Memory, Internal Context:, Types:, Components , Reconsolidation in Human…
Physiological State: Drugs and Alcohol:
- :smoking: Marijuana (Eich, 1975) P smoked M or a placebo, then learned categorised words before test 4 hours later
- Overall M impaired recall
- CD effect for free recall but not category-cued recall = cat names gave P enough cues that they did not need other cues
- :tropical_drink: Alcohol (Peterson 1977) CD effect for free recall but not in category-cued recall
- Eich (1980)
- Reviewed 57 studies and found the type of retrieval test is the best predictor of state-dependency effect in drug studies
- :coffee: Keleman and Creeley (2003) Drive with caffeine v placebo, studied 40 word pairs, cued recall test, state dependency effect found
:runner::skin-tone-5:Physiological State: Aerobic Exercise:
- Miles and Hardman (1998) Words learned during exercise were better recalled during exercise, those learned at rest better recalled at rest
- Schramke and Bauer (1997) Old and younger adults, walking vs resting before learning words
- Context dependency effect but only in free recall
:flag-ru: Cognitive State: ideas, thoughts, concepts at encoding also constitute "context"
- Marian & Neisser (2000): Russian-English bilinguals talked
about their lives in response to prompts
- Responding in Russian generated more mems encoded in Russian lang context (64% of all mems)
- Responding in Eng = only 35% of Russian-encoding mems and vice versa
- Conversing in one lang "mode" e.g., Russian = retrieval of mems acquired in that mode
- Mood congruent memory: Tendency to produce mems with the same emotional tone as current state (e.g., more happy mems when you are happy, and sad mems when you are sad)
- Mood dependent memory: Recall enhanced when same mood at encoding and at retrieval
= evidence is mixed
= effect mostly shows with internally generated mems, e.g., autobiographical mems
- Limited attention
- 2 slave systems = VSS and PL, specialized for speech-based and visuo-spatial input, respectively.
- Selects and manipulates material in each
Coordinated the activities of the three sub-systems in memory
- Monitors incoming data from senses and LTM
- Makes decisions
- Allocates processing resources to those tasks
- Allocates the slave systems (other components)
- Limited processing capacity
- Can only process info from any sensory modality
- Logie (1995)
- Passive inner cache for storing visual and spatial items
- Active inner scribe for rehearsal
Processes visual and spatial info
- Often called our "inner ear"
- Visual info = What things look like
- Spatial info = Physical relationship between things (e.g. visualising the number of windows in your house to count them)
- Limited capacity of 3 or 4 objects
- Logie (1995) subdivided VSS into:
The visual cache: Stores visual data
The inner scribe: Records arrangements of objects in the visual field
- Contributes to our understanding of 'visual semantics' (meanings of objects in our visual environment)
- For example, "think of an object you can sit on," we retrieve an image of a chair or sofa from LTM
- Acoustic or speech-based items
- Passive store of limited capacity
- Active articulatory rehearsal process to prevent decay
Semantic dementia: A progressive neurodegenerative disorder characterized by gradual deterioration of semantic memory.
- Semantic dementia involves a severe loss of concept knowledge from semantic memory even though their episodic memory and most cognitive functions are reasonably intact at least in the early stages of the condition. Semantic dementia is associated with damage to the anterior frontal temporal lobes. = Normal intelligence, problem solving, episodic mem, STM, impairment in knowledge of objects and concepts = involves brain injury to the aterior temporal lobe
Lee, Graham, Simons, Hodges, Owen, and Patterson (2002): brain imaging (PET) study with healthy participants
- Presented names of animals: lobster, tiger, fox
- Or names of non-living things: hammer, whistle,
- Participants produced information about each one, either perceptual, e.g., “lobster is red with a hard shell”, or non-perceptual, e.g., “lobster is often very expensive”
- Each participant did all four conditions. Items
counterbalanced between participants.
Findings = Same regions activated whether
living or non-living thing
- Perceptual and non-perceptual
activated different areas = Perceptual - left posterior inferior temporal
lobe = Non-perceptual - middle temporal lobe
Pobric, Jefferies, & Lambon Ralph (2010):
Concepts represented by both abstract core (hub)
and modality-specific information (spokes)
- Transcranial magnetic stimulation (TMS) applied during object naming of three categories (living, manipulable, non-manipulable things)
- TMS at anterior temporal lobe (ATL) => longer
naming times for all 3 categories
- TMS at inferior parietal lobule (involved in processing of actions on objects) => longer
naming just for manipulable objects
War of the Ghosts:
- Told participants a Native American legend called The War of the Ghosts. The participants in the study were British so the study was filled with unknown names and concepts, and the manner of which the story was developed was also foreign to them.
- Bartlett found that participants changed the story as they tried to remember it, this is known as distortion. They also used assimilation, leveling, and sharpening
- Story became consistent with cultural expectations
- :red_cross: Low ecological validity = approached ppl in random settings, Bartlett’s instructions may have led them to construct rather than remember, did not show accumulative results, more accessible to a modern audience
- :check: Suggests mem used for reconstructive process
- When replicated ...... Used strict and lenient instructions, examined whether the initial instructions and their effect on a first recall test would carry over to a later test
Bransford and Johnson:
- Either gave participants a title for a passage or did not title was "washing up", gave P passage and asked to recall as accurately as they could
- When participants were given the title before they recalled 5.8 units after = 2.6 idea units, which is similar to those who did not receive the title at all = at encoding
- Schemas play an important role in reading and listening as they allow us to fill gaps and make inferences e.g., the washing up passage.
Brewers and Treyens (1981)
- Participants waited in an office for 35 seconds that contained with office objects e.g desks, typewrite, calendar but there were other that seem to not fit with objects e.g. skull, pair of pliers. 2. The participants were told to write everything they remember.
- Participants were unable to remember objects incongruent with their schema, causing them to use past experiences to predict what would be found in the office. This shows that we are active processors of information and that we reconstruct our memories based on our schema.
- = schematic mem affect accuracy or mem at retrieval
Confirmation bias: Distortions of memory caused by the influence of expectations concerning what is likely to have happened.
- Lindholm and Christianson (1998). Swedish and immigrant students watched a simulated robbery = wounded cashier with a knife
- finding was that both immigrant and Swedish eyewit-nesses were twice as likely to select an innocent immigrant as an innocent Swede. Immigrants are overrepresented in Swedish crime statistics, and this fact probably influenced participants expectations about the likely ethnicity of the perpetrator.
