2239 Flashcards
Study of cognition aim
understand the workings of the human mind by
studying human behaviour
Attention has 2 parts
― Selectivity: Focused auditory attention
+ How effectively can we select one input and avoid being distracted by others?
― Control: Divided attention: Dual task performance and automaticity
+ How do we perform dual (multiple) tasks at the same time?
+ Practice makes a task automatic
selective attention
Focus on one stimulus input and ignore others
Dichotic listening task (cocktail party problem)
Participant asked to shadow (repeat back) message played to one ear and ignore (not be distracted by) the other, were later asked about the content of the unattended message
Cherry’s (1953) findings
― Subjects could report if the speaker was male/female,
soft/loud, high/low voice
― But did not notice when it changed to foreign speech or reversed (backwards) speech
- Only the physical characteristics, and not the meaning, are processed in the unattended message
Moray’s (1959) findings
subjects did not notice repetition of the same word 35 times
=> Meaning not processed in the unattended message
* Only about 1/3 of participants detected their own name in the unattended message
=> Meaning is processed in the unattended message
Bottleneck models
They all assume the multistore model of memory architecture
― assume transfer of information from sensory register to short term memory store
― Sensory register has a large capacity; STM store is limited in capacity: -> a bottleneck
They differ in where they regard the bottleneck is, and the nature of the bottleneck
Broadbent’s filter model
- Stimuli gain access in parallel to a sensory register
- Selective filter (all or none) blocks processing of unattended information to prevent overloading of limited-capacity STM store
― bottleneck as an all-or-none filter - Input remaining in STM not blocked by the filter (i.e., attended) undergoes semantic processing -> if unattended, only the physical characteristics of the stimuli are processed (same as cherry’s findings)
Treisman’s attenuation model:
- an attenuator turns down the amount of processing of unattended information, bottleneck is not all-or-none
- The thresholds of context-appropriate stimuli are lower, top-down processing can lower the threshold of conscious awareness, leading to “breakthrough of the filter”
- Own name has a low threshold due to high salience (personal significance) -> noticed in unattended channel (Moray’s finding)
- Meaningful context lowers the threshold
Deutsch & Deutsch’s late selection model: Assumptions
Argued that attenuator is redundant; only the idea of different thresholds is necessary
* Information is analysed fully (physical, semantic), even
for unattended message
* Bottleneck is late: at selection for action
Early vs. late selection (Treisman & Riley’s (1969) experiment)
Pps were presented with two messages to two ears
―Shadowed one message
―At the same time, made a tapping response to a target word in either message
Prediction and result:
- Attenuation model
* Attenuated processing of unattended (nonshadowed) message
->Target detection should be worse in the nonshadowed message
Late selection model
* Full processing of attended and unattended message
-> Target detection should be equal in the two messages
=> Target detection 87% in shadowed message; 8% in non-shadowed message (early selection)
* But the shadowed message was more important because it required two responses (shadowing + tapping) -> late selection
Johnston & Heinz (1978) flexible bottleneck view
Ss were presented with a list of words dichotically
― Instructed to detect a target = member of a semantic
category (e.g., musical instrument) in either ear
― “Non-target” = word presented coincidentally with target
Critical targets = ambiguous meaning (e.g., ORGAN)
* Meaning interpretation biased by non-target
― Appropriate: church-ORGAN (musical instrument)
― Neutral: paper-ORGAN
― Inappropriate: kidney- ORGAN
* If unattended message is processed to the level of
semantics, meaning of non-target should influence detection of ambiguous target word
* Focused attention condition: pps told which ear targets would arrive =>non-targets were NOT processed to the level of meaning
* Divided attention condition: pps did not know which ear targets would arrive =>Meaning (of non-target) was
processed (late selection) when attention was divided over two ears but not when attention was focused on the other ear (early gating)
- The more stages of processing (physical -> semantic),
the greater the demands on attentional capacity
* Selection occurs as early in processing stages as
possible to minimise attentional demands
Dual task performance
The degree of similarity of the two tasks: similar tasks
interfere result in how well to duel tasks
