final Flashcards
script
knowledge of structure for an event or situation
heuristic
rule of thumb, guideline based on knowledge
schema
mental structure that organizes info
can be used with heuristics to tell us what to expect
Bartletts repeated reproduction technique
first to report that memory retrieval is reconstructive, fragile, reencoded every time its retrieved
we use general knowledge and expectations from experience to organize memories (schemas)
- pts reproduce (verbally or by drawing) something over and over again from memory and it will almost always turn into something more familiar and culturally relevant as it is reencoded
Deese-Roediger-McDermott (DRM) procedure
semantically related lists of words lead pts to falsely remember semantically related words that were NOT on the list
If attention is divided, effect is seen less bc there is less attention available to activate schema
Support for Activation-monitoring theory, a type of source monitoring error in which we use a schema to come to a conclusion and attribute it to a memory
Activation monitoring theory
proposed that we activate a schema and falsely attribute the source to an episodic memory - source monitoring error
Support: DRM procedure
Also, if there is less available attention, there is less source monitoring errors bc less attention available to activate schema
misinformation effect
people tend to recall what they were told about an event rather than their experience of the event - for other people
phases of the misinformation effect
- encoding - show info
- Post event info - either consistent with reality or misinformation (can be subtle or intentional)
- recognition task - which is the correct image?
Implanted memories
false memories regarding yourself
“shopping mall study”
harder to implant but still very effective - easier with photoshop
procedure of shopping mall study
- write down 3 real stories from childhood and add a fake story about being lost in a shopping mall as a child
- send to pt and ask them to add detail repeatedly over several iterations - not prying for more info, just repetition
- When brought in and asked which were fake
Outcome: 25% guessed the wrong story and after 3 iterations, many ppl added inc detail to the fake story
Problems with eye-witness testimony
- Weapon focus and change blindness (attention) - will remember weapon more than person holding it
- Misinformation effect from post event interrogation
- Source monitoring errors - may attribute blame to a bystander they just remember well
- Confirmation bias -confirmation like “everyone said they same story” will increase confidence in memory
knowledge = ___ memory
semantic
category vs concept vs exemplar
category: group of objects that belong together
concept: mental representation of a category
exemplar: an item in the category
classical view of categorization
categories are defined by a list of necessary and sufficient features
necessary vs sufficient
N: must have all
S: no other features required
problems with classical view of categorization
- not all categories have a list of defining features - impossible to agree on a list - maybe we are just bad at coming up with the list but they do exist?
- Graded categories - we rate things on a scale of how “in” they are
- Typicality effects
Typicality effects
- rt faster for typical than atypical
- generate typical more often than atypical
- typical more affected by priming (lexical decision tasks faster for typical pairings - pts decide if something is a word faster if it comes after a semantically paired item)
Prototype theory
Rather than defining features (classical), exemplars have characteristic features (common features that are not required for the category)
We determine category membership by matching item with prototype stored in memory (every member shared at least 1 feature with another member)
Compare stored prototype with exemplars
central tendency for exemplars
categories have a central tendency in which exemplars with the most characteristic features are found
All category members share ____ even if they are atypical members
family resemblance
Levels of categories
superordinate: broad category, e.g. mammal, plant - distinctive but not that informative
basic: moderately specific, e.g. dog, tree - informative and distinctive
Subordinate: specific, e.g. poodle, maple - informative but not distinctive
exemplar theory
“opposite” of prototype
We store exemplars and create a prototype if necessary
Research supports this theory more
Pro: allows access to atypical exemplars thru memory
more characteristics = close to the
prototype
Similarity theories
Exemplar and prototype theory
Problems with similarity theories
- People may give typicality rating bc they are just “playing along”, so we cant use that as reliable evidence
- These theories don’t define which features are important to categorization
- no input of experience
Explanation based theories
categorization based on implicit ideas about categories you learn as you grow up
Based on psychological essentialism based on past experience
Accounts for why some features are more important than others
Psychological Essientialism
The quality of an item - a bird is “birdy”
Essential nature of a cat is a cat but the essential nature of a coffeemaker is to make coffee
Semantic Network Models
Collins and Quillian hierarchical model and Collins and Loftuses Semantic Relatedness model
Collins and Quillians Hierarchical model
nodes contain info and they are connected by ISA and property pathways that are activated via spreading activation
hierarchy based on specificity - superordinate at the top subordinate at the bottom
Important feature: property inheritance
Property inheritance
Important feature of C&Q
As you move down the hierarchy, concepts inherent properties from concepts higher in the hierarchy - efficient
Evidence for C&Q
Property inheritance could be demonstrated thru RT of properties directly related to a word - faster if directly related, slower if related to a related concept (have to mentally travel up the hierarchy to find the feature)
Problem: atypical exemplars are not consistent (chicken=animal is faster than chicken=bird) and so cannot account for typicality effects
Collins and Loftus’ Semantic Relatedness Model
No hierarchy, instead semantic relatedness determines length of connection (and therefore RT)
Nodes still contain info and are connected by ISA or property pathways
Typical exemplars have shorter pathways and you can have multiple nodes for the same thing, therefore accounts for typicality effects
Different ppl can have different pathways which makes the model unfalsifiable and so unfavored
Neural Network Models
