Cogs 101B Midterm 1 Masterdoc-2 PDF

CLASS NOTES THE MIND the mind creates and controls mental functions such as perception, attention, memory, language, etc., and creates representations of the world around us internal self that makes sense of the external self how do we study the mind break down something to its constitue...

CLASS NOTES THE MIND the mind creates and controls mental functions such as perception, attention, memory, language, etc., and creates representations of the world around us internal self that makes sense of the external self how do we study the mind break down something to its constituent parts to study the thing we want to study ○ breaking down mind difficulty: you can only experience your own mind ○ mental representation: internal symbols or cognitive structures that the mind uses to represent and process external information; form basis of cognition (images, sounds, abstract ideas used in recall or reason) computational capacity to support this–symbols that have as their meaning something in the world information processing: manipulating these symbols like computer manipulating code mental representations are patterns of neurological activity but we experience their information analyze in terms of their information and relation to each other, and not in terms of their neural basis ○ neuroscience cannot tell us all the answers; recording every single action potential from every single neuron allows us to understand the brain but not necessarily the mind david marr: trying to understand perception by studying only neurons is like trying to understand bird flight by studying only feathers: it just cannot be done Marr’s levels of analysis level 1: computational (goal of system) – what the system does and why it does it; involves defining the problem and understanding the purpose of the process ○ what is the problem being solved? eg. the problem is an arithmetic one; the calculator needs to multiply x by y and get xy ○ receiving the communication, the act of comprehending level 2: representation and algorithm (steps to achieve goal) – how of the process; the algorithm or rules used to solve the problem defined at the computational level  ○ what are the steps and representations used to solve it? eg. based on binary logic, the calculator represents each number in binary code and turns on or off each position to produce the final answer ○ not always aware of the steps you use to reason something, how are you understanding and comprehending something level 3: hardware implementation (hardware to actualize goal)– physical realization of the process, answering how the algorithm is implemented in hardware or biological substrate ○ how are these steps implemented in physical or biological hardware? eg. integrated circuits house transistors which can have on or off states, representing the 1s and 0s of the binary code ○ neuroscientific aspect of reasoning how you’re comprehending holistic understanding of complex mental processes from what to how history wilhelm wundt “we can study mental processes” introspection (late 1800s) ○ structuralism: breaking down mental processes into basic components ○ unreliable because there’s no way to objectively test claims and unconscious thoughts ○ analytical introspection: conscious mental events and meticulous training in how to describe and record mental experiences individuals report mental experiences in detail ○ contribution: study behavior in controlled environment generated first psychology labs and students john b watson “we can’t study mental processes” behaviorism (early to mid 1900s) ○ staunchly against talking about mental events because they’re subjective ○ observations, objective; abandoning study of mind  ○ observable behavior should be main focus stimuli and behaviors, interaction with environment classical conditioning–neutral stimulus paired with stimulus that elicits response → neutral stimulus elicits response (learning by developing implicit associations between events) ○ implicit association: subconscious associations between concepts and evaluations or stereotypes operant conditioning–how behavior is shaped by reward and punishment ○ b.f. skinner: language is learned through operant conditioning (learning by consequences and reinforcement) ○ cons: mind is a black box; limit vocabulary to observable what is happening internally cannot be observed no vague intermediate mental constructs (thoughts, concepts) cognitive revolution (1940s and 50s) studies cognition rather than behavior, how humans think and process information ○ behavior is best explained by positing rich internal states tolman’s maze (1948): rats form cognitive maps, there’s a map layout inside rat’s mind when navigating the maze quick adaptation to changes in the maze and choosing new routes to reach goal when familiar paths were blocked learning layout of maze even without immediate reinforcement→ learning occurred before demonstration in behavior noam chomsky (1959): there must be an internal abstract notion of grammatical rules re-evaluated b.f. skinner’s behaviorist language acquisition via rewards–kids produce all sorts of speech they never hear and are never rewarded for language