Cognitive Psychology Ch. 34

1
Cognitive PsychologyCh. 3-4
2
? Dept. Cognitive Seminar

• Join us this Friday for a cognitive seminar
  double-header
• At 11 AM, a special talk will be given by Dr.
  Mike Dixon (University of Waterloo) on
• "Roses are red and sixes are blue The colourful
  world of grapheme-colour synaesthesia"
• Harry Hickman Building (formerly the Centre for
  Innovative Teaching), Room 116 .
• At 230 PM (our regular time), Michelle Arnold
  will present her latest work on "Objective versus
  subjective experience in the knew-it-all-along
  effect."
• David Strong Building, Room C113.
• To view previous and upcoming talks please go to
• http//web.uvic.ca/psyc/cognitive/seminar.html

3
Colour Attributes

• Three main physical properties along with their
  corresponding mental qualities
• Wavelength Corresponds to experience of hue
• Intensity Corresponds to experience of
  brightness
• Purity Corresponds to saturation (richness)

Source http//home.wanadoo.nl/paulschils/04.03.ht
ml For a historical chart of colour theories,
see http//home.wanadoo.nl/paulschils/08.00.html

4
Transduction in Vision

• Retina contains light-sensitive cells that react
  to light by creating neural impulses
• Rods Sensitive even in low light (only to shades
  of grey) and movement
• 120 million in retina, concentrated in periphery
• Cones Sensitive to fine detail (visual acuity)
  and colour
• 6 million in retina, concentrated in the fovea
  (100µ)

• Photopigments create chemical reaction to light
• These break down in bright light, regenerate
  after time in low light (this enables dark
  adaptation)

5
Consciousness as a mongrel concept

• Toulmin (1982) distinguishes four senses of
  conscious
• being aware
• being attentive
• being articulate
• knowing together
• i.e., con-scious a form of collective, group,
  class or cultural awareness
• Block (1995) four meanings of consciousness
• access consciousness
• the manipulation of representations which has the
  potential to influence your reasoning,
  communication, or behavior
• phenomenal consciousness
• a subjective awareness of what our mind is
  currently doing
• monitoring consciousness
• one's ability to reflect on one's own thinking
  processes (metacogntion)
• self consciousness
• one's general knowledge about themselves

6
Subliminal Perception backmasking

• research indicates that messages embedded in
  various media (e.g., print ads and rock music)
  are actually creations of the lookers' or
  listeners' imaginations.
• top-down processing (expectation) determines
  message, not information in stimulus
• Vokey and Read (1988) made backward recordings of
  Lewis Carroll's Jabberwocky and the 23rd Psalm
• through "creative listening, they identified six
  sequences (per recording) that could be heard as
  something meaningful e.g. "Saw a girl with a
  weasel in her mouth "I saw Satan etc.
• for one passage, participants were to listen to
  see if they could hear the messages designated by
  the experimenter
• for the other passage, they listened without any
  prior prompting about what they might hear
• Results
• participants were very good at hearing the
  backmasked message, but only if they were told to
  expect it none of the control sequences were
  found

7
Subliminal Self-Help?

• presenting subliminal messages on tapes in order
  to offer assistance with everything from weight
  loss, to alleviating depression, to curing acne,
  etc.
• difficult to distinguish effects of subliminal
  perception from motivational effects
• Greenwald, Spangenberg, Pratkanis, and Eskenazi
  (1991) conducted a double-blind study on the
  effectiveness of subliminal message tapes
• gave participants one of two tapes with an
  embedded subliminal message
• the message was designed to improve either
  self-esteem or memory
• but the label on each tape was randomly assigned!
• participants were instructed to listen to their
  tape every day for one month, and were given
  pre-tests and post-tests of memory and
  self-esteem
• Your prediction?

8
Subliminal No-Help, cont.

• Greenwald, Spangenberg, Pratkanis, and Eskenazi
  (1991) (cont.)
• Results
• the post-test showed no improvement over the
  pre-test measures in any of the conditions
• an embarrassing subject-expectancy effect was
  found
• participants believed they had improved in
  whatever quality was on the tape label (even if
  the tape had been labeled incorrectly!)
• Note stimuli did not meet the minimum stimulus
  conditions necessary for an (objective-threshold
  ) subliminal influence
• participants were unable to distinguish between
  tapes that did and did not have embedded messages

9
Subliminal Stroop Effect

• Marcel (1983) used a Stroop task to study
  subliminal perception
• Subjects were asked to name the colour of a
  presented colour patch, but interference was
  created by presenting a conflicting color name
  before the patch GREEN (MASK) ?
• the word presented before the color patch (called
  a prime) was (on some trials) presented
  subliminally (by presenting it for an extremely
  short period of time and masking it after
  presentation)
• masking involves following a stimulus immediately
  with a jumble of other stimuli, effectively
  obliterating the stimulus that was just seen
• the length of time between the prime and the mask
  had been determined previously for each subject
  individually
• adjusted to a point where they achieved only
  slightly better than chance performance in
  guessing whether the prime was present or not
• Your prediction?

