Exam 1 Flashcards
(127 cards)
1
Q
Neisser (1967) basic definition of cog. psych. and 6 intervening processes
A
Cog. psych. is the study of all processes by which a stimulus is:
1) Transformed
2) Reduced
3) Elaborated
4) Stored
5) Recovered/Retrieved
6) Used
and then produces behavior
2
Q
Our job as cog. psychologists is to figure out what is going on between the _____ and ____
A
IV, DV
3
Q
Organism ____ a stimulus, not just processes it
Ex: if you see a dog in the pattern, your visual system will be tuned to find it quickly again if you see the pattern later
A
TRANSFORMS
Ex: word superiority effect (word-level representations affect how you see a stimulus)
4
Q
What is reduction?
A
Filtering out of other possible stimuli to focus on the most relevant stimulus
*It’s not truly possible to attend to 2 things simultaneously (ex: Neckert cube, changes perspectives)
5
Q
Elaboration: all human memories are ____
A
contextualized (remember not just Balota’s words, but his characteristics too!)
ELABORATE ON INFO. in addition to reducing it–you’ll remember later where you were sitting in class and other traces of memory
6
Q
What is storage?
ANY stimulus ____ neural connectivity
A
As you attend to something, that something is stored so you can access it later!
CHANGES
low-level cellular change (consolidation), re-retrieval of info. INCREASES storage
7
Q
Recovery/Retrieval
“need the right ___ to unlock a memory”
A
can store lots of info but need a RETRIEVAL SITUATION consistent with something that occurred earlier
key
8
Q
Paul Kolers experiment
A
reversed letters in words (eht god=the dog)
had people read words aloud
long-retention studies–brought participants back in –“do you remember reading this sentence a year ago? THEY COULDN’T REMEMBER, yet they read the previously read sentence faster
Need the appropriate test to access a memory trace
9
Q
What is “used”?
A
BEHAVIOR
you measure what is used (reaction time, recall decision making, etc.)
neuroimaging is NOT more powerful measure than simple recall!
10
Q
Kuhn (1962) and scientific revolution
~Built ___ of science starting with ____ (philosopher?)
A
history, Copernicus (Earth is not the center of the universe)
11
Q
Kuhn’s scientific revolution results in scientific paradigm (4 parts)
A
1) Methods & Procedures
2) Assumptions
3) Analogies used
4) Subject matter
12
Q
2 Roots of pressure on paradigm
A
1) Within psychology
2) Other outside disciplines
13
Q
Wundt (1879, Leipzig Germany): _____(what school of psych?)
A
Structuralism!
First school of psych!
Developed out of philosophy
14
Q
Wundt’s/structuralism topics:
Can figure out these topics by _____ (taught people to ____)
A
Language, Memory, Perception, Attention, Emotion (sound familiar?)
introspection, introspect
15
Q
Problems with Introspection (3) (DTV)
A
1) Demand Characteristics (Orne): act of observing influences phenomenon that you’re thinking about
2) Thinking about thinking: can’t get outside yourself, influences thought
3) Some processes can’t be verbalized: they are unconscious! don’t have access to all underlying mechanisms
~ICEBERG theory of consciousness
16
Q
Behaviorism–who?
What was its goal?
A
John B. Watson
Fix problems with introspection
17
Q
Behaviorism placed an emphasis on ____
A
experimentation (manipulate stimuli, measure responses)
18
Q
Black Box Metaphor for Behaviorism
A
Stimulus–>Black Box–> response
no inference of what’s happening inside black box–focus on how I can control the response
19
Q
Ultimately, cog. psych. benefits from:
______ of structuralism
______ of behaviorism
A
Topic of study of structuralism (language, memory, etc.)
Methods of behaviorism (sci. methods)
20
Q
Scientific revolution: we can do both and investigate the topics of ____ with the methods of _____
A
Both! (Investigate Wundt’s topics with behaviorist methods)
21
Q
Cognitive models take ____ from the environment and make _____
A
information, inferences!
