Cognition III

category-based induction

This type of induction depends on inferences and happens when people lack sufficient information to answer the question. It depends on Similarity and typicality. • Given that whales have an omentum, what is the probability that wolves have one?  Not so good, because • Whales are not similar to wolves • Whales are not typical mammals.  Given that whales have an omentum, what is the probability that wolves have one?  Given that wolves have an oemtum, what is the probability that all mammals have them? The greater typicality of wolves makes 2 more likely than 1 Experts use knowledge to reason rather than following this logic.

base-rate neglect

Tom who sounds like a engineer- dry secluded person who is good at technology, people say he is in the engineer field rather than humanity, which is a much better field. Ignoring the statistic/ actual size of the field, choosing the one that how similar Tom is to an engineer.

interactive theory of comprehension

Use both syntax and semantics in parallel.

conditional syllogisms

[remises that has the form of if p (antecedent) then q (consequent). 1. affirming the antecedent- if p, then q 2. denying the consequent- if not q, then not p. 1&2 are valid, but people often judge 2 as incorrect.

Effect of the prefrontal cortex in decision making

1. planning and preservation- to switch between options. 2. problem solving- because of planning and preservation 3. understanding stories- making sense and connect parts of stories 4. reasoning

backwards blocking-Gopnik

A and B present, blicket light up. repeated for multiples times. Then B is present, consequence follows. subjects show reduction in the association between A and consequences.

Incorrect belief that small samples are more representative.

A flips the coin 10 times a day, B flips 20 times a day. People think A has a chance of 50/50 chance, where as A is more likely to have a skewed data because of its small sample.

speech is special?

A. special: people are especially good at speech perception, and its a uniquely human thing, so no other animal has this ability. Evidence: McGurk effect- motor theory of speech perception; Eimas- study of gars. B. not special: Chinchilla differentiate ba and pa/

Another example of expected utility theory/ framing effect

A. win 400 with 0.8 chance, win 0 with 0.2 chance B. win 300 for sure. People choose B when they should choose A. C. loss 400 with 0.8 chance, loss0 with 0.2 chance d. loss 300 for sure. people choose C when they should choose D.

Lord, Lepper&Ross- biased evaluation of evidence experiments- confirmation bias

Capital punishment studies- is it good or bad? o Subjects got both kinds of studies, one agreeing & one disagreeing with subjects' opinions. o Results: Subjects were better able to identify the design flaws in the study whose conclusion they disagreed. And people come out more convinced by their previous beliefs. o When we have a piece of evidence that is against our beliefs, we spend more time analyzing the evidence whereas when we have a piece of evidence that agree with our beliefs, we just accept it. The fact that people identified the flaws in the evidence that disagreed with their opinions strengthened their opinions: I knew that was wrong.

pragmatic reasoning schemata

Cheng&Holyoak. generalized sets of rules defined in relation to classes of goals that could help people predict the outcome in a given situation. permission schema is an example. tested the idea that permission schema is involved in reasoning about the card task. Checking disease and entering, making checking as a way to see if the traveler has permission increased the accuracy on which two cards to pick.

Practice makes perfect examples.

Crossman's study of cigar rolling(physical skills) • Workers at a cigar packaging factory become faster and faster at cigar rolling. Balckburn's study of addition problems- (mental skills)improvement with practice in time taken to add two numbers decrease as practice increase. You will always increase, but the rate of improvement decreases.

Omission bias

Fisher and coworkers. Personal health decision. people are likely to do nothing, even though it doubles the chance of dying. this is the tendency to do . Ho to avoid having to make a decision that could be interpreted as causing harm. However, people often recommend other people to receive the treatment.

the importance of coherence of a category

If one creature can fly but have not wings, but the other has wings but cannot fly. Then we think the one has wings is more likely to be a bird. If looking at a feature structure, then a feature of having wings is associated with the bird category, even though the link between them might by 0.8/ 80%. On the other hand, if the creature has no wings, than it is not coherent with the feature of the category.

screening off

In causal induction, If one causes is already present, then attribute the effect to that cause and not other possible causes. if there is a common cause D that cause F1, F2, and F3, or a common effect D caused by F1, F2 and F3, then the presence of D and F1 makes people think that F2 and F3 exist. even though any one of F alone could cause the D. e.g. Different causes of HIV. If you know that unsafe sex leads to HIV, then you ignore all other possible causes of HIV.

word superiority effect

People take less time to recognize if they saw Fork rather than Frok. Top-down processing.