- eyewitness memory would often be distorted to conform to the relevant schemas.
- Tulving & Pearlstone (1966): Participants studied 48 words with their category names (2 words per category) e.g., bird-sparrow, fruit-melon, fruit-guava
- Half given a free recall task
- Half given given the 24 category names and asked to recall the words e.g., bird? fruit?
- Cued recall = better
Distinguished between availability and accessibility of mem traces: inability to retrieve a mem does not mean the mem is not stored
- Availability: Mem is stored and could potentially be retrieved, given the right cues
- Accessibility: The mem can be retrieved in current conditions
- Sensory stores = do to with perception
- STM = Brief, 30s or less, communicated both ways with the LTM
- Limited capacity
*Limit to how much info can be held
- Short duration
*Items decay from STM as a function of time and have to be rehearsed to be maintained
- Digit span task = number of items recalled immediately in correct order
- Miller (1950) = most people can remember around 7 +/- (for eng digits)
“A system that is capable of temporarily holding and manipulating information... a mental workspace.” Baddley 2015
- A mental workspace used for both storage and manipulation of info
- Allows us to perform complex cognitive activities such as reasoning, learning, comprehension
- Other models see working mem diff e.g.
- "Cognitive processes that retain info in an unusually accessible state" (Cowan 1999)*
- Backward digit spans = read some numbers and have to recall them in the opposite worder to which they were read
- OSPAN (Turner and Engle, 1989) = participants have to read aloud arithmetic calcuations, e.g., IS (2 X 3) + 1 = 7?, then verify them (i.e., say "Yes" or "No") and read aloud the word that follows the calculation, e.g., CLOCK. After a series of these (which might range from 2 to 9 or more), they are required to recall the words in the correct order. This is used as a measure of working memory capacity, which can then be correlated with other measures, e.g. intelligence or ability to resist interference.r
Processes info in terms of sound
- Preserves the order in which the info arrives, includes both written and spoken material
- Contributes to our learning lang
- Accesses LTM to store and retrieve info
- Develop our vocab as children in foreign langs
- Divided into:
The phonological store: Stores the words you hear like an inner ear
The articulatory process: Used for words heard or seen, allows maintenance rehersal. Capacity 2 secs worth of what you can say
- Items in the store are stored in a phonological
- The phonological store has a limited capacity
- The articulatory rehearsal process:
- Maintains items in the store
- And can recode visually presented items into a
- Auditory items have direct access to the
- Visually presented verbal items (e.g., words, letter, digits) have to be recoded into a phonological form before entering the store
- Conrad (1964): Similar-sounding letters = lower recall e.g. BCPTV vs ABIFO
- Baddeley (1966) wondered the effect of words rather than letters
- Similar sounding words = lower recall e.g. mad, man, mat, cap, cad, can vs cow, day, bar, few, hot, pen
- Supports the proposal that items are stored in a phonological code
- Phonologically similar words less well recalled, even when presented visually and not read out = evidence for the articulatory rehersal process where words are recoded into a phonological form
- PS effect occurs at retrieval
- Visual (formal) similarity does not have the same effect (Baddeley 1966)
- Conducted experiment where words have diff form but same sound (bought, sort, taut, caught, wart) = heard to recall. (rough, cough, through, dough, bough) = easier to recall despite similar form
- It is only the sound that matters = evidence for phonological code
- Short words are better remembered than longer ones: sum, harm, wit, bond, yield, worst, twice vs association, opportunity, representative, organisation, considerable, immediately, universal
- Recall = the number of words that can be articulated in about 2 secs
- Short sounding words better remembered than longer ones, incl when presented visual e.g. scratched easier than utopia
- Evidence for recoding in a phonological form, as well as for the limited capacity of the phonological store
- Prevents rehearsal e.g. counting or saying the over and over again
- Items in the phonological store need to be maintained by a process of subvocal rehearsal
- Articulatory suppression occupies the articulatory loop = leads to lower recall of spoken items, suggesting it disrupts maintenance of items in store
- Prevents recoding of visually presented items
- PS effect disappears for visually presented items but not for auditory presentation (Baddeley, Lewis & Vallar, 1984)
- These data suggests that auditory info has direct access to phonological store, while visual, info needs to be recoded via the articulatory loop
- Auditory items have direct access to the phonological store = phonological similarity effect
- With articulatory suppression, no diff btwn recall of phonological similar and dissimilar words when visually presented - equally diff for these items to enter the store = very low recall for both
- Case study on PV = had a stroke with left here with a digit span of just two items, but was otherwise unimpaired
- She was completely unable to learn words in an unfamiliar lang
- PV was unmpaired relative to controls when learning pairs of familiar words but showed no ability to learn novel vocabulary items - even after 10 trials, her performance was as poor as it had been on the very first trial. Baddeley et al. concluded that the phonological loop was necessary for new language learning. PV was unimpaired in her use of her native Italian, because she had acquired her deficit in adulthood, after she had already learned her native language.
- Baddeley compared her with a control group, asked them to learn words that were familiar in their lang
- Words in an unfamiliar words, controls got better and better of learning the words as they were presented to them more and more = evidence of learning
- PV remains constant throughout at zero = no evidence of learning
Chunking: The process of combining a number of items into a single chunk typically on the basis of long-term memory.
- chunking is taking advantage of cues from prosody, the natural rhythms that occur in speech and that make its meaning clearer by separating into coher-ent phrases the continuous sequence of sounds that make up the normal speech stream
Irrelevant sound effect: A tendency for verbal STM to be disrupted by concurrent fluctuating sounds, including both speech and music.
- That longer words are more complex and this leads to more interference
- Explored the role of the phonological loop in language learning. Used an articulatory suppression task (repeating an irrelevant sound) as a way of occupying the phonological loop. They found that learning of pairs of native language words was unaffected by repetition of irrelevant sounds but learning of foreign words was disrupted, suggesting that the phonological loop has a role in vocabulary learning.
- Gathercole and Baddeley investigated the origin of language deficits in children with Specific Language Impairment (SLI) by developing the Nonword Repetition Test, in which participants have to repeat nonwords varying in length of syllables.
- compared them with a typically developing (TD) group matched for chronological age and a younger TD group matched for language level.
- The children with SLI performed poorly on the nonword repetition test compared with both the age-matched children and the language-matched children.
- The fact that the language-matched children performed better than the SLI group suggests that the phonological loop, which is needed when repeating nonwords, was defective in the SLI group.