22
Q
Roots outside Psych.
1) Linguistics (____ vs. ____ (people) late 50s)
Highly ___: Production vs. _____
A
Chomsky (language in GENES) vs. Skinner (language in ENVIRONMENT)– nature vs. nurture: how do we get language?
Structured: Production vs. Comprehension
23
Q
Roots outside Psych.
2) Neuroscience (____ vs. ____ (areas), H.M.), Neuroimaging
H.M. Case?
Taking chunks out of the ___ ____
A
Broca’s area (production) vs. Wernicke’s area (comprehension)
H.M.: Amnesia, removal of hippocampus, NO declarative memory, same IQ
Breaking up the Black Box, H.M. isolates memory so we can study it separately
24
Q
Roots outside Psych.
3) Computer Science: ___ ____ Metaphor for how brain works
A
Information Processing
25
Levels of Description: Binary code to Google--find a useful level to describe. Which level is NOT most useful?
Neural level not most useful
26
Turing Test
Can computers think?
Interrogator attempts to decide whether machine or person is answering them
27
Any behavior that can be ____ can be mimicked by a computer
QUANTIFIED
28
Chinese Room Problem
Trying to figure out if computer knows Chinese, feed it characters and what they correspond to. Question is: Who knows Chinese? Computer ISN'T thinking: using syntax to coordinate input/output
29
Searle's idea of Syntax vs. Semantics
Semantics=knowing MEANING
Syntax=coordinating input/output
Searle: Computers DON'T have semantics. What do we have that reflects semantics?
30
Churchland & Churchland Reply to Searle
Depends on level of analysis
Where is semantics in these levels of analysis?
We can't find where we have semantics--->we can't dismiss the possibility of computers having semantics!!
31
Neal Miller's Theoretical Approach
- Assumes underlying intervening constructs between IV and DV. Make an inference about what's going on that might tie IVs and DVs together
- Something in common must be going on between the variables
- All IVs are influencing same thing
- All IVs tied to DV
- ex: thirst as a motivator between salty food IV and speed to run maze DV
32
Skinner's Atheoretical Approach
- Assumes each relation is critical (9 relationships, with 3 IVs and 3 DVs)
- Not worrying about underlying mechanisms
- Looking at relationships
- Separate ties between each IV and DV
33
Theoretical Approach allows for __ (#) relationships
Atheoretical Approach allows for __(#) relationships
Theoretical: 6--> it's simpler
Atheoretical: 9
34
Advantages of Theories (3) (PPN)
1) PARSIMONY: decreases the number of relationships that need to be explained
2) Can make predictions
3) Details something about the nature of an organism (able to comment on behavior)
35
Parsimony
- decreases the number of relationships that need to be explained
- simplest explanation
- favors explanations that can be generalized and hold true across a variety of events
36
Disadvantages to Theories (2)
1) May be premature in some cases to build theories (data isn't quite strong enough)
2) Can delay more applied breakthroughs (searching for the best explanation can detract from seeking ways to apply findings)
37
What is critical is the ____ of theoretical and applied approaches
INTERACTION
38
Converging Operations
Used to discriminate amongst alternative theories
Any set of 2+ experimental operations that allows one to select amongst alternative accounts of IV-DV relations
39
Example of Converging Operations with perceptual defense studies based on Freudian theory
Taboo vs. non-taboo words presented, viewer response.
~Are p's more reluctant to repeat the words because perception blocks upsetting material?
~Test: Manipulate stimuli (when you see shit, say igloo, etc.)
~NO: people are simply resistant to say bad words aloud in an experimental context
40
John Platt's Strong Inference
Tree metaphor and discriminating amongst major branches
~Lots of possible "branches" to take: strong inference experiments allow to discriminate between classes of theories
~The lower you go on the tree, the STRONGER the inference because there are fewer possible branches that you can take
41
Pattern Recognition
remarkable ability to recognize distinct sensory stimuli as the same pattern (ex: can read words written in different handwriting)
~Difficult to build machines that can recognize patterns like we can
42
Big Problem for Speech Perception
Perception of phoneme depends on the context in which it's presented (Difference between "rider" and "writer")
43
Phonological Variability Problem
No one-to-one correspondence
Ears receive variation in phonemes, yet you can still understand
Can totally eliminate voicing via whispering and your perceptual system still adapts!