- Similar stimulus modality e.g., reading text message (visual) interferes with driving (visual) more than listening to phone message (auditory)
- Similar response modality e.g., sending text (tapping = manual) interferes with driving (manual) more than talking on the phone (vocal)
automaticity
- are fast
― require little attentional capacity
― inflexible (once learned, difficult to modify) = habits
― unavoidable (occur without intention)
― unavailable to consciousness
Inflexible (difficult to modify) - Unavoidable
Shiffrin & Schneider’s memory search experiment
- pps memorize 1, 2, 3 or 4 targets (memory set)
- pps are shown a display containing 1,2,3 or 4 items (display set)
- Respond as quickly as possible whether the display contained a target from memory set
- Consistent mapping (CM) condition: Target set and distractor set do not overlap from trial to trial
- Varied mapping (VM) condition: Target on one trial may be a distractor in the next trial
=>CM training led to automatic processes
Criticisms of the original multistore model
-It over-emphasized the structural aspects of memory
- STM store and LTM store are not unitary: each store does not operate in a single, uniform fashion, but has
subcomponents -> STM replaced by working memory; LTM replaced by episodic memory and semantic memory
- STM is not the gateway to LTM
1.The memory systems are interconnected:
– Short-term memory tasks make use of knowledge in long term memory
2. Rehearsal may not be as crucial to learning
– Maintenance (rote) rehearsal does not result in durable memory
* Impairment of STM does not necessarily lead to impairment of LTM
Memory: Architecture and Processes
- Architecture (structure): The way in which the memory system is organised
- Processes: The activities occurring within the memory system
Stroop interference effect
- Word reading interferes with colour naming
― Demonstrates the automaticity of word reading:
Unavoidable/obligatory
Logan’s instance theory
- “Automaticity is memory retrieval”
― Instance representation: Each encounter with a stimulus is encoded, stored and retrieved separately as a memory episode
― With practice, performance becomes automatic when it is based on a single-step direct-access retrieval of solution from a memory episode, rather than application of rules or algorithm which is time consuming ie multiplication
Phonological Loop System: (verbal/speech-based) memory
Passive phonological store
* Holds verbal information (capacity limited by articulatory duration ~ 1.5-2 sec)
* Auditory presentation of words has direct access, bypassing articulatory rehearsal
process
* Visual presentation only has indirect access
Articulatory control process
* Like subvocal speech
* Maintains (“refreshes”) phonological information by
rehearsal
* Provides access to the phonological store for visual input
-> Used in temporary storage and manipulation of
phonological (speech-based) information
-> Involved in learning new word forms (e.g., learning of
foreign vocabulary; new words)
Baddeley & Hitch’s (1974) working memory model
The central executive resembles attention – a control
system which co-ordinates the peripheral storage
systems
* Episodic buffer holds and integrate diverse info
* All components of the system are thought to be
limited in capacity and relatively independent of
each other
― If two tasks use the same component, they can’t be
performed successfully together
― If two tasks use different components, they should be able to be performed similarly together or separately
Phonological similarity effect
- Immediate serial recall (= memory span) of verbal material is reduced for lists of items that sound similar
Conrad (1964)
― tested immediate serial recall of visually presented
letters, recall errors were mainly acoustic confusions
resembling letter names heard against noisy
background e.g., T (recalled as) -> “D”, S -> “F”, B ->
“P”
=> short-term memory uses a phonological code, even
for visually presented stimuli
Baddeley (1966)
― Immediate serial recall is worse with a phonologically
similar list, compared to a dissimilar list, even when
words are presented visually
=> Suggests that we use speech-based rehearsal processes within the phonological loop
Word length effect
Memory span is reduced for lists of longer words
- Memory span (immediate serial recall) is worse for words taking a long time to say
-capacity of phonological loop is determined by articulatory duration (not phonological complexity/number of syllables)
Articulatory suppression (concurrent articulation)
- Stops the silent repetition of the material that needs to be remembered.
- Silent repetition helps keep the memory trace from fading away.
-eliminates/reduces the phonological similarity effect for visually presented lists
― Auditory presentation allows direct access to the
phonological store e.g., “tee, vee, ..”
