PSYC 2143 Final Review PDF

PSYC 2143 Final Review PSYC 2143- Lecture 21 Chapter 9 Learning and Knowledge The Relationship between Learning and Knowledge o Learning pertains to the process required to acquire new information When we discuss learning, we tend to focus the discussion on the mechanisms...

PSYC 2143 Final Review PSYC 2143- Lecture 21 Chapter 9 Learning and Knowledge The Relationship between Learning and Knowledge o Learning pertains to the process required to acquire new information When we discuss learning, we tend to focus the discussion on the mechanisms allowing us to compute new information o Knowledge pertains to the learned information: once learned, it becomes knowledge (semantic memory) When we talk about knowledge, we tend to focus on the structure of the learned information is in the brain  Why do we need to learn? o During the memory lectures, we discussed the different types of memory This, of course, coincides with the different types of information: Semantic memories hold fact information Episodic memories hold event information Procedural memories hold skill information ….. o But we aren’t just sponges that collect whatever type of information hits us o Rather, our learning is goal-directed  Common Highest-level Goals o Our highest-level goal, of course, is survival. This suggests that our learning mechanisms may be particularly adapted to situations that threaten our survival Perceptual evidence: threatening stimuli are processed automatically Learning evidence: food poisoning You only need to learn it once Prioritize our survival goal  Different Types of Learning o We learn To expect and prepare for significant events (classical conditioning) To repeat acts that bring rewards and avoid acts that bring unwanted results (operant conditioning) By watching what other people do and say to fulfill our purpose (observational learning)  Classical Conditioning o Identify Four Components: US, UR, CS, CR o Identify Three Stages: Before, During and After Conditioning Acquisition: association being formed o Stimulus Stimulus Normally CS comes before US during conditioning The CS predicts the US, and this predication elicits CR in preparation for the US o Most classical conditioning involves S S associations Acquisition: association being formed o Contiguity learning Learning occurs when a CS and a US occur near each other in time Not a critical factor o Contingency learning Learning occurs as CS predicates the US, then causing CR (S S associations) Driven by deriving this cause–effect relationship o Conditioned association (acquisition): takes time to develop Stimulus Generalization o Once response has been conditioned, other stimuli similar to conditioned stimulus (CS) elicit similar responses (CR) Stimulus Discrimination o Learned ability to distinguish between conditioned stimulus and other irrelevant stimuli sit speak Extinction o Association weakens in the absence of any US (i.e. CS alone) A type of forgetting  Spontaneous Recovery o After a time, the CS is presented again The CR, which was extinct, reemerges Yet the CR is not as strong as before o Savings: less time is required to learn than during the first time PSYC 2143- Lecture 22 Chapter 9 Learning and Knowledge  Skinner’s Operant Box o Learning by associating a behaviour with its consequences Reinforcement: increase the behavior e.g. press the lever get the food pellet press Punishment: decrease the behavior e.g. poke the nose get a shock stop poking Operant Conditioning: Principles o Reinforcement Strengthens a response, making it more likely to occur again in the future Operant Conditioning: Principles o Punishment Decreases a response, making it less likely to occur again in the future  Let’s put them together … o Does the consequence increase or decrease the likelihood of the behavior recurring in the future? Reinforcement Increasing behavior Punishment Decreasing behavior o Is something being presented or taken away? Positive something is given Negative something is taken away Quick recap of terms: http://www.youtube.com/watch?v=imkbuKomPXI Drawbacks of Punishment Physical punishment does not replace the unwanted behavior Punished behaviour is suppressed, but not forgotten Punishment can teach fear Punishment alone (no reward) does not teach children the correct behavior Punishment teaches discrimination among different situations Physical punishment may increase aggression, as children may model it to cope with problems  Reinforcement Schedules o Fixed-ratio: A reinforcement schedule that reinforces a response only after a specified number of responses. Reinforced every 100 responses (blue line)  Reinforcement Schedules o Variable-ratio: A reinforcement schedule that reinforces a response