- Used a non-word repetition rather than digit span task as a control, so there would be no semantic meaning of the words to the children
- Different languages also have diff digit spans e.g. welsh and mandarin
- As more sound patterns become familiar to children, the larger the connection they can make with them
Digit span is the number of words that can be recalled in the correct order in immediate serial recall. For English digits, the typical digit span is around 7 digits. Digit span is longer in Mandarin and shorter in Welsh. There is some evidence to suggest that this is due to a word length effect - digits tend to be shorter in Mandarin and longer in Welsh
- Memory for facts, general knowledge, vocab
- Not associated with spatial or temporal context: Knowledge is independent of where and when it was learned
- Know rather than remember
- Arranged in a hierarchical structure
- Nodes e.g. animals, birds
- Properties = features e.g. is yellow
- Organisation based on 'cognitive economy = properties stored once only e.g. birds can fly = stored under birds only
- Used a sentence verification task measured time taken to decide if a statement was true or false
- This model predicts faster reaction times for items such as a canary is yellow because the concept (i.e. canary) and the property (i.e. is yellow) are stored together at the same level of the hierarchy and shorter for a canary has skin because there are two levels sepa-rating the concept and property.
= Reaction times slower as distance between elements increased, supporting 'cognitive economy'
- P decide as rapidly as they can whether sentences are true or false
= Should take longer if the concept and property are separated by one level in the hierarchy e.g., a canary can fly and a canary has skin
- :red_cross: Frequency/familiarity also effects reaction times (Conrad 1972) = investigated through organising the frequency of descriptions of items and ordered them into categories of high, moderate, low frequency items e.g. high = a banjo has strings, medium = a cherry has skin, low = an onion has vitamins
- Put these items in a sentence verification task
- Found when you control for frequency and you no longer get a relationship for how far something this in the hierarchy and reaction time
- :red_cross: Typicality: Atypical items take longer to process, e.g. a penguin is a bird (rosch 1973) but have the same relationship in the model as typical ones
- :red_cross: Fuzzy concept boundaries (McCloskey and Glucksberg 1978) = is a pumpkin a fruit? = is a stroke a disease?
- Found there wasn't agreement amongst participants and after a month, some participants changed their mind
= if items were sorted the way collins and quillian suggested you would not get differences over time and between ppl
- Based on semantic relatedness
- Semantic relatedness can be measured by asking people to decide how closely related pairs of words are.
- Whenever a person sees, hears or thinks about a concept the appropriate node in semantic mem is activated and spreads more strongly to closely related concepts
- Distance between concepts reflects the degree of relatedness
- Activation spread to related concepts
- Can account for typicality effects = less typical takes longer
- related concepts are connected by paths in the network
- Meyer and Schvaneveldt (1976) P decide rapidly whether a string of letters formed a word = given a target word (butter) preceded by a semantically related word (bread) or unrelated word (nurse) according to the model activation should have spread from first word to the second only when they were semantically related and this activation should have made it easier to identify the second word = butter is faster when preceded by bread than by nurse = facilitation or semantic priming effect for semantically related words
- :red_cross: Does not make precise predictions but is more flexible = everyone's network will be different = difficult to falsify
- :red_cross: The model implies that each concept has a single, fixed representation. In fact, our processing of any given concept is flexible = assumes concepts are represented in one way
- The notion that activation spreads among semantically related concepts = central importance to Dell's theory of speech production
- When we plan an utterance, this leads to activation of several of the sounds in the intended sentence before we start to speak
- Speech errors occur whenever an incorrect word is more activated than the correct one
Category-specific deficits: Disorders caused by brain damage in which semantic memory is disrupted for certain semantic categories.
Semantic priming is when participants are faster to respond to an item when a semantically related item has just been presented. e.g. faster to say LEMON is a real word when preceded by CITRUSMeyer and Schvaneveldt (1971, 1976)
- Measured reaction times in a lexical decision task
- Lexical decision task = decide whether the word is a real word or a nonword
- Semantic Priming: P faster to decide that "nurse" was a word when preceded by "doctor" than when a word was preceded by an unrelated word (e.g. elbow followed by butter = slower to respond to butter than you would nurse)
- Semantic priming is a facilitation (demonstrated by faster reaction times) or identifying words when they are preceded by related words
- McNamara (1992) = more facilitation when words are very closely related than when they are more distantly related - e.g. after presentation of red participants were faster to identify roses than they were to identify flowers
- Barsalou: Representations are not abstract, stable or similar for diff ppl
- Wu and Barsalou (2009): Processing of concepts influenced by current context or goals
- Concepts stored in relevant perceptual and motor areas
- Bub, Masson and Cree (2008) P learned movements in response to colours (red = poke, blue = open grasp)
- Object words presented in colours, e.g. doorbell in red font = congruent = colour is congruent with the action performed for the word e.g. ringing a doorbell
- Calculator in blue = incongruent = not action you associate with a calculator
- Pliers in blue = congruent...
- Nutcracker in red = incongruent
- Slower responses to incongruent that congruent words, suggesting that the concepts represented by the words evoked the relevant movements, which then had to be overridden
- :red_cross: Role of perceptual and motor features may be overstated: No
evidence that concepts do not possess a stable, abstract core.
-- Much concept knowledge does not involve perceptual and
motor features (e.g., abstract concepts, such as peace, virtue)
-- We can recognise associations between concepts that do not
share perceptual/motor features (e.g., raspberry/banana).
- :red_cross: Perceptual and motor processes may not be critical for
concept processing –may only occur after perceptual meanings have been accessed
-- . (Vannuscorps et al., 2016) patients
with damage to sensorimotor areas could still process action-related concepts e.g. tools
Evidence from brain damaged patients
- Warrington & Shallice (1984): patient JBR = Could identify pictures of non- living things
- But very impaired with living things (6%
- Similar difficulty when giving definitions e.g. briefcase = “ a small case used by students
to carry papers” vs daffodil = a plant
- A minority of patients show the opposite
pattern – better at recognising living than nonliving things
Gainotti (2000): review of 44 patients
- Impairment for living things mainly associated with damage to anterior, medial, and inferior parts of the temporal lobes.
- Impairment for non-living things associated with damage to fronto-parietal areas.
- Alternative proposal = Farah and McClelland (1991): sensory-functional
- Living things distinguished from each other on basis of
- Non living things on basis of functional properties
- 3 times as many visual units in semantic system as functional units, i.e., things more often described in terms of visual features
- Sensory vs functional distinction more important
than living vs non-living
- What exactly is stored in the hub?
