Cognitive Final

• Paying attention requires effort
• Attention functions like a spotlight
• Some things automatically attract attention: they "Pop Out"
• There is limited-capacity route to attention

Task 1: press a button to a probe tone --> use foot pedal to disambiguate from hands OR say "high" or "low" to high or low tone

Task 2: press button 1 if an "X" is in the string of letters, press button 2 if no "X" appears; letters can be clear or fuzzy

Both tasks: require 3 stages of processing: Stimulus recognition, Central attention and response selection, response execution

• Increased difficulty (fuzzy letters) = stimulus recognition takes longer
• Some processes cannot be shared = 1st stimulus must complete processing before 2nd can access stage processes
• Wait time produces RT2 increase, the PRP effect

• Task 1: search for and identify designated target among distractors
• Task 2: Indicate if probe letter appeared in the stream

• If T1 is correctly identified --> decrease in performance across next 6/7 items
• If T1 is omitted, there is no performance deficit

Appearance is that of a "blink" in attention

• Information input is critical: blank after T1 disrupts effect, Masking is necessary--limits info on T1 adds additional info to be inhibited
• Distinctiveness/Similarity of targets and distractors: Symbols less effective than digits at disrupting letter processing, in figure: legend indicates type of distractor immediately following T1 and T2

• Resource Availability
• Level of Automaticity
• Baseline activation of information
• Motivational Salience

• Automaticity occurs after practice
• Practice may make some stages more efficient, but does not allow parallel processing

Same memory set every trial

Different memory set every trial

• Biases attention to new locations
• If searching, biases search away from previous (unproductive) locations

• General IOR function observed reduced IOR for LVH stimuli in OCD
• Suggests lateralized deficit in volitional attention

Alerting, Orienting, and Executive network

Achieving (maintaining) state of heightened sensitivity to incoming stimuli

• Selecting attentional focus
• Moving and engaging the spotlight

Maintains and sustains attentional focus

• Locus Coeruleus: Norepinephrinergic distribution through cortex, increased arousal
• Frontal Cortex
• Parietal Cortex

• Responsible for overt and covert attentional selection
• Parietal lobes: Disengagement of attention, Indicate that other stimulus is requesting attention, system is damaged in neglect syndrom, over activation may lead to excess distractibility

Main function is eye movement control, tracking

• Selects the new target in the spotlight
• Engages attention

• Anterior Gyrus
• Lateral Prefrontal Cortex
• Maintains, sustains attentional focus
• Damage produces rumination

• Reduces complexity of the environment (chunking)
• Allow rapid, easy identification of objects
• Reduces need for constant new learning
• Evaluate relationships among objects
• Quick response selection

• Emphasis on true forms
• Mind was a shadow of the external truth

• Mind contains ideas
• thoughts, images, sensations
• Ideas are the fundamental unit of consciousness

• 1.) The fundamental unit of thought for modern philosophers
• 2.) Concepts organize information, similar to categories, define categories
• 3.) Simple concepts can be expressed by language, but the linguistic definition is NOT identical to the concept

• Information can be part of conceptual knowledge but not necessary for definition
• Linguistic definitions are NOT conceptual categories

• Linguistic definitions
• Hypothesis testing
• Developed by questions/answers with adults

• Single: cards with red; cards with circles
• Conjunctive: cards with red circles or cards with 2 borders and 2 shapes
• Disjunctive: cards with either a read shape or a square of any color

• Wholist (reception): take first positive instance, note attributes, each subsequent instance eliminate any attribute that doe not recure
• Conservative: Find 1st positive instance, test additional cards varying by one dimension from criterion card

• Categories with fuzzy borders
• artificial categories designed as unambiguous
• Natural categories have ambiguous 'sort of' members
• Caminalcules

Summaries of concept

priming

Typically a basic level: level at which people most commonly processes information

• Commonalities among members
• Differences with non-members
• Alignable differences: based on a commonality
• Nonalignable differences: not based on commonality

• 2 basic categories, same superordinate: (bed-couch, horse-cow)
• 2 basic categories, different superordinate: (bed-horse, couch-cow)
• 2 superordinate categories (furniture animals)

• Tend to organize new information based on previous info
• Generate arbitrary criteria
• Prior knowledge affects category formation (Piaget redux)

Dictionary: internal, mental representation, known words

• Lexical decision: is stimulus a word (y/n)
• Word naming: say word quickly

• Frequency
• Familiarity
• Shorter RTs for frequent or familiar words

• Present word
• Repeat item on subsequent trial = shorter RT

Present word, activation flows to nearyby nodes, access to activated nodes appears as shorter RTs

Concepts represented by nodes, nodes connected via links

• Teachable-language-comprehended, by R. Quillian
• represented as a hierarchical tree
• Prediction: number of nodes to intersection will be associated with search time
• Problems with TLC: doesn't account for the effect of basic categories or prototypicality and fuzzy borders