after an unpredictable number of responses. Reinforced in a variable pattern (green line)  Reinforcement Schedules o Fixed-interval: A reinforcement schedule that reinforces a response only after a specified time has elapsed. Reinforced every 10 min (red line)  Reinforcement Schedules o Variable-interval: A reinforcement schedule that reinforces a response at unpredictable time intervals. Reinforced in a random interval (purple line) Contrasting Classical and Operant Conditioning o Classical o Operant Learning associations between Learning associations between our events we do not control (passive behavior and its consequences learner) (active learner) Responses are involuntary Responses are voluntary Salivating Pressing lever Fear Cleaning room Observational Learning o Observational learning (social learning): learning by observing others o The Famous Bobo Doll Experiment Notice how the children’s actions directly imitate the adult’s behavior  Social Learning Theory o Children watched a film of an adult playing with a Bobo doll Adult was either aggressive (used a mallet) or not o The kids were later brought into a room with toys Albert Including a Bobo doll & mallet Bandura o Kids who saw the aggressive adult modelled their aggressive behaviour Footage of the original studies: Observational http://www.youtube.com/watch?v=byhLPxT_FJQ Learning  Mirror Neurons o Mirror neurons in macaque premotor cortex Firing: make a movement or watch an experimenter perform the same movement Understand other individuals’ thinking and actions Provide biological foundations for observational learning PSYC 2143- Lecture 23 Learning and Knowledge Chapter 9 Why Categories Are Useful o Categories help make knowledge well-structured o “Pointers to knowledge” in our everyday life Categories provide a wealth of general information about an item Allow us to identify the special characteristics of a particular item Definitional Approach o Determine category membership based on whether the object meets the definition of the category Categories are defined by necessary and sufficient features e.g. features that must be present -> the item belongs to the category Nonessential features are not needed Definitional Approach “A place to sit” o Does not work well for natural and human-made objects They have complex and variable features Cannot fit neatly into ONE rigid definition Prototype Approach o Prototype An average representation of the “typical” member of a category Not an actual member Contains the most salient features but allows a degree of variation Category decisions are based on how "close" an example is to the prototype  Prototype Approach High prototypicality: a category member closely resembles the category prototype “Typical” member For category “bird” = robin Low prototypicality: a category member does not closely resemble the category prototype For category “bird” = penguin  Prototype Approach: Smith’s Sentence Verification Highly prototypical objects are judged more rapidly Lowly prototypical objects are judged more slowly Typicality effect: prototypical objects are processed differently (Smith et al., 1974) Prototype Approach: Naming Name objects in a category Highly prototypical objects are named more rapidly Exemplar Approach o Concept is represented by multiple examples (rather than a single prototype) Examples are actual category members (encountered in the past) To categorize, compare the new item to the stored examples  Exemplar Approach o Similar to prototype view Do not follow a rigid criterion (necessary features) as in the definitional view o Different from prototype view Representation is not abstract, consisting of specific examples Category decisions are based on how 'close' a new item is to all the examples  Prototypes or Exemplars? o May use both Exemplars may work best for small categories Prototypes may work best for larger categories o Learning effect Novices: use prototypes = averaged typical member Experts: use exemplars (including some atypical cases or exceptions) PSYC 2143- Lecture 24 Learning and Knowledge Chapter 9  Levels of Categorization o Categories help make knowledge well-structured o To fully understand how people categorize objects, one must consider Is this concept a broad concept? Is this concept a specific concept? How are different levels of concepts organized? Levels of Categorization o Measure RTs in the sentence verification task: To understand how semantic knowledge is organized and accessed Superordinate: animal; furniture Broadest Basic: dog, chair; snake; table Instance Subordinate: Samoyed; beanbag chair; Subtype rattlesnake  The bigger the category, the more searching A beanbag chair is a piece of furniture (superordinate). A table is a piece of furniture (superordinate). A Samoyed is a dog (basic). A rattlesnake is a rattlesnake (subordinate). RT (superordinate) > RT (basic) > RT (subordinate)  The more levels, the more searching A beanbag chair is a piece of furniture. 