Lec 7/ TB Ch 7

Question 1

• Selective attention
• why we need attention
  
  • Bottlenecks
  • 3 types: Sensory, cognitive, motor
• why can attention help

Answer 1

• Selective attention refers to a set of cognitive brain mechanisms that enable one to process relevant inputs, thoughts, or actions while ignoring others that are less important, irrelevant or distracting. (the focus)
• Arousal: a global state of the brain reflecting an overall level of responsiveness.
  
  • Sometimes when we talk about attention, it is confused w/ arousal
  • • Why we need attention.
  • Bottlenecks: It is impossible to process everything at once.
  • There are Sensory – cognitive – motor bottlenecks
    
    • IOW: sensory – sometimes there is so much info, and it is difficult to dfind waldo (at the hippo)
• IOW: Cognitive - we need to think about things one at a time
• IOW: Motor: we have only 2 hands, we have to do things one at a time
  
  • Ex. robot on Mars have to take picture then collect sample (can’t do 2 things at once)
• Why can attention help?
  
  • Ex. Lion roaring at the right side, birds chirping on the left side
  • Attention helps you pay attention to the lion sound first
  • You want to ensure you are safe first b4 listening to birds
  • IOW: attention helps us survive
  • There’s a lot of into in the world, not all info are equally important
• Where does attention play a role?
  
  • Attention to vision
  • • Attention to audition / touch / smell
  • • Attention across modalities
  • • Attention to thoughts
  • • Attention to motor tasks

Question 2

Studying attention

• Can we measure attention directly
• 5 ways
• perceptual thresholds
• perceptual biases
• 3 ways to study attention

Answer 2

Studying attention

• How can we measure attention? – mainly indirect measures
  
  • RT
  • Perception:
    
    • perceptual thresholds: luminance contrast and influence attention
    • perceptual biases: ex. biased to look at black thing on ice when playing hockey
  • Motor precision / accuracy
  • Eye movements: overt shifts of attention but not covert shifts of attention.
    
    • Attention -> change eye movement
  • Brain activity (fMRI, EEG)
• How can we study attention?
  
  • Cues influence (hv cues)/bias attention
    
    • Posner’s attentional cueing paradigm (spatial based)
    • Natural biases (spatial based)
    • Feature-based cueing (non-spatial)
  • Visual search
  • Attention in time (ex. RSVP)

Question 3

Cueing

• cueing task
• methods: 2 steps
• RT: visual vs auditory tones
• Posner’s method - extra step
  
  • Valid vs invalid
  • stimulus driven vs voluntary
    
    • 4 other names for stimulus driven
    • 3 other names for voluntary
• Cueing effect
  
  • condition to achieve this effect

Answer 3

Cueing as a tool for examining attention

• Simple probe detection experiment measures RT (or perceptual thresholds)
  
  • # 1: fixate eyes on *
  • # 2: probe appears (can be visual – red dot, or auditory signals) -> press key
  • We respond faster to auditory tones (by 30ms)
• Posner: adding a cue
• Cue: A stimulus that might indicate where (or what) a subsequent stimulus will be: valid vs. invalid vs. neutral => helps measures cueing effect
  
  • Cue: red square
  • Cues – valid or invalid
    
    • Valid: cue shows up on the same the probe is -> faster RT; invalid = vv
• • Another dimension: Behaviour can be stimulus-driven or voluntary
    
    • Involuntary – kicking reflex; voluntary – sibling kicking for revenge
• Stimulus-driven cues: info (cue) conveyed through previous events at the same location.
  
  • Aka: (involuntary = reflexive, peripheral (always away from fixation point), exogenous)
• Voluntary cues: (spatial) info conveyed through cognitions & memory, often based on language or other symbols. (symbolic, central, endogenous)
  
  • Need to infer the meaning of the symbol (arrow) -> symbolic
  • Usually presented near fixation point -> central
  • Need to process the cues -> endogenous
• • This is an invalid trial: test probe shows on the other side of the voluntary cue
• Cueing effect: the difference (in RT, brain activity, etc.) between the effect of a valid and an invalid cue.
  
  • To have cueing effect, at least 50% of trials are valid ones

Question 4

Cueing cont

• Stimulus onset asynchrony (SOA)
• voluntary vs involuntary SOA
• peripheral vs symbolic cue curves
• Inhibition of return (IOR)
  
  • peripheral vs symbolic cue curves
  • Sun analogy

Answer 4

• What’s the difference between stimulus- driven/peripheral and voluntary/symbolic?
  