• Each concept is a node in the network
• Nodes related by links, not hierarchy
• Related nodes close together (central members of categories)
• Unrelated nodes far apart (peripheral members of categories)

• Subjects reject "school is a bus" or "school is a house" quickly
• In SAM, school and bus close
• In TLC nodes far apart--requires more time to reject

• propositions or a relationship among concepts
• Predictions: responses to questions should be quicker within propositions compared with between propositions (concepts + relationships)

• concept-like info about events and related concepts
• a way of organizing concepts and info into practical, useful sets
• Proposition template

• A concept regarding a typical sequence of actions in common events
• Proposition template
• Derived from schemata, organizes info about an event
• Makes events familiar and comprehensible

a reconstruction with schema filling in the details

• Info stored in semantic network
• Events represented as propositions
• Categories or groups of events may represented in schemata: scripts detail event sequences with schemata
• Episodic info has modality-specific nodes
• Knowledge can be stored in networks of neurons
• Knowledge must be acquired and accessed: the process of learning and memory

• Modality specific conceptual systems
• Semantic networks: relatedness: semantic assoc., semantic memory
• Propositions: relatedness: contextual organization

• Changing modality increased sentence verification times
• Semantic association alone did not reduce RT

• Perceptual neural circuits encode stimulus
• Memory involves top down reactivation of same code (neural patterns)
• Neural circuits-metaphor underlying neural network models

• Node: units, simulate neurons, receive input from multiple sources, connected for output to multiple target
• Implemented as software

• 2 input nodes
• 1 output node
• no intermediate (hidden layer)

the weighted sums of the inputs

• Familiarity, Memory, and schemas
• Larger objects perceived more quickly

Learning has occurred when there is a relatively long term change in behavior that results from experience

• Both require changing behavior: Learning: process of acquisition, Memory: implies learning
• Each is evidence of the other: Learning implies memory, and memory implies learning

Decrease in the strength of the response following repeated presentations

• Unrelated stimulus increases response strength
• Habituate to gunshots @ firing range...attractive range instructor

• Increase in strength of response following repeated presentations
• Soldiers sensitize to nearby artillery shells
• Sensitization generalizes PTSD
• High intensity: sensitization
• Moderate intensity: habituation-sensitization

• Both processes evoked
• Example: opponent-process often lead to after effects (visual after images, emotional, etc.)

• Appetitive: Meat powder conditioning in hungry dogs, US is pleasant, satisfies a need or motivation
• Aversive: Eyeblink conditioning, US is unpleasant, normally avoided, causing discomfort

• Habituation
• Acquisition
• Extinction
• Reacquisition

• Pair taste with nausea
• novel food in the environment - nausea
• Lab: novel taste in water with injection of LiCl
• ISI = 2 hours

• CS may have some US value
• present set of CS-only trials
• Decreases US-UR association
• Latent inhibition

• CS-US association trial
• Intersperse CS-only test trials to assess development

• Repeated CS-only trials
• Assess stability of newly acquired conditional effect

• Assess savings
• spontaneous recovery
• phase may delayed (hours, days, weeks)

• Most typical procedure
• CS precedes US
• CS and US co-terminate

• CS terminates prior to US representation
• more difficult
• requires hippocampus

• CS+ predicts US
• CS- does not
• CR appears to CS+, but not to CS-

• Reactive to external stimuli
• Tend to withdraw
• Condition easily

• Less reactive
• Condition less easily

• CS2 through association with another CS1 produces CR
• Wasp sting --> fear to sight of wasp
• Tool shed --> wasps
• Tool shed --> fear, anxiety

• Like higher order
• CS2-CS1 association precedes CS1-US association
• Tool sheds (already assoc. with wasps) --> fear, anxiety

• A behavior followed by satisfying consequences will be 'stamped in'
• A behavior followed by an annoyer will be 'stamped out'

• Emitted behaviors operate on the environment
• Reinforcers increase the probability that an emitted behavior will be produced again
• Punishers decrease the probability
• Skinner Box- the standard laboratory equipment of the 50s and early 60s

• Learning results from association of stimuli
• Contiguity produces association

• Garcia effect no predicted
• Free-US training
• Blocking

• A cognitive approach to learning
• Learn that CS predicts (is informative regarding) US
• Response depends on info content of stimulus
• US does not proved additional (predictive) info regarding US

• Emphasize attention to CS
• Elements available for association
• Associative strengths determine conditioning

• Learning results from formation of Cell Assemblies
• Covactivation of 2 neurons-synapse between them is strengthened
• Synapse where change occurs is known as a Hebbian synapse
• Hebbian and Rescoral-Wagner approaches are not necessarily mutually exclusive