― Visually presented stimuli are re-coded into phonological form by means of subvocalization
Visuo-Spatial Sketchpad
Used in the temporary storage and manipulation
of spatial and visual information
* Logie (1995) proposed that it consists of two
components:
― Visual cache
. Stores information about visual form and colour
― Inner scribe
. Processes spatial and movement information
. Involved in rehearsal of information in visual cache
. Transfers information from the cache to the central executive
Episodic Buffer Baddeley (2000)
― The components in the original model were too separate in their functioning
― Provides temporary storage for integrated information
coming from visuo-spatial sketchpad and phonological loop
― A process of active binding: An example is visuo-spatial
bootstrapping of verbal memory
Verbal memory can be bootstrapped (supported) by visuo-spatial memory
* Immediate serial recall of a sequence of random digits (e.g., 3-2-4-5-1) is better when they were presented on a keypad display than as a single item
―>Visual/spatial information (supported by long-term knowledge of keypad display) bound with verbal information
Central executive
- Initially conceived of as a limited capacity pool of
general processing resources related to control of
working memory and its slave systems
Norman and Shallice’s view of attentional control
Contention scheduling
― co-ordinates (prioritizes) several action schemas on the basis of strength of schema
― Practice strengthens the schema (e.g., word reading is more practiced than colour naming)
– routine habits prevail
Supervisory attentional system (SAS)
― Fully conscious control
― Flexibly controls selection of action schemas according to goal of the task
― Involved in novel situations where routine control is insufficient (e.g., name the colour, not the word)
-If damaged Utilization behaviour, Dysexecutive syndrome, difficulty with novel situations
Executive Functions
Inhibition Function:
*Stop prepotent/habitual responses
*Resisting distractor interference
*Stroop task, anti saccade task
Shifting Function
*Task switching
* shifting between multiplication & division; colour & shape
Updating function
*Updating and monitoring of working memory representations
*Letter memory task
Operation span task
- Read the maths problem, say correct or incorrect, say
the letter - 12 items – recall the letters in order
- Working memory capacity correlates better with measures of higher cognitive functions and intelligence than short-term memory span tasks (e.g., digit span).
-> Working memory: storage AND cognitive control (central executive)
Episodic memory
memories for events in specific temporal and spatial context
Semantic memory
conceptual knowledge, linguistic knowledge, memories for general facts
- impairment➜ can’t comprehend meanings
sentence verification task ( test sematic )
- Subject-predicate: A canary is a bird
- Sentence types
Set inclusion
A canary is a bird (true)
A whale is a fruit (false)
Property-attribute
A canary has feathers (true)
A whale has seeds (false) - Have provided important data used to develop theories of (lexical) semantic memory
― Typicality effect (on positive decisions)
― Similarity effect (on negative decisions) - Limitations of sentence verification data
― Data can be explained equally well by theories with very different assumptions (localist vs. distributed representations)
― Do the data reflect the structure of semantic memory, or the task process
― Stimuli are words - is recognition of word meanings the same as recognition of objects?
Hierarchical network model (Collins & Quillian, 1969)
Concepts are organized in a hierarchy
* Cognitive economy: Property attribute is stored uniquely at the highest (most general) level
* Concepts are represented by nodes (localist representation)
* Relationships between concepts are represented by links
Problems:
Conrad (1972)
* Showed that RT (reaction time) data are better explained in terms of frequency of co-occurrence of concept and property rather than the levels
* RT varied with the subject-property frequency (associative strength: determined from norms) within a level
* RTs do not always mirror hierarchical relationship
― e.g., A dog is an animal < A dog is a mammal
* Does not explain within-category typicality effects
― e.g., A canary is a bird < An ostrich is a bird
* Does not explain negative (“NO”) judgments are not
faster for closer concepts; in fact the opposite
e.g., A canary is a salmon < A canary is an ostrich
Spreading activation model (Collins & Loftus, 1975)
- Concepts are organized nonhierarchically
-> explains lack of hierarchical effect - Distances between concepts vary in associative strength
-> explains the typicality effect (bird- canary < bird-ostrich) - Activation of a concept spreads to other concepts linked to it
-> explains the semantic priming effect (Response to a word is faster following a
semantically related word)
Feature comparison model (Smith, Shoben & Rips, 1974)
- assume knowledge is represented within a
