letter span was 7.3
Miller argued span of STM was 7 +-2 units
concluded chunking helps improve capacity(can recall five words/5 letters)
DURATION
Peterson and Peterson
– 24 undergrads, eight trials each
– each trial given a consonant trigram.
– count backwards from a three digit number told to stop (up to 18 seconds)
Found:
– After 3 seconds: 80% correct
– after 6 seconds: 50% correct
– After 18 seconds: less than 10%
Concluded:
- capacity of approximately 18 seconds
- If unable to rehearse, not passed to LTM
SHORT TERM MEMORY
Ao3
JACOBS
STRENGTH: findings confirmed in many later studies despite age, easy to replicate
LIMITATION: didn’t account for individual differences – better recall ability from STM depend depending on age and individual brain structure
MILLER - LIMITATIONS
–Cowan concluded capacity of STM only 4 chunks
– Simon suggested capacity shorter for larger chunks such as phrases
PETERSON AND PETERSON
STRENGTH: sometimes remeber mundane things e.g. phone numbers so gives some insight
LIMITATION: stimulus material artificial – consonant trigrams don’t reflect real life memory activities
LONG TERM MEMORY
Ao1
CAPACITY
Assumed to be infinite
DURATION
Bahrick - longitudinal study
– 392 American pps, age 17–74
required to identify schoolmates from yearbook either by free recall, or by matching photos to names
Findings:
Photo recognition: 90% within 15 years, 70% after 48 years.
Free recall: 60% within 15, 30% after 48
LONG TERM MEMORY
Ao3
BAHRICK
STRENGTH: used every day real life meaningful memories, reflect more real estimate.
LIMITATION: real life research = confounding variables not controlled, e.g. pps may have looked at the yearbook
CODING IN STM/LTM
Ao1
3 formats: visual, acoustic, semantic
Baddely study
PROCEDURE:
– Gave word lists to 4 groups: acoustically similar/similar, romantically similar/dissimilar
– Article either immediately/20 minutes
FINDINGS
– In STM , acoustically similar worse, suggesting STM coded acoustically
– In LTM, semantically similar worse, suggesting LTM coded semantically
CODING IN STM/LTM
Ao3
STRENGTH
– Identified two memory stores, idea of different coding stood test of time, supported by Tulving
LIMITATION
– Used artificial stimuli rather than meaningful material, wordless had no personal meaning cautious about generalising on memory tasks. Might be semantic even for STM.
MULTI STORE MODEL
Ao1
Atkinson and Shiffrin
SENSORY REGISTER
– environmental stimuli from all senses (ionic, echoic)
– capacity unlimited, constantly receiving information. duration = milliseconds.
attention sends to:
STM
– Duration less than 30 seconds
– 5 to 9 pieces of information, disappear if displaced, maintenance rehearsal
Prolonged rehearsal sends to
LTM
– potentially permanent, capacity unLimited
– semantically encoded
Can be recalled by retrieval to STM, remembered from STM e.g. for exam
Sensory register tested by Sperling
– saw grid of digits/letters for 50 ms
– asked to write down 12 items or one row
– whole grid = 42%, one = 75%
Shows info decay rapidly – if more input it cannot be retained
Glanzer and Cunitz
– Showed pps list of 20 words, asked to recall
– Primacy and recency effect (1st and last words = best)
Occurs because first words = rehearsed and transferred to LTM, last words fresh and still in STM
MULTI STORE MODEL
Ao3
STRENGTHS
– Case study of HM: after hippocampus removal STM remained intact, however LTM damaged as was unable to form new memories – couldn’t transfer info to LTM. Supports different stores
– Baddely found STM acoustic, LTM semantic – show separate and independent stores.
LIMITATIONS
– MSM states STM is unitary store. However, KF amnesia case study shows digits memory = poor when listened, but improved when read
– Research surrounding MSM use meaningless info e.g. digits/letters. in reality memories related to meaningful info e.g. faces, names etc – may not be valid for every day life
TYPES OF LONG-TERM MEMORY
Ao1
Tulving said MSM too simple; proposed 3 different LTM stores:
EPISODIC
– events from life, people/object/places/event. E.g. last b-day party.
– time stamped and relative
– single episode includes several elements
– conscious effort to recall
SEMANTIC
– Knowledge of the world, facts/concepts. E.g. meaning of words, taste of orange
– not time stamped, less personal
– Less vulnerable to distortion than episodic
– recalled consciously/deliberately
PROCEDURAL
– actions/skills e.g. driving, tying shoelaces.