acquisition relies on innate cognitive structures ○ the digital computer computers as an analogy for the mind computers can be understood functionally (computational and algorithmic) without ever mentioning the underlying hardware ○ solving complex problems by manipulating symbols or representations information processing approach trace sequences of mental operations involved in cognition mental processes as computations manipulating symbols through a set of rules → cognitive models of how humans think and solve problems donald broadbent (1958) proposed the first flow diagram for the mind to represent how attention might work  1956: birth year of cognitive science (program computer to mimic operations of human mind?) logic theorist (newell and simon 1956): first AI program, mathematical proof ○ new approaches mental imagery (rotation of shapes) mental imagery people are slower to respond if two figures are rotated further from each other ○ perfectly linear evidence ○ mental rotation: manipulating mental images takes time; people mentally rotate images in their head → mental representation angle of rotation: greater angle disparity between two objects, longer to respond quantitative measure of mental imagery – how mental events can be studied scientifically ebbinghaus 1885 pioneered empirical study of mental processes first quantitative measure of memory ○ forgetting curve ○ memorized nonsense syllables in order to eliminate the effects of prior familiarity participants were more likely to remember the first few items and the last few items in the list memory can be quantified and behavior could be used to describe operations of the mind cognitive revolution two key ideas: ○ we MUST study the mental world to understand behavior ○ we cannot study the mental world directly  transcendental method (kant) ○ “inference to the best explanation” ○ reason backwards: observable effects from unobservable causes ○ kick start of how to approach mental event scientifically inferences from the visible to the invisible ○ analogy: detective solving a crime cognitive models representations of structures or processes that help us visualize or explain the structure or process ○ models aren’t meant to be perfect representations process models: show processes that are involved in cognitive mechanism, with boxes usually representing specific processes and errors indicating connections between a starting point for research; attempt to make complicated systems easier to understand and provide a starting point for research ○ CLASSICAL CONDITIONING implicit learning: classical, operant, statistical CS: conditioned stimulus – learned stimulus that was previously neutral that can eventually trigger a conditioned response US: unconditioned stimulus – automatic reaction to that stimulus NS: neutral stimulus – initially elicits no response constraints that guide learning critical period: most favorable period in development to learn a behavior ○ imprinting: forms the basis of young animals attachment to parents ○ bird song ○ human language learning ○ humans have a critical period of learning language, ability to learn language drops as we get older  biological preparedness: certain associations are learned more readily than others (intrinsic tendencies) ○ animals hardwired to approach desirable things and avoid undesirable things ○ learning also happens more readily for certain cues-consequence combinations taste aversion is often “one trial” learning ○ phobias: people are biologically prepared to learn to fear objects and situations that threatened the survival of the species throughout its evolutionary history delayed conditioning: conditioned stimulus is presented, and remains present, for a fixed period (the delay) before the unconditioned stimulus is introduced ○ eg. dog is trained to sit before being given a treat nonassociative learning change in response to a stimulus that doesn’t involve pairing it with another ○ habituation: a decrease in the strength of a stimulus to elicit a response after repeated presentations occurs when we learn not to respond to a stimulus that is presented without change, punishment, or reward ○ sensitization: an increase in strength of stimulus to elicit a response after repeated presentations behaviorism can’t talk about mental representations nurture > nature classical (pavlovian) conditioning involuntary, reflexive type response that occurs in response to a stimuli acquisition is the process of developing and strengthening a conditioned response  ○ happens more rapidly for more intense neutral stimulus (loud noise) and unconditioned stimulus (preferred food) → asymptotic asymptote of acquisition: association between the conditions stimulus and the unconditioned stimulus is the strongest, in that no further conditioning can increase response after many pairings, there’s a point at which the dog salivates as much as it ever will in response to the bell alone→ beyond this point additional bell-food pairings won’t increase the salivation level explains ceiling effect of conditioning: response is stable and unlikely to increase without a change in conditions neutral stimulus: a stimulus that does not produce a