2 levels Superordinate: animal; furniture A table is a piece of furniture. 1 level Basic: dog, chair; snake; table A Samoyed is a dog. 1 level A rattlesnake is a rattlesnake. 0 level Subordinate: Samoyed; beanbag chair; rattlesnake RT (2 levels) > RT (1 level) > RT (same level)  Hierarchical Organization People tend to organize concepts into global (superordinate), basic, and specific types (subordinate), supporting the idea that concepts are organized hierarchically Evidence That Basic Level Is Special Going above basic level → large loss of information Going below basic level → little gain of information Basic categories are used more often Semantic Networks Concepts are arranged in a hierarchical organization network  Semantic Networks: Cognitive Economy Shared properties (e.g. can fly) are only stored at higher-level nodes Exceptions (e.g. can sing) are stored at lower nodes Semantic Networks: Spreading Activation Activation is the arousal level of a node When a node is activated, activity spreads out along all connected links Concepts that receive activation are primed and more easily accessed from semantic memory  The Connectionist network Nodes are “neuron-like units”, and intended to represent individual neurons Knowledge represented in the distributed activity of many units Parallel distributed processing  The Connectionist network How learning occurs Network responds to stimulus When an error signal occurs Back propagation: error signal transmitted back through the circuit Weights should be changed/arranged to allow output signal to match the correct signal Process repeats until the error signal is zero PSYC 2143- Lecture 25 Chapter 11: Language What is language? Webster’s Dictionary: the expression or communication of thoughts, feelings, ideas and experiences by means of sounds or symbols, and combinations of such sounds or symbols, that are formed around proscribed rules or grammars; this primarily refers to human speech  Summary: What is Language The hierarchal nature of human language “Small components that can be combined to form large units” e.g. Letters -> Words -- > Phrases -- > Sentences -- > Text -- > A story The rule-based nature of human language ”We follow the grammar rule”  Is language innate or learned? Short answer: Both! Some aspects of language appear to be innate Others are definitely learned through experience  The innate nature of language o Noam Chomsky (1957) Noam Chomsky Suggests language is both innate and learned Human language coded in the genes Children produce sentences they have never heard and that have never been reinforced The capacity for language is hardwired into the brain, allowing children to learn any language they are exposed to The innate nature of language o A specific gene, known as FOXP2, has been shown to be relevant to language development o KE family: a heritable language disorder Mutation of FOXP2 Impaired language acquisition Reduced gray matter volume in MRI scans of affected KE family members The learned nature of language o Critical Period Noam Chomsky Childhood is a critical period for mastering certain aspects of language o If not exposed to either a spoken or a signed language by age 7 Children lose their ability to fully comprehend and use language Adults who try to learn a second language spend a longer time acquiring it PSYC 2143- Lecture 26 Chapter 11: Language Understand Words Not all words are generated equally: they differ in frequency o Word Frequency Effect High-frequency words are used more often: home, history Low-frequency words are used less: astrophysics, entropy We respond more quickly to high-frequency words (like home) than to low-frequency words (like astrophysics) Sometimes words can be ambiguous o Lexical ambiguity Words often have multiple meanings “rose”  Lexical Priming (1) The noun-noun condition: Con1: She held a rose (noun) -- > probe: flower (noun) Participants: read the probe word as quickly as possible “rose” Use context to determine the word meaning (i.e. rose means flower) Tanenhaus (1979)  Lexical Priming (2) The verb-noun condition: Con2: They all rose (verb) -- > probe: flower (noun) Participants: read the probe word as quickly as possible “rose” Tanenhaus (1979) Priming effect is still there but smaller (context not working?)  