Exam #2 Review Sheet Psych 240
Lecture 9: Working Memory
Thinking involves working memory:
Raven's Progressive Matrices
used in IQ tests
holding and manipulating and maintaining it - working memory
Mental arithmetic
Individual differences:
measure of working memory span
Measure:
reading span- how many words can you hold onto while reading
reqadin a series osf sentences and thenrecalll the final words of each sentence in order
span = number of final words you can hold
Raven's, aging, and working memory
raven scores worse fo older people and raven score is high when WM is good
WM scores are worse for older people
Raven’s get worse because WM gets worse
evidence: same score for MW as ravens score at any age
Interference: r
random number generation and syllogistic reasoning:
interfwerenecne for doing boht these tasks at once
Reading comprehension
popel with high WM have better comprehension of text
Evidence for WM/LTM distinction:
Dissociations: Anterograde amnesia (LTM disorder) vs. working memory disorder.
Anterogade amnesia: inabilty to form new memeories
damage to the medial temporal lobe spcffiall the hippocampus (patient H.M.)
impared LTM but WM is good
Serial position curve:
primacy effect
remmebre the beginning of the lsit
When you test right after a working memory task you remember the beginning of the list more than the ending of the list
recency effect
remembering the nd of the list
evidence: different speeds help with recenenvy effect recall
Double dissociation logic (as applied to lesions, behavioral dissociations, and neuroimaging)
Paiteint K.F:
normal LTM and impaired WM
Baddeley's 3-part model.
Phonological loop: buffer and rehearsal
short term storage
subvocal rehearsal in the mind
evidence
span test : Phonological coding: acoustic confusion (acoustic similarity effect, what about visual similarity or similarity in meaning?)
inreference between somiimlar acoustics letters
confusion when they sound alivke but ono when they have similar meaning
Articulatory suppression.
repeating saing y the while seeing a list inhibits accuracy with recall and causes more errors
Visuospatial sketchpad
Central Executive.
Phonological storage capacity:
chunking
grouping things to make them moe menaingull makes them easier to remember
errors thinks that are mmore meaningful to you are easier to remember ( ex: 2005).
Time effects:
word length:
remember as many words you can say in 2 seconds
meaning you can store more shorter words than longer words
theses effect diaapper with aurticualory surepression
speed of speech
WM span is large for: words that are pronounced quickly,
people who speak quickly, and languages where words can be pronounced quickly
they hold more information and have better memory span
Phonological loop neuroimaging evidence:
rehearsal process activates left hemisphere (Broca's area) and not right.
fMRI/ PEt
2 back test- letter nothing pattern, does the tletter match the letter two back?
the search task is there this letter?
subtract them to fined therehersal processes in 2 back
PET scan subtraction revels the brian areas that a correlated with rehearsal
frontel and partial
Visuospatial buffer:
from vision or long term memory
devoted to visual memory
image then bacn be treated as percept
Visuospatial sketchpad:
Behavioral double dissociation:
Brooks letter-scanning task or sentence task coupled with pointing responses or vocal responses.
Pattern of interference.
Visuospatial sketchpad: neuroimaging evidence: visuospatial WM activates right frontal lobe and not left (prefrontal cortex)
. PET double dissociation between phonological loop and visuospatial sketch-pad.
Central executive: supervised attention
frontal lobe syndrome: damage
low concentration
perseveration:
fail to stop inappropriate behavior
distractibility
Modern views:
distributed representation,infinite buffers?
sensory recruitment: different for each sense
Baddeley Article
Dissociation between long-term and short-term/working memory, recency effect, acoustic
coding, Modal model (three-stores model): sensory memory, short-term memory, longterm memory, levels of processing (depth of processing)
Individual differences in WM: working memory span (reading span).
Tripartite theory of WM: central executive, phonological loop, visuospatial sketchpad, Frontal lobe syndrome (central executive dysfunction): perseveration, distractibility (utilization behavior).