- How are modality-specific (spokes) and modality-independent (hub) information
integrated with one another?
- How many spokes are there and what is their
Verheyen and Storms (2013) = vagueness = individuals may use different cutoffs to separate members from non-members of a category
- Schema = organised mental representation created from past experiences
- Memory as a constructive process = relies on meaning and is constructed out of our expectations and what is actually happening
- At encoding: Interpret perceptual input on the basis of past experience, and store the interpretation
- At retrieval: Reconstruct the memory according to our schemas
- Whilst schemas can be useful by allowing us to interpret large and overwhelming amount of information available to us, they can also alter our memory and distort our recollection of a stimulus.
- They typically place emphasis on our pre-existing beliefs and ideas, and can be changed by new information being incorporated to pre-existing schemas, or they are changed as the individual has new experiences and learns new information.
- Loftus and Palmer = hit or smashed
- During a crime: eyewitnesses can use their crime schema to process and attempt to understand the event
- Afterwards, during the interview process, may use this schema to help them retrieve details of the crime
- Tuckey and Brewer (2003)
- When do witnesses draw on schemas to interpret events?
- Showed p a video of a bank robbery and later interviewed them
- Stimulus ambiguity manipulation: e.g., robber pointing a bag which may have had a gun in it vs bag hanging limp by his side (unambiguous condition)
- Participants made more schema consistent errors for ambiguous vs unambiguous details (said they saw a gun)
=> Schemas biased the perception of ambiguous stimuli towards a schema consistent interpretation
- Loftus and others suggested that original memory was overwritten by misinfo, M and Z = original memory is intact - recognition test which includes studied item and misinformed item leads to appearance of poorer memory for studied item in misled
- Some participants will not remember the original item (not encoded). = control p = no misinfo = guess correctly 50% of time
- Misinformed P = used the misinformation instead = recalled less than 50% correctly
- • In addition, participants who DO remember the original information and also the misleading info may assume the experimenter must be
right and select the misinformation anyway.
=> Lower correct responses for misled participants than control
- Participants saw a series of slides depicting an
- "Did another car pass the Datsun while it was stopped at the stop/yield sign?"
- Forced choice recog test: One slide depicting a stop sign the other a yield
- • Correct recognition: 75% for consistent question,
41% for misleading question
Why do misleading q lead to incorrect responses?
- Unlearning of OG mem (i.e, mem weakened or destroyed)
- Response competition at retrieval (i.e., mem intact but competes with newer mem for retrieval = forgetting as retrieval failure)
- Devised a recog test which did not include the misinformed item at all
*Study = mean holding hammer
*Misinfo = Man held screwdriver
*Test = hammer or wrench?
- No opp to pick the misinformation = if it is the case that the studied item has been completely overwritten by a new mem, P should be no more familiar with the hammer than they are with the wrench = both should be equally new
- If the reason for misinfo leading to lower recall is because of response comp then participants should pick hammer as screwdriver is not ablt to be picked
- If of mem trace has been overwritten, misled P should perform at chance on recog test, while controls will be above chance
- Modified test procedure => No difference btwn misled and control groups, suggesting that of trace has NOT been overwritten
- Overall, warnings reduced effect by more than half = Blank and Launay, 2014 meta analysis
- Least effective type of warning = "social discrediting" i.e., casting doubt on competence/neutrality of misinfo source (saying something like 'that info was supplied by someone with an agenda)
- Most effective "Enlightenment" i.e., explaining to P why misinfo was introduced
- Lost in a mall = Loftus (1993, Loftus and Pickrell, 1995)
- Imagination inflation: False photo plus guided imagery => planting of false mems after several sessions = got ppl to believe they had been on a trip on a hot air balloon as a child = Wade, Garry, Read, Lindsay, 2002)
- A follow-up study (Garry & Wade, 2005) found that misleading narratives led to even more false memory than misleading photographs.
- Young children under 10 = more susceptible = Bruck and Ceci, 1997 = maybe cos of social compliance and difficulty distinguishing imagination from reality (faulty reality monitoring)
- Older adults also more susceptible = poor source memory (mem for context of an event) so may confuse one event with another
- Delay btwn event and misinfo increases susceptibility = event mem weakened over time so discrepancies harder to detect
- Loftus (2005): malleability of memory system allows us to update our memories with new information, correcting errors.
- Reinstate context, report everything, change perspective, reverse the order
- Maximises overlap of retrieval and encoding cues, and encourages generation of many potential retrieval cues
Cognitive interview (Geiselman et al., 1985):
- Compared with standard police interview = led to recall of more correct info but also led to slightly more incorrect info
Self-administered interview (Gabbert, Hope, & Fisher, 2009)
- To enable early capture of witnessed information
- Minimised schematic errors or other info seeping in
- Developed in collaboration with UK police and adopted by other countries, including Norway, Netherlands, Sweden, Germany
- Structured recall format
- Includes most effective elements of Cognitive Interview (report everything and mental context reinstatement)
- Horry et al., (2020) meta-analysis found the SAI to be an effective tool for increasing accuracy at subsequent recall attempt.
- involves a failure to detect changes in an object
Change blindness blindness:
- Individuals’ exaggerated belief that they can detect visual changes and so avoid change blindness
Dud effect: An eyewitness’s increased confidence in his/her mistaken when the lineup includes individuals very dissimilar to the culprit.
“The process of recovering a target memory based on
one or more cues, subsequently bringing that target into
awareness” (Baddeley et al., 2015)
- Cue = piece of info that allows u to access a memory
- Target = The mem we are seeking; or the to-be-remembered (TBR) item in a mem experiment
- Spreading activation in an associative network of mem traces: Cues in the environment activate related traces, which in turn activate traces related to them, until the target is reached
- Activation level of a memory varies and determines it's accessibility at a given time, through:
-Spreading activation from another trace
- Memory as a set of features
- :arrow_right: A particular mem = activation of a subset of features
- :arrow_right: Cues in environment match some of the features
- :arrow_right: Activation of features spreads to related units
- :arrow_right: Reinstating the original pattern
- :arrow_right: Leading to pattern completion
- The match allows you to reexperience that memory
Retrieval success is determined by the amount of overlap (match) between the conditions present at retrieval and the conditions present at encoding = Tulving
- Can a weak cue lead to successful recall, as long as it was present at encoding e.g., emblem or soar as a cue for EAGLE
- Studied single words with associated cues (emblem - EAGLE) and were later tested with no cue, the same cue (emblem) or a different cue (soar)
- Recall of targets = better when the og cue was present then when it was not
- Presence of different, equally weakly associated, cue (soar) at test but not at study, impaired recall relative to no cue i.e., emblem-eagle at study, soar-? at test worse
than simply ? at test = better off given nothing than being given a different cue
- Support for ESP
- Will strong cues lead to successful recall if they were not present at encoding
- P told to study cues in relation to targets (train - BLACK) to help remember targets later.