Allais paradox

Problem 1 Which of the following situations would you prefer: Situation A $100.00 for certain Situation B A 10% chance of winning $500.00 An 89% chance of winning $100.00 A 1% chance of winning nothing Problem 2 Which of the following situations would you prefer: Situation C An 11% chance of winning $100.00 An 89% chance of winning nothing Situation D A 10% chance of winning $500.00 A 90% chance of winning nothing What happens here is that two problems are in fact identical in terms of utlitity. In both, the second situation is better by $39.00. People however, fail to see it when they make their choice and wrongly chose the worse option in problem 1.

spectrogram

Represents how the sound's frequencies change over time. used to study how speech sounds related to the phonemes that is heard.

syntax first in comprehension theory

Syntax first- first identify the major phrases and clauses, and then input this to the semantic strategy.

apple and Towel- interactionist theory of speech perception.

Tanenhaus and coworkers. one apple condition: put the apple on the towel in the box. eye movement: fixed point, apple, the other towel,back to apple, box. two apple condition: put the apple on the towel in the box. eyemovment: fixed point, apple on the napkin, apple on the towel, box.

Brian Ross- content vs. principle experiment.

Taught learners how to solve 6 types of probability problems, each with an example (rolling dice, matching photographs, etc) Then he tested them on new problems that either • Matched in content the original one (e.g. both dice) • Had new content, not used before • Or mismatched the content (i.e. matching a different kind of problem) o Subjects had to say how to solve the test problem, but didn't have to solve it. • Eg. Learning about Canadian paper using one permutation concept, test about Canadian paper using a new concept. Same content, but different probability principles. o Results: • Matching content: 76% correct • New content: 43% correct • Mismatching content: 22% (switching the contents and concepts. ) o One's ability to solve problems is attached to the content of the example they learned. o If people had ben using only the abstract principle, they would have scored identically in all three types, this is what statistical experts do o Novices tend to rely on superficial aspects of problems (e.g. dice vs. photographs) to figure out how to solve them, even if these are irrelevant.

Kim & Ahn- clinicians and causally central features.

The causal status effect is said to occur when category features causally central to an individual's theory of that category are treated as more important in categorization than less causally central features. If symptom A causes symptom B in a clinician's theory, then A is more causally central than B, and A is thereby predicted to have greater diagnostic importance than B.

transitional utility

a good deal. consumer's decision to purchase a product depends not merely on how much the consumer need the product but also on whether the consumer perceives the purchase as a good bargain. Eg. Coupons. Buy things that you don't need.

Status Quo bias

a preference for the current state of affairs. o Ex. Change to self-serving gas station at NJ. People are really upset of the loss if the change happen, then it ends up that nothing is changed. o Reasons: • Loss aversions • The current situation is considered to be normal, and is not questioned.  Organ donation, the country that has everyone signed up has a much higher rate of organ donation than the country that ask the citizen to sign on their own. When they have to do it on their own, they worry about the loses that might happen to them. However, if they are signed up automatically, they don't question about it.

deductive logic

a way of determining the validity of arguments. Premise: if As are Bs, All Bs are Cs. Conclusion: All As are Cs. Logic follows this can be false- leading to a false conclusion. Problems: Logic is a matter of form, yet the content of an argument sways people's judgments. People tend to agree with arguments that have true conclusions or that sound plausible. e.g. . If it is sunny, the picnic will be held.  If not sunny→ picnic will not be held  Sounds valid, but its actually not!  If p, then q. If not p, not q. ----invalid  The content makes people interpret the argument in a causal rather than a logical way. When people hear the statement, they are most concerned about the picnic, not about the logic of the statement. Valid but not true

causal induction

a way to link two things, one caused the other

Delta P measure

a way to measure possible causality. = P (effect I cause present)- P (effect I causes absent) Problem: other factors are not taken into account.

how language form

a. deciding the massage that you want send b. function processing: lexical selection (select words) and function assignment (what is the S.V.O.) c. positional processing: constituent assembly(making clauses and phrases) and inflection (singular/plural, tenses) d. phonological encoding- find the phonemes associated o *this process happens quickly. However, People don't say them after thy have completed this process completely. They may begin the processing while planning out the next step. This is why people sometimes stumble. "the things.. that thing, you know?"

inductive logic

arguments are likely but not guaranteed to be valid. • the sun has risen every day in my lifetime, so it will rise again tomorrow. Often involve going from a specific premise to a more general premise. judge how strong the argument is based on a. representativeness of observation: how well the observation represent all other members of the category. b. number of observations. sun rising is strong b/c there are many observations. c. quality of evidence: scientific evidence makes it stronger.