• If 2 neurons (nodes) are active simultaneously
• Then the synapse (weight) should be strengthened

• type of learning
• network learns by contiguity
• but contiguity not necessary for learning
• Outcomes of behavior affect subsequent behavior
• alternative strategies for 'teaching' a network have been devised

• Associative strength between 2 stimuli is a function of the difference between the present strength and the target (expected) strength
• A behavior produces suprising outcome --> rapid change in association
• A behavior produces the expected satisfying outcome --> little change in associations
• Principles implemented in connections networks as the Delta Rule

• learning is proportional to unexpected results
• learning does not occur in a single cycle

• Storage not in any one, single location
• no one memory trace or engram, each cell contributes to many memories
• memory is a property of how all the nodes are connected with each other
• network operates as in accordance with Lashley's principle of mass action

• Δwit: is the change in the weight connecting nodes at twolayers (e.g., input and output)
• lr: is the learning rate, typically assumed to be thereciprocal of the number of input nodes
• ai: is the activation level of the input unit
• at: is the correct activation level of the target output unit
• (i.e., You know if the two nodes (or neurons) are active or not)

• Used nonsense trigrams in study of memory
• Wanted to study pure memory
• How to exclude influence of prior experience
• Trigrams could be totally novel
• Did not require ‘habituation’ as often used inconditioning procedures

• Relearning takes fewer presentations than originallearning
• Savings = (trialsoriginal – trialsrelearning)/trialsoriginal
• If perfect savings: (20 – 0)/20 = 1.0
• If no savings: (20 – 20)/20 = 0
• If some savings: (20 – 12)/20 = 8/20 = .4

• Most forgetting occurs in1st hour after learning
• Rate of loss is not constant0.81
• Counter to beliefs of thetime
• Rate of loss decelerates
• Relearning is alwayseasier than the original
• Something is alwayssaved

• Decay
• Interference

• Decay: Associations become weaker --> occurs in perceptual buffers, short-term storage, access links to stored info decay with disuse
• Memory traces fade: Ebbinghaus’ explanation
• Memory strength depends on: Original strength (original learning), Time since learning (savings)
• Interference: Memory cannot be retrieved because --> occurs when new info is placed in perceptual buffers, checking access to long-term storage

• Primary memory: Knowledge of which we are currently aware
• Secondary memory: Information available for recall, Things we could be aware of
• Although not a direct antecedent, is similar toa number of modern models including AS3S,the modal model

• Three sequential registers (or buffers): Each can store information, Differ on characteristics of capacity, duration, etc.
• Information not transferred to the next store is lost
• Information maintained in STM by Rehearsal
• Executive control areas have access to memory stores
• Response selection, attention, etc.
• 

• Based on Sperling’s work
• Modality specific
• Iconic – visual
• Echoic - auditory
• Fairly large capacity
• 12 – 20 items ( or more)
• Short Duration:.25 - .5 s for the icon, 2 – 4 s for the echo (??)
• Exhaustive retrieval available(with appropriate cues) )
• Vulnerable to masking and decay

• Based on Miller’s and the Brown &Petersons work
• Contains exact sensory features
• Small capacity: 7 ± 2 items, Items may be ‘chunked’
• Duration of 18 - 30 s
• Exhaustive retrieval available(within time limits)
• Vulnerable to both interference and decay
• New items displace old

• Provided letters for recall: Example: J V G M T
• Asked to count backwards out-loud by 3’s: Begin with different number each trial, Decreases ability to rehearse letters, Time for counting varies: 3, 9, 18, 30 s
• Recall letters
• Results: Poorer recall withgreater interval, Almost no recall at 30 seconds

• What most people refer to as‘memory’
• Stores semantic information: Gist, abstracted material, NOT exact sensory information
• Very large (infinite) capacity
• Very long (infinite) duration
• Partial retrieval: depends on cue, reconstructive
• Retrieval failure due to poor cues, interference, brain injury

• List learning: 36 words, Presented 3 s per / word, Recall words
• Serial Position Effect: Position of item in listpredicts likelihood ofsubsequent recall
• Items at end of listremembered better
• Usually the first to be recalled: Immediate awareness, Dumped onto responsesheet ASAP
• Result of STM: If 1 minute delay with fillertask (spoken arithmetic) --> effect reduced

• Items at the beginning of listhave higher probability of recall
• Usually receive the most rehearsal
• Most likely to beconsolidated
• Result of LTM
• If quick presentation rate (1word / s) to preventrehearsal--> primacy effect reduced

• Most common amnesia: anterograde – inability toform new long-term memories: STM intact – digit span, word span are normal, Normal on tests of implicit memory/procedural memory, LTM retrieval is normal, Cannot learn new information for long term storage
• Retrograde amnesia is the inability to retrieveprevious LTM information: Never occurs without anterograde amnesia