Lexical Priming Add a delay between the end of sentence and the probe Con1: The noun-noun condition: priming effect existed Con2: The verb-noun condition: no priming Now context can significantly differentiate the two meanings of “rose” It takes some time (e.g. 200ms) to influence word meaning processing Tanenhaus (1979)  Meaning dominance Some words are used more frequently than others Biased dominance When words have two or more meanings with different dominance e.g. tin: Frequency(a type of metal) > Frequency(food container) Balanced dominance When words have two or more meanings with about the same dominance e.g. cast: Frequency(members of a play) = Frequency(plaster)  Processing words are complicated o Without prior context, processing ambiguous words (multiple meanings) is affected by meaning dominance only Biased meaning: fast response Balanced meaning: slow response o With the context to aid in our understanding Our ability to access the ambiguous words depend on a combination of meaning dominance and context  The Brain and Language o The specialization of certain functions in one hemisphere of the brain more than the other o Lateralization Exclusivity  The Brain and Language Paul Broca Left frontal region: Broca’s area (discovered by Paul Broca) Damage to this region leads to severe impairments of speech production e.g. “Go ahead and do it if possible” -> “Go … to do it” Intact language comprehension  The Brain and Language Left temporal region: Wernicke's area (discovered by Carl Wernicke) Carl Wernicke Damage to this region leads to severe impairments of speech comprehension e.g. Invented nonsense words; Difficulty in naming people or objects PSYC 2143- Lecture 27 Chapter 11: Language Understand Sentences and Texts Understanding Sentences: Parsing Garden Path Sentences “After the musician played the piano was wheeled off the stage.” Garden path sentences Sentences that begin by appearing to mean one thing, but then end up meaning something else “Leading a person down the garden path” “oh no, that’s a wrong path!”  Garden Path Model of Parsing (1) o Late Closure People tend to attach new information to the current clause they are processing, rather than starting a new clause [After the musician played]… [After the musician played the piano] … [After the musician played the piano was wheeled] … Late closure may make you get lost in the garden!  Garden Path Model of Parsing (2) o When this happens, we reconsider the initial parse, and make appropriate corrections [After the musician played] [the piano was wheeled off the stage] [The old] man [the boat] o Some researchers question whether the correction is too late (i.e. determine parsing until the meaning is obvious o Other factors in addition to syntax?  Constraint-based Approach to Parsing o Constraint-based approach to parsing combines: Syntax: relationship among words or phrases Word meaning Story context Scene context Memory load and prior experience with language o Those Information help us to make predications about how the sentence should be parsed Predication in Language “The boy will move …” The ball? The train? Or the cake? (unclear) Altmann and Kamide (1999) Eye-movements to target occurred 127 ms after hearing the word “cake” “ The boy will eat …” Eye-movements to target occurred 87 ms before hearing the word “cake Conclusion: People make predications as they were reading a sentence  Anaphoric Inferences o An inference that connects an object/person in one sentence to an object/person in another sentence Coco, the famous poodle, won the dog show. She has now won the last three shows she has entered. What does she refer to?.. we really love to … go down to our ranch … I take the kids out and we fish. And then, of course, we grill them. What did they grill?  Instrument Inferences o An inference about tools or methods that occurs while reading text or listening to speech “William Shakespeare wrote Hamlet while he was sitting at his desk” You likely pictured him writing with a quill pen rather than a laptop  Causal Inferences o An inference that events in one sentence were caused by those in the previous sentence Emily took an Advil. Her headache went away. What caused her headache to go away? Sarah watered her plants. They started to look healthier. What caused her plants to look better?  