44
Co-Articulation Effects
Speech production changes as a function of upcoming word
Ex: Say "Tackle Lou" differently than "Tackle Jim"
45
Based on co-articulation effects, we know that the perceptual system doesn't parse speech ___ by ____ in a ___ form!
Word by word in a pure form!
It's been difficult to train speech recognition systems to do this
46
Pattern Recognition
Step 1
Figure Ground Segmentation
What's the pattern to recognize? Lower-level perceptual process
47
Gestalt Principle of Proximity
Group things that are physically close together
Used for figure ground segmentation
48
Gestalt Stage of Similarity
Group things with similar visual characteristics
49
Gestalt Stage of Closure
Visual system wants to see objects--wants to close forms even when closure isn't visible
50
Models of Pattern Recognition
1) Template Matching
Match stimulus to internal representation
Set of receptors map onto presented stimulus (template)
Match physical stimulus to best matching neurons--like a bar code!
51
Problems with Template Matching (2)
1) variability in patterns, yet constancy in perception--no templates have that much flexibility!!
2) variability in context (effects of context)
52
Models of Pattern Recognition
2) Feature Analytic Models
Primitive features analyzed first. Pattern recognition is the convergence of these features
Ex: Primitive features for recognizing the letter "b"
53
Gibson & Gibson work on reading development
Critical in ANY domain to identify the features of importance for discriminating patterns within that domain
Major component in perceptual development: "What are the important characteristics to discriminate in this category?"
Brain "tunes" itself to particular features
54
Visual features should do 3 things:
a) survive size differences
b) survive orientation/shading differences
c) be critical in discriminating between features
55
Features in Letter Recognition
Who made the big chart with all the features of letters (vertical lines, symmetry, etc.)
Break letters down into parts that can be identified
Gibson & Gibson
56
Evidence by Gibson et al. for features
feature overlap predicts reaction time
~Speeded same/different recognition task
~Feature overlap predicts reaction time: KY is slower to reject than EF because KY was 5 overlapping features vs. 4 for EF
~Compare with template model where EF is more similar
57
Pandemonium Model by Neisser and Selfridge
One of the first computational models of perception
~"Demons" in pattern recognition
~"Image demon"= "iconic representation" when stimulus is present
-->Activate set of "feature demons" that recognize features
--> "Decision demon" listens to which letter is screaming the loudest
58
Serial and Parallel Characteristics of Pandemonium Model
Serial: Features--> Letters--> Decision Demon
Parallel: When "R" is represented: features are simultaneously activated
59
Problem with Pandemonium Model
No account for context effects
Purely a bottom-up model
60
Converging Evidence from Neuroscience in visual system
Mapping of millions of photoreceptors onto gangleon cells in each eye
~Convergence of information to neurons--> receptive field=window into visual world that a neuron is receptive to
61
Evidence for ability to fatigue receptors and receptive fields
Afterimages after consistently stimulating a neuron
62
Who found the experimental evidence for this ability?
Slow drift
Physiological Nystagmus
Pritchard et al.!
Stopped images--held eyes locked on a certain image, observed slow drift( if you fixate at a location, your eyes will actually drift a bit) and physiological nystagmus(high frequency "jittering" of eyes due to inconsistency in muscle tension: PHORORECEPTORS CONSTANTLY ACTIVATED
63
Pritchard et al. study continued--tried to stop eye from moving
What did they find?
Used a contact lens that acted like a slide projector--whenever eye moves, projected image moves too
Found that pieces of the stimulus went away but the whole stimulus never disappeared when these receptors were fatigued (contrary to what template theory would suggest).
