Memory

multi-store model of memory date and people

atkinson and shiffrin 1968

MSM outline

• info flows through a number of separate unitary storage systems in a linear fashion

• 3 main storage systems: coding capacity duration

• info from SR enters STM via paying attention

• information can remain in STM by maintenance rehearsal

• prolonged rehearsal transfers it to LTM

• info lost via decay or displacement

memory

liners flow of information through different stores

sensory register

• first store which holds the sensory information is relieved through all the senses for a very brief period

• e.g brief retention of an image after you’ve looked away

• can be iconic sensory memory (visual), echoic (sound) and haptic (touch) memory

• capacity is large but very brief 

iconic sensory memory example

type of sensory register

brief retention of an image after you’ve looked away e.g still being able to see the shape of a firework for a fraction of a second after it dissapears 

short term memory

• temporary store for holding small amounts of information for brief periods of time and disappears unless rehearsed 

• coding is acoustic

• capacity is limited to 7 ±

• 2 individual items

• duration is 18-30s

acoustic coding

changed into sound to be stored

long term memory

• memory for past events that can last the lifetime of a person

• capacity is unlimited

• coding is semantic

• capactiy is unlimited and duration is lifelong

semantic coding

according to the meaning of the information

roles of rehersall

maintenance and prolonged rehearsal

maintenance rehearsal

• involves repeating information again and again to keep it in short term memory (STM)

• inneffective at transferring to long term memory (LTM)

prolonged rehearsal

• deeper form by giving something more meaning and linking it with prior knowledge

• more likely to transfer info from STM to LTM

research supporting MSM AO23

• Murdock (1962) found that if participants were allowed to immediately recall a list of one syllable words they’d been asked to remember, they could remember the words at the beginning and the end of the lists, but not the words in the middle. This is known as the primacy-recency effect (first and last).

• This supports the MSM because words at the beginning of the list are rehearsed and therefore transferred to LTM, but the words in the middle arequickly displaced and not rehearsed enough to enter LTM. Whereas the words at the end of the list are remembered because they remain in STM.

• These findings support the distinction of STM and LTM and the role of rehearsal in passing information from STM to LTM

primary recency effect

remembering first and last words on a list but not the middle words

research criticsing MSM for failing to explain transfer to LTM without prolonged rehearsal- name, date, method

Craik and Tulving 1975

• gave participants list of nouns and asked question that involeved shallow or deep processing 

• remembered more words in task involving deep processing

MSM fails to explain how info can be transferred to LTM without prolonged rehearsal AO3

• Craik and Tulving (1975) gave participants a list of nouns (e.g. shark) and asked a question that involved shallow or deep processing i.e. asking whether the word was printed in capital letters (shallow) or asked whether the word fitted in a sentence (deep). The participants remembered more words in the task involving deep processing rather than shallow processing.

• This suggests depth of processing, rather than amount of rehearsal, can play a significant role in transferring information to LTM.

• This challenges the MSM by showing meaningful processing can aid recall rather than maintenance rehearsal (verbal repeating). Casting doubt over some of the assumptions of the multi-store model.

2 researchs saying MSM shouldn’t be considered unitary stores- names, dates

Blakemore’s Clive Wearing case 1988, KF case study 1974

Blakemore study: Clive Wearing

1988,

• after virus damaged hippocampus, he had very little long time memory

• yet he could still remember skills like piano, music reading, writing in a diary

• LTM not unitary

• demonstrated LTM may store and process episodic and procedural differently

KF case study

• damaged his short term memory for information spoken to him

• but could retain information presented visually

• suggests STM not unitary

• suggests stm has separate visual and acoustic components

evidence to suggest that both the STM store and the LTM store should not be considered unitary stores AO3

• Clive Wearing case study by Blakemore (1988) after a virus caused damage to the hippocampus, Clive Wearing had very little long-term memory for events that had happened in his life however, he could still remember skills such as playing piano, reading music and writing in a diary, suggesting the LTM is not unitary.

