Cognitive Psych - Exam 3

Power Law of Learning

How practice relates to performance. Initial gains in skill are larger; however it takes longer to get from an intermediate level to an expert level.

Cognitive stage

• one of the stages of skill

• Declarative knowledge 

• Commit facts to memory 

• Rehearse as you try to perform 

• Requires attention - can’t do second task 

• May be independent of skill: the best teacher may not be the most skillful, but rather someone who knows how to describe this stage well

Associative Stage

• one of the stages of skill

• Strengthen connections that lead to desired result 

  • Feedback is important: see which actions lead to desired result

• Get rid of actions that lead to errors 

Automaticity

• one of the stages of skill

• Fast

• Executed with less attention/consciousness 

• Less verbalization

  • Less dependent on verbalization

  • Declarative knowledge less available (harder to describe what you are doing at this stage. Golf example in class)

Proceduralization

Take declarative knowledge and turn it into productions

Composition

take several productions and join them into one

Serial Reaction Time Task

• Four boxes on a screen 

  • Each one matches up to one of your fingers 

    • As soon as one of the boxes lights up you are supposed to press that button 

• You could have them light up randomly or have them light up in a sequence 

  • Ecplicit training - “look for the sequence”

  • Implicit training - “it’s all random” 

• Even if its just random you get better from repeating the task 

  • However, there will be an added benefit if it is a sequence 

• When tested implicitly you dont know that you have learned the sequence

  • Your performance is better whether you know that you learned a sequence or not

Multiple Learning strategies that operate in parallel

• Fast system(s)

  • Large amount of learning per trail that saturates quickly 

  • Requires extra time, cognitive resources for processing

  • Accessible to awareness and conscious intention

  • Flexible 

• Slow system (s)

  • Small, incremental amount of learning per trail 

  • Learns automatically without effort

  • Impenetrable to awareness, operates independent of conscious strategies 

  • Inflexible (habits) 

Response Chaining

• Feedback from one movement triggers the next one

  • Feedback from the first response is the stimulus that tells you to respond to the next one

• Originally a behaviorist notion

• typing speed provides evidence against this idea

Motor Program

• Representation of the plan for movement and movement sequences 

• Fast doesn’t require feedback 

• Abstract 

• Hierarchical 

  • Abstract high level 

  • Specific low level 

• Composed of subprograms 

  • Less abstract representation of movement subparts 

Signature looks similar for small and large versions

Provides evidence for Abstract motor representations and hierarchical representation

Cerebellum

• Known to:

• Coordinate motor movements (e.g. aiming towards a target)

• Control the muscle tone, the base level of tension in a muscle

• Prime the motor system just prior to the onset of a movement 

• Contribute to motor leaning (e.g. classical conditioning)

Basal Ganglia

• Set of nuclei located in the cortex

• Involved in directing the movement of limbs and in coordinating programs for automatic action

• Impairment of the basal ganglia -> parkinson’s disease

Speed-accuracy tradeoff

• Movement time increases more as one increases the distance from one to two inches than when the distance is increased from 10 to 11 inches 

• As precision is increased, movement time increases

• Movement time increases as distance between trafets increased and their width decreased 

• Subsequent research (Keele, 1989) confirmed the generality of expression for a wide range of movements and targets 

Fitt’s Law

• The greater the distance to the target, the greater the number of component movements and adjustments necessary 

• Collectively, the submovements are assumed to optimize motor performance, much as problem-solving activities are optimized given one’s limits in memory and info-processing capacity 

Properties of Language

Communicative

• the purpose of language is to communicate

Language is Symbolic 

• Language creates symbols that make reference to things, ideas, processes, relationships, and descriptions, etc. 

Language is (mostly) arbitrary 

• Why is it “dog” vs “perro”?

• Why does “dog” mean what it does and sound the way it does 

• It is not the physical combination of sounds/symbols that are giving it meaning 

• It is just a common set of conventions that speakers of a language adopt 

• Mostly… (kiki and Bouba)

Productive (generative)

• Within the limits of a linguistic structure, language users can produce an infinite number of novel utterances 

• The possibilities for creating new utterances are virtually limitless 

• Say something entirely novel right now!