  • Partially independent neural structures. (each activate some common neural structures)
  • Stimulus onset asynchrony (SOA): the time between the onset of one stimulus and the onset of another.
    
    • IOW when the cue shows up then the probe shows up
    • Different time courses of SOAs (b/w voluntary vs involuntary); slower effects for voluntary cues.
  • Y-axis = cueing effects (RT difference b/w valid vs invalid cue)
  • X-axis = stimulus onset asynchrony
    
    • when the cue shows up then the probe shows up
  • Here: peripheral cue
    
    • Cue -> 10 ms -> target
  • Symbolic cue
    
    • Cue -> 100 ms -> target
      
      • This makes sense b/c you need to convert the symbol to its meaning to interpret it
    • • X
    • Inhibition of return (IOR)
      
      • Symbolic cue
        
        • If SOA increases (ex. 1000 ms
        • IOW: you see a cue -> 1000 ms waiting for probe
        • You might give up for some time
      • For peripheral cues
        
        • When SOA increases, the cueing effect decreases -> -ve
        • The RT for valid trials is slower than invalid trials (very counterintuitive)
    • IOW: once you attend a location -> nothing shows -> attention wanes
    • Next time there’s a cue that points you there -> more difficult to direct attention
  • IOW: you shift your attention to somewhere else
  • Ex. you go outside, the most salient stimulus is the sun
  • Since it is a involuntary stimulus -> you pay attention to the sun
  • The sun is there all the time, but staring to it all the time is no good -> you need to look somewhere else
  • Here, IOR allows you to shift your attention away from something that is very salient
    
    • -> then look at smth still very interesting (but less interesting than the sun)

Question 5

Cueing cont

• A type of cue between stimulus-driven and voluntary
• 2 steps
• Joint attention
• Overt shifts of attention vs covert

Answer 5

• A type of cue between stimulus-driven and voluntary
  
  • # 1: fixate at centre face
  • # 2: the gaze of the smiley face shifts to right. Your attention follows it
• This type of cue is somewhat voluntary (you learn to shift your gaze) and involuntary (you can’t help it)
  
  • Aka joint attention
• X
• Overt shifts of attention: A shift of attention accompanied by corresponding movements of the eyes.
  
  • Ex. presented a face -> your eyes look at the person’s eyes, nose, mouth
• Covert shifts of attention: A shift of attention in the absence of corresponding movements of the eyes
  
  • Ex. you see * -> and you fixate on it

Question 6

Cueing cont

• perceptual biases
  
  • Ex. faces & FFA in RH
• Line bisection task
  
  • 2 steps
• Grating scales
  
  • 2 steps
  • L vs R brain bias
• Subway sandwich case

Answer 6

Natural biases

• Perceptual biases: Asymmetries in perception between the left and right side of a stimulus.
  
  • Vary with task, e.g. listening to speech
  • Ex. when you listen to someone talking, your left brain is more activated, so you are more receptive on the right side
  • Ex. #1: you see 2 faces
  • # 2: which face is more M vs F?
  • # 3: the faces are mirrors of eo, and a blend (Top: L = M, R = F; Bottom vv)
  • Rationale: when we perceive faces, we tend to use are RH more (FFA is bigger in RH) -> perceptual bias to the left (same for age)
  • • Line bisection task
  • # 1: you see a bisection
  • # 2: indicate if this is exactly at the middle, slightly left or R
  • For some trials, the bisection is shifted to L or R; for others, it is exactly at the middle
  • When we are presented w/ those that are bisected exactly at the middle, we think they are biased to the L or R
• Gratings scales:
  
  • (the gratings increase in thickness/spatial f for L -> R or vv)
  • # 1: see a bunch of visual gratings
  • # 2: indicate Which bar has more of the thinner/thicker stripes?
    • NOTE: there are no differences b/w the 2 gratings
    • Ppl prefer to state there are more thin stripes when the skinny lines are on the left
    • • Gratingscales: fMRI study: greater activation of attention networks in the right hemisphere
  • In the RH: there are more orange patches located at the occipital and parietal cortex, interparietal interoccipital sulcus
  • This suggest there is RH dominance in spatial tasks
  • • Real world setting
  • When you go to sandwich places
  • Ppl are biased, right-handed
  • When they cut a new loaf of bread, the LS is shorter than the RS and they put the LS away
  • Moral of the story: make them cut a new loaf of bread, so you get a bigger bread (on the right)

Question 7

Cueing cont

• Feature-based cueing of attention
  
  • how it works
• How is it disadvantage

Answer 7

Non-spatial cueing

• Space-based cueing of attention
• Feature-based cueing of attention: attention is guided based on non-spatial information about features.
  