Take-home message: making inferences Inferences create connections that are essential for creating coherence in texts Making these inferences can involve creations by the reader We’re not passive reader: add information from our knowledge of the world to the information provided in the text Situation Models o Mental representation of what a text is about Simulates a perceptual or motor characteristics of the objects and actions in a story RTs were faster when the picture presented matched the situation described in the sentence  Supporting Situation models ERPs recording while reading a story N400: unexpected word elicits a larger response (BAKE > EAT) We create a situation model while we’re reading Metusalem et al. (2012)  Situation Models and Brain Activations Physiological mechanisms during creating situation models (1) move fingers/legs (2) read ”action words” like pick or kick Neural representation of body-related concepts (right) is organized to mirror the body’s sensory or motor regions (left) in the brain Hauk et al. (2004) Knowledge representation is grounded in the sensory and motor systems of the body (the embodied approach) PSYC 2143- Lecture 28 Chapter 13 Reasoning and Decision making  Reasoning o Reasoning is a process we draw conclusions Through observation Based on the available evidence Which one is correct?  Deductive Reasoning vs Inductive Reasoning From general principles (premise 1) to specific Deductive Inductive Reasoning Reasoning cases (premise 2) General General If the premises are true and the logic is valid The conclusion must be true Specific Specific Determine whether a Draw general conclusions conclusion logically follow based on specific from statement observations and evidence  Deductive Reasoning vs Inductive Reasoning Examples of Inductive Reasoning: Deductive Inductive Observation: In a study of 500 older people (60-90 Reasoning Reasoning years), they suffered from memory loss. General General Conclusion: Memory loss is common among older adults Specific Specific Observation: My friend’s small dog barks a lot. Determine whether a Draw general conclusions Conclusion: Therefore, all dogs bark a lot. conclusion logically follow based on specific from statement observations and evidence  Inductive Reasoning: from observations to general conclusion o Inductive reasoning does not guarantee truth but suggests a plausible conclusion o To make Inductive reasoning is valid 1. Representativeness of observations 2. Number of observations 3. Quality of evidence o Used to make scientific discoveries Hypotheses and general conclusions o Used in everyday life Using past observations to predict future outcomes  Heuristic vs Algorithm o Algorithm Strict scientific procedure (slow) Guarantees a solution to a problem o Heuristic Simpler thinking strategy (fast) Mental shortcut More error-prone  The Representative Heuristic Make judgments based on how much an event resembles the others People make judgments about the probability of an event based on how similar it is to a prototype or stereotype they have in mind  Base-rate Fallacy Use base rate information if it is all that is available Use descriptive information if available and disregard base rate information Making judgments based on a description resembles the stereotype or prototype in your mind  The Law of Large Numbers o Customer reviews o Psychology Research 5 participants vs 100 participants Large sample size: more represent the population  Conjunction Fallacy Conjunction fallacy is making judgments with a failure to consider the conjunction rules It occurs when people mistakenly assume that specific and detailed scenarios (e.g. feminist bank tellers) are more representative than more general ones PSYC 2143- Lecture 29 Chapter 13 Reasoning and decision making  The Availability Heuristic o Make judgments based on how easily an event comes to our mind o Available Events: put less demanding on our cognition Recent Frequent Negative and Extreme o Ignore other information (short-sighted) Illusory correlations A correlation appears to exist (two events co-occur), but either does not exist or is much weaker than assumed Related examples come to mind easily (availability heuristic) Temperature!!!  Attitudes can Affect Judgments: Conformation Bias Tendency to selectively look for information that Conforms to our hypothesis Overlook information that argues against it Are we rational decision makers? o Expected Utility Theory People are basically rational Utility: we chose a desirable outcome because they are in our best interest If we have all the relevant information, we will make a decision that results in the maximum expected utility (i.e. achieve a person’s goal) Are we rational decision makers? o Denes-Raj and Epstein (1994) Participants made a choice to maximize their chances of getting the red beans Choosing the large bowel people do not always make rational decisions Emotion Affects our Decisions o Ultimate Game $10 split by a proposer and a responder Proposer: how do we split the money Responder: accept (get the money) or reject (neither of them gets the money) How people evaluate “fairness”: how emotion affects our decision making Emotion Affects our Decisions o Rational decisions Always accept the proposer’s offer o But we are not… People reject more low offers more proposed by human partners (they became angry that offers were