Instead, FEATURES of image faded in and out
64
Letvin et al. Study with frogs
What feature detectors did they find?
Single cell recording in frog optic nerve--significance of feature detectors
1) Net dimming detectors: when light incr/decr., optic nerve areas fire
2) Net edge detectors: detect where to jump
3) Net convexity detectors: lots of output when something jerky goes across visual system, like a bug
65
What did Letvin notice about these specific feature detectors?
They all detect necessary components of a frog's world!!!
66
Hubel & Wiesel cat experiments
Found specific neurons sensitive to ____?
~Recorded from visual cortex
~Trace optic receptors by inserting electrode into cat visual cortex
~Present patterns to visual system and observe how RECEPTIVE FIELD in brain "lights up"
~Can find neurons that are particularly sensitive to patterns
~Specific neurons sensitive to specific angle, location, orientation
~Verified sensitivity of neurons in visual cortex
67
What 3 types of cells did Hubel and Wiesel find in the visual cortex?
1) Simple cells: Sensitive to lines in specific orientations
2) Complex cells: Represent larger visual receptive fields and are sensitive to motion with preferred direction
3) Hypercomplex cells: Sensitive to angles and ends of lines
68
Area MT in humans mapped through these experiments
In temporal lobe
Lesions here--> can't see continuous motion
Neurons in MT area have to fire in appropriate sequence to perceive motion
69
Converging Evidence from Neuroscience (Disappearing Dot Illusion)
Afterimages and detector mechanisms
~Spiral illusion and motion-induced blindness (Disappearing Dot Illusion fatigued feature detectors
~Expectancy from motion of what comes next--> images "go away"
70
Context & Letter Perception: Cattell (word superiority)
Letters in briefly presented strings are reported better in WORDS than NONWORDS
Conundrum: if we must recognize letters to figure out the words then how can word-levels influence our perception of the letters?? TOP-DOWN INTERFERENCE
71
Alternative accounts for Cattell's theory (2)
1) Possibly people quickly recognize words and report letters from spelling knowledge (not really associated with letter perception)
2) Maybe people are just good at guessing based on approximations to English ORTHOGRAPHY (what's the likelihood of one letter following another?)
72
Zietgeist & Reicher-Wheeler Try to remove alternative accounts
What did they find?
Create situation where people can't guess as easily
Word followed by pattern, then a mask (Condition 1: WORD; Condition 2: OWRD; Condition 3: _ _ _D)
~Forced choice: _ _ _D or _ _ _K (WORD or WORK?)
Found that people were better recognizing "D" in context of WOR than OWR or _ _ _D--> word superiority effect still applies!
73
New Modeling Approach
McClelland & Rumelhart's ____ _____ framework
Interactive Activation
Similar to Pandemonium Model: Bottom-up & Top-down processes
74
Cascadic Processing
Continuous flow of activation that constantly updates information as opposed to staged processes
75
Within-level inhibition
Inhibition in each level of processing
76
Interactive Activation & Top-Down Reinforcement
As word representations get activated, there's also activation from word-level representation in top-down support for the letters that form them--> serves as REINFORCEMENT
Perception is ALSO from higher levels influencing lower levels
77
Bottom up: from ___ to ____
Top down: from ____ to ____
letter to word
word to letter
78
IA Model has similarities to which model?