- Difficulty = train – BLACK may be studied, but participants may be
thinking of white in order to help them remember target
BLACK, or it might just come to mind due to strong pre-existing association
- Therefore, P told to relate cue to target adn took several cued recall tests with the weak cues e.g., practiced train - BLACK before they took final test
- FInal test used strong cue (white - BLACK), no cue or weak cue
- When targets were studied and repeatedly
tested with a weak cue (train - BLACK) switch to no cue at final test (free recall) produced an equally large drop in recall
- Switch to strong cue at final test (white?) produced equally large drop in recall
- But when targets were studied with no cue, testing with a strong cue improved recall, suggesting P were generating own cues at encoding
== All consistent with ESP
- Divided attention at retrieval impaired recall of targets (auditorily presented words) more when distracting task was word based (e.g., press a key when you see 3 consecutive man made objects) than digit based (3 consecutive odd numbers)
- Divided attention at encoding => large impairment of recall of targets regardless of type of distracting task
- Activation given to a concept increases with attention. If so, dimin-ishing attention might make a cue less useful and lead retrieval to fail
- Supperadditive effect of extra cues: Rubin and Wallace 1989) = 2 cues a lot better than one.
- Provided semantic and rhyme cues improved target generation considerably more (97% said ghost = when told it mythical being and rhymes with post) than either cue alone (14% said ghost= when only told it was a mythical being and 19% said ghost = when told it rhymes with post )
- Elaborative encoding = generates many cues which can later be used at retrieval
- Context can provide numerous cues
Pyc and Rawson (2010)
- Repeated testing => generation of many cues (mediators) both increasing number of cues and quality of cues
= e.g., study wingu-cloud
= wingy reminds of wings = birds with wings in clouds
= on every retrieval generating cues (bird, sky, cloud) creates extra cues that can be used on the next test
Encoding specificity principle: The more similar the cues available at retrieval are to the conditions present at encoding, the more effective the cues will be
Retrieval mode:The cognitive set, or frame of mind, that orients a person towards the act of retrieval, ensuring that stimuli are interpreted as retrieval
- Recall enhanced if retrieval occurs in the same context as encoding
- Context can be
-Internal: Mood, state, internal thoughts
- Godden and Baddeley (1975):
- Got divers to learn material either on dry land or underwater =
- Recall was worse when it occurred in a diff. context to coding than the same context
- E.g. recall words learnt underwater recalled better when underwater than on land
- No main effect of learning or retrieval context = not statistically significant
- Interaction between encoding context and retrieval context: Items learned in dry context, better recalled in dry, items learned in wet better recalled in wet
- Overton (1972):
- Participants to learn material when either drunk or sober
- Found recall was worse when participants were in a different internal state at recall than internal state at coding
- E.g. recalling info learned drunk better if learnt drunk
Thompson, WIlliams, L'Esperance and Cornelius (2001)
- Experiment 1: Skydivers performed poorly when encoding or retrieving while skydiving (in the air)
- Context dependency effect only for land (encoding on land and retrieving on land = better than retrieving in the air)
= Stressful situation when in the air? Difficult to concentrate?
- Experiment 2: Context dependent mem effect found when skydiving was replaced with watching a video about skydiving = less stress
- Key diff = the stress
- environmental and/or mood cues are unlikely to become encoded or linked to newly acquired information and thus cannot serve as cues to retrieval.
- Results can be applied to understanding variations in context-dependent memory in occupations (e.g., police, military special operations, and Special Weapons and Tactics teams) in which the worker experiences considerable emotional stress while learning or recalling new information.
You don't get these effects during a recognition test:
- P recalled material better when tested in the same room in which they had studied it than when they took the test in a diff room
- P recalled material better when tested in the same room in which they had studied it that when they took the test in a diff room
- Could be using the room as a cue = episodic and not semantic memory
- Experiment 2: Simply imagining og learning context = mental reinstatement of context = also improved recall "imagine yourself back in the study context"
- Experiment 3: Studying in many different rooms made mental reinstatement harder, leading to greater context dependency effects
- Recog test provides with cues so there is less need to rely on context cues
Noisy vs silent (Grant et al, 1998)
Male and Female voices (Geiselman and Glenny, 1977):
- Studied educationally relevant materials (an article)
- Noisy (uni cafeteria) vs silent conditions for learning and retrieval
- P performed better on a short answer test (cues recall) when tested in similar conditions
- Stimuli presented in male and female voices at study
- Items better recognised when test voice was same as at study
- Extent to which context is/isn't encoded with the target mem
- Stress at encoding = lack of attention to study material
- Paired associate learning = attention directed to cue-target association rather than to background context
- Lack of distinctiveness or prominence may lead to context not being encoded
- Extent to which encoding context is/can be reinstated at retrieval (mental reinstatement hypothesis):
- Use of appropriate strategy = imagine og context if physically unavailable
- Many similar contexts = more diff mentally reinstating context
- Many items associated with the same context = cue overload (Tulving and Pearlstone (1966) and long retention interval
- Extent to which context is needed for successful retrieval (“outshining” hypothesis, Smith, 1988)
- If other cues are powerful enough to elicit retrieval, no need to draw on context, e.g., recog tests - provide strong cue
- Use of context reinstatement in cognitive interview
Crime reconstructions: recreating cues present at time of crime may help witnesses to recall other details
- A time-dependent process by which a new memory becomes stabilised after initial acquisition
- Two types:
1. Cellular/synaptic consolidation
- Occurs within the first few hours after learning
- Involves changes to synaptic connectivity
2. Systems consolidation
- May take months or years to complete
- Gradual establishment of memories in neocortex, independent of hippocampus
- Introduced term 'consolidation' = mems are initially fragile and sensitive to disruption (interference)
- Over time they become strengthened and resistant to interference
- Are integrated into a network of pre-existing memories
- Temp protection against interference = no lasting benefit
- Protects against interference so consolidation can proceed more
efficiently than during wakefulness = lasting benefit
- Active consolidation of declarative mems = lasting benefit
Ellenbogen et al (2006)
- Does sleep play an active role in consolidation of declarative memories?