10
Marcel (1983) Subliminal Stroop, cont

• results
• the color prime words influenced the speed of
  color-naming, even if the primes were presented
  at a level where participants reported not being
  aware of them (i.e., below their subjective
  threshold)
• if the prime word was consistent with the color
  of the patch, it facilitated responding
• if inconsistent, it slowed responding
• Subjective and Objective Thresholds
• subjective threshold in terms of the
  participants' phenomenological experience
  (self-report of whether the stimulus was
  perceived or not)
• that is, sometimes, we are not aware that we are
  aware, not aware that we have perceived that is,
  we report report nothing but behave differently
• (it may be that a response bias is involved
  here, as per signal detection theory)
• objective threshold the level at which
  performance on some task (not participants'
  self-report) indicates that the prime was not
  perceived
• e.g. we score no better than chance on a forced
  choice question about the prime
• Below the OT, will there still be ( or -)
  priming of response time?

11
Subliminal Semantic Activation

• Cheesman and Merikle (1984) used the same task as
  Marcel (1983) but varied conditions of threshold
• after presentation of a prime, participants were
  asked if they had seen anything
• almost always reported a lack of awareness of any
  prime word
• then presented with four color names and forced
  to choose which they had seen (if there was one,
  what colour was it?)
• if guessing performance matched self-report, they
  shouldn't have been able to guess the color more
  than 25 of the time
• But subjects guessed the color they had seen at a
  level much higher than 25, indicting that some
  info about stimulus must have been processed
• procedure (as in Marcel,except.)
• objective-threshold condition presented primes
  under conditions that had led to chance levels of
  performance in the pre-test
• subjective-threshold condition presented primes
  under conditions that had led to 55 accurate
  guessing in the pre-test
• Your prediction?

12
Below the Objective Threshold?

• results
• bars indicate difference (savings or cost) in
  reaction time comparing primed vs. non-primed
  trials
• primes presented below an objective threshold
  (25 is chance) failed to influence performance
  -- no priming occurs.
• primes presented below a subjective threshold
  show small but reliable priming (usu. around 1/20
  s) Subliminal Semantic Activation
• i.e., slowed response time (RT) when incongruent
  with color patch, and speeded RT when congruent.

13
Standard Model Atkinson Shiffrin (1968)

• External world as meaningless physical energy
  proximal to sensation is also without meaning,
  and totally unlike the experience
• Meaning emerges from process, not at start
• Transduction converts energy into information for
  use. Sensory receptors pass this initial
  information on to a sensory store.
• Sensory store keeps copy (or trace) of input to
  allow for STM information processing it is
  precategorical (not meaningful)
• Consciousness and meaning emerge through process
  (to far right)
• Influence of this model has been great but is now
  diminished

14
Visual Sensory Memory (iconic memory)
A Q 6 8 T P W 1 2 Y 6 L

• Sperling (1960) Whole-report vs. Partial Report
• whole report technique
• subjects saw a display of letters for 50 msec
• subjects then asked to recall the items
• results subjects recalled 4 items
• partial report technique
• subjects saw display for 50 msec
• 50 msec later subject heard a tone indicating the
  row of the display to be recalled
• high pitched tone--top row
• medium pitched tone--middle row
• low pitched tone--bottom row
• results subjects recalled 3-4 items (regardless
  of the row)
• also, partial report advantage disappears if
  stimulus is masked
• Implication
• 9-12 items available in iconic memory - we see
  more than we remember!