satisficing- Simon

choosing the first acceptable alternative you find. Not anything option, the acceptable option. A is not, B is good. Than one stops without keeping thinking about C. In some situations, one is forced to this because of time string. Eg. A car is coming at you, should I jump to the left or right? • this is because a. time b. mental effort required searching c. decision is not important- order food from a manu d. the option of changing our decisions later on/

incidental emotions affect decisions

clouds make nerds look good. another e.g. people in sadness are willing to sell their things at a lower price because they have need to expel things and a need for change.

o Availability heuristic/ salience heuristic

decide how likely something is by how easily it comes to mind. 1. Which city is bigger? Depends on how much you know about each cities. Hong Kong vs. Guangzhou 2. Gave subjects one of two lists of names: • 19 famous men, 20 less famous women • 19 famous women, 20 less famous men  subjects then had to estimate whether there were more men or women on each list. • Reported More men than women • Reported more women than men.  Other subject took a surprise memory test on the lists: • They recalled 12 of 19 famous names • But only 8 of 20 less famous names 3. cause of death. plane crush vs. car crush. 4. Asking people whether there are more words in the alphabet that have a K as their first or third letter, most people choose the first letter. The reason people overestimate this is because it is more easier to think of examples that start with a K.

Watson card-selection task

each card has a letter on one side and a number on the other side. Your task is to indicate which cards you would turn over to test the following rule: if there is a vowel on one side, then there is an even number on the other side. four cards: E K 4 7. 53% choose E, affirming the antecedent- correct. only 4% choose the other correct one- 7- denying the consequent.

syntax

grammar, which resolve ambiguity. subject verb agreement on number and gender.

expected value

how much money something will bring or take away from you.

expected utility

how much value something has for you. Very subjective. Including sentimental values.

framing effect

how the choices are described in the context they are put into affects which one is preferred. The medical example. Doctors tell their patinets they have x% of chance of living/ being cured instead of the Y% of dying

anaphoric inference

inference that connect an object or person in one sentence to an object or person in another sentence. Riffifi won the dog show. She has won many so far.

instrument inference

inferences about tools or methods. Shakespeare wrote Hamlet while he was sitting at his desk. We assumed he was not using a laptop.

causal inference

inferences that the events described in one clause or sentence were cause by events that occurred in a previous sentence. Sharon took an aspirin. Her headache went away.

Illusory correlations.

when a correlation between two events appears to exist, but in reality there is no correlation or it is much weaker than it is assumed to b. e.g. stereotype. selective attention to the stereotypical behaviors makes these behaviors more "available "

representativeness heuristic.

judge the likelihood of something based on similarity or typicality. can be misleading reading a profile of a person then asked to rank different fields that this person is most likely to be in.  People choose the one that he is most stereotype of rather than the ones that are the largest. Engineering vs. humanities and education.  Base rates tend to be ignored. Base rates= overall probability of different outcomes.  In fact, people give the same answer if you tell them that this description is from a group of descriptions that contain 70 engineers or lawyer. This is because representativeness heuristic.

semantics

meaning. Basic semantic strategy: • Identify content words and use them to build propositions that make sense • Note that this requires you to use your knowledge of the world to build the right kinds of propositions  A top-down strategy • People have trouble understanding things that don't make sense  E.g. I went to bed and took a bath. People make sense out of these sentences by inferring what happened. So people are using what they know about the words- what they mean- to interpret the sentences.

Chi, Feltovich& Glaser; physics problem, novices vs. experts.

o Asked subjects to categorize physics problems • experts: spend much time analyzing the type of problem it is • novices: generate equations that might be used to solve the problem. o To see through superficial similarities, one has to • Solve lots of different kinds of problems • Superficiality, the problems will not be similar, but the underlying principles will presumably be the same.

opt-in vs. opt-out procedure

whether if the decision require the person to take an active step. Organ donation

• Bock, Nicol & Cutting- study language production

o Complete-the-sentence task- so you don't tell people what to say, which is not production bur rather repetition, you tell them to complete the sentence however they want. • E.g. the actors in the soap opera.....(plural subject, singular local noun) vs. the actor in the soap operas...(singular subject, plural local noun). Possible completion: was/ were overpaid/horrible.... • 1% errors in first case but 10% in second. Plural can overcome singular but not vice versa. This is because plural words have more features that you can mixed up with, whereas singular words have no other features, so there is less opportunity for it to be mixed up. o Also studied about the collective noun: cast, class, family—words that have implicit plural meaning. Do these cause errors. • In fact, American speakers do not make these errors, proving that people take this as grammatical. Its not the meaning that determine plurality, but the fact the there is a s/es that makes a word plural.