Phonological loop: verbal store, articulatory loop, subvocal rehearsal,
phonological similarity effect (same as acoustic confusability), irrelevant speech effect, word length effect, articulatory suppression.
Lecture 10: Amnesia and the Neuropsychology of Memory
Explicit:
declarative knowledge
recall
recognition
conscious recollection
implicit
unconscious recollection
procedruakl knowledge
Amnesia:
psychogenic
like in movies
probably fake
organic
due to brain damage
anterograde
unable to learn new things
retrograde(temprarly gradeded)
unable to recall old memories prior to injury
temporarily graded
Patient H.M.,
hippocampus,
where memories are made
spared implicit memory: mirror reading
Participants
Korsokoff’s amnesics and N.A. versus normal
Severe Anterograde Amnesia (Can’t learn new things)
• Methods
Experiment included 50% repeated words across 4 days
Non-repeated words: implicit
Repeated words: implicit + explicit
• Results
For new words, Normals and Amnesics improved about the same (implicit only)
For old words, Normals were better than amnesics (implicit + explicit).
Tower of Hanoi.
procedural memory is sytilll intact in amnesiacs
Priming:
word fragment completion
Priming = complete more old fragments than new. (prior experience influences present performance)
Result: amnesics show normal priming, but poor recognition memory.
Amnesics' implicit performance (completing more old words than new) vs. explicit (recognition).
PET studies of healthy individuals: word stem completion. Explicit task & results vs. Implicit task and results.
Behavioral double dissociation:
1. Modality of presentation: implicit test and results, explicit test and results.
2. Depth of processing: implicit test and results, explicit test and results.
Taxonomy of long-term memory: types of implicit/explicit memory and associated brain regions
Ogden Article
Patient H.M., anterograde amnesia, retrograde amnesia (temporally graded).
Dissociation between immediate (working) memory and long-term memory, dissociation between remote and anterograde memory, dissociation between implicit memory and explicit memory.
No dissociation between verbal and non-verbal memory, no dissociation between semantic and episodic memory.
Lecture 11: Long Term Memory: Representation/Categorization
Explicit vs. Implicit.
Explicit memory: conscious, verbalize
Semantic vs. Episodic
Semantic memory: general knowledge, memory of facts
not tied to time or place
fact about the world
Episodic memory: Personal episodes
how did you get to campus today
first kiss?
Specific time and place form you own POV
Categorization:
we treat similar things the same
Allows inferences about members of the class
Use by Pigeons
4 categories
pigeon peck on one of the four keys depending on the stimulus(image given)
trained for 30 days
become good at identifying what images go in what category w original stimuli
flaw: could be by associations
Gave new examples/stimuli and catorized and pigeons were still pretty good.
Experiment on physical similarity vs. conceptual knowledge (milk vs. mashed-up food experiment).
categories are not always similarity based
ex: hawks are birds but they are more similar to bats
Children 4 and older can identify that the black bird(the hawk) mashes up food like the flamingo even though they are not similar looking instead of give milk like a bat
Classical view:
defining properties:
necessary and sufficient( defined and mutually exclusive)
Problem: what defines "game." - is hard to define because there are not clear backgrounds
Modern probabilistic view:
characteristic properties: properties and features that are in common
fuzzy and probabilistic
similarity between members- some have more or less characters
Typicality evidence:
Ratings.
exemplars with more charteritiic sproeprties are seen as more typical in that catagory
Sentence verification
more faster to verify typical exemplars then less typical exemplars
Hedges
Categorization on the basis of similarity:
To exemplars
exmaples fo a agatogries
what is closest to the example in your storied memory
To prototype
average example, only a prototype
compared to similarity of a prototype
Geometric approach:
similarity-rating task
rate relations to catagory and make a map stored in our mind
3 Metric Axioms : ALL OF THESE ARE VIOLATED
Minimality
dissimilarity between a concept and itself must be the smallest possible
DISPROOF: apple to apple is more similar to pomegranate pomegranate
Symmetry
distance pomegranate to apple versus apple to pomegranate is different.