  • Cued feature becomes more “visible” throughout the visual field = outside the focus of attention.
  • Ex. focus on red -> see oval
  • Ex. focus on horizontal bars/blue -> see diagonal line
• Feature-based attention can be a disadvantage
  
  • If you are presented w/ a pic w/ lots of stripes
  • Looking at stripes only won’t really help you find the target (waldo)

Question 8

Visual search

• example
• Define visual search
• Target
• Distractor
• set size
  
  • influence on efficient search
  • define efficient search
  • influence on inefficient search
• Define feature search
  
  • Graph: RT vs set size
• Define conjunction search
  
  • # 1 vs #2 vs #3 box
  • Graph: RT vs set size
  • Red vs blue curve
    
    • amplitude
    • each item & RT
• Spatial configuration search define
  
  • Task
  • Red vs blue curve
    
    • amplitude
  • Why is Target present searches are faster than target absent?

Answer 8

Visual search

• Ex Among the bunch of bars. Is there one that is unique?
• When there’s more stuff -> even harder to find smth unique
• • Visual search: Looking for a target in a display containing distracting elements
• Target: the goal of visual search
• Distractor: any stimulus other than the target
• Set Size: the number of items in a visual display (ex. # of bars in total)
  
  • Has no influence on search time for “efficient searches”.
    
    • Efficient search: ex red bar clearly pops out among blue bars
      
      • Here, it doesn’t matter if you have many or few total bars, you can still locate the red one easily
  • Set size Impacts search time for “inefficient searches”.
    
    • Small set size -> fast; large set size -> slower down
• How much time does it take to perform a visual search task, i.e. to tell whether a target is present or absent? – It depends (of course…).
• X
• In a feature search: need to find red vertical bar, you can use only 1 feature to find the target
  
  • Easy to spot the target among distractors
  • The target and distractor differ by at least 1 feature (color/orientation)
  • RT vs set size
    
    • set size #1 -> #2 -> #3 = 5 -> 10 -> 15
    • Set size does not change RT
    • Also red = target absent; blue = target present
    • The RT is the same for both cases
• In conjunction searches: you need to use both (conjunction) of red and vertical to find the target
  
  • 1^st box: you can locate a red vertical var
    
    • Here there are other bars that are red, and other bars that are vertical
  • 2^nd and 3^rd bar -> harder
  • As the set size increases (as seen from box 1-3), the RT increases w/ set size
  • red = correct target is absent
  • The red curve is 2x the height and 2x the steepness of the blue curve
  • Red curve: for each item you add to the set -> RT increases by 10-30 ms
  • Ex. set size 10 -> 15 items; it takes you 50 ms longer (10ms x 5 items)
  • Blue curve: For each item you add to the set -> RT increase by 5-15 ms
  • Ex. set size 10 -> 15 items; it takes you 25 ms. Longer (5 ms x 5 items)
  • -> inefficient search
    
    • -> increase set size -> increase RT
• Spatial configuration search (even more inefficient)
  
  • Need to find T and its orientation among Ls
  • RT curve is steeper
  • The red curve is 2x the height and 2x the steepness of the blue curve
  • Target present searches are faster than target absent searches b/c only half the items need to be checked
    
    • Correct target present: Once you found the target -> task completed
    • Correct target absent: you need to search through the entire field

Question 9

Visual search cont

• feature searches
  
  • basic features - 4
  • less basic feature
  • Top vs bottom image
  • is it efficient
  • Salience
  • parallel search
• conjunction search
  
  • is it efficient?
  • // search?
  • real world example of conjunction search?
• Is conjunction search serial?
  
  • Yes: Serial self-terminating search - define
  • No: // processing limited
    
    • “adjustable spray nozzle” model
  • Final verdict?