Pandemonium Model
79
2 Types of Pathways in IA Model
Activation, Inhibition
Interaction of bottom up and top down activations
80
IA Model: when "WORD" is presented
K is still activated before "D" appears because WORK is acceptable, too
81
Phonemes
Building blocks of language
Speech based on 44 phonemes as opposed to 26 letters
82
Human articulatory system
Mouth/nasal cavity, tongue, lips, teeth
Notice place of articulation for particular sounds
83
Bilabial, labiodental
Bilabial: "b" or "p" (bring lips together)
Labiodental: "v" or "f"
84
Stops/Plosivity vs. Fricatives
Stops: totally close vocal tract (p, t, k)
Fricatives: bring upper teeth with lower articulator but don't totally stop sound ("f" sound)
85
Features of vowels (2 sets of 3 varieties)
High, middle, low
Front, center, back
86
Features of Consonants (3) IMPORTANT
1) Plosivity: totally close vocal tract in an abrupt release of air (p, t, k, b, d, g) vs. Fricatives: don't totally close, air produced throughout phoneme (v, f )
2) Place of Articulation: bilabials, alveolar ridge (t and d)
3) Voicing: synchronicity of release of the air in vocal chord vibrations
~"ba"= voiced, cord vibration synchronized with air
~"pa"= voiceless, delay after release of air
~"p", "t",= voice less; "b", "d," = voiced
87
Evidence for Feature Detection
____ Speech Perception vs. ____ Speech Perception
What type of experiment?
Categorical Speech Perception
~Speech synthesizer, manipulate interval between release of air and voicing--strict boundary between hearing "ba" and "pa"
Continuous Speech Perception
~NO strict boundary between voice onset time in "ba" and "pa"
88
Liberman et al.: Evidence for ____ Speech Perception
Problem with this type of experiment?
Categorical Speech Perception
~Participants detected cutoff at 30 ms. between "ba" and "pa"
No detection between 60 ms. and 30 ms. "Pa"-- which is the better "Pa"? Is there a way of testing if people can hear difference?
89
Converging Evidence for AB, and X Paradigm
Main Point?
A or B match test stimulus X? Does not rely on labeling, truly perceptual
Provides evidence for feature and voiceless detectors--can't hear difference within categorical boundary
Main Point: Can't hear 30 ms. voice onset difference when WITHIN categorical boundary (voiced/voiceless). Evidence that we have detector that is either ON/OFF at one time
90
Can you fatigue a feature detector between "Ba" and "Pa"?
Who tested this?
Finding?
Present "ba, ba, ba, ba..." repetitively
Eimas and Corbit: repeat "ba" at beginning of stimulus
*Before you give repeated ba, critical boundary between Ba and Pa is 20 ms! After adaptation to repeated Ba, boundary decreases to 12.5 ms! You hear "pa" at a VOT where you used to hear Ba because the Ba voiceless detector has been fatigued!!!"
91
Problem with this fatiguing study?
If this is voice detector, you don't repeat "ba" but should repeat something else like Da. Want to fatigue voice detector, NOT the phoneme detector for Ba
Effect occurs even after "da" repetition adaptation (15 ms)
92
Elaborative Encoding
Any sort of memory technique/mnemonic
Assigns meaning to stored information
93
Bizarre Image
powerful type of elaborative encoding: make image look weird (imagine melting Dr. B)
94
Method of Loci
Interactive imagery technique-- create memory palace with things in it that you should remember, give "tour" upon recall
95
Memory Experts Characteristics
NOT high IQ or attention; they just practice a lot!
96
Should space study out when you have ___ time, cram when ____ time
more; less
Retention Interval is critical to bear in mind!!
97
5 steps in identifying words (basis of language comprehension)
1) Features are detected
2) Phonemes are detected
3) Phonemes combined into syllables
4) Syllables--> words
5) Word meaning retrieved from memory
98
Spectrogram
Analyze characteristics of speech
Plot frequency of produced sound across time--difficult to see distinctions between individual words
99
Problems with Speech Perception (3)
1) Words aren't neatly segmented
2) Difficult to identify phonemes (coarticulation effect where phonemes are affected by neighboring phonemes)
3) Speaker differences: pitch affected by age, gender, dialects
100
2 Models of Spoken Word Identification and People
1) Cohort Model (Marslen-Wilson 1989)
2)Trace Model (McClelland & Elman 1986)
101
Cohort Model
Uniqueness Point?
Speech unfolds across time
~Ex: "Speed": First receive "s"--> activate lots of possible "s" words--> sp words--> spee words--> speed
~Uniqueness point: Point that identifies a word above all others!!!