DISPROOF: an unfamiliar category is judged more similar to a familiar category
Triangle
if one concept is similar to another concept and that concept is similar to a shirt consent then the first and last should be similar
DISPROOF: a lemon is close to orange, an oragne is close to apricot, but a lemon is not close to apricot
Feature-based measure.
Tversky's feature comparison (contrast) model: Similarity(I,J) as a weighted function of features common to I & J - features unique to I -features unique to J. How violations are accounted for.
Smith article
Coding experience by category, categories allow inferences, greater similarity among items within category than between categories
Measurement of similarity: geometric approach, metric axioms, violations of metric axioms, featural approach to measuring similarity, Tversky's contrast model, contrast model's account of metric violations
Similarity and categorization: typicality effects, typicality as similarity. [YOU CAN SKIP SECTIONS 1.4 and 1.5]
Theory-based categorization: results showing that sometimes adults (but not kids) categorize based on theoretical features and reasoning, but not similarity.
Lecture 12: Long Term Memory: Representation and Semantic Networks
Teachable language comprehender (TLC - Collins and Quillian semantic network model).
is a links connect two categories
can make a tree out of this
Hierarchical network structure: family tree like
cognitive economy
highest node that still can take on that definition
save amount of info that needs to be saved
inheritance.
children inherit the characters from their parents
Feature storage (highest node).
Sentence verification task.
uses links to verify statements (category membership)
look into node characters (fratures and properties)
look into links and their node chareecters (inheritance of characters)
Distance effects
more links, more time
Problems:
reverse distance
a dog is an animal versus a dog is a mammal
better at the first statement than the opther deprive the dog-mammal relationship being closer
typicality
some exemplars are more common representations but the TLC model says all exemplars of a category are on the same level
basic-level effects.
most people default tot eh most basic level
Ex: dog versus german shepherd
Revised TLC model with spreading activation
Connectionist model (nodes and connections).
a node is activated hwen a persons are activated or they think about a concept
this activation then spread to connectin/adjacent nodes
Structure:
Not hierarchical.
Links vary in strength.
Explicit information about relations
Intersection search:
spreading activation:
when activation intersection it decides if the itnersection is true
How it accounts for:
reverse distance
links have determined distances that allow for spread to be slower or faster in all directions
priming
McClelland Article
Hierarchical structure, privileged categories, category prototypes.
Quillian’s Model (TLC in lecture): taxonomic hierarchy, the predictions shown to be incorrect.
Semantic dementia and language development as evidence for hierarchical structure (Table 1.1). Illusory correlations: Attributing characteristics from a broad category to an object even though it contradicts perceptual experience.
Basic level: maximizes informativeness and distinctiveness (Table 1.2). Children learn basic labels first. Basic level info retrieved first. Subordinate (more specific) information and labels used more often for atypical objects.
Expertise effects on basic level and in their domain of expertise (Table 1.3). (e.g., use more specific names, equally fast at verifying group membership for basic and more specific categories)
Lecture 13: Long term memory: encoding and retrieval
Verbatim information: vs. gist information:
we remember the gist of the information not the exact wording
when more ideas are stung t-gher the moer likely we are to remember them
In-class demonstration of sentence memory and data.
Semantic vs. syntactic information:
semantic information: meaning
sytantic information: grammatical structure
Sachs (Galileo paragraph) study.
people are good and identifying meaning but not as good with changes in wording
Remembering wording.
DRM paradigm
false memory for gist for semantically related words
Central vs. peripheral information:
central information
if its important ad encoding you will remember it otherwise you wont remember it
Rating importance.
Children extract central info implicitly.