– occurs through repetition and practice, without conscious awareness
Clive Wearing case study
– musician, who had Viral infection that damaged hippocampus
– Episodic memory damaged (forgot parts of his life and family members) but procedural fine (could still play piano) and semantic (still understood world)
TYPES OF LONG-TERM MEMORY
Ao3
STRENGTHS
– Evidence from case studies HM + Clive wearing. Show damaged episodic memories but intact semantic
– real world application, old age memory loss. Specific to episodic, Belleville devised intervention to target this
LIMITATIONS
– conflicting findings about brain areas, some studies find semantic = left and episodic = right PFC. Tulving found opposite, challenges neurophysical evidence.
– 2 or 3 types? Cohen and Squire argue declarative (consciously recalled) and non-declarative (automatic)
WORKING MEMORY MODEL
Ao1
Baddely and Hitch , addition to MSM, expanding on STM open (an active store)
CENTRAL EXECUTIVE
– supervisor role, monitors incoming data
– doesn’t store/process, allocates to 3 slave systems
1) PHONOLOGICAL LOOP
Auditory/acoustic information, preserves order.
– phonological store: words you hear?
– articulatory process: allows maintenance rehearsal, capacity two seconds worth of speech
2) VISUO SPATIAL SKETCHPAD
Limited capacity (3–4 objects)
– Visual cache: visual data
– inner scribe: arrangement of objects
3) EPISODIC BUFFER
Added later, temporary store integrating visual, spatial and verbal info.
Maintains time sequencing, recording events – capacity around 4 chunks
Links working memory to LTM
Dual task performance
– 2 visual tasks = hard
– 1 auditory 1 visual task = easier
working model explains this as there are two different stores
WORKING MEMORY MODEL
Ao3
STRENGTHS
- case study of KF supports (poor STM for verbal, but good when presented visually)
– Baddely conducted dual performance task, found 2 visual = difficult, visual + verbal = less difficulty. Supports existence of separate systems.
LIMITATIONS
– Central executive is vague and untestable, added later as model doesn’t make sense without. however, little understanding/evidence of CE.
– Braver conducted brain scans, tasks that involved CE. Found greater activity in left PFC, especially when task = harder. Scientific evidence supports concept.
EXPLANATIONS FOR FORGETTING: INFERENCE
Ao1
idea the one memory disrupts another.
Proactive: passed interferes with new
Retroactive: new interferes with past
McGeoch and Donald effects of similarity.
Tested retroactive interference.
PROCEDURE:
– Six groups pps learned word lists to 100% accuracy, then learn new list.
– second word lists varied in similarity too fast (synonyms, antonyms, numbers)
FINDINGS:
– Found more similar, the more interference
EXPLANATIONS FOR FORGETTING: INTERFERENCE.
Ao3
STRENGTHS:
- Real world example, rugby players who had played most games forgot more names of opponent teams (most interference)
– consistently demonstrated in lab studies, often have high control over extraneous variables. Confident that interference is valid explanation.
either?
– Tulving recreated McGeoch and Donald experiment, but with cues. Recall rose to 70% so it can cause temporary loss but can be overcome by cues.
LIMITATIONS
– only explains forgetting when memories are similar, doesn’t explain every day instances of forgetting
EXPLANATIONS FOR FORGETTING : RETRIEVAL FAILURE
Ao1
– Memories aren’t lost, just need to trigger to access.
– “Encoding specificity principle“ meaning Q must be present at time of learning and retrieval
CONTEXT DEPENDENT FORGETTING
One environment at learning is different to recall
Studies:
Gooden and Baddely
learn and recall word lists either online or underwater
Found significantly higher recall in matching conditions
Abernethy
Do a task in same/different rooms, by teacher/stranger
Student students in same room by teacher = best results
STATE DEPENDENT FORGETTING
Internal state at recall is different to learning
Studies:
Carter and Cassaday
Learn and recall word list on/off antihistamine drugs
Performance better in matching conditions.
Goodwin
Learn Andrew cool word lists when drunk/sober.
Sober = sober better
Drunk = drunk better
RETRIEVAL FAILURE
Ao3
STRENGTHS
– Huge amount of lab studies and real world evidence to support retrieval failure. Eysenck argued perhaps the main reason for forgetting
– Face validity: go into room but forget, original room remember. So can use principles in real world
LIMITATIONS
– Context effects not as strong in real life as in studies e.g. underwater versus on land is a big contrast.