Like Artificial neural nets
ANNs
Computing models based on neurons in a brain - connectionist rather than semantic network model
Nodes are like neurons connected by weighted connections (-1-1, inactive, excitatory, inhibitory) making up input, output, and hidden layers
Knowledge is stored in the pattern of activation (distribution of weights across nodes), not individual nodes
Hidden layers
Layers of ANN nodes between input and output, like cognitive processing neurons, we dont know how they work just that they produce the output
Unknown = hidden
Output nodes and input nodes are synonymous to…
motor and sensory neurons
ANNs are consistent with which theory
embodied cognition and memory trace
ANN facilitated development of
AI, self driving cars, cog psych (make models to represent outcomes of a hypothesis and see if ppl respond the same way)
rationalism
priori truths
deduction
priori truths
born knowing
empiricism
posteriori truths
induction
posteriori truths
gain thru observation
deduction
remember and/or apply
prinicple to instance
induction
observe and combine
instance to principle
inferences about what will likely happen - impossible to reach logical conclusion bc we can never be sure of the future
categorical syllogisms
deductive reasoning
draw conclusion from two statements using quanifiers
definite are easier to solve, negation are harder
cant draw logical conclusion = indeterminate
solve using mental models of ALL possibilities
what are mental models limited by
WM, prior knowledge, visual imagery skills
conditional reasoning
deductive
if, then propositions and draw conclusion
propositions made up of…
condition statement is made up of antecedent (if) and consequent (then)
valid reasoning for conditional
affirm the antecedent, therefore affirm the consequent OR
deny the consequent, therefore deny the antecedent
Wason selection task
4 cards, have one thing on front and one on back, how few do you need to turn over to validate the rule?
*remember directional so you must affirm ante or deny consequent (OPPOSITE OF WHAT IS STATED)
wason selection task affected by….
confirmation bias (ppl dont look for information to refute) and harder if abstract
solution to abstract reasoning tasks
pragmatic reasoning schemas - concrete examples to reduce resources required to solve
belief bias effect
relying on knowledge rather than reasoning to come to an incorrect conclusion
inductive reasoning is limited by
WM, prior knowledge, imagery abilities
Expected Utility Theory
assumes ppl make rational decisions based on subjective utility and probability
ppl don’t actually make all decisions perfectly methodically
NORMATIVE
subjective utility and subjective probability
our personal perception of how much utility we will gain from something and how likely it is
cognitive approach/heuritiscs + biases approach
we make decisions by using short cuts to limit cognitive resources, time, environment
short cuts like heuristics and biases
more realistically what we actually do
Framing and Prospect theory
framing influences decision making:
- our current state (now) is the reference point
- we choose gains over losses
- losses are more important than gains
If framed in terms of gains, ppl are…
risk averse
if framed in terms of losses, ppl are…
risk taking
representativeness heuristic
if similar to population, more likely
heuristic fails with small samples and base rates
most logical conclusion is the most likely
ppl fall for this even with lots of education on the topic
small sample fallacy
falsely assume small samples are representative of a population when in fact they are inherently more variable
plays a role in prejudice - assume one person from a group represents the whole group
base rate fallacy
when judging category membership, we ignore base rates and only use representativeness
solution: ignore description, only use probability
conjunction fallacy
relying on description rather than the knowledge that a single event is always more likely than 2 events
availability heuritic
if easier to remember, we think its more frequent
Things that effect ease of retrieval (recency, familiarity, saliency) will affect how much we think it occurs
simulation heuristic
related to availability
easier to imagine a future hypothetical, the more likely we think it is that it’ll come true
e.g. missing the bus by 30 sec is more frustrating than 3 min bc you easily imagine what you could’ve done when its driving away from you even tho outcome is the same
dual process theory for decision making
two systems are involved in decision making
system 1 (fast, automatic, implicit)
system 2 (slow, controlled, conscious)
sys 1 vs sys2
1: high capacity, uses heuristics, based on past experiences and biases
2: serial (low capacity), abstract, normative reasoning
Everyone defaults to 1 but can override it by pausing, inhibiting, and using resources to make logical decision
Sys2 is not bad, its based on experience, often that is the most logical
experts are more likely to use ____ than novices
heuristics/sys1
bc they have a lot more experience
experimental evidence for source monitoring error, misinformation effect, and implanted memories
DRM procedure, car crash wording, shopping mall
prototype theory terms
characteristic features, central tendancy, family resemblance
how are exemplar and prototype theory similar
membership based in similarity of features, explain typicality effects
Hockett’s universal characteristics of language
semanticity - meaning
arbitrariness - symbols arent drawing
flexibility and naming - everything has a label and the label can change
duality of pattering - signs can be broken down into units
productivity - infinite meaning from finite signs
displacement - we can talk about things that arent there
phonology
sounds - words
morphology
words
semantics
meaning - words, sentences, stories
syntax
rules of language - sentences and stories
pragmatics
how we use language everyday - stories
phoneme
smallest unit that can change meaning without having meaning itself
lack of invariance problem
based on the rules of a language (accent, speed, etc.) we perceive different sounding phonemes as the same. We perceive invariance when there is in fact physical variances between the phonemes
Influenced by the fact that phonemes almost always overlap each other (coarticulation)
e.g. aspirated and unaspirated t sound the same to english speakers but not for some other languages
Segmentation problem
we hear clearly separated words even when there isn’t any clear acoustic break between words
the mcgurk effect
we use vision to help differentiate phonemes
- say ba, play ga, people will hear ga
when vocal cords vibrate, a phenome is… when they don’t, its….