• KF case study by Shallice and Warrington (1974) KF damaged his shortterm memory for information spoken to him, but could retain information presented visually, suggesting the STM is not unitary.

• Blakemore’s study demonstrates that the LTM may store and process episodic (memory for events) and procedural (skills) differently. The KF case study suggests the STM has separate ‘visual’ and acoustic’ components. Therefore, the MSM is unlikely to be an accurate model. (newer models of memory have non unitary stores to reflect this evidence).

limitations of MSM oversimplified AO3

• oversimplified: doesn’t account for the complexity of memory processes like emotion, context, coding

who researched sensory register and date

Sperling 1960

sperling sensory register aim

to investigate the capacity and duration of the sensory register, particularly iconic memory

sperling sensory register method

• participants shown a grid of 12 letters in 3 rows for 1/20th of a second

• in whole report condition participants asked to recall from the whole grid

• in partial report condition participants heard high, medium, low pitched tone immediately after grid disappeared, each tone corresponding with top, middle, bottom row and asked to only recall that on

sperling sensory register findings

whole report condition, participants could only recall 4-5 letters

partial report condition, participants could recall 3 or 4 from cued row, suggesting they had access to all 12 letters briefly

sperling sensory register conclusion

sensory register has a very large capacity but iconic duration 1/2second, echoic 3seconds

information decays rapidly unless attention is paid to it and it moves into short term memory

research for STM coding and date

Baddeley 1966

Baddeley STM coding aims

to investigate how information is coded in STM, whether it is coded acoustically or semantically (by meaning)

Baddeley STM coding method

• participants divided into 4 groups (independent group design) and each group given different lists of words to remember

• acoustically similar, dissimilar

• semantically similar, dissimilar

• asked to recall the words immediately (testing STM)

Baddeley STM coding findings

• participants more difficulty recalling acoustically similar words than acoustically dissimilar ones when tested immediately

• no significant difference in recalls for semantically similar vs dissimilar words in STM condition

Baddeley STM coding conclusion

STM primarily coded acoustically as similar sounding words caused more confusion

means that acoustic coding is the preferred way of trying to store information in STM

supports idea format of coding in STM is based on sound not meaning and having to recall immediately

STM capacity research and date

Jacob’s 1887

Jacob’s STM capacity aim

to investigate capacity of short term memory specifically how many digits or letters people can hold in STM

Jacob’s STM capacity method

• participants presented with a sequence of digits or letters increasing one item at a time, digit span task

• asked to recall the items in the correct order immediately after presentation

• pace controlled at half second intervals using a metronome

• digit span determined by the longest list the participants could recall correctly at least 50% of the time

Jacob’s STM capacity findings

• digits recalled more than letters

• average digit span was 9.3 items

• average letter span was 7.3 items

• digits were recalled better than letters possibly because fewer digits (0-9) than letters (26)

Jacob’s STM capacity conclusion

• STM has a limited capacity around 7+- 2 items supporting millers later theory

• capacity can vary slightly depending on type of material (digits vs letters)

• and age (8yr old girls 6.6 and 19yr old 8.6)

theory Jacob’s STM capacity supports

Miller magical No.7

• outlines how frequently humans use 7

• concluded as jacob’s that 7+- 2 is number of items the human memory can hold

• we also used chunking (grouping info to reduce number of individual items)

research for STM duration and date

Peterson and Peterson 1959

peterson and peterson STM duration aim

investigate duration of short term memory when rehearsal prevented

peterson and peterson STM duration method

• participants given trigrams of 3 letter consonants

• immediately after seeing it they were asked to count backwards in 3’s from random number

• this was inference task done for set amounts of time (3,6,9,12 secs) to prevent maintenance rehearsal