how syntactic priming is studied in general

o Mixed sentence-picture displays o People read sentences, which contain primes o They describe pictures o Measure the effect of the prime on picture description. o E.g. • Sentence: The governess made the princess a pot of tea. • Picture of a boy handing a guitar to a man. • Does reading the sentence earlier influence you one says the sentence themselves. • There is no content overlapping, so there is only the structure of the sentence. • Dative sentences-  Double object construction agent-verb-indirect object- object. The governor made the princess a pot of tea.  Prepositional construction: agent-verb-object-prepositional phrase. The governor made a pot of tea for the princess. • Transitive sentences  Active: I broke the glass  Passive: the glass was broken by me.

mental accounting

o Rather than considering overall finances or resources, people group transactions into categories. • Jacket $125 • Calculator $10 • Would you walk 15 minutes to save $5 on this purchase? Depending on how much do you need the $5. People would more likely to save 5 dollars on calculator because it is much larger portion on 10 than 125. you buy a movie ticket that cost 10, then lost it, fewer people would want to buy another ticket. you lost 10 before you decided to buy a movie ticket, more people would still buy the ticket. You budget for the movie ticket is 10, losing the ticket would empty your movie ticket, so people would less likely to buy another ticket. However, the budge of how much cash you carry with you would not effect movie ticket budget.

Video is more salient than stats. Hamill, Wilson, & Nisbett

o Subjects viewed interviews with a "prison guard". Guard was either • Humane • Inhumane- prisoners are brutal, and ruthless o Subjects are told that the guard is • typical - this person represent most other guards • or atypical- this person more brutal or humane than other guards. o Later attitude test about guards, regardless of the typical question • Saw humane interview→ guards are pretty nice • Saw inhumane interview-> guards are brutal o This is because interviews is always salient or vivid compare to statistics (typical)

Ferreira et al.

o Task: • Neutral sentence, neutral, neutral....prime sentence (the priming sentence), target picture (if the subject use the same syntax as primed), foil sentence (change one or tow word of the prime sentence to test explicit memory.) • Also manipulated lag: the spaces between the prime sentence and the target picture. • Subjects: both amnesic and normal subjects. • For amnesiacs: they don't remember the sentence as well as control- worse on recognition test. But they still show the same priming effect. SO this priming is procedural. o Conclusions • Syntactic priming must be procedural  Long lasting, like implicit memory effect (based on the lag manipulation)  Even though it's an abstract form of priming, because there's almost no overlap of words • Wait, wasn't amnesia supposed to harm relational memory? And isn't syntax relational? • Well, anterograde amnesia just can't learn new relations. These sentence structure are not really new to them. So professor think this does not contradict to how amnesias can't make relational memory. • Understanding a complex skill like language also requires us to understand underlying cognitive process of perception, attention, memory, skill acquisition, reasoning and inference. You can't just study the subject itself.

optimistic theory of speech perception (not a real name)

o There is supposedly a on-to-one relationship between sounds and phonemes. o So each phoneme has a single sound, and each sound pick out only one phoneme. o Problems to match sounds to phonemes: • Individual differences in speakers:  Voice differences based on gender, size, age,  Accents. • Coarticulation-->speech segmentation

Expected utility theory

o To compare tow options. o Consider the possible outcomes associated with each option o Estimate the probability of each outcomes. o Estimate your utility of each outcome o Multiply probability by utility for each outcome and add up all the outcomes. o the total is that option's expected utility. o P(crab salad is good)* utility of good crab salad o However, in real life, people are not always rational

how to make better decisions.

o Watch out hwen you find yourself relying on one or two examples or very salient information • E.g. college visit, personnel interviews (salient to other factors. ) o Watch out for stereotypes o Seek out better information • Statistics (e.g. surveys, Consumer reports • Recommendations from experts • Not anecdotes from you roommate, sister-in-law... o Use of heuristics often arises from laziness in failing to get actually valuable information.

motor theory of speech perception

o believes that "speech is special". Speech recognition is an evolved human trait, we are built to do it which is why its not hard for us. o speech production is implicitly used in speech perception, to solve the articulation problem. In order to understand speech, we use our own knowledge about how speech is created using our own motor memory. o Instead of the words, people try to find the motor/ the movement your mouth when one makes sounds. Then only the animal that use speech can understand language- special! This o Evidence: McGurk effect