Answer 9

Visual search cont

• Feature searches are efficient (= differs by one feature)
• Basic features: colour, size, orientation, motion
• Less ‘basic’ features (it seems) are yet efficiently searched: lighting direction
  
  • Top image: interpreted in 3D, can see lightning directions. -> “Pop-out”
  • Bottom image: nope
• Salience: the vividness of a stimulus relative to its neighbours (feature contrast “clearly” above JND threshold)
  
  • Ex. red vs blue bar
  • Ex. inefficient = purple & another purple that is one shade lighter
• When we look at a scene, we can process a feature (lighting direction) for all items in parallel
• Parallel: the processing of multiple stimuli at the same time
• Is there serial search -> maybe (in conjunction search)

Conjunction search is inefficient

• No single feature defines the target (EX there is red or vertical bars)
• Defined by co-occurrence of 2+ features
• Conjunction search Fairly inefficient: b/c larger set size ->slower RT
• Searches are no longer parallel: you need to look at each item and determine if this is a target
• Real-world conjunction search
  
  • Go through parking lot to find your own car (ex Silver Toyota matrix)
  • There are many matrixes and many silver cars

Is conjunction search serial?

• • YES
  
  • Serial self-terminating search: items are examined one after another until target is found or until all items are checked
  • Attention shifts are similar to eye movements scanning a scene; but it is faster (because we don’t use eye-movement).
• • NO
  
  • Limited capacity in parallel process:
  • You have a complex scene, you process everything in parallel
  • Since there are so much stuff going on it takes you longer to process
  • IOW: you have a limited attention or resources, since there’s so much stuff going on, your resources are spread thin
  • “adjustable spray nozzle” model
    
    • If you water a small Garden, you can use nozzle is smaller you mind and water is more concentrated
    • if you water larger pot of grass, you use a wider nozzle and the water is more thin (this takes more time)
• • COMBO?
  
  • Neurophysiological evidence that both mechanisms (serial & // processes) co-exist.

Question 10

Visual search cont

models

• feature integration theory
  
  • 2 main stages
• Guided search theory
  
  • 2 steps
  • binding problem - illusory conjunctions - define

Answer 10

Visual search cont

Modelling visual search

• # 1: Feature Integration Theory (Treisman and Gelade):
  • 1.Preattentive Stage: parallel processing of basic features across entire visual field before selective attention is deployed
    
    • Smth pops out -> attention (red bar among blue bars)
  • 2. Attentive Stage: when you look at the scene, you don’t see the conjunction (EX red + vertical bar)
       * You need to use spatial attention (shift attention from 1 item to the next), each time it binds together features (ex. red and vertical) for one item at a time, serial search
• # 2: Guided search theory (Wolfe): Attention can be restricted to a subset of possible items on the basis of information about the target’s basic features.
  • # 1: if you only know there is only 1 red horizontal bar -> you pay attention to the red items first (as it is more salient)
  • # 2: then among the red bars -> look for horizontal bar
  • -> direct attention -> bind feature (shape + color)
    
    • But this is very simplistic

The binding problem

• illusory conjunctions: Can you recall the letters that are present? Can you recall their color?
  
  • It is easy to recall only letters or only color
  • Recalling the color + letter is tricky: you may recall a purple Y
  • Since you don’t have enough time at the task, you can’t direct attention to the specific location and bind the 2 features together

Question 11

The Attentional Blink: (In-)attention in Time

• RSVP - define
• normal speed for 100% accuracy
• When there’s 2 targets → Attentional blink - define
• RSVP task
  
  • Overall process
  • # of targets
  • Lags
• Graph → what does it chow
• Marvin Chun’s fishing metaphor for attentional blink
  
  • Case 1: F1 → F2
  • Case: F1 & F2 come together

Answer 11

The Attentional Blink: (In-)attention in Time

• RSVP: Rapid Serial Visual Presentation is a method of displaying information at one location in which each piece of information is displayed briefly in sequential order
  
  • Normal understanding with 250 words per minute.
  • 650 wpm (speed things up): 20% reduction. (understanding)
  • Special case: Visual search in time for a target.
  • Ex. press a button when you see a # only
  • AVYWLNF4RUXFHVX
    
    • We can see 8-10 items per second and detect them accurately
  • RSVP tasks has 2 targets, T1 & T2
• Attentional blink: difficulty in perceiving the second of two targets within a rapid stream of distractors; depends on whether the observer responded to the first target presented 200-500 ms before.
  