Immediate Report 4 correct 33 of
array Immediate Cue of Row -3 correct implying
75 of array available
15
Sperlings Results Partial Report and Iconic
Memory

• So why are only 4 items recalled in whole report
  technique?
• One idea the image of the items fades so
  rapidly that a person can only report 4 items
  before the rest of it fades from iconic memory

• To test this hypothesis, Sperling varied the
  delay between the offset of the letter matrix
  display and the onset of the tonal cue
• although 80 of the matrix could be reported when
  the cue was presented immediately, only about 50
  could be reported after ¼ of a second
• after about ½ second, recall had fallen to the
  rather meager levels seen in the whole-report
  condition
• the entire matrix was available initially, it
  faded rapidly, and was gone within about 500 ms

16
Characteristics of Iconic Memory

• Large Capacity
• Up to 17 letters, up to 2 seconds (although
  duration is typically briefer, 500 milliseconds
  to 1 second)
• Capacity depends on stimulus conditions
• dark fields enhance, bright pre and post
  fields cut down duration
• Spontaneous Decay and Potential to be Erased
• Loss of info due to spontaneous decay (rather
  than interference)
• Decay begins at onset of target, not offset
  (DiLollo, 1980)
• Information can also be lost due to a following
  stimulus acting as a mask (Breitmeyer Ganz,
  1976 Turvey, 1973)
• Display followed by location cue ( _ or ?)
  circle wipes out target
• Precategorical Representation
• Initially thought to be a physical representation
  as selection of items on physical characteristics
  (ex. color) possible but not on semantic
  characteristics (digits vs. letters) (Sperling ,
  1960).
• This has been challenged there may be some
  semantic info in iconic memory

17
Echoic Memory

• Auditory variety of sensory memory.
• Experimental Evidence
• Massaro (1970) identification of pitch of a
  short tone that was masked by a random auditory
  stimulus.
• If the tone was delayed, identification improved.
  Echoic memory duration was estimated by Massaro
  to be 250 milliseconds
• Efron (1970a) estimated echoic duration of 30
  ms tone to be 130 ms
• Neural Representation
• Lu, et al. (1992) found that volume of a test
  tone tended to average, as the delay before
  recall indicator tone increased
• Echoic memory is the lifetime of neural activity
  in auditory sensory cortex (as measured with
  magnetoencephalography, similar to EEG)

18
More on Echoic Memory

• Darwin, Turvey, and Crowder (1972)
• three different sequences of mixed letters and
  digits were presented
• one sequence was presented in each ear, while the
  third sequence was presented to both ears, giving
  rise to the perception that it was coming from
  midway between the ears
• whole report condition participants were to
  report each of the letters upon receiving the
  recall cue in addition they were to report the
  location of each letter (e.g., right ear,
  midpoint, etc.)
• partial report condition participants were cued
  to report just one of the three sets of stimuli
• results paralleled those for iconic memory
• partial report led to greater recall and its
  advantage became increasingly small as the delay
  between presentation and cue increased, as in
  iconic memory
• the partial report advantage lasts about 2
  seconds for echoic memory, considerably longer
  than iconic memory--500 msec
• note whole report condition required
  participants to report the letters and their
  location but the partial report condition didn't
  have this extra requirement
• may have inflated the difference found between
  whole and partial report
• consistent with this, a partial report advantage
  in echoic memory has been a difficult finding to
  replicate

19
Modality and Suffix Effects

• modality effect
• when presented with a short list of items (e.g.,
  numbers or letters) in either visual or auditory
  mode, recall of the last few items is much better
  for the words presented in an auditory mode than
  for words presented in the visual mode
• larger recency effect for auditory material has
  been taken as evidence for an "echo" that lasts
  for several seconds after stimulus presentation
• auditory lists feature the echo, while visual
  lists do not
• larger recency advantage for words spoken aloud,
  relative to words read silently (although
  findings here are not consistent)
• suffix effect
• the finding that an auditory signal or suffix,
  presented at the end of a list wipes out the
  modality effect the suffix wipes out the echo on
  which the advantage is based
• Suffix effect thought to confirm precategorical
  nature of echoic memory (but this has been
  disputed categorical cues can yield partial
  report advantage!)

20
Precategorical Acoustic Store (PAS)

• the PAS is a sensory storage system capable of
  holding a few pieces of auditory information for
  a few seconds following presentation
• this relatively intact auditory trace could be
  consulted in tests of immediate recall, leading
  to the an advantage for the last few items
  presented
• presenting another stimulus at the end of a list
  interferes with the information in the limited
  buffer, thus obscuring the recency advantage for
  auditory info
• problems
• recency advantage was obtained for non-auditory
  (lip-read) stimuli
• since the stimulus list was not auditory there
  should not have been an echo after presentation
  and hence no recall advantage for the last couple
  of items
• suffix effect found to be sensitive to conceptual
  aspects of the suffix, but the PAS should be
  insensitive to meaning (that is, if it is truly
  pre-categorical).
• backward masking
• auditory version involves presenting a high or
  low tone, which is immediately masked by another
  tone the task is to classify the pitch of the
  first tone
• accuracy in classification increases as the
  presentation of the mask is delayed, up to a
  delay of about 250 ms at this point, accuracy
  levels off
• inference an auditory image of the target
  stimulus exists for about 250 ms, after which it
  becomes unavailable for further processing about
  250 ms for the duration of echoic memory, an
  estimate that fits well with the estimates of
  iconic memory