– when underwater test replicated with recognition rather than recall, no effect. May only reply to recall not recognition.
EYEWITNESS TESTIMONY: MISLEADING INFORMATION
Ao1
LEADING QUESTIONS
Loftus and Palmer
– Participants watched car crash video and asked about speed of car when “ contacted/smashed/collided/bump/hit“
– Found main speed estimate
40.8 for smashed, 31.8 for contacted
response bias expl- doesn’t change memory, only way you respond
substitution expl - changes memory
in a follow up study, students asked a week later if they had seen any broken glass
‘smashed’ - 32% saw glass
‘ hit’ – 14% saw glass
No mention of speed – 12%
POST EVENT DISCUSSION
Gabbert
– pps so clips of crime from different perspectives, unique elements each
– Well then able to discuss with each other before being individually tested
FOUND:
71% mistakenly recorded aspects they haven’t seen, control was 0%
affects EWT due to
- memory contamination, combine misinformation with own memories
- Memory conformity, NSI/ISI
EYEWITNESS TESTIMONY: MISLEADING INFORMATION
Ao3
STRENGTHS
– Loftus interviewed pps about effectiveness of Disneyland adverts featuring Bugs Bunny. When asked about childhood experiences of Disneyland, 36% recalled meeting Bugs Bunny (not at Disneyland)
– good real world application, often called into court of law to explain to jury problems with EWT. Has led to improvement in the way police question witnesses.
LIMITATIONS
– Studies find participants recall central details better than peripheral ones, even when asked misleading questions. Central details not distorted, not predicted by substitution explanation.
– Much research comes from lab studies, demand characteristics, might guess answers to appear helpful
EYEWITNESS TESTIMONY: ANXIETY
Ao1
NEGATIVE IMPACT ON RECALL
Johnson and Scott
– pps believed we’re waiting for lab study, loud argument stage next door
Low Anxiety condition: greasy pen
High Anxiety: bloody knife
– asked to identify from 50 photos the perpetrator
Findings:
- 49% pen, 33% knife
TUNNEL THEORY - WEAPON FOCUS
POSITIVE IMPACT ON RECALL
*Yuille and Cutshall**
- 13 witnesses of Vancouver shooting
– Interviewed 4/5 months after, compared with original police interviews
Found high anxiety = 88% accurate
Low anxiety = 75%
CONTRADICTORY EVIDENCE - Yerkes-Dodson Law
lower arousal = poor recall
Optimum arousal = most alert
Higher arousal = stress/anxiety, impaired recall
EWT: ANXIETY
Ao3
STRENGTHS
– London dungeons attraction measured heart rate, asked to identify actor they met. High anxiety = 17%, Low anxiety = 75%. Supports anxiety negative
LIMITATIONS
– Pickel found eyewitness accuracy poorer in high unusualness condition (hand guard, scissors, wallet, chicken) contradict weapon focus, not anxiety but shock
– Anxiety very difficult to define and measure accurately, as many elements (cognitive, behavioural, emotional and physical) limits lab studies
IMPROVING EWT: THE COGNITIVE INTERVIEW
Ao1
Fisher and Geiselman
Issues with police interviews, developed cognitive interview based on psychological principles.
1) REPORT EVERYTHING
– Encourages reporting every detail, even irrelevant. Interconnected, might be a cue for something else.
2) REINSTATE THE CONTEXT
Encourages to mentally re-create the physical and psychological environment of incident. Could be a cue to help retrieval
3) REVERSE THE ORDER
Find alternative ways through timeline of incident, e.g. start from last thing you remember. Prevent reporting expectations/untruthful.
4) CHANGE PERSPECTIVE
Recall from other peoples perspectives, e.g. from witness/perpetrator
To disrupt effect of schema on recall
ENHANCED COGNITIVE INTERVIEW
Fisher developed additional elements to focus on social dynamics.
– interviewer knows went to establish eye contact/relinquish
– ideas e.g. reducing eyewitness anxiety, minimising distractions, open-ended questions
COGNITIVE INTERVIEW
Ao3
STRENGTHS
- in meta analysis, CI produced on average 41% more correct information than standard interview.
– study found that each individual technique alone produced more information than standard interview. Even parts better than none, shows effective.
LIMITATIONS
– Kohnken found also increases amount of inaccurate information
– Much more expensive and time-consuming to train and conduct, not practical for everyday use