specific response, but may cause someone to focus their attention eg. sound of bell information value: the conditioned stimulus must be informative for learning to take place ○ blocking: failure to learn association between stimulus and outcome due to presence of another stimulus that already predicts that outcome added the light provided no new information about when shock should be expected conditioned response can be weakened and eliminated via extinction spontaneous recovery is the reappearance of a conditioned response after a rest period from extinction sometimes a conditioned stimulus (CS) elicits ○ behavior that is identical to the unconditioned response eg. dog’s salivation to food → then to bell ○ different than the unconditioned response rat jumps to shock, freezes to buzz that predicts shock preparatory response theory: purpose of the CS is to prepare the body for the unconditioned (US) in whatever way is most adaptive compensatory response: an automatic response that is in a direction opposite to the effect of an expected response  ○ one reason for drug tolerance: brain's withdrawal reactions in response to the presence of a drug ○ allows maintenance of homeostasis drinking in an unfamiliar environment→ compensatory response isn’t activated because alcohol was consistently paired with a familiar home environment→ gets drunk quicker OPERANT CONDITIONING operant: voluntary behavior that operates on the environment to produce rewarding and reinforcing effects more voluntary and are controlled by their consequences future of probability of a behavior is affected by outcomes ○ positive outcome → occurs more vice versa involves a behavior and consequence that is further increased or decreased in the future due to consequence it is associated with negative (taking away) and positive (adding something) are consequences ○ in operant conditioning, positive and negative do not mean good and bad ○ instead positive = adding, negative = taking away punishment (punisher) and reward (reinforcer) describe the way consequences affect behavior ○ reinforcer → increased behavior, vice versa ○ both negative and positive consequences can be used as either punishers/reinforcers  discriminative stimuli: a behavior is reinforced or punished in the presence of a particular stimulus, but is not reinforced or punished for other stimuli, even if they’re similar ○ a signal for the opportunity to respond sets up the occasion for when the specific behavior is to occur ○ eg. green light ques go in traffic, but green street lights don’t reinforce driving forward behavior in the same way schedule of reinforcement: different types of contingencies for when a behavior is reinforced; can have dramatically different effects ○ continuous: every instance of the behavior is rewarded ○ intermittent: only some instances of the behavior is rewarded ○ ratio, interval: the reward is either dependent on number of behaviors (eg. lever presses) or time (eg. every 5 mins) ○ fixed, variable: describes either a fixed ratio/interval, or a variable one ○ cognitive perspective on behavioral data learning beyond behaviorism cognitive perspective ○ in order to understand behavior, it is essential to talk about the mind ○ people are more complicated reinforcement acts as a source of information about relations between actions and consequences (but doesn’t compel behavior) but we also learn from watching others, even when it’s not reinforced our theories and expectations dominate observations and in turn influence how we infer causation strict behaviorist perspective ○ only talk about stimuli-response relationships, not the mind ○ learning consists entirely of forming stimulus-response relationships, which are formed automatically ○ tend to view causal learning as simple associative learning - dependent on the frequency of events and their outcomes many observations require richer explanations eg. ○ distinction between learning and performance not everything learned is immediately manifested in performance learning isn’t just the performance of a learned behavior, it is a change in the ability and potential to do the behavior although learning is often inferred from performance, the absence of learning may not be inferred from the absence of performance ○ expectations matter expectancy hypothesis: the tendency for behavior to occur in a given situation is a function of the individual’s expectations, not necessarily which behaviors have been reinforced in the past rescorla: animals and humans learn not just through associations, but by expecting one event to follow another test: choose between two cards→ correct cards are rewarded → secretly set up so never guessed correctly on some high reward trials high value cards that had never been rewarded were chosen over lower value cards that had been rewarded even though the person can never guess correctly on high-reward trials ○ rule learning participants learned the more complex rule, but often failed to learn the simple rule people learn the rules associated with the hypotheses they evaluate, not just associations ○ higher sensitivity to probabilistic relationships over pure associations  rescorla (1967): only in conditions