  • Reduced attentional blink for smaller/larger time differences between T1 & T2
• Method: see fixation point -> D -> blank screen -> M …
• Target 1: find a letter different in color (ex. white) -> push a button
• Target 2: Also need to push a button when you see X
• Lag 2 = from L to X, there’s a lag of 200 ms
  
  • In this scenario, when there is a lag -> you have trouble seeing X
• Y-axis: probability of getting T2 correct given T1 is correct
  
  • Lag 1 = performance is good
  • Lag 2 = near chance
  • Lag 3 = bad
  • Lag 4 = better -> recovering
• Marvin Chun’s fishing metaphor.
  
  • RSVP: try to get F1
  • See f1 -> scooped out fish
  • F2 passes by
  • Put the net back, the fish passed
  • Catching = awareness
  • • IOW: you will miss F2 if there is not enough time b/w F1 and F2
  • If F1 and F2 comes together -> you can catch both fish at the same time
  • -> explains the attention blink for lag 1 is not as bad compared to lag 2 (the curve)

Question 12

The physiological basis of attention

• early stages of processing
  
  • characteristic
  • Condition 1: Task
    
    • 2 steps
    • main result
    • Issue:
  • Condition 2:
    
    • 2 steps
    • major difference compared to condition 1
    • result
  • Activation pattern in later stages of processing (stronger/weaker?)
• EEG study
  
  • 2 steps
  • ERP graph
    
    • blue curve & red curve
      
      • define
      • major difference
  • Skull graph activation
    
    • LHS vs RHS
• Endogenous vs exogenous attention
  
  • how the results differ b/w endo/exo attention
• IOR → results?

Answer 12

The physiological basis of attention

• Examples of Physiological Areas Involved in Attentional Processing
• Early stages of cortical processing: influenced by attention in a retinotopic manner
  
  • Brefczynski & DeYoe (1999)
  • # 1: Showed a grating
  • # 2: fixate in the middle point of the screen while looking at a segment that flickers (orange thing)
    • Then you see different segments in different blocks of the trial
    • They differ on how far they are from the fovea
  • Results (left)
    
    • Red = activation
    • Blue = deactivation
    • # 1 box: flicker near the fovea -> v1 activates
    • # 2-4 boxes: flicker near the periphery -> another v1 area is activated
    • Thus v1 is activated in a retinotopic manner: the further away the stimulus is from the retina, another part of v1 is activated
    • Here you see 1 thing at a time, and pay attention to 1 thing at a time
    • -> visual stimulation -> change your attention
    • IOW: you don’t know if the activation is caused by you paying attention to the stimulus or due to flickering of the stimulus
  • The authors added a 2^nd condition
    
    • # 1: you see the whole bull’s eye, everything is flickering
    • # 2: you are cued/told to pay attention to a segment near the fovea or in the periphery
  • The areas activated in this condition is similar to that of the previous condition
    
    • activation depends there’s smth visually present + attention is directed there
  • Thus, attention activates area v1 in a retinotopic manner
• This is also seen increasingly stronger effects from V1 to extrastriate areas (V2, V4, LOC etc.)
• X
• Use EEG
  
  • # 1: have fixation point in the middle of the screen
  • # 2: smth is flashing at the periphery at random times -> this causes a chain of reaction in your visual system
  • ERP graph
    
    • Y-axis = voltage
    • X-axis = time
      
      • T = 0: stimulus presented
      • T = 120 ms: 1^st +ve peak (p1)
      • T = 180 ms: 1^st -ve peak (n1)
    • Blue curve = attention (LHS pic: yellow circle) is away from the location of stimulus
    • Red curve: vv
      
      • The p1 and n1 is more pronounced: Spatial attention amplifies ERP responses.
      • Greater amplitude = gained modulation
      • -> Gain modulation (no change in map) of P1 and N1. (also seen in skulls)
      • • The skull
      • LHS: do not pay attention to the stimulus
      • 90-130 ms (aka p1)
        
        • There is activation in left posterior area when the stimulus appears on the RS
        • (vv)
      • RHS: you pay attention to the stimulus
        
        • Activation in the same location but more strongly
  • • Endogenous & exogenous
  • You see a gain of modulation in both endogenous and exogenous shifts of attention
• IOR: reduced P1 and N1. (reduced modulation)
  
  • You have attention cue
  • Stimulus shows up at 700 ms (IOR: target shows up really later after the cue)
  • P1 and N1 reduced in amplitude
• X

Question 13

The physiological basis of attention cont

• Shows blended stimulus of face and house
• How does attention influence brain activation?
• 3 ways responses of a cell can be changed by attention
• x
• v4: receptive fields → ???
  