  • We dont learn to imitate, we learn to generate 

Dynamic (constantly evolving)

• Selfie 

• What slang terms do I not know? 

  • Body is tea 

  • Rizz

Structured at multiple levels (hierarchical)

Infant directed speech (motherese)

• High spitch, slow rate, exaggerated intonations 

• Falling pitch and pausing signals phrase boundaries 

  • Aids parsing 

• Infants prefer to listen to this

  • Head turning 

• Might help kids acquire word boundaries 

Holophrastic Stage

• One word utterances

• No syntax, need context (gestures, affect) to disambiguate

• Undergeneralization and overgeneralization for first ~ 75 words 

  • Ex: Dog = other small mammals

• Do understand some phrases 

  • Can only produce one word

Telegraphic Stage

• Two word utterances 

• Correct use of word order:

  • Subject-action 

  • Action-object 

• Can convey a lot of information succinctly (like a telegraph)

Problem

• Some initial state where you start out and some goal state that you are trying to get to

  • And some methods that help you get to the goal state

• consists of some initial state in which a person begins and a goal state that is to be obtained, and a non-obvious way of getting from the first to the second 

Two types of problems

• Well-structured (well-defined)

  • Completely specified starting conditions, goal state, and methods for achieving the goal 

    • EX: geometry proof 

• Ill-structured (ill-defined)

  • Finding a mate 

  • Choosing a career

  • Writing the best novel 

• Most of the problems we encounter in our lives 

  • Difficult to study 

Problem Space

Whole range of possible states and operators, only some of which will lead to goal state 

Operators

Actions that move between states

Analogies

• Retrieve a representation of a problem from memory that is similar to the problem you currently face

• People tend to miss deep similarities between problems, because they tend to focus on surface similarities

Functional Fixedness

• see an object as having only a fixed, familiar function

• EX: Dunkers’ candle problem and Maier’s Rope Problem

Getting “Stuck in Set”

• People will continue to use the same way to solve the problem, even if there is an easier solution (if they have done it enough times in a row) 

  • If there is a situation where only the easier one works and the other one does not, you will struggle 

• Top-down hindrances  

  • If you think a problem should be solved in a particular way it could hinder your ability to find/use easier solutions

Algorithm

• Completely specified sequence of steps that is guaranteed to produce an answer

  • Usually guaranteed to produce the correct answer 

  • But may be slow or laborious 

Heuristics

• Short cut/”Rule of thumb”

  • Never guaranteed to produce correct answer

  • But usually quick and easy 

Difference Reduction (hill climbing)

• At any point, select the operator that moves you closer to the goal state: is new state more similar to goal? (never choose an operator that moves you away)

Means-end Analysis

• Identify the largest difference between current state and goal state

• Set as a subgoal reducing that difference

• Find and apply an operator to reduce the difference 

  • If operator can’t be applied, new subgoal=remove obstacle that prevents applying the operator

• can involve working backwards

Expert Problem Solving

• Rich, organized schemas 

  • Lots of well-organized declarative and procedural knowledge

  • More sophisticated representations

• Spend more time on representation

  • Experts take longer to start a solution, but less time to complete it 

• Recognize subcomponents 

• Less means-end analysis (using fewer heuristics)

  • Pre-stored solutions in long-term memory 

  • Fewer demands on working memory 

• Move forward, not backward 

Deterministic/deductive

• General to specific 

• Detecutive: theory -> hypothesis -> observation -> confirmation

Probabilistic/inductive

• Specific to general 

• Inductive: Observation -> pattern -> hypothesis -> theory 

  • What we use most in day to day life 

Normative theories of reasoning

• How one ought to reason

  • E.g.; rules of logic (Bayes' theorem)

Descriptive theories of reasoning

•  How people actually reason 

  • E.g.; biases, heuristics 

Base Rates

• When this is 50/50 people tend to pay too much attention to it and get anchored there