concept node (localist representation) - don’t represent similarity between concepts well
e.g., car – truck; car – street – both are associated, but car and street are not similar - assumes concept is represented as a collection of distributed features
- e.g., concept of “sheep” represented in terms of: Visual features, Tactile, Auditory, Gustatory
Two-stage decision model
- Assumes that decisions are made by comparing the similarity of
features of subject and predicate terms - Assumes that features can be two types
― Defining features: Features essential to a concept
― Characteristic features: Features that are characteristic, but are not criterial to the concept - Offers a natural explanation for:
― Typicality effect (on positive decisions): faster to say YES
― Similarity effect (on negative decisions): lower to say NO
-Problem: Clear definition of defining and characteristic features is lacking
Neuropsychological studies of semantic memory
. Selective impairment of categories (living things/biological kinds vs. nonliving things/artifacts)
- Selective impairment of living things together with preserved knowledge of non-living things is more common than the opposite
. Even more specific (and seemingly contradictory) category-specific impairment is also observed
- Preserved knowledge of body-parts together
with impairment of living things
- Impaired knowledge of musical instruments (e.g., piano, violin)
together with preserved knowledge of non-living things
Perceptual-functional theory
. category-specific impairments reflect different types of property
― Living things are distinguished from each other on the basis of perceptual (visual) properties
― Non-living things are distinguished from each other on the basis of functional properties
* Explains more common impairments with living things in terms of visual properties being more frequent
* Also explains category-specific impairments WITHIN manmade things where members are distinguished by visual properties
Problems:
― Some patients do not necessarily show impairments of one type of knowledge (only perceptual, only functional)
― Some properties are neither sensory/perceptual nor functional
Distributed-plus-hub theory
-There is a hub (in Anterior Temporal
Lobes/ATL) for each concept in addition to distributed modality-specific information
- provide an efficient way of integrating our knowledge across different sensory modalities of any given concept
- They make it easier for us to detect semantic similarities across concepts differing greatly in their modality-specific attribute
-predicts occurrence of item specific and modality-independent deficit of semantic
memory
Craik & Lockhart’s levels-of processing framework
- A stimulus may be processed via a continuum of progressively deeper processing visual -> phonological -> semantic
- Retention is a function of the depth (level) of
processing
-Problems: - Response type (YES/NO) interacts with the type of orienting task
- Words should be processed to the same level whether the answer is YES or NO in the orienting task
- Spread of processing: Amount of processing of a particular kind (within a level) is also important
- LOP effect is more limited for nonverbal items (pictures of doors and clocks) than for menu items
Incidental encoding instruction
Subjects not told of an upcoming memory test
Orienting tasks
to manipulate the type of encoding used
Visual e.g.,
- Is the word in capital letters?
- Does the word contain a letter e?
Phonemic e.g.,
- Does the word rhyme with TRAIN?
Semantic e.g.,
- Is it a type of animal?
- Is it a synonym (mean the same as) of ROAD?
Note taking may be
― Generative: summarizing, paraphrasing, concept mapping = deeper processing
― Nongenerative: verbatim copying = shallow
processing
* Laptop use encourages verbatim transcription of lecture
content
Recognition failure of recallable words Tulving & Thomson (1973)
- Usually recognition is easier than recall, but showed the opposite
1. Study phase: pps learn a list of weakly associated cue-target pairs
2. then given strong associates of
target words and asked to generate associates
3. were asked to circle target words that were presented earlier (recognition test)
4. Ss were asked to recall target words given cues from Stage 1 - Cue-dependent forgetting: Forgetting occurs when there is a poor match between memory-trace information and information available
at retrieval - Memory performance is better when the context at encoding (study) and the context at retrieval (test) match
- Encoding specificity principle states that the match between encoding and test context is important
Environmental context(Godden & Baddeley, 1975, 1980)
- At encoding, Ss listened to a list of 40 words on land or under water
- Memory for the words tested on land or under water
Match/mismatch of environmental context affected recall, but not recognition
Recall is more reconstructive than recognition:
― In recognition, the items serve as retrieval cues
. “is this the man that was driving the car?”, “Is this a word that was on the list?”