voiced, unvoiced
when vocal cords start to vibrate its called…
voice onset time
categorical perception
we perceive phonemes as discrete categories rather than normal sounds
- if you gradually change the VOT of ba and pa (the only difference between ba and pa) there will be a steep change when ppl detect ba instead of pa or visversa, as opposed to a gradual curve that would be expected from non-phonemes
phomenic boundary
the point at which you stop hearing one phomene and start hearing another
what does the categorical perception of phonemes explain
lack of invariance effect (we easily group sounds together into categories)
accents
morpheme
smallest meaningful unit of language
table is 1, snowman is 2
Free vs bound morphemes
have meaning on their own vs contribute to a words meaning but cant be on their own
mental lexicon
mental dictionary in LTM that stores the meaning of words
Lexical access
retrieving the meaning og a word from your mental lexicon
effects of context on lexical access
- respond faster to high frequency words
- notice errors better when in predictable context
- recognize letters better in a word than by itself
- phonemic restoration effect
phonemic restoration effect
ppl cant tell if there is a sound missing from a sentence bc of the importance of context
context affects comprehension AND perception
homophones
words that sound the same but mean different things
lexically ambiguous, use context to disambiguate
Is lexical access bottom up or top down?
bottom up immediately (access all meanings) and then a couple hundred milliseconds after top down (access depending on meaning)
Evidence from lexical priming and lexical decision task - when priming word was presented at the same time as related target words, rt was fast (BU). When priming word was right before target words, rt was only fast for the word that was contextually relevant aka we only accessed words that were related/primed for (TD).
lexical decision task
is this a word? yes/no
discourse
stories - units of language larger than sentences
often studied in the context of reading bc we cant control what ppl say
reading is dependant on…
lexical access
orthography (graphemes - written symbols)
phonology (phonemes)
3 levels of discourse
- surface level (your memory of EXACT WORDING intonation, etc. forgotten quickly)
- Proposition level (important info is maintained in a propositional network for longer than surface)
- Situational (elaborated representation of the story, with your own added inferences - lasts longest)
Levels of discourse applied
- exact words
- propositional network
- propositional network + inferences from context
Local coherence
coherence between sentences
necessary for successful discourse
relevant for surface level
sentences make sense together but not as a whole
global coherence
coherence of all sentences to a theme
necessary for successful discourse
relevant for situation level
____ is necessary for local coherence unlike ___ which is helpful for global coherence but not required
simple reference/bridging interference
explanatory inferences
parts of a simple reference
antecedent - the term
anaphor - term that refers to the antecedent
Maria - antecedent, her - anaphor
FASTER THAN BRIDGING
bridging reference
without explicit antecedent, we make a bridging inference
take more cognitive load and time to process than simple references
explanatory inferences
instrument and causal
help explain events and help establish global coherence, not required to understand whats going on but automatic for most ppl
instrument inferences
inferences about tool used
causal inferences
inferences about what was caused by what
what do situational models contain
temporal and spatial information that is assumed but not explicitly said
updating a situational model is reflected by…
increased reading time
If time moves, new characters introduced, or new info is added in a story, we take longer to read it
what makes phonemes special
lack of invariance, segmentation, using vision to help hear, categorical perception
imagery
mental representation (visual or otherwise) of a perceptual experience
different from symbols bc symbols are arbitrary
Paivios dual code theory
thoughts can be represented as words (symbols - arbitrary) OR images (analogue - resembles the thing), one wasn’t more important or primary
everything can be represented verbally (symbolically) but not everything can be imaginal
Kosslyn functional equivalence hypothesis
most knowledge (imagery and words) is stored as imagery (analogue code) but some is propositional
same as paivio but emphasis on images as primary
Pylyshyn propositional theory
knowledge is wholly stored as propositions - images are converted to propositions not stored as images. if images are produced, they are produced from propositions
proposition
smallest unit of knowledge that can be verified, whether true or not
“this is a phone” but not “this phone”
proposition = the idea, imagery/language = method of communicating the idea
imagery debate
debate of what came first/what is primary: propositions or images? AKA do images come straight from their source or are they converted into propositions and then into images. Not, whether or not they exist or not.