- If so, memories will be resistant to interference following sleep but not following
- • Sleep following learning of AB paired associates (e.g., BLANKET-VILLAGE) = A = cue, B = target
protected memory against interference from AC pairs (e.g., BLANKET-RUBBER)
- Had four groups, two learnt at 9pm and went to sleep, when they woke up they were split into another two groups
- One was tested on the initial list, the other group learnt the AC list immediately before being tested on the AB group
- Other set learnt the list at 9am, stayed awake and were tested at 9pm
=> If consolidated memories are resistant to interference, and if sleep plays an active role in declaritive mem consolidation then those mems would be resistant to mems after sleep but susceptible after wake
- Sleep group did better in all conditions but did significantly better than the wake group in the interference condition
- Experience initially encoded in parallel in hippocampus and cortical areas
- Reactivation of hippocampal network reinstates the activity in cortical networks
- Repeated replay of og experience across hippocampal cortical networks, leads to strengthening of cortico-cortical connections
- Allowing new mems to become independent of hippocampus and integrated with pre-existing cortical mems
Can occur as a result of:
- Interference from new learning (Muller & Pilzecker, 1900)
- Electroconvulsive shock (ECS) =. mem y impairment if given
immediately after training, but not at a later stage (Duncan, 1949).
- Injection of protein synthesis inhibitors after memory
acquisition => memory impairment in non-human animals (Nader, 2000)
- Misanin, Miller and Lewis (1968)
- Reactivated a consolidated mem (fear of white noise that had been paired with footshock in rats)
- ECS (electroconvulsive shock) given after reactivation of the memory => disrupted memory
- No disruption to memory when ECS given without memory reactivation
= Not just the ESC making the mem disappear but only when the white noise is paired with the ESC that you see an impairment to mem
- Reactivation (a “reminder”) returns memory to a fragile (“labile”) state, similar
to the state it is in when first acquired
->Mem then needs to be 'reconsolidated' to stabilize it again
->While in this fragile state it is vul. to disruption: modification or even unlearning
- Finger tapping task:
- 5-element sequence (e.g., 4-1-3-2-4)
- Repeated tapping for 30s, then 30s rest, for 12 trials
- Groups learned 2nd sequence on day 2
- Overnight improvement in speed and accuracy when sequence
tested on Day 2: consolidation
- Learning 2nd sequence (S2) on Day 2 did not interfere with memory for S1 when tested on Day 3, which showed mem enhancement
- But brief rehearsal (reminder) of S1 on Day 2 immediately before learning S2, impaired memory for S1 on the Day 3 test: reactivation => disruption of reconsolidation
:check: First convincing evidence for reconsolidation effect in human mem = using new learning as amnestic treatment = not involving emotional arousal
:red_cross: Hardwicke, Mazi, and Shanks (2016) failed to replicate Walker et al's finding of a reconsolidation effect in procedural mem= 4 direct replications incl using og software, 3 conceptual replications
- If testing benefits memory, but retrieval may make memory fragile, does testing make memory more susceptible to interference from new learning or can it protect mem against interference?
- Potts and Shanks (2012) learning Finnish words on Day 2 impaired mem for Swahili words learned on Day 1 when there was no reminder test
- But reminder testing immunized mem against interference => memory was as good in the group who had a reminder followed by
interference as it was in the group who had the reminder test alone
== no evidence of disruption to reconsolidation in vocabulary-learning task, testing protected mem against interference
== Some researchers (e.g., Forcato et al., 2009, Hupbach et al., 2007) suggest that only a brief (or incomplete) “reminder” will allow modification of the memory to occur, not a full retrieval
Reconsolidation as mechanism for incorporating
- On day 1 P learned a set of objects placed in a blue basket
- Blue basket briefly mentioned on Day 2 (the subtle reminder) before learning a new set of items, test on day 3
- No diff in total recall between groups, reminder group showed more intrusions of List 2 into List 1 = updating of list 1 = episodic mem of this experience (blue basket) so objects associated with the blue basket interfered with the list
new information into existing memories?
- Does taking a test before receiving misinfo eliminate susceptibility to misinfo? Chan and LaPaglia (2013)
- Cued recall test ("what does the terrorist use on the flight attendant") and control (tetris)
- Audio recap of plot with misinfo
- T/F test = the terrorist used a hypodermic syringe on the flight attendant, T/F?
- Taking an initial cued recall test before hearing the misinformation led to greater susceptibility to the misinformation: "Retrieval Enhanced Suggestibility" (RES) = did not have a protective effect
:red_cross: True and misinformed items mutually exclusive = allowed updating to occur?
:red_cross: P did not know they were misinformed, so may have accepted narrative as true, without overwriting of mem
Reactivation = making someone recall the mem again
Reconsolidation = Process that happens after reactivation and a mem going through another consolidation process to be stored in the same form as before
Retrieval Induced Forgetting
- The RIF effect is the finding that retrieving items from mem impairs subsequent recall of related but untested items, by comparison with unrelated untested items
- Inhibition account = lasts up to 20 min = item is suppressed while mem of e.g., mango is suppressed
= banana may become activated and intrude during fruit-or.. in retrieval practice = to facilitate the retrieval of another fruit a different fruit may be inhibited
- Blocking account = route from the item to the category is affecting (fruit-mango) = because practiced words are more easily remembered at test, they may occupy a "response channel" in memory and effectively prevent unpracticed–related words in memory from being remembered
- process by which retrieving an item from long-term memory impairs subsequent recall of related items expressing one thought to remember another retrieval can change subsequent memory (adding things that weren't there)
- RIF occurs when an individual is trying to retrieve a specific memory. Closely related words and memories 'compete' with the memory you are trying to remember causing a temporary forgetfulness.