• after delay they were asked to recall the trigram

peterson and peterson STM duration findings

longer time delay the worse the recall of trigrams

• STM has very limited duration of 18-30 when rehearsal prevented

peterson and peterson STM duration conclusion

without rehearsal info in STM quickly decays

LTM coding research and date

Baddeley 1966

Baddeley LTM coding aim

investigate how info is coded by LTM whether it is coded semantically (by meaning) or acoustically

Baddeley LTM coding method

• participants divided into 4 groups (independent groups design)

• given either acoustically similar/ dissimilar or semantically similar/ dissimilar

• after 20min delay to test LTM participants asked to recall words

Baddeley LTM coding findings

more difficulty recalls semantically similar words than dissimilar

Baddeley LTM coding conclusions

LTM prefers using semantic coding

research for LTM capacity and date

Konkle 2010

Konkle LTM capacity aim

investigate capacity of visual long term memory

Konkle LTM capacity method

• participants shown 2,900 images of various objects from different categories

• images shown for 3s each and participants given a short break

• then given forced choice recognition test and had to choose which image they’d previously seen

Konkle LTM capacity findings

96% accuracy with identifying what images they’d seen and which were new

Konkle LTM capacity conclusions

capacity for LTM unlimited

research LTM duration and date

Bharik 1975

Bahrick LTM duration aim

investigate duration of long term memory particularly how well people can remember info from many years ago e.g names,faces of classmates

Bahrik LTM duration method

• 392 US participants 17-74 who’d graduated from high school

• tested in their memory of old classmates in 4 ways

Bahrik LTM duration findings

free recall test:

15yrs had 60% accuracy

48yrs had 30%

Bahrik LTM duration conclusion

recognition memory is strong, free recall declines more sharply over time, info can remain in LTM for lifetime

research working memory model and date

Baddeley and hitch 1974

features of each store is experimental, extraneous variables controlled AO3

• Peterson and Peterson’s study of duration of the STM,the distraction task meant that participants were prevented from using techniques such as rehearsing the trigram. Therefore the variable of duration could be manipulated in order to establish cause and effect.

• In addition, the lab set up means that procedures are standardised and the research can easily be replicated and checked for reliability.

• Therefore, this study has good internal validity due to the isolation of the IV, the control of extraneous variables and the replicable procedure.

features of each store research has highly artificial tasks AO3

• For example, trying to memorise strings of digits (Jacobs) or consonant syllables (Peterson and Peterson) does not reflect memory in everyday life. Although it must be noted that we do occasionally try to remember things such as phone numbers or passwords.

• March et al (1997) showed that if people were not expecting to have to recall information, STM had a maximum duration of four seconds.

• This matters because the poor mundane realism of the studies may lack ecological validity and therefore make it very difficult to generalise the conclusions to explain the complexity of memory in the real world. They simply tell us how people recall digits, trigrams or word lists under very strict laboratory conditions.

additional evidence supporting limitations of features of each store being artificial task with Jacob’s

March et al 1997

march et al 1997

if people’s were not expecting to have to recall info, STM has a maximum duration of 4 sec

features of each store real world benefits AO3

• info about capacity for STM is useful for teachers to know cognitive limitations

• helps teachers plan lessons that reduce amounts of new info held in STM

• postcode system uses Millers research of clunking

limitation of features of each store for sensory register

limited understanding as it is difficult to study due to fleeting nature, lots inferred

central executive (CE)

• directs attention to particular tasks determining how the brains resources are allocated to tasks

• attend and monitors incoming information from the senses of LTM

• this info is sent to other sub systems (VSS, episodic buffer, phonological loop) for processing and temp storage

• little capacity

• no capacity for data so uses episodic buffer as storage system

• modality free (can deal with all the senses)

phonological loop

• limited capacity

• store 2 seconds worth of speech based sounds in chronological order

• 2 components are phonological store and articulatory control process

phonological store

allows 2 seconds of acoustically coded items to be stored, storing speech-based sounds so chronolgically

nickname inner ear

articulatory control process

allows sub vocal repetitions of items in phonological store, similar to maintenance rehearsal

nickname inner voice

visuo spatial sketch pad (VSS)