where there was a contingency did the bell become a conditioned stimulus contingency: how well CS predicts the US relative to US appearing without the CS co-occurrence: how often the CS and US happen together positive contingency: learning is strong if CS reliably predicts US in which US occurs more frequently when the CS is present than when it’s absent ○ if every time the bell rings, food follows, the dog will strongly associate the bell with food → even when co-occurrence is low, the predictive value of the bell is high because food never occurs without it zero contingency: no strong association if CS and US occur together but the US also happens randomly when CS is present ○ dog doesn’t learn that the bell predicts food because food sometimes appears without the bell negative contingency: CS inhibits response if US occurs more often when CS is absent ○ if bell rings and food is less likely to follow, the dog might learn that the bell signals the absence of food pairing of a conditioned stimulus (CS) and unconditioned stimulus (UCS) does not always produce learning it is not the absolute frequency of the pairing that is important, but the general probabilistic relationship between (how well CS predicts US) two conditional probabilities: the probability of the US given the CS and the probability of the US in the absence of the CS learning would only occur if there was a difference between the two conditional probabilities observational learning learning to respond in a particular way by watching others, who are called models. also called “vicarious conditioning” because it involves learning by watching others quire responses through classical or operant conditioning → learning by observing events and their consequences observation of skilled performances is much less beneficial than observation of the learning process ○ observer cats faster at solving puzzle after observing other cats learning how to do the puzzle compared to cats that already knew how to solve the puzzle but didn’t observe; speed increased with the trials watched imitation is major source of learning in children social learning theory behaviorist explained it away as “vicarious reinforcement” and “vicarious punishment” bandura’s social cognitive theory of learning (1960s) ○ bobo doll experiment (1961): kids watch human demo rewarded or punished for physically abusing the bobo doll kids more likely to pursue physically aggressive behavior if exposed to aggressive model more aggressive to same sex models (boys to male models) ○ people learn from each other via observation, imitation and modeling ○ “cognitive” in that the theory discussed attention, memory, mental representation, motivation over imitation: children have a tendency to instrumentally irrelevant actions ○ appears to be uniquely human and increases with age animals quickly reduce their behavior to the minimum that will produce the desired outcome ○ humans will imitate even when there is no obvious reason for the behavior infer hidden reason (?) – highly cognitive and hard to explain through behaviorism generalized imitation: imitation itself is rewarded ○ individual learns to imitate reinforced behaviors but also continues imitation of non-reinforced and unprompted behaviors that resemble previously reinforced behaviors imitation becomes learned general skill ○ if a child is reinforced for imitating a specific action like waving, they might later imitate other gestures, such as nodding or shaking hands, even without direct reinforcement for those new actions acquisition of “opaque cultural behaviors” ○ imitate things that you don’t know the purpose of STATISTICAL LEARNING learning the mechanism that facilitates the detection of structures, patterns, and regularities in the world  ○ not learning about statistics first shown in infant language acquisition (saffran, aslin, newport 1996) ○ how do infants learn to segment speech in order to segment speech into words, infants must have some way to break the speech stream into word-like units ○ recognizing syllable pairings ○ ○ transitional probabilities: some syllables are more likely to precede others eg. “bu la do” – “ba” often precedes “la”, “la” often precedes “do” higher probability of one syllable coming after another higher, probability of combining as a word start to draw boundaries in speech streams based on what is likely in absence of understanding language, this is a kind of ‘bootstrap’ method used to start to make sense of the continuous stream of language a baby hears babies implicitly learn what syllables form part of the language and what syllables go together to form words ○ preferential looking: habituation to a stimuli, babies look longer at novel syllable combination babies look longer at non-words, suggesting learned statistical structure potential mechanism for how infants learn to draw boundaries in speech stream not unique to language acquisition ○ likely a response to all structured input ○ domain-general learning mechanism: the idea that humans are born with mechanisms in the brain that exist to support and guide learning on a broad level, regardless of the type of information being learned pattern recognition cognitive