unfair) People reject few low offers proposed by computer partners (less angry with an “unfair” computer) What is Loss Aversion? o Loss aversion Prefer avoiding losses over obtaining gains of the same value The pain of losing something valuable can be more intense than the pleasure of gaining that same item The fear of incurring losses make us choose very safe choice (avoid any risks) Why it happens? Loss aversion: a mixture of our neurological makeup, socioeconomic factors, and cultural background Our brains Socio-economic factors Culture  In our Brain o Amygdala Process fear (fear and loss are closely connected) Loss aversion: stronger amygdala activities o Striatum Prioritize avoiding losses over obtaining equivalent gains Loss aversion: stronger striatum activities  In our Brain o Insula React to disgust Work with the amygdala to make individuals avoid certain types of behaviors Insula region lights up when responding to a loss The higher the prospect of loss, the more the insula is activated  Social hierarchy: a good indicator of an individual’s level of loss Socio-economic factors aversion o People in power show less loss averse o Wealthy people show less loss averse Due to their wealth and network, they’re in a better position to accept a loss (more resources) Give less weight to losses and they are willing to take on risky decisions  Culture o Cultural values: affect an individual’s perception of losses compared to gains o Individualist culture People from Northern American countries tended to be the most loss averse o Collectivist culture Individuals from African countries being the least loss averse PSYC 2143- Lecture 30 Chapter 13 Reasoning and Decision making  What is decoy effect? o When we are choosing between two alternatives The addition of a third less attractive option (the decoy) can influence our perception of the original two choices o Decoy: make us do a comparison Completely inferior to the target (big size; same price difference) Make the target very attractive! Target How decoy affects our decisions? How to avoid decoy effect o Beware of sets of three Set of three: competitor (cheaper), decoy and target (most expensive) Try to notice when things crop up in groups of three Framing Effect o Decisions are influenced by how a decision is stated Which yogurt do you want to buy? Framing Effect o Decisions are influenced by how information is presented Highlight some aspects of the situation (e.g. 200 people will be saved, 80% fat free) Deemphasizing other aspects (e.g. 400 people will be lost, 20% fat)  Anchoring Effect o People make estimates starting with an initial number (the "anchor") and adjust it, but the final estimate is still influenced by the anchor Rely too heavily on early information that is presented during decision-making  Anchoring Effect o Real World Examples (Tip boards in two restaurants) $125.00 $125.00 Anchoring point 15% 20% 25% 12% 15% 20% $18.75 $25.00 $31.25 $15.00 $18.75 $25.00  Anchoring Effect o How does this work? Activate the concepts around the anchoring point Affect decision-making e.g. Average tip starts from 15% -- > fancy restaurant, nice service -- >they deserve a big tip Average tip starts from 12% -- > long wait time -- >they deserve a small tip e.g. Original price was $20-- > very great lemonade-- > buy it as it is now only $1 PSYC 2143- Lecture 31 Chapter 13 Decision-Making Mechanism  Neural basis of decision making Behavioral Results Rejected low offers because they became angry that offers were unfair from humans Less angry with an “unfair” computer Sanfey and coworkers (2003)  Neural basis of decision making fMRI Scanning Participants more likely to reject unfair offers More activation in the anterior insula (connected with emotional states) Activation of PFC was the same for offers being rejected or accepted Sanfey and coworkers (2006)  Iowa Gambling Task: Healthy People o Healthy people: after about 40-50 selections, they are fairly effective at identifying and sticking to the good decks  Iowa Gambling Task: PFC-damaged Patients OFC-damaged patients: keep choosing at random, which means they could not differentiate good and bad decks LOSE $ 200 !  Iowa Gambling Task: PFC-damaged Patients vmPFC-damaged patients: choose outcomes that yield high immediate gains regardless of higher losses in the future (preserve to bad decks) LOSE $ 200 ! Summary: Iowa gambling task o PFC is involved in decision-making process o Without the function of PFC, people cannot make optimal decisions (e.g. risk-seeking behavior) o Applications The Iowa Gambling Task is used with fMRI to study brain activation in healthy volunteers and clinical groups (e.g. obsessive-compulsive disorder) to assess risk-taking behavior and impulse control