• processes and stores mental images and spatial info in terms of what they look like and their place in the visual field

• independent from phonological loop

• limited capacity of 3-4 objects

• used when planning a spatial task (moving from 1 room to another)

• nickname inner eye

episodic buffer (sub system)

• added in 2000

• general store (can hold 4 chunks of info for central executive)

• allows both sound and visual info to be bound together

• integrates info from other systems and offers sense of time sequences

• records events that happen and send info to LTM

• limited capacity

• modality free

evidence to support the limited capacity and separate nature of the sub systems AO3

• Gathercole and Baddeley (1993) found that participants had more difficulty doing two visual tasks (simultaneously tracking a moving point of light and describing the letter F) than doing both a visual and verbal task at the same time. Two simultaneous visual tasks must compete for the same sub system (VSS) which has limited capacity, whereas when doing a verbal and visual task simultaneously, there is no competition.

• In addition, Baddeley, Thomas and Buchanan (1975) found the ‘word length effect’; fewer long words can be recalled, and more short words suggesting ‘capacity’ of the phonological store is ‘2 seconds worth’ of acoustic information.

• The evidence indicates that the VSS and the PL are separate sub systems with limited capacity

Gathercole and Baddeley

• found participants had more difficulty doing 2 visual task than doing both a visual and verbal task at the same time

• 2 simultaneous visual tasks must complete for the same sub system (VSS) which has limited capacity

• whereas verbal and visual has no competition

Thomas and Buchanan

• found word length effect

• fewer long words can be recalled and more short words

• suggests capacity of phonological store is 2 seconds worth of acoustic info

lack of clarity over CE exact role in WMM AO3

• Baddeley acknowledges this flaw in the model when he stated in 2003 “The central executive is the most important but the least understood component of working memory.”

• Evidence to support this argument was conducted by EVR case study by Eslinger and Demasio (1985) They reported on the case study of EVR who had a cerebral tumour removed. He performed well on tests of reasoning, which suggests his central executive was intact; however, he had poor decision-making skills, which suggests his central executive was not wholly intact.

• This suggests thatthe original concept ofthe central executive needs to be more clearly specified than just being simply ‘attention’ and that perhaps it may consist of separate components.

research for the limitations of CE role explanation and date

EVR case study by Eslinger and Demasio 1985

eslinger and demasio

• reported in case study of EVR who had cerebral tumour removed

• he performed well on tests of reasoning suggesting CE intact

• poor decision making skills suggests CE not wholly intact

• suggest CE may consist of separate functions

impact of WMM comparing favourably to MSM, explains how memory functions AO3

• lead to greater understanding of cognitive dysfunctions eg ADHD which linked to impairments of working memory

• led to methods developed to help focus of ADHD kids

• many measurable benefits in contrast to limited practical applications of MSM

strengths of WMM

• case studies provided detailed data but they are idiographic so lack generalisability

• internal validity due to controlled lab conditions

idiographic

data based on a specific individual

interference theory

• forgetting in LTM caused by 2 sets of info getting mixed up with eachother

• most likely when info is similar and happens close together in time

proactive inference

new memory is forgotten due to interference with old memory e.g teacher cannot remember new names as she has learn too many before

theory about why proactive inference occurs

Jacoby et al 2001

• argued it occurs because of competition between the strength of old learning and weakness of new learning

retroactive inference

old memory is forgotten due to interference with new memories e.g not being able to remember old password

research to support proactive interference AO3

• Underwood (1957) showed that participants, who were repeatedly asked to learn series of word lists made far more errors as the number of lists increased, compared with when they first began the study.

• As the participants made more errors in the word lists shown at the end of the study, this demonstrates how the old information (the first word lists) was interfering with the newly learned information (the most recent list).