To determine, we look to see if perceptions act like images, if they do, suggests imagery is primary (functional equivalence)
Support for functional equivalence
mental rotation, image scanning, image scaling, perceptual interference, FMRI
how does mental rotation support functional equivalence
rt pattern for identifying a physically rotated object is the same as a picture of an object - we rotate the object in our minds in the same way we rotate a real object
perception = imagery`
how does image scanning support functional equivalence
we take longer to mentally scan the distance between two objects that are spatially far away than close by
if it was propositional, the distance would be the same regardless bc we would just know the distances between objects`
how does image scaling support functional equivalence
takes longer to answer questions about the details of small imagined things than large imagined things, just like if you were to be looking at them both in front of you (smaller would be less visible)
used the same subjects as the small/large thing to control for specific effects
how does perceptual interference support functional equivalence
if imagery = perception, real and imaged stimuli should interact aka if they are the same, should use the same brain functions and would not be able to use them independently at the same time
Displayed visual and auditory stimulus at the same time as imagined visual and auditory stimulus, found matching modalities slowed RT = interference. opposite modalities did not show interference
how does fmri data support functional equivalence
FFA only responds to seeing faces, PPA only responds to places
When asked to imagine faces and places, the same areas lit up although at a lower level (makes sense so we don’t think fake things are real)
viewing = imagining = same brain mechanisms
what evidence supports propositional theory
difficult and ambigious figures
how do difficult and ambigious figures support propositional
when asked if there was a parallelogram in a davids star, majority said no even tho there is UNLESS they were looking at it
imagery was worse than perception
we made davids star into a propositional shape of several triangles so we don’t see a parallelogram when asked
conclusion to imagery debate
mental images are ANALOGOUS to real objects (functional) unless relatively complex objects or assigning meaning is required
majority of evidence supports functional but it doesn’t really matter what came first bc we use both, what matters is application
picture superiority effect
improve LTM by imagining interactive images
bizareness effect
weird things like interactive images improve memory
concreteness effect is influenced by….
imagery, when imagery is supressed, concreteness effect diminishes
dual coding is probably what determines concreteness effect, as you have 2 copies of the object (visual and verbal)
how to improve LTM
generate info yourself
imagery (dual code theory)
interactive images
problem with functional equivalence and propositional theory and imagery in general
both representational = grounding problem
functional equivalence support could also be embodied cognition support
imagery debate should instead be: how much of knowledge understood thru embodiment?
Barsalous situated simulation theory
embodied cognition version of functional equivalence
1. no representation, gain knowledge thru body and sensorimotor system interacting with the environment
2. cognition requires simulation of the sensorimotor system
3. only representation is the distributed activation of sensorimotor neurons, not abstract/amodal
3. knowledge is flexible and goal driven
support for situated simulation theory
brain activity when moving part of your body is in the same areas as thinking about moving the areas of your body
however this is complicated by hub and spoke model/sensory functional hypothesis
semantic dementia
damage to anterior temporal lobe (ATL) causing difficulty with semantics (knowing what things are or how to use them), suggesting ATL stores semantic info, not varied areas like simulated theory suggests
HOWEVER, when there is further damage to specific areas, knowledge is further impaired, suggesting these varied areas still hold some info
problem for situation simulation theory
semantic dementia/hub and spoke model/sensory functional hypothesis
sensory functional hypothesis
combo of situated simulation hypothesis (activity is only in varied areas) and data from semantic dementia (ATL stores semantic info)
hub and spoke model
hub and spoke model
sensory functional hypothesis
a hub, modality-independant center (ATL), with spokes, modality specific sensory/motor areas across the cortex that contain supplemental “embodied” info
hub vs spokes
properties and features (semantic)
that are invarient
sensory and motor info depending on modality
hub and spoke evidence
when TMS “shuts off” hub (ATL), naming was slowed for both living and nonliving this. when lesion motor spoke (IPT), naming was slowed jus for nonliving
When lesion hub, naming for both high and low manipulability objects was slowed. When lesion IPT, only high was slowed
AKA ATL is a hub that stores knowledge of everything, IPT is a spoke that stores specific knowledge about grabbing
TMS
virtual lesion thru magnetic field which makes neurons behave different
individual differences in imagery abilities
aphantasia (cant picture anything) and hyperphantasia
measured by self report questionarres and performance tasks