Simon; novices vs. experts- chess study

o gave experts and novices chess boards to look at for 5 seconds o subjects then had to recreate them on a real board- a real game o some of boards had randomly-placed pieces. This is to test if expert's memory is much better than novices when there is no meaning in the chess position- so pieces appear in positions that they are supposed to. o Results: • Actual game: beginners did really bad, Class A did better, and Master did the best—master have good memory • Random positions: beginner did as bad as in the actual game. Class did almost as well as the beginners, and masters did even a bit worse than beginners- masters do not have good memory for everything. o Analysis: • Chunking- you can form big chunks of information if you engage in pattern matching. • In fact, chess masters do not consider many more alternatives than novices do, nor do they necessarily calculate more deeply into the move sequence. Instead, they seem able to recognize a position as a good or bad, and what its strengths and weaknesses are. They divided the whole board into parts that make sense to them. Seeing multiple things as a chunk instead of memorize them independently. • Experts generally know more of the underlying principles of the domain.  In chess, this is principles of strategy and tactics

Rational decision

one that follows certain basic principles of consistency and coherence. E.g. if you were offered 25 dollars and and a personal, you would choose the 25 dollars as a rational decision.

permission schema

our ability to catch cheaters who don't have permission to do what they are doing. 19-year-old booze drinker problem. Evolutionary psychology approach i. Early humans lived in groups, and so social exchanges were important for success. ii. Issues such as reciprocity would be important. Groups must punish freeloaders and cheaters, who take but don't give their own share. Cheaters must be found and punished- cheating detection has been selected for. This is why booze drinker problem is easier than the red card and vegetable problem.

circular preferences

people prefer A rather B, B rather C, BUT C rather than A. A>B>C, but C>A. Eg. I prefer Apples than banana, banana than orange, but orange than apples. Irrational

syntactic priming

people's tendency to repeat the types of sentences that they use during language production.

coarticulation

phonemes are always influenced by the phonemes that proceeded before them and the phonemes that come after them.  Pronounce C, freeze your mouth, then pronounce s  Pronounce SU, freeze you mouth, then pronounce S.  The S don't sound the same!

lexical priming effect- Swinney

playing someone reading a sentence to subjects. As they hear the word bugs, they see a word on the screen, and are asked to identify whether or not it is an actual word. Words for bugs are : ant, sky, spy. Hearing the bugs prime ant and make them quicker at identifying ant. Spy uses bug to surveillances, which is meaningfully related to bugs, so people are supposed to be quicker at identifying spy as well. § Result: so people are fasted at ant, faster at spy, and longer at sky. So hearing the word bug also active the concept of spy! § Context does not immediately help comprehension (ant not that much faster than spy) but does work after delay. proving that the meanings of ambiguous words such as bug are briefly accessed before the context takes over.

categorical syllogisms

premises and conclusion describe the relation b/t two categories, using sentences that begin with all, no, or some. e.g. All birds are animals, all animals eat food. So, all birds eat food.

attraction/ context effect

prevent people from being rational. two choices unchanged, but the third choice that no chooses affect the decision between the first two. This is what happens instead of expected utility theory. people want to be able to justify their decisions. Harvard has a better reputation and its financial support is not as bad as Dartmouth.

priming

repetition/ associative/ syntactic

McGurk effect

research look at movies of people saying things and spilled the wrong syllables into the film. what the people were seeing influenced what they heard.

garden path sentences.

sentences that are grammatical but syntactically misleading.  The editor read the story was not amused.  Nonmisleading: the editor given the story was not amused.  Verbs that take a direct object or sentence complement. • Bill believed the argument was mistaken. • Bill believed that the argument was mistaken. (not misleading)

morphemes

smallest units of language that have a definable meaning or a grammatical function. Table has two phonemes, but only one meaning. Trucks have two meanings!!

Chinchilla experiment- Kuhl &Miller

speech is not special. Playing sounds on the spectrum from ba(voiced) to pa(unvoiced). People are pretty good at differentiating ba and pa. Chinchilla can do this too, they can tell the difference between ba and pa.

Bock, Nicol, Cutting

speech production errors. The actors in the soap opera was.. The actor in the soap operas were... people are 10 more likely to make errors in the second sentence, because we are more prone to the s of the operas, plural is a feature, and singular is default. so we pay more attention to plural.

syllogism

statements that are used in deductive reasoning. include premise and conclusion.

parsing

the grouping of words into phrases. two ways, syntax-first or interactionist.

word frequency effect

the more frequent words that we use in everyday life, like home, are easier to be recognized.

Normative decision

the objectively right decision based on some rule or principle. People might not agree on the rule.