• These results are exactly what the proactive interference aspect of the theory would predict.

research supporting retroactive inference AO3

• McGeoch and McDonald (1931) gave participants an original set of words to recall. Then a second list – which varied in similarity to the original list.

• The more similar the later list – the harder the original list was to recall.

• This supports retroactive interference and the role of similarity – the more similar the info, the more interference it causes, as predicted by the theory.

mcgeoch and mcdonald supporting retroactive inference

gave participants og set of words then second than varied in similarity to recall

more similar the list the harder the original to recall

lab study of retroactive and proactive inference limitations AO3

• lab studies, while this can be seen as a strength as extraneous variables are generally well controlled improving internal validity, it is also an artificial setting, which lacks the context of real-life memory use.

• For example, learning list of unconnected words (as in the two studies above) is not how we tend to use memories in real life, instead we recall Faces, events, dates, feelings etc…

• This means that lab findings might not tell us much about the real-world use of memories – we don’t know whether interference occurs outside of controlled conditions.

• Therefore, findings that support interference theory can be argued to have low ecological validity

field study supporting inference and still supports and date

Baddeley and hitch 1977

Baddeley and hitch field study supporting interference theory AO3

• Baddeley and Hitch (1977) [the same researchers who came up with the WMM] asked rugby players to recall the names of the teammates of teams they had played over the season.

• Their recall was not affected by how much time had passed but by how many other teams they had played in the intervening time.

• This study shows retrospective interference in a real-life setting, this goes some way to answering criticisms aboutthe low ecological validity of most research into interference theory.

strengths of proactive and retroactive inference

• brain scan studies: show that similar memories activate overlapping brain areas, causes confusion so biological support

• wide applicability of theories: interference effects found in memory tasks

limitations of proactive and retroactive inference

• doesn’t explain all types of forgetting: only when similar info competes, can’t explain forgetting due to lack of cues, repression, decay

• individual differences: diff factors like working memory capacity

context cues

cues about your environment that are recorded about at the same time as info

can help prompt recall if present at time of retrieval

if not there can be context dependent forgetting

how absence of cues causes forgetting

when info placed in memory their associated cues are stored at the same time but if not available at time of recall then info forgotten

suggests info can’t be accessed until correct cues are in place

endcoding specificity principle (ESP)

Tulving 1983

• principal that cues present at coding (when learned info) must also be present at retrieval

• if cues are different then context or state dependent forgetting will occur

internal state

• cues about how you are feeling

• e.g mood can be recorded at time of learning info

• if not state dependent forgetting

evidence supporting role of context to explain why we forget AO3

• Godden and Baddeley (1975) studied deep water divers and the effects of context dependent memory. Participants had to learn lists of 36 words (two or three syllables) either on land or underwater and then recall them either on land or underwater.

• Recall was better when learning environments match recall environments. In mismatched learning and recall environments there was a 40% drop in accuracy.

• This supports context dependent forgetting because if the cues at coding are the same as the cues at retrieval; recall is more likely. However, if they are different, recall is less likely

Godden and Baddeley supporting forgetting method

• studied deep water divers and effects of context dependent memory

• had to learn list of 36 words in land or underwater and 4 conditions:

• learn on land recall on land, learn underwater recall on land, learn on land recall underwater, learn underwater recall underwater

Godden and Baddeley supporting forgetting results

• learn underwater recall on land: 23%

• learn underwater recall underwater: 32%

• recall better when learning and recall environments matched

when mismatched dropped 40% accuracy

evidence supporting role of internal state in learning and recall and date

Carter and Cassady 1998

Carter and Cassaday supporting role of internal state in learning and recall

• gave antihistamine drugs (cause drowsiness altering awareness) or placebo to students

• participants asked to learn and recall info in 4 different conditions

• showed mark decrease in accuracy of recall in a memory test when the participants internal state did not match at time of encoding and retrieval

• increase performance when internal states did match

• supports concept of state dependent forgetting as cues were different during coding and retrieval, recall of a list of items decreased