- Caused by physical damage to the brain
- Severe, disabling and irreversible
- Psychologically caused
- Involves repression of disturbing memories
- Alzheimer's Disease (AD)
- Degenerative disorder: begins with memory impairment, develops into general dementia affecting all aspects of cognition
- Korsakoff's Syndrome:
- Due to thiamine deficiency = often associated with chronic alcoholism
- Tends to make things up that haven't actually happened
- Gradual process cannot pinpoint
- Herpes Simplex Encephalitis
- Clive Wearing = viral infection
- Sudden onset = date of onset of symptoms can be precisely identified (unlike degenerative diseases)
- Temporal Lobe Surgery
- Rare, date of onset, location of lesions precisely known
- Strokes, tumours, head injuries, anoxia, HIV, Huntington's Chorea, Parkinson's Disease
- Damage to hippocampus and surrounding medial temporal lobes
- Korsakoff’s patients often damage to diencephalon (thalamus and hypothalamus)
- Hippocampus = important for storing new memories
- Parts of diencephalon = important for processing and retrieving memories from storage in temporal lobes
- Characterised by:
- Anterograde amneisa and may also be retrograde
- Inability to recall memories acquired before onset
- Varies in severity
- Retrograde amnesia refers to loss of access to events that happened in the past, typically before the onset of the disease. The densely amnesic patient HM is a classic case of anterograde amnesia because his capacity for new learning was greatly restricted and his ability to recall events from before his operation was far from perfect
- Two patients by Baddeley and Wilson (1986) both had dense but pure amnesias with high and well-preserved intelligence, but one appeared to have excellent retrograde memory and could, for example, talk in great detail about his wartime experiences, whereas the other had at best only a hazy memory of his past.
- Deficit in encoding, storing or retrieving info acquired since onset
- Symtoms: Impairment of episodic memory post-onset
- Preserved language and intelligence
- Intact working memory = normal digit span and recency effect = remembering the most recent items on a list
- Intact implicit memory, intact implicit mem = Cohen & Squire (1980): distinction between declarative and non-declaritive (implicit) mem
- Amnesics can learn procedural learning, classical conditioning and priming
- Huppert and Piercy They took advantage of the fact that people are very good at recognizing pictures that they have previously been shown, demonstrating first that the performance of amnesic and control patients can be roughly equated by giving the amnesic patients longer to encode the pic-tures than the controls. Their study involved presenting pictures either once or twice on each of two successive days. After the second days presentation, participants were shown a sequence of pictures and asked to say whether they had seen each picture. If they recognized a picture they were then to decide on which day that picture had been shown
- mnesic patients were more likely to say that items presented twice on day 1 had in fact been presented on day 2, presumably because the degree of familiarity was greater. The controls showed exactly the opposite pattern, being more accurate in assign-ing items to day 1 if they had been presented twice.
- the absence of the link to context provided by episodic memory, the amnesic patients had to rely on a general feeling of familiarity. This did not allow them to distinguish between greater familiarity resulting from two presentations, and that resulting from a recent experience
- Rome is the capital of Italy
e.g., the nurse-doctor experiment
- Graf, Squire, & Mandler (1984)
- Amnesic patients and controls made judgments about words e.g., DEFEND, "does it share same vowels as previous words?" and liking rating = how much do you like this word
- Word stem completion test, e.g., DEF___
- Implicit task = think of first word that comes into head = patients performed as well as control
- Explicit task = complete the stem of the studied word = Patients impaired relative to controls
=> implicit task reflects activation level of primed word = spared activation in amnesia
=> Explicit task requires more info (e.g., the recollection of context) = impaired in amnesia
- Mandler (1980), 2 routes to recognition
- Familiarity: Feeling of knowing without recalling specific details
- Recollection: Remembering particular details of the experience
- Huppert and Piercy (1978) presented pictures once or three times on two consecutive days
- Day 1 = 80 pics, Day 2 = 80 diff pictures Day 2 test: Recency and frequency judgements = today or yesterday? Once or three times?
- Korsakoff patients misinterpreted frequency as recency – responded “today” to items seen 3 times on Day 1 as often as to items seen once on Day 2
- Hippocampus important for == binding items to their context and for mem consolidation
- Sanders and Warrington (1971) photos of famous people from different eras
- Earlier mems better preserved than more recent ones: Ribot's law = the temporal gradient of retrograde amnesia
- Other methods: TV shows, news events... or autobiographical mem validated by relatives
- presented their patients with photographs of people who were famous for a limited period at different points in time, finding that their amnesic patients typically performed more poorly on this task than controls. They also observed that ear-lier memories were better preserved, so-called Ribot’s law. This asserts that older memories are more durable than those acquired more recently
- Memories initially encoded in hippocampus (HPC)
- Also stored in neocortex – over time, become independent of hippocampus
- Patients with HPC damage may therefore remember remote but not more recent mems, recent memories are not yet independent of the HPC.
- Only declarative memory relies on HPC, so procedural memory is unaffected
- Nadel & Moscovitch (1997, 1998)
- HPC always involved in retrieval of episodic mems
- Episodic memory trace consists of ensemble of hippocampal and neocortical neurons
- Upon each retrieval of the memory, a new trace is encoded in HPC with context of retrieval
- More traces => more opportunity for retrieval
- Neocortical structures extract gist (schemas) and also semantic memory from episodes
- Memories depend on HPC for as long as they retain episodic features
- Schematic version of memory develops in neocortex, without contextual details
- Dynamic interplay between 2 types of memory – one or the other dominates according to relative strength and conditions at retrieval
- Memories transformed from context-dependent to context-independent
- HPC not involved in retrieval of semantic mem
- Tests often tap semantic memory (e.g., memory for famous faces/ TV programs) => temporally graded RA for semantic mems but ungraded amnesia for episodic mem
- Recent memories more likely to be episodic, so rely on HPC => temporal gradient
- One involved shallow visual processing (Is this word in upper or lower case? TABLE), one was phonological (Does this word rhyme with dog? Log), and the deepest required semantic processing (Does the word field fit into this sentence? The horse lived in a —).
- They asked to recall the words mixed up with other words and asked them if they are studied words or random words
- Yes items better remembered than no items and semantic words were better remembered (as they create a mental image in our mind)
- Elaborative encoding hypothesis: Designed another experiment where they prevented rich images from being created and equated the opportunity for elaboration between the difference in recall yes and no answers e.g. is it bigger than a chair? (HOUSE) is it bigger than a chair? (PIN) = no elaboration for both
- Varying sentence complexity = differing degrees of recall for yes response
- More complex sentences that resulted in better imagery were better recalled than the simple sentences
Elaborative processing hypothesis = Semantic processing creates a more elaborate memory trace - i.e., a trace with more associations. More associations = more routes back to mem at retrieval
Morris, Bransford and Franks
- Same test with semantic and rhyme judgement condition
- But changed the final test, instead of having one recog test they had two where the word was either...