Grice's conversation maxims

the rule that you are supposed to follow to make a conversation successful. It's like the red light—its there, but sometimes people violate the rule. o Quality • Do no say what you believe to be false • Do not say that for which you lack sufficient evidence • Although people lie sometimes, it only works when people do tell the truth most of time. If nobody abides to this rule, we would not believe anything anyone says, including the lies. o Quantity • Make your contribution as informative as is required • Do not make your contribution more informative than is required o Relation • Be relevant • For conversations like this: A- how do you like your new apartment? B- I wear earplugs every night when I go to sleep. Superficially, there is no connection in between these two sentences. In order to understand these sentences, we assume the maxims are true which says that information is relevant. So we start to make inferences that this person is probably saying he is not liking his new apartment. o Manner • Avoid obscurity of expression • Avoid ambiguity • Be brief- avoid unnecessary prolixity • Be orderly o When maxims are violated, it often means that the speaker is trying to tell you something. Listeners must try to infer what that something is. But in general, people formulate their speech so that they follow the maxims. • E.g. when reading a uninformative letter, then it is probably a bad letter. The fact that the recommender writes little about the person probably means that there is nothing good about him to be said. "this is student is always neatly groomed. "

phonology

the study of speech sounds. o different languages have different rules about what sounds can be used and how they can be combined • e.g. in English, you cant star ta word with [zb..] o rules are based on properties of sounds such as • voicing: [ba] vs. [pa] • place of articulation: [pa][ Θa][ta][ka]

mental models

they are used for reasoning. Like a real model in that it reproduces the important parts of the object, so you can "run" if and see what it does. Look up.

late closure strategy of comprehension

this strategy leads to the garden path sentence. As you are reading or hearing the sentence, you are going to to add on the new information to the already existing phrase for as long as you can, instead of starting a new phrase. e.g students believed the teacher was on the third floor.

falsification principle

to verify a rule, you have to look for situations that would falsify the rule. denying the consequent in conditional deduction reasoning

ultimatum game

two people playing the game, one person gets certain amt of money however the way he wants with another person. The other person can either accept the offer and get parts of money rewards, he can also reject the offer and neither gets any money. Rationally, if the first person give any money to the second person, he should take it. In real life, the second person reject, sacrificing the money he might get to punish the first person. However, when playing with a computer. people are more likely to accept the offer b/c they don't get angry with the computer.

Shafir- justification for actions experiment.

two vacation reservation options.  Option A: bland: 6 ok features.  Option B: mixed: three good, three bad  Half subjects asked to option to keep: half asked which to cancel (the same questions)  Justification: they both choose B, because it gives people reason(either the good features, or bad features) to either keep or cancel B. Another example. choice behavior and knowledge of exam outcome. more people want to delay decision making about the trip until they found out whether they had passed or failed the exam, even though the other two groups indicate that there is no difference in the actual decision about the vacation packages. people who pass saw this as a reward, people who failed saw this as a consolation--just need justification.

situation model

we have a mental representation of what the text is about. He hammered the nail into the wall vs. he hammered the nail into the floor. pictures of two nails, one vertical and the other is horizontal. People take less time to recognize the picture when it matches the sentence. Brain scan shows that reading about an action causes activity in the brain that is similar to the activity that occurs when carrying out the action.

voice onset time (VOT)

• Amount of time between onset of voicing and the beginning of the sound (opening of vocal track) • In English, voiced consonants start voicing at about the same time as the rest of the sound (VOT=0). • Unvoiced consonants don't start voicing until 40 ms or so afterwards eg. ba- voicing start right away because b is voiced. Pa- voice start when a is pronounced, so voicing start later because p is unvoiced. Think of this ba and pa as the end of a spectrum, then there is a point in the middle where its hard for people to tell whether is ba or pa.

framing effect and risks

• An unusual disease is expected to kill 600 people this year. Two alternative program have been proposed to combat the disease. Assume that the exact scientific estimate of the two programs are as follows:  A. under this program, 200 people will be saved, 400 people die  B. under this program, there is a 1/3 chance that 600 will be saved, and 2/3 that not people will be saved.  People choose A- like a guaranteed gain. Risk aversion  C: under this program, 400 people will die, 200 will not die.  D: under this program, there is 1/3 chance that no people die, 2/3 all people will die.  People choose D- gamble to avoid loss.  Result: people like certainty when it comes to gains. They will take a small sure gain over a gamble for a larger amount. People also prefer risk when it comes to losses. They will take a gamble on a larger loss over a sure smaller loss.

associative priming

• E.g. reading doctor helps you identify nurse as a word. The bug example

repetition priming

• E.g. reading notebook helps you see that word in a briefly-presented display • Seeing a word makes one faster at lexical testing. priming from implicit memory

talent vs. practice

• Ericsson et al. suty of violin players  World class players: 7410 hours of practice  Very good players: 5301 hours of practice  Problem: this is a correlational study, no causality. • Winner- "most children cannot stand practicing instruments, but highly gifted children cannot stand not practicing" • Limits of practice- only good practice counts.