process of matching information from a stimulus with statistical information from memory more abstract ○ top down processing using knowledge, models, ideas, expectations to interpret sensory information prior information; interpretation method change the way you experience raw sensory input  ○ context effect you read as “the cat” even though the second “letter” is the same bottom up processing: based on “raw data”--the information that reaches the sensory receptors ○ unconscious inference (von helmholtz 1821-1894) perceptions are the result of unconscious assumptions that we make about the environment we perceive those objects and events that under normal circumstances would be most likely to produce the received sensory stimulation think as result of interaction between bottom-up and top-down processing ○ bayesian inferences formalize (use calculation) unconscious inference reality she could bt 25, 45 – data is ambiguous; system makes best guess based on information ○ word superiority effect letters within a word are processed more rapidly than single letters memorize “k” more accurately when shown the word “fork” as compared to only shown the letter “k” non-words also produce effect if they conform to standard spelling patterns (eg. klane) but not if words don’t conform to standard spelling patterns (eg. jbkrt) frequent and primed words are more easily recognized cake > ladle more specific theories that have been proposed ○ feature nets ○ template theory we have templates for every object ○ recognition by components objects are recognized by identifying their component parts (geons) all objects can be constructed from ~40 geons which are viewpoint invariant choice of shape may seem arbitrary but its based on non-accidental properties–they can be recognized across a variety of different perspectives ○ modules: face perception  tendency to perceive faces in things that aren’t meant to be faces fusiform face area (FFA) holistic processing configural information hypothesis: ○ unlike most objects, faces are processed primarily with configural properties (eg. relationship between features) ○ representation of face is holistic, not dependent on individual features but on an integration across all features; less sensitive to changes in individual parts ○ ATTENTION subjectively: "everyone knows what attention is. it is the taking possession by the mind, in clear and vivid form, of one out of what seem several simultaneously possible objects or trains of thought…it implies withdrawal from some things in order to deal effectively with others…” – william james what it’s for ○ direct the focus of our awareness ○ select relevant perceptual information ○ enhance certain perceptual information ○ perceive a coherent world feature binding ○ sustaining behavior ○ select appropriate actions attention and its limits on information selective attention: directing awareness to relevant stimuli while ignoring irrelevant stimuli in the environment we need attention to limit the amount of information that is processed why are there limits on the amount of information we can process  ○ attention is there to act as a bottleneck for information coming from sensory inputs we pay attention because there are capacity limitations ○ for example in vision: the amount of information coming down the optic nerve - around 10^8 ~ 10^9 bits/sec = far exceeds what the brain is capable of processing ○ after parallel processing (all at once) of incoming sensory signal → serial (one after another, higher level) bottleneck of information processing selective attention types ○ change blindness inability to detect changes in a scene that you’re looking directly at even if information is perceptually available we need attention to process information in the right way in order to notice changes detecting changes between two images that seemingly look the same ○ inattentional blindness people fail to see what they are not expecting failure to notice an unexpected, but fully-visible item when attention is diverted to other aspects of a display invisible gorilla experiment ○ attentional blink when the second target occurs too quickly, the brain misses it completely because it’s still processing the first target second of two targets cannot be detected or identified when it appears close in time of the first attention isn’t only selective spatially, but across time too ○ covert attention: attending without moving your eyes eg. a driver might be overtly attending to the road ahead, their covert attention might pick up a pedestrian trying to cross the street from the side ○ dichotic listening study selective attention and auditory processing test: individual receives two different audio streams simultaneously one in each ear; individual is to focus on and repeat the message coming into one ear while ignoring the other people focus on the message and filters out the other → limited awareness of unattended information classic models of selective attention early selection model (broadbent’s filter model): information is filtered based on physical characteristics such as rate of speaking, pitch accent early in the processing stream ○ filter eliminates before processing for