102
Revolutions occur when:
something in funky in current scientific paradigm--inconsistencies!
103
word superiority effect
word-level representations affect how you see a stimulus
104
Demand Characteristics (Orne):
act of observing influences phenomenon that you're thinking about
105
Thinking about thinking:
can't get outside yourself, influences thought
106
Some processes can't be verbalized:
Some processes can't be verbalized: they are unconscious! don't have access to all underlying mechanisms
~ICEBERG theory of consciousness
107
chomsky
language in genes
108
skinner
language in environment
109
2 Models of Spoken Word Identification and People
1) Cohort Model (Marslen-Wilson 1989)2)Trace Model (McClelland & Elman 1986)
110
slow drift
if you fixate at a location, your eyes will actually drift a bit
111
physiological nystagmus
high frequency "jittering" of eyes due to inconsistency in muscle tension
112
Simple cells
Sensitive to lines in specific orientations
113
Complex cells
Represent larger visual receptive fields and are sensitive to motion with preferred direction
114
Hypercomplex cells
Sensitive to angles and ends of lines
115
Cohort Model Uniqueness Point?
Speech unfolds across time~Ex: "Speed": First receive "s"--> activate lots of possible "s" words--> sp words--> spee words--> speed~Uniqueness point: Point that identifies a word above all others!!!
116
Early Uniqueness Point?
Fast to recognize a word
117
Audition demands _____ processing, not ____ processing
Serial, NOT parallel processing
118
Tanenhaus 1998 analyzed speech perception with ___ _____
Eye trackingTask: Images on screen; hear "point to the beaker": do your eyes get pulled over to "speaker" or "beetle" due to cohort effects?Moment by moment evidence of what's being activated
119
Tanenhaus Findings
From 200-400 ms: eyes likely to move to both "beetle" and "beaker", small activation for "speaker"System processes phonological characteristics of competitors in displayConsistent with COHORT MODEL
120
Trace ModelSimilar to ___ ____ model? Why?
First-level features picked up and unfold across time. Feed into phoneme layer--> feeds into word layer~Similar to Interactive Activation Model because of within-level inhibition and suggestion of top-down influence from word-level to phonemes
121
Top-Down Effects in SpeechPhonemic Restoration: Which study?
Warren and Warren~Played a sentence to subjects where portion of speech was replaced by a cough. Asked subjects to identify which phoneme was missing~Word "legislature" applies TOP DOWN INFLUENCE--you think that the "s" sound is there because of knowledge of the word even though the cough is covering it
122
Top-Down Evidence when you hear background noise
Rely on top-down info about words to more easily fill in the sounds that you can't hear
123
Verbal Transformation Effect
Syllable/word played over and over: Word CHANGES over timeFatigues feature detection and representations; other detectors take over and you hear something that isn't there. Lexical representations "take over"
124
Effect of Semantic Context StudyWho?MAIN POINT?
Pollack & Pickett~Recorded several conversations~Subjects had to identify words in the conversation~When words were spliced out of the convo. and presented auditorily, subjects identified the correct word ONLY 47% OF THE TIMEClarity of individual words within speech is an illusionCan understand what's being said in a convo, but you need the context
125
Alternative Descriptive ModelMotor Code Theory of Speech PerceptionWho?Evidence of tie between ____ and ____
Liberman & Cooper~Relies on motor codes of speaking~Recognize the codes you would use to produce speech~Evidence of tie between production and perception--when hearing speech, produce codes linked to how you would produce it
126
Motor Code Theory: think of ___ as a consequence of ____
Perception as a consequence of productionPart of your perception of the word "dog" is how you would say it
127
Pattern Recognition in Vision and AuditionAnalysis by ____We ____ perception
Analysis by SynthesisUse lots of sorts of info.What we perceive is synthesized--not purely provided by the worldWe CREATE perceptionBottom-up & Top-down processes at work