Prior knowledge facilitating comprehension and retrieval (laundry and balloons examples).
prior knowledge tlets you organize memory during encoding and make links with what you know
Prior knowledge hindering comprehension and retrieval: War of the Ghosts.
the brotsih students used their propior cultutrls ideas to frame their thoughts and recall
Schema: general knowledge, meaningfully organizes information, what to expect/what to infer
Event schema (Script) and evidence,
script is ually agreed about
recalled in order
faster reading when in that order
more likely to to recall information from the schema that weren't true
Scene schema and evidence,
remember things in a scene that are part of you schema
spent less time looking at expected things
false memory for thinf that were not there bu where in the schema
memory is not goof if no expectations
Story (Narrative) schema.
story has an order ‘
memory when it follows this script
non-example: oppenheimer-out of order hard toremerber
Schacter Article
Transience: gradual (long-term) and rapid (short-term).
Absent-mindedness: lapses of attention, depth of processing effect, change blindness.
Blocking: tip-of-the-tongue and interference at retrieval, pronounced in old age, nonretrieved items inhibited by retrieving related items.
Misattribution: source confusion, cryptomnesia, false fame effect, Roediger & McDermott experiment, frontal lobe important for monitoring and damage leads to errors & false recognition.
Suggestibility: see Loftus article, but know that it’s one of the sins! Bias: consistency. Persistence: directed forgetting and PTSD
Lecture 14: Long Term Memory: Encoding and retrieval
Using prior knowledge to make inferences.
Logical inference: logiclaly must have it
Spatial relations example.
people make inferences with spatial organization
peiople do not make incorrect inferences
Pragmatic inferences: somehow useful/helpful and maybe true but not right necessarily
in comprehension (fixing the bird house):
didn’t directly say hammer
still made an inference do to the “pounding” word
In advertising:
assertion
asserting a fact
implication
ex: Tabcin pills get through eh winter without cold, take tabcin pills, but it never says the two are related but we still make the inference that they are
never explicitly saying something I related
hedges,
things that imply something si really good and something but not guaranteed or factual
comparisons.
unclear
ex: bananas make you healthier! -> healthier than what?
When are inferences made:
encoding
giving context affects wha tis encoded and then later recalled
storage
native american story from lecture 13
recall of teh sotry led to more distorted 4 moths fater
change in the memory dring storage
schemas affect the memory
forgetting happens
retrieval.
Helen Keller experiment
a sotry was said about a person
at retrieval it was told that it was about h;ankellar for 50% of the people
peeple were then tested if the sentence” she was deaf and blind” was in the text
the helen kellar group asaid yes more often due to prior information
Pathologies:
Misinformation experiments (Loftus): hit vs. smashed.
video of car crash
1 groups question had hit and the other had smashed
asked if there was broken glass to all
poeple who saw smashed thogh the ere was broken class and were influenced by misinformation
Yield sign vs. stop sign.
video of car crash
1 groups question had yeild sign and the other had stop sign
picture of stop sign and yield sign and asked which one is right
poeple who were told stop sign thought it was a stop sign
Overwriting hypothesis: misinformation overwrites memory
Hammer experiment.
video of a man carrying a hammer
group 1 control: man was carrying tool- the correct hammer
group 2 misinformation: The man was carrying a screwdriver - mostly right for hammer identification
group 3:Man was carrying a screwdriver. Hammer or wrench? - people chose hammer. if it was overwritten it should be 50/50 but it was not
Misinformation acceptance.
cognitive interview: let eyewitness tell story uninterrupted, ask questions about events in reverse order, use multiple interviews rather than one long interview.
Encoding specificity: memories are tied to context
Sequential vs. Simultaneous lineups
simultaneous
more likely for false positive
could have a problem with signal detection and have a false alarm
seqauatail
less likely to say yes to anyone
Hypnosis is not very useful
not good at pulling on memories
more eager to corporate
generate a lot of information not necessarily right
Proactive interference
old info affected learning of new information
retroactive interference(inhibition)
new information interferes with the retrieval of old information
ex: remembering your old address after moving
Loftus article
Problems illustrated by Brewster case. Planting false childhood memories (lost-in-the-mall study). Effects of imagining fictitious events. General impairment vs. suspect-bias variables in eyewitness identification. Problem of relative judgments in lineups; solving the problem with sequential presentation.