  • no attention → what we see? (ex bunnies)
  • Attention → how does v4 behaves (to bunnies)

Answer 13

The physiological basis of attention cont

More high lv areas

• O’Craven & Kanwisher, 1999:
  
  • # 1: showed sandwiched stimuli
    • You can focus on house or face
  • PPA or FFA light up depending on which layer is attended
    
    • Ex. attend to house -> activate PPA; vv
• • Attention and single cells
• Three ways responses of a cell could be changed by attention
• 1. Response enhancement (Treue & Martinez Trujillo, 1999)
     * Ex. neuron prefers vertical orientation
     * When the monkey pays attention to the stimulus -> the amplitude of tuning fx increases
• 2. Sharper tuning (Lu& Dosher, 1998: noise exclusion)
     * Psychophysics
     * If the monkey pay attention to the stimuli (vertical orientation), more strongly prefer to the vertical orientation
• 3. Altered tuning in space (Moran & Desimone, 1985)
     * If you pay attention to the oblique stimulus, the tuning fx/ the neuron prefers from vertical to oblique orientations

Moran & Desimone, 1985: V4: receptive fields zoom in/shrink

• Tuning for location in space
• V1 receptive field is large
  
  • All the bunnies fit in the receptive field of the neuron (top square)
  • We dunno what the neuron is responding to
• If monkey is cued to respond to a specific bunny
• The v4 neurons behave like the bottom image
  
  • Attention -> Receptive field shrinks to focus on 1 bunny

Question 14

Attention: how do we perceive whole scenes?

• Scene perception: good vs bad
  
  • Task → present 16 scenes → how many can we recognize?
  • Change blindness
• Global vs local scene recognition
  
  • Study: computer classified scenes based on which 2 dimensions?
  • What does this suggest?

Answer 14

Attention: how do we perceive whole scenes?

Picture memory and change blindness

• Present 16 scenes -> which one hv you seen b4
• We can correctly remember very large numbers of photos scenes.
• Potter (1975, 1976) present the scenes fast -> recognize many
  
  • fast RSVP for scenes.
• X
• Change Blindness: failure to notice a change between two scenes; perception depends on meaning of change
  
  • If smth meaningful is removed (ex. trebuchet that attacks the castle)
• • Suggests our scene perception is very poor.
• • How does that fit together?

Local and global approaches to scene recog

• Spatial layout of a scene: description of the structure of a scene
• • Global vs. local scene analyses
• X
• • Global: Oliva & Torralba (2001):
  
  • Wrote computer programs
  • Computer that extracts global scene info from photo
  • scene classification based on a few easy-to-process scene dimensions:
    
    • – Spatial frequency
    • – Openness
    • – Naturalness
    • – Roughness
• X
• Scenes organized according to 2 dimensions.
  
  • Depth (ex. no depth = lack vanishing point)
  • Openness (ex. highway)
• • Scenes with similar meaning tend to group together.
  
  • Even the algo did not extract pics based on meaning
• This help w/ Fast scene understanding; might result from spatial frequency analyses

Question 15

• We can’t see global or local/details?
• Ex. 4 cases we missed the details…

Answer 15

What do you actually see?

• Change blindness might result from an inability to “see” more than one item at a time.
• • The impression of a rich percept of the word around us is an illusion.
• We don’t see the indiv details
• # 1: Demo - Ex. basketball & gorilla
• Attention is so powerful that the gorilla is blocked from entering conscious awareness
• # 2: Misdirecting: magic tricks
  • Ruber band/runny noise trick
    
    • Band is attached to wrist
    • Performer makes hand gestures that draws our attention away from the rubber band to the nose -> illusion we are snapping at the nose
• # 3: Misdirecting attention: ball disappear in thin air
  • Magician gazes upwards (we tend to follow ppl’s gaze) while his hand holds the balls -> looks like balls disappeared in thin air
• # 4: Misdirecting attention: pickpocketing
  • Confederate is doing tricks
  • Ppl are focused on the tricks
  • Pickpocketer stand among the crowd, gazing upwards -> pickpocket
  • Gazes upward b/c if others are looking at him, they tend to follow their gaze (upwards) rather on their pickpocketing action
• X
• Las vegas entertainer
  
  • Touchy
  • Hold women’s hand – direct attention there
  • Touch women’s shoulder (less obvious – puts the coin there)
  • -> ppl don’t see him removing the watch