meaning → attended input is privileged from the start so that unattended input receives little analysis (never perceived) only physical characteristics (like pitch or loudness) of unattended information are processed, unattended information is filtered out before it reaches deeper, semantic processing ○ problem: cocktail party effect ability to selectively attend to particular auditory input while filtering out a range of other stimuli BUT… people sometimes hear important semantic information like their own name from the unattended channel → semantic processing occurring before filtering stage we must process the fact that the sound was our name in order to let it through conscious awareness → thus higher level processing listeners hear their own name about ⅓ of the time even when focused on another conversation → semantic content processing from unattended streams ○ listeners also report coherent sentences combined from words from both streams ○ problem: meaning of ambiguous word is biased by content in the unattended channel case 1: unattended channel: “river, stream, forest, water” / attended channel: “they were throwing stones at the bank” case 2: unattended channel: “money, wealth, financial” / attended channel: “they were throwing stones at the bank”  the meaning of "bank" in the attended channel is influenced by the content (semantic cues) in the unattended channel → semantic processing in unattended channel treisman’s attenuation model: builds on broadbent’s ○ more flexible filtering approach; unattended information is "attenuated" or weakened but not entirely blocked by the filter, allows some semantic processing important or contextually relevant information in the unattended stream (eg. hearing name) can still reach awareness if surpassed threshold of importance accommodates findings like the ambiguous "bank" interpretation late selection models: all input receives relatively complete analysis and the selection occurs after analysis (processed at semantic level before filtering). either ○ just before it reaches conscious awareness or ○ it makes it to consciousness but is quickly forgotten perceptual load theory of attention (lavie and tsai 1994): integrates concepts of early and late selection by proposing that attentional selection depends on load of primary task ○ attentional resources are limited in capacity task relevant stimuli are processed first, then task irrelevant stimuli are processed until all attentional resources are used up ○ flexible allocation of limited resource (limited capacity) processing capacity: limited amount of information that people can process perceptual load: difficulty (load) of task affects how much processing capacity is available low load (easy) primary task leaves excess resources for remaining perceptual capacity→ can process irrelevant information  ○ available to process task-irrelevant stimulus so it gets processed more information can be processed → slows down responding distraction/interference or competition for attention with primary task eg. recognizing familiar object is easy thus attention not fully engaged so you might notice other things in environment high load (hard) primary task leaves no resource for perceptual capacity ○ irrelevant stimuli can’t be processed and have little effect on performance of task, thus filtered out ○ eg. finding something small makes you focus entirely on the task, leaving less chance for distractions to interfere with processing spotlight metaphor ○ spatial attention has been compared to a spotlight beam that can shine anywhere in the visual field → attention selectively focusing on specific locations in visual field find the vertical red bar amongst green bars: “pop out” effect (visual capture)--unique target is present among homogenous distractors it captures attention in a bottom-up fashion; we’ve analyzed everything in parallel ZAPS IMPLICIT ASSOCIATION TEST implicit bias attitudes outside our awareness, stored in memory as shortcuts for understanding the world ○ mind access immediately can influence explicit decisions ○ eg. the gender gap between leadership roles in companies → implicit association between men and leadership IAT is a method to measure implicit bias explicit bias attitudes and associations we are aware of and can readily discuss with others implicit association test (IAT) measures implicit bias by analyzing differences in response times sort the target word into one of the two categories that it best belongs into as quickly and accurately ○ eg. sort “daniel” in the “man” category; sort “boss” in the “career” category stereotype consistent vs inconsistent tasks stereotype consistent: sorting words into category of “woman family” or “man career” (only one [gender or role] has to match) ○ women are stereotypically associated with family and men are stereotypically associated with careers stereotype inconsistent: sorting words into category of “woman career” or “man family” people who implicitly associate women with family should theoretically have a slower response time in the stereotype-inconsistent trials than in the stereotype-consistent trials ○ stereotype consistent associations are readily accessible in people’s