― Recall involves self-generating retrieval cues
. Matching environmental context may help generation of retrieval cues
Implicit memory: Repetition priming
- Refers to facilitation (better
performance) as a result of prior
exposure
-Better able to complete the fragment
when the word had been presented in
study
Amnesia
Causes
* Korsakoff’s disease: (chronic alcohol abuse)
* Alzheimer’s disease
* Encephalitis
* Stroke
* Closed head injury (most common)
* Surgical removal of hippocampus/limbic system
Graf, Mandler & Squire (1984)
- Amnesic patients (Korsakoff patients) vs. nonamnesic hospital patient controls
- Study phase: learn a list of words
- Test phase
- Explicit tests: free recall, cued recall, recognition
- Implicit test: word completion
=>Amnesiacs performed worse than controls on all conventional (explicit) memory tests, but not on word completion (implicit
test)
Schacter, Church & Bolton (1995)
Study phase: Subjects heard words spoken by six different speakers
* Implicit memory test: auditory word identification
=> “Amnesics lack the ability to
bind voices with specific studied
words” = binding the item (word)
with context (voice)
Huppert & Piercy (1976)
- Korsakoffs and controls were shown pictures on Day 1 and different pictures on Day 2
- On Day 2, 10 minutes after picture presentation, pps were asked to respond only to pictures shown on Day 2
=>performed equally well
Korsakoffs are responding on the basis of
familiarity of pictures” not recollection of
context
Binding of item and context (Diana, et al., 2007)
- Recognition memory can be based on
FAMILIARITY or RECOLLECTION
― RECOLLECTION: recognition memory accompanied by retrieval of context (correct source judgment)
― FAMILIAIRTY: recognition memory based on
assessment of memory strength - Perirhinal cortex: Receives information about specific items
-> “What” info needed for familiarity judgements
-Para hippocampal cortex: Receives information about context
-> “Where” info useful for recollection judgements
-Hippocampus: Binds “what” and “where” information
-> Forms item-context associations
Reading: linguistic processing
- Orthography: The spelling of words (letter identity and letter order)
- Phonology: The sound of words
- Semantics: The meaning of words
- Syntax: Rules for combining words e.g.,
- Discourse processing: Making inferences
Research methods used to study reading
- Interference and priming tasks: word processing is studied indirectly by studying the influence of distractor or prime word (response isnot required) on target: Compared to control condition
- Eye movements during reading: eye tracking measures
- Neuropsychology – study of patients: types of errors
- Neuroimaging
Masked priming
- A prime is presented briefly and masked so that participants cannot identify it
- Different prime-target relationship can be used to study different aspects of
reading
Frost (1998): Strong phonology hypothesis
– Phonological representation is a necessary
product of processing printed words
– Phonological processing is mandatory (obligatory) in reading
* Evidence:
1. Homophone effects ( Words with same pronunciation but spelt differently) in semantic
categorization
2. Masked phonological priming effect
Rastle and Brysbaert (2006): Masked Phonological Priming
- Word processing faster when preceded by phonologically identical nonword primes than
by primes similar in orthography but not phonology - Even when the prime is masked and brief, so it is not consciously perceived
– Suggests phonological processing occurs rapidly and automatically
Word Superiority Effect
Identification of a letter is better when the letter is presented in a word context
McClelland and Rumelhart’s (1981) Interactive Activation Model
- Is the basis of almost all models of reading
- 3 levels (layers) of detectors: feature -> letter -> word
- Bottom-up excitation/inhibition and top-down excitation/inhibition
- Within-level inhibition: different features/letters/words are mutually inhibitory
. explains the word superiority effect in
terms of the top-down feedback from the word level to the letter level
Grapheme
letter or letter cluster that correspond to a
single phoneme
Deep/opaque Orthography
- Some spelling-to-sound mappings are highly
ambiguous/inconsistent - Some words have “irregular” mapping (odd
one out)
Shallow/transparent orthography
- Mapping from grapheme to phoneme is
predictable/consistent
Dual-Route Cascaded Model (DRC)
- Individuals use both the nonlexical (Route 1) and the lexical (Routes 2 and 3) paths
when reading aloud - Familiar words depends on
the lexical route
– Sensitive to word frequency
->predicts regularity effect (naming advantage for regular words) for low frequency words
Surface dyslexia
- Intact nonword reading; poor at reading irregular words
- Regularisations
– PINT read to rhyme with “mint”; YACHT with ”hatched”
Phonological dyslexia
- Intact word reading; poor at reading nonwords
Deep dyslexia
- Like phonological dyslexia, plus semantic errors
Connectionist triangle mode
. All information (orthography, phonology, semantics) is used to read both words and pseudowords
* Two pathways in reading aloud:
– Direct orthography to phonology
– Indirect pathway through word means
– It does not assume an orthographic input lexicon which stores knowledge of spelling of (familiar) words
* Semantic knowledge has largest impact on inconsistent words because of longer processing time
Schemas (e.g., word reading, colour naming)
- are series of co-ordinated action sequences (habit patterns)
- are triggered by cues in the environment (colour, word)
- several action schemas may be activated simultaneously