- Standard = saw words that they had studied earlier (in either semantic or rhyme) but would be mixed up with new words and P had to decide if it were a studied or new word
- Rhyme = not showed words they had studied earlier but either rhymed or did not rhyme with words shown earlier .
- Added an important qualification to the levels of processing hypothesis that the levels of processing effect depends on exactly how memory is tested.
- Support TAP as learning more efficient when tested the same way as learned
Transfer Appropriate Processing = Likelihood of recall affected by match between processes engaged at encoding and processes engaged at retrieval
- Craik and Lockhart (1972) Depth of processing determines how well it will be retained in mem
- Success determined by type of operation performed on stimulus at encoding
- Stimuli first processed according to perceptual features (e.g. visually, acoustically) then semantically
- Semantic is a deeper level of processing
- Orienting task = the instructions push you to investigate something else e.g., acoustic v semantic
- Difficult to define and measure e.g. processing speed
- LoP (features) are not processed in a serial order but simultaneously
Deeper is not always more memorable => TAP
Another experiment done by Haline E. Schendan and Marta Kutas present the neurophysiological evidence for transfer appropriate processing. They verified that memory is best recalled when the situations are very similar to one another. In this experiment two different studies were done. The event-related brain potentials (ERPs) were recorded as a means for information during a memory test. According to this specific study as well as other transfer-appropriate processing accounts, there will be significantly more memory recalled when things are continually grouped together on a perceptual level. Kutas and Schendan showed that there is neurophysiological evidence that if the correct transfer processing of study takes place, then the test experiences will show a difference in memory reactivation. This will occur even if there are some small visual differences within the setting
- Had P study short passage (one about the sun and one about sea otters)
- In the second phase they restudied one of them and took a test on the other (free-recall test)
- They were then divided into three groups and asked to recall both texts either 5 mins after, 2 days or 1 week
- Mixed design = within subject elements and between studies elements
Act of retrieval itself strengthens mem
- Retrieval is a memory modifier and changes the representation of info in memory, making future success more likely
- Testing is a desirable difficulty as the more effort needed to retrieve the item, the greater the boost to it's memory strength and accessibility e.g. if you just retrieved an item it is more likely you will retrieve it 10 mins later
- Conditions that make learning difficult or effortful lead to better subsequent memory for the learned material
- Spacing > massing (e.g. Cull 2007)
- Recall test > recognition test = max number of cues (Bjork and Whitten 1974)
- Fewer cues at initial test = greater benefit (Carpenter and deLash, 2006)
- Longer time to retrieve at initial test = greater benefit (Gardiner, Craik and Bleasdale, 1973)
Baddeley 1966- Found that phonologically similar words (sim. sounding) are more poorly recalled than different sounding words phonological traces are similar = likely to get confused in our store, explaining the worse performance for phonologically similar words due to the fact that phonological is based on sounds, it demonstrates that our phonological store is based on auditory information
Baddeley, Thomas and Buchanan (1975) - as the no. of syllables increases, recall decreases. why? articulatory loop rehearses information, so with 1 syllable words you can fit more in as opposed to longer words which take longer to say, and therefore less can be rehearsed.. short words get more rehearsal which means that more can be maintained in phonological store
Murray 1965- saying 'blah blah blah' during the task, worsens performance especially if stimuli are visual. why? the articulatory loop is preoccupied saying blah blah blah, so the words in the task which flash on the screen are not being recoded (because they are visual and not auditory that goes in automatically) and put in the phonological store; technically as they are not entering the store, you shouldn't remember any words at all
Deese-Roediger-McDermott (DRM) task
- is a procedure in cognitive psychology used to study false memory in humans
- The procedure typically involves the oral presentation of a list of related words (e.g. bed, rest, awake, tired, dream, wake, snooze, blanket, doze, slumber, snore, nap, peace, yawn, drowsy) and then requires the subject to remember as many words from the list as possible.
- Typical results show that subjects recall a related but nonpresented word (e.g. sleep), known as a 'lure', with the same frequency as other presented words. When subjects are asked about their experience after the test, about half of all participants report that they are sure that they remember hearing the nonpresented word, indicating a false memory - memory for an event that never occurred.
- subjects are asked whether they remember previously presented words, as well as related (but never presented) critical lure words ('doctor'). Typically, the critical word is recognized with high probability and confidence
- Roediger and McDermott are quick to point out that their results could be explained by an associative model of memory. That is, the presentation of associated words could spread activation through an associative network to the absent lure word, and thus the false recognition of words could be due to residual activation. This model explains the higher false recognition in the second experiment as a product of more activation being spread due to the longer word-lists used
During reconsolidation, a memory can be strengthened or updated
Unless you wait a considerable amount of time (like 2 days) before introduction misinformation, then you can be influenced by the misinformation and not remember the full story for those things not recalled before
Propanol is used to inhibit reconsolidation
Since memories are subject to change during reconsolidation, you could make them less memorable or manipulate the traumatic memories to be less so
- Tulving encoding specificity principle = we generally store away contextual information about events, and memory should be maximal when the information available at the time of retrieval (including context) matches that in the memory trace
- Priestley, Roberts, and Pipe (1999) con-sidered the possible beneficial effects of rein-stating context on childrens memory for an event. Children between 5 and 7 years of age participated in an event that involved pretend-ing to be a pirate and their memory was then tested 6 months later. Children in the context condition were tested with the pirate props present whereas other children were tested in the absence of relevant context. The key find-ing was that children in the context condition recalled approximately 40% more items of information.
Anterograde amnesia: A problem in encoding, storing, or retrieving information that can be used in the future.Retrograde amnesia: A problem accessing events that happened in the past.
Anterograde amnesia and HM:
- demonstrated that despite his amnesia preventing him from learning new declarative information, procedural memory consolidation was still possible, albeit severely reduced in power.
- He, along with other patients with anterograde amnesia, were given the same maze to complete day after day.
- Despite having no memory of having completed the maze the day before, unconscious practice of completing the same maze over and over reduced the amount of time needed to complete it in subsequent trials.
- From these results, Corkin et al. concluded despite having no declarative memory (i.e. no conscious memory of completing the maze exists), the patients still had a working procedural memory (learning done unconsciously through practice).
- This supports the notion that declarative and procedural memory are consolidated in different areas of the brain
Morris Bransford and Franks
- The deep condition involved semantic processing, for example “Does the word that follows fit the gap in the sentence, ‘The — ran into the lamp-post= car
- whereas the shallow condition involved a judgment of rhyme such as “Does it rhyme with fighter? Writer.”