minds VISUAL SEARCH pop-out effect the brain quickly detects simple features (like color), allowing for easy identification of distinct objects ○ happens early in perceptual processing ○ eg. finding a yellow circle in a sea of blue circles conjunction search searching for a combination of features (e.g., a man + sunglasses) is called a conjunction search requires higher-level processing, which takes more time because multiple features need to be integrated feature integration theory begin a visual search by automatically identifying simple features—such as color, shape, size, and movement—within an environment parallel processing our mind uses separate systems to analyze the different visual features of objects at the same time ○ attend selectively to one feature by effectively blocking the further processing of others salient features our attention is drawn to high contrasts and movements because they are more likely to grab our attention salience sorts out what to look at and pay attention to first visual search task task 1: find blue circle in a sea of orange circles and squares ○ feature search for the single color blue ○ uses preattentive processing → don’t need to deliberately focus to process sensory information due to high salience of target ○ reaction times stay the same regardless number of distractors due to pop-out effect task 2: find blue circle in a sea of blue squares and orange circles ○ conjunction search for shape and color, need to deliberately focus ○ reaction times increase as number of distractors increases STROOP EFFECT a task designed to show the automaticity of language ○ you can’t help but read words in front of you when you look at the word you don’t see a meaningless string of symbols when you see a word such as “cookie” ○ reading is fast and automatic behavior identifying color is also a quick process but it is still slow compared to words ○ two automatic processes–the identification of color and the identification of word meaning—conflict with each other test: ○ identifying color of the font as quick as possible eg. “pig” → red ; “black” → blue ○ result: congruent (words and colors match): “blue” → blue / incongruent (words and colors don’t match): “blue” → red congruent trials where the color of the font and label match show a faster reaction time on average; incongruent trials where the font and label are mismatched typically take longer to make a response theories stroop interference conflict between the color label and the written word on an incongruent trial ○ processing the word text conflicts with color recognition process → slower reading ○ competition of two answers for the response race model two sources of information (color and meaning) are running next to each other in mind and the faster runner wins ○ since your brain has a well-worn path for comprehending words, information can travel quickly along this path but there is now a second, less automatic path for naming a font color, and it is also racing to the finish line ○ slower on the incongruent trials because two sources of information show up to the finish line with different answers ATTENTIONAL BLINK even if you look right at something, you can still fail to attend to it experiment: ○ naming the alphabet with the alphanumeric sequence flash ○ types 1 number between target letters (180 ms between the appearance of the first and second target letters) 3 numbers between target letters (360 ms between the appearance of the first and second target letters) → second letter most missed 7 numbers between target letters (720 ms between the appearance of the first and second target letters) No second target letter ○ the attentional blink is the failure in reporting the second of two targets when it’s presented in close succession after the first target even when looking directly at stimuli, the brain processes unreported targets to a late stage, but without conscious awareness the blink effect occurs even when participants know about it, implying it is difficult to override with top-down control people can be unaware of unattended stimuli, even though these stimuli are processed by their sensory systems ○ eg. driving–if your attention is on your phone, you can miss hazards on the road, potentially leading to serious consequences attentional blink can be reduced under specific conditions, suggesting it is more about how attention is tuned rather than an absolute capacity limit selective attention focus on important stimuli while filtering out irrelevant ones bottom-up (exogenous): automatic attention, such as a sudden noise or the timing of targets (like in the attentional blink) top-down (endogenous): conscious control of attention, such as choosing to focus on specific information and ignoring other distractions ○ eg. focusing on one conversation while ignoring other conversations happening around divided attention dividing attention between multiple tasks decreases efficiency and increases errors → reduce performance and awareness of critical information participants missed only a few important events when focusing on one superimposed video, but made eight times more detection errors when dividing attention between two videos