Lecture 4 - Memory Flashcards
1
Q
LTM classification
A
- explicit - we are aware of what we have learned and when we are learning it. (episodic & semantic)
- episodic - events. can time travel
- semantic - info & general knowledge. can be acquired on a single occasion
- amnesia most often affects episodic
2
Q
LTM stores & dissociations
A
- brain lesion to hipp = impaired episodic memory formation
- brain lesion to fronto-temporal lobe = semantic dementia & impairment
- maybe separate but clearly interdependent for their types of info.
- retrieval activates same neuronal networks however (frontal, temporal & parietal) - interact.
- e.g. semantic dementia may not impair recall of recent autobiographic memory but impair recall for remote events
3
Q
does meaning facilitate episodic long-term learning
A
- naturalistic approach with material that has meaning
- Bartlett (1932) free recall of complex unfamiliar folk tales. found remembered stories are shorter, more coherent and fit with personal viewpoints more.
- culturally & socially developed schemas make sense of and integrate new material
4
Q
Levels of processing hypothesis
A
- info is encoded and processed at varying depths. deeper = more durable memory trace
- craik & Tulving 1975
> 3 conditions: visual judgement of letter case, phonological judgement of word or semanttic judgement within sentence. recognition best for words processed semantically (deep). pp’s also better at recognising words when the answer was yes - some meaningful processing here - criticisms: how is depth processing measures? deeper processing not best for immediate recall & instead SSSS is.
5
Q
transfer appropriate processing
A
- processing requirements of test have to match the processing conditions at encoding (test bias) for best performance
- Fisher & Craik (1977) - if words are learned in associative (semantic - cat-dog) context recall is facilitated by associative cues. same shown for shallow rhyme context (cat-bat) & shallow cues.
- but seems to be adv for deep encoding overall as leads to richer memory code more readily retrievable
- craik & Lockhart (1972) 2 types of rehearsal:
1. maintenance - continued processing at one level
2. elaborate - link rehearsed item with other material in memory. enhances LT memory
6
Q
how is semantic memory organised - hierarchical network model
A
- semantic = many hierarchical networks.
- related to nodes (major concepts) are other types of concepts each with assigned properties
- feature info stored as high as possible to min storage space so does not need to be stored again if relevant to most things
- collins & quillian (1969) - manipulated sentence links to tap into different levels of links. faster RT when links are closer together in model hierarchy both on concept and properties levels.
- criticisms:
> conrad (1972) - when sentences are controlled for familiarity hierarchical distance has little effect on reaction times.
> Rosch (1973) - typical category members possess more concept characteristics than less typical ones
> too inflexible & confounded by familiarity
7
Q
how is semantic memory organised - spreading activation model
A
- organised based on semantic relatedness
- smaller link length are more closely related.
- once a concept is activated, activation will spread to the model and most strongly to semantically related concepts
- Meyer & Schvaneveldt (1976)
> decide whether letters form a word. subsequent words are semantically related or unrelated.
> RT faster when a word is preceded by semantically related word.
> conc: semantic priming facilitates word recognition by pre-activating semantically related concepts. spreads through our memory and pre-activates other concepts that pre exist & are related. - eval:
+ more flexible & explains familiarity & typicality effects - assumes concepts are abstract, stable & rep by a single node
8
Q
units of semantic memory: hub-and-spoke model
A
- concept processing involves perceptual and motor systems, contextual info and dif brain regions
- spoke = modality specific brain area for sensory & motor processing. 6 dif spokes (olfaction, sounds, verbal, visual, somatosensory, praxis)
- hub = modality independent semantic concept rep that provides platform to integrate info from dif spokes. sits in ATL
- support: Pobric et al (2010)
> name item from picture (either praxic/malleable). TMS condiitons either in ATL (modality independent hub putting together info) or IPT (modality specific spoke)
> inhibiting ATL (Semantic processing) inc naming times for both. inhibiting IPT (action processing) only inc times for praxic/manipulable objects not low manipulable - inhibiting ATL - general category impairment
- inhibiting IPT - selective impairment for manipulable objects (category specific)
- support:
+ supports assumption that LTM units are combo of semantic and modality specific info
+ supports idea units of LTM not rep by single area of brain
9
Q
implicit learning
A
- non-episodic learning of complex info in incidental manner without awareness. forms implicit knowledge (abstract rep)
1. procedural
2. priming
3. CC
10
Q
implicit learning - procedural
A
- motor skills, once learned not thought about.
- Nissen & Bullemer (1987) - serial RT task either random or repeated sequences. random sequence has similar RT across trials but repeated sequence reduced RT
- repeated sequences implicitely learned w/o instruction
11
Q
implicit learning - priming
A
- exposure to stimulus modulates the response to a subsequent stimulus
- recent rep enhances processing
- typical task: stem & fragment
- Graf et al. (1984) - gave pp’s 4-7 letter word lists & had to either explicitely free recall or implicit stem completion. amnesic patients do poorly on free recall but no deficit on stem completion = implicit learning can be preserved in amnesic patients = evidence of 2 stores
12
Q
implicit learning - classical conditioning
A
- Pavlov - NS (bell) + reflex (salivation) = learning
- Stuart et al (1987) - CS (toothpaste) + pleasant pictures of sunsets etc. CS = new brand of drinks, soaps etc presented with neutral pictures. found after single presentation, toothpaste is more likely bought than random control new brand.
- NS (toothpaste) paired with pleasant feelings = CR of buying
13
Q
how do we know there are subsystems in LTM (explicit vs implicit)
A
- Bechara et al (1995)
- SM (damage to amygdala) WC (hipp damage) RH (damage to both)
> task: coloured slides followed with aversive stimuli & measured skin conductance. pp’s asked what colour they saw
> SM failed to show CC but could remember colours = explicit episodic memory but not conditioning.
> WC evidence of CC but unable to report seeing slides
> RH no conditioning and no memory