Erbert Simon- optimal decision

• Ideal decisions require massive computations. There are many possibilities that need to be thought about. The time you put into making the decision might not be worthy of the quality of the decision you make. • Decision making itself involves a cost, in time and effort, which must be factored into the benefit of making better decisions

Loss aversion- Tversky

• Imagine that you have to make two decisions. Examine both decisions and then choose the option you prefer for each. • Decision1. Choose between  A. a sure gain of 240  25% chance of gaining of 1000, 75% chance of gaining nothing • Decision 2. Choose between  C a sure loss of 750  D. 75% of losing 1000, 25% of losing nothing. People have risk aversion because of their inaccurate prediction of their emotions. They don't take their ability to deal with adversity into account.

The endowment effect

• Once you have something, you value it a lot more than you would if you didn't have it • Owner of something think it is worth more than potential buyers of something think it is (this is what happened in the real estate) • This is because of the loss aversion  Sellers is losing the object, buyer is gaining it  Losses are worse than the same gain is good.

the conjunction fallacy

• P (A+B) must be no greater than P (A) or P (B). • E.g. P(Linda is a bank teller)> P (Linda is a bank teller and helps organize anti-war protests). Linda is more fit to the feminism bank teller than just bank teller.  But people make the opposite judgment b/c of representativeness • Which situation is more likely?  A. In the next 5 years, the US gets involved in an all-out war with Russia  B. In the next 5 years, through a conflict with a third country such as Iran or Georgia, the U.S. gets involved in an all-out war with Russia.  The reason makes B more likely. But the likelihood of B is less than A. if draw two circles representing the probability of the two events, there is no way the intersection of the two is greater than any one circle.

sound exchange errors- Garrett

• Sn-sh exchange: so, while you do the cooking, Bill snovels Show, right? (Bills shovels snow) • T-f excahgne: children interfere with your nife lite. (night life) • These sound exchange happens at beginning, middle and the end of sentences, and between all letters.  This tend to occur between words that are next to each other, and happen to the corresponding part of words. Support the idea that people think of words in terms of syllables. It also more likely happen to words that are similar. Also, the speech errors are almost grammatical, proving that people are thinking in terms of grammar. • Switching of voicing- Glear Plue Sky (Clear Blue Sky). o Conclusion: • words are assembled from scratch every time they're said.  We don't just have blue stored in our memory, and we don't just say them. We don't have the repertoire already fully formed, but rather we have phonemes stored, and we make words as we go. We make words from scratch.  They must not be retrieving an entire motor sequence, or else they couldn't make these exchanges. • Errors are based on both meaning and sound  When people make word errors, the words typically sound similar (Justin→ justice)  However, the words are often related in meaning too- more than predicted by chance • Tour de France commentator: "the battle for the green journey (jersey )" currently on ice. (meaning) • This supports interactive theory, as in connectionist models.

Blicket detection- Gopnik

• The blicket detector lights up when a blicket touches it. • Tested 3- and 4-year old children. And their job is to decided what is a blicket • Do children use the screening off principle to assign causality? • In one condition- one-cause condition  Two objects A and B  A cause lights up  B does not cause lights up  A+B cause lights up  Screening off applies here- A cause it to go off • Another condition- Two-cause condition  A cause light up 3 times  B cause lights up 2 times  B does not cause lights up one time  Screening off doesn't apply here, but same # of on and off associated to each block • Results: In 1st group, kids say A is the blicket 90% of the time, B is the blicket only 16%. In 2nd group, kids say A is the blicket 97% of the time, b is the blicket 78% of the time. Even though the B is associated with going at 66% of time in both situations. • Backwards blocking: A&B are presented twice: blicket detector goes off, A presented: blicket detector goes off  Question: is A a blicket or B is? • Note that both objects were only associated with the blicket detector going off  A is a bliket 99% of the time, B is a blicket 31% of the time. Screening-off

critique on avaliability/ representativeness huristic- Gigerenzer & Todd

• Too much emphasis on errors • Evolutionary argument  We are still alive, so the decision we make must not be that bad  Perfect decision is impossible in the real world. Heuristics have "ecological rationality"  Probably, human decision making is as good as it can be, i.e. optimal. o Critique of critique • People don't make optimal decisions. Doing drug, wrong careers • We can in fact make better decisions if we recognize that it could be improved. If it's optimal already, then nothing can be done.