21
What is Iconic Memory For?

• Haber (1983) argues against Iconic Memory as a
  useful concept, as ecologically invalid
• Icon supposed to maintain stable image of the
  world through eye movements
• Not supported by Irwin et al. (1983), who
  showed iconic information is not useful in
  integrating across saccades (rapid eye movements)
• Haber suggests the only thing the icon would be
  useful for is trying to read at night during a
  thunder storm
• Loftus (1983) argues that the iconic experimental
  paradigm is a useful tool and the real world
  argument of Haber is not appropriate as in many
  sciences lab situations dont mirror the real
  world but give us insights
• Reframing Auditory Sensory Memory
• auditory sensory persistence refers to the
  rapidly decaying trace of an auditory stimulus
  that remains for about ¼ of a second after the
  stimulus is gone

22
Perceptual Memory

• a more stable and (relatively) long-lasting
  representation of information that retains the
  perceptual characteristics of a stimulus
• an auditory perceptual memory might range
  anywhere from replaying the last thing said by a
  friend in conversation a few seconds ago (a
  short-term memory), to replaying your favorite
  song from the Broadway show you saw last month (a
  long-term memory)
• a visual perceptual memory could take the form of
  visually running through the last few places
  youve looked for your lost keys (a short-term
  memory), to recalling the horrible sight of the
  9/11 attacks on the World Trade Center (a
  long-term memory)
• similarities and differences between perceptual
  memory and sensory memory
• both retain modality-specific characteristics
  present when the event was encoded
• sensory memory is an extremely fragile and
  transient replica thats present for just a
  fraction of a second
• perceptual memories can last for seconds, days,
  even years

23
Chapter 4 Attending to and Manipulating
Information
24
Ch 4 Attention, Automaticity, Working Memory

• I. Attention, Theories of
• 1. Attention as a Gateway
• 2. Attention as Capacity
• 3. Multimode Theory of Attention
• II. Automaticity
• A. The Stroop Effect
• B. Automaticity with Practice
• C. Re-evaluating Automaticity
• D. Costs of Automaticity Action Slips
• III. Short-term Memory
• A. Limited Duration
• B. Limited Capacity Chunking and Word Length
• C. Coding in STM
• D. Forgetting in STM Decay Interference
• IV. A Modular Approach to STM Working Memory
• A. The Articulatory Loop and Articulatory
  Suppression
• B. Visuo-Spatial Sketchpad
• C. Central Executive
• D. Evaluation of Working Memory

COGNITION AND CONSCIOUSNESS Failure of
Early-Selection Theory COGNITION AND INDIVIDUAL
DIFFERENCES Sex Differences in Visual
STM COGNITION AND NEUROSCIENCE PETscanning the
Articulatory Loop
25
I. Attention

• Features
• allows for voluntary control of how we deal with
  incoming information
• basis for the conceptualization of attention as a
  gateway into working memory
• we can choose which aspects of the environment we
  wish to attend to or not attend to
• when dealing with multiple streams of
  information, we have two choices
• selective attention monitoring and responding to
  one event at the exclusion of others
• divided attention monitoring and responding to
  multiple events.
• limited in capacity we simply cannot
  (effectively) monitor all of the events occurring
  around us simultaneously
• Contrast large capacity of sensory memory
  register
• basis for conceptualizing attention as a
  reservoir of mental resources
• there is a limit in the attentional resources we
  have to devote to performing tasks

26
Theories of Attention

• Attention selects from uninterrupted info in
  sensory memory
• Attention may be a filter, gateway or bottleneck
  for sensory info
• Study with selective attention tasks
• Alternatively, attention can be seen as limited
  in capacity
• Study with divided attention tasks
• These two kinds of theories differ as to when
  stimuli can be identified

27
Early Selection Theories Problems

• Early Selection Theories
• multiple streams of information all make it into
  sensory memory
• at this point, the information sources are
  filtered on this, with only one source selected
  for identification and further processing (i.e.,
  meaningful identification)
• this initial processing encodes each source of
  information in terms of its physical
  characteristics
• after the selection of the attended message, the
  other messages are essentially discarded
• Problems with Early Selection Theories
• cocktail-party phenomenon when selectively
  attending to one conversation, if someone in