• people tend to ignore this when it is NOT 50/50

Availability

• People estimate frequency or probability by the ease with which instances or associations can be brought to mind 

  • May be biased because highly publicized illness may be easily “available” because of the media attention it receives 

Representativeness

• This heuristic says that the probability of an event is estimated by the degree to which it fits an existing cognitive stereotype 

• leads doctors to ignore base rates, make pseudo diagnostic judgements, to misinterpret random events and to misunderstand statistical regression

Gamblers Fallacy

•  the unwarranted expectation that every sample must represent the population mean

Hindsight Bias

•  occurs when our present knowledge is allowed to influence our estimates of the likelihood of previous events 

  • Suggests that doctors may be reasoning backwards (start with diagnosis and adjust the predictive value of the signs accordingly)

Anchoring and Adjustment

• Recommends that probabilities be estimated by first beginning with an “anchored” probability value and the adjusting of this value according to the features of the specific case 

• The initial placement of the “anchor” can have an unduly large influence on final judgements 

• Closely related to availability and representativeness 

Conjunction Fallacy

• Occurs when people mistakenly believe that a conjunction of events (hot and sunny) is more probable than a single even (hot). 

  • The ing and n example is an example of this (avaliability heuristic)

How a problem is framed

• One is framed as a gain, the other is framed as a loss

• Gain frame -> risk-averse (avoiding risk)

• Loss Frame -> Risk-seeking 

Sunk Cost Fallacy

• Tendency to continue an endeavor if we have already invested time, effort, or money into it, even if the costs outweigh the benefits 

  • You might as well continue to make the jet if it is already 90% completed, even if you won’t be able to sell it 

Quantifier (categorical syllogism)

•  (how much of some particular group is in a particular category) 

  • Some businessmen are wealthy 

  • All wealthy people are powerful 

  • Therefore, some businessmen are powerful 

Comparative (Linear)

• Mighty Joe Young is more powerful than Godzilla

• King Kong is more powerful than Mighty Joe Young

• Therefore, king kong is more powerful than Godzilla

Conditional

• If you write a good research proposal, then you will get funded 

• You write a good research proposal

• You will be funded 

Quantifier types

• All doctors are rich (Universal Positive) (positive = they have some property)

• Some lawyers are dishonest (Particular positive) 

• No politician is trustworthy (universal negative) (negative = they do not have a property)

• Some actors are not handsome (particularly negative)

• In this situation, positive does not equal good 

Validity Affect

• More than likely to accept valid than invalid syllogisms

  • Implies that we are sensitive to validity in some way 

Atmosphere Effect

(Universal vs Particular)

• SOME A’s are B’s

• SOME B’s and C’s

  • Therefore, ALL A’s are C’s

• Invalid 

• Usually rejected because the atmosphere of the premise does not match the atmosphere of the conclusion

• ALL A’s are B’s

• SOME B’s are C’s

  • Therefore, SOME A’s and C’s

• Invalid 

• Accepted 

• If the atmosphere matches, they are more likely to say valid and if atmosphere does not match then they are less likely to say valid 

(positive vs negative)

• Some A’s are B’s 

• Some B’s are C’s 

  • Therefore, NO A’s are C’s

• Invalid 

• Rejected 

• NO A’s are B’s 

• ALL B’s are C’s 

  • Therefore, NO A’s are C’s

• Invalid 

• Accepted 

• Note: the premises make up the “atmosphere”

Continuos Reinforcement

• Every time the animal does something that you like then you give it some form of reinforcement. Best for teaching an animal a new thing that it has never done before

Ration Reinforcement

• Reward given after action is performed more than one time (some fixed number of times)(Ex: every 3rd time). 

• Fixed ratio (every 3rd time)

Variable Ratio

•  (somewhat random where you give the reward) - will make the animal do the behavior most frequently. Will continue to do the behavior because it does not know when it will get the food. 

Near Miss Effect

• Areas of the brain that encode for actual wins still light up when you have a near miss 

  • The pattern for brain activity is the same when you get a near miss to when you get an actual win 

• Near miss similar to full miss