John Anderson's three stages of skill learning

• Use declarative knowledge (slow) • Develop procedural knowledge from the declarative (automatized, fast) • Refine and enlarge (chunk) the procedural knowledge (slowly increase speed)  How long does this refining process go on? Forever. Herbert Simon claims that an expert is someone who has seen 10,000 examples of something. • After a while, experts loss declarative knowledge after it become automatic. When ask how to drive a shift car.

The phonemic restoration effect-Warren-warren

• recording like this are played to subjects- "the state governors met twith their respective legi*latures convening in the capital city" • where was the cough? What words were missing? • People heard the s even though they did not hear it. Because of their knowledge about he word legislature. • It was found that the *eel was on the shoe. (heel) • It was found that the *eel was on the table (meal) • It was found that the *eel was on the orange (peal) • Based other information in the sentence, people was sure they heard the word in parenthesis. • So, just like in visual perception, we are influenced by what we know and expect in our speech perception- top down perception

rule of plural

• these words that end with unovicing letters takes [s]  cat, dock, cup, cliff, bath • these words that end with voicing letters take [z]  cab, cad, crag, day, idea, sea, sigh, car • these words that have a noisy ending take [ąz]  catch, class, hedge, clash

Eimas et al- study of infants.

 Babies sucking to make the noise happen. If they suck a lot, its an indicator that there is something interesting/ new going on. Played the noises on the spectrum between ba and pa. The difference between the noises are the same, but at some point, the sound cross the point where the sound goes from ba to pa. • Group one: start with [ba], sucking drops off after a while, then change something that still sounds like [ba], sucking continue drops off. • Group two: start with [ba], sucking drops off after a while, then change to [pa], then sucking goes right back up • Babies are able to detect different phonemes in every language. By the time they are 1 years old, they can only differentiate the phonemes that they use in their language. Janpnaese babies can detect the difference between r and L, but lose it after a while.

Garnsey et al

 Plausible versions: the historian read the manuscript had been destroyed in the fire. • Readers slow down when they get to "had been". This shows evidence o the syntactic strategy to take the noun phrase after the verb as being the direct objects. Measuring people's eye movement, people slow down and go back to the beginning of the sentence.  Implausible version: the historian read the statue had been destroyed in the fire. • Readers do not slow down when they get to "had been" in the implausible versions. The syntactic strategy has been overruled by meaning. Statue cannot be read! People do not make the syntax structure when they read the statue. Measuring people's eye movement, people just read through.  Plausible and implausible only refer the noun as the direct object.  Apply syntactic strategy here first does not work because both sentences have the same syntax.  The syntax is the same across sentences, but the plausibility of the direct-object interpretation apparently influenced the sentence structure readers perceived. This give evidence for interactive or parallel view of comprehension.

speech segmentation

 Speech segmentation: looking at spectrogram, the gap between the auditory sounds are not the gap between each words.  It happens even across word, not just letters  We do not put spaces between words when we speak. We don't stop between each words  Its hard to say when one word stops and the next one begins. Huge problem for computer voice recognition. • "I like her" vs. "I'll eye cur" • "and to the republic for which it stands" vs. "for Richard stands" our knowledge of the meanings of words in a language and other characteristics of speech, help us define when a word begin and end. top-down processing

Tompson, Reyes & Bower- jury decision task

 Subjects read the testimony from a bogus trial  The prosecution or defense evidence was made vivid or pallid, depending on condition. Eg. • Vivid- the suspect flipped a bow of quacohomoli which stained the carpet • Pallid- the suspect flipped a bow of presztel  Subjects made their decisions • Immediately, or o Manipulation made little difference • Next day o More made decision based on the vivid evidence. This is because of availability heuristic.

syntactic strategies of comprehension

 Use function words to guess the structure • Articles, prepositions, auxiliary verbs.  Use syntax to guess what's coming next • E.g. article-adjective-???, perhaps noun, another adj, not verb. • E.g. article-noun-???, perhaps v? E.g. cast iron sinks quickly rust. Garden path sentence because after the noun phrase- cast iron, people are expecting the verb phrase to be very close to the N. phrase, so after sinks, people think it is the verb.

Hsee experiment. justification of actions.

 asked subjects whether they would prefer to receive a Beatles CD or Barbra Streisand CD for being in his experiment ()  other subjects were told that they could do one experiment for 90 points or another for 50 points. • The Beatles CD cost 50 points, and Streisand 90  Now subjects preferred to get the Streisand CD, even though they preferred the Beatles. The 40 more points made people prefer the Beatles CD even though they had no preference at the beginning. This violate the expected utility theory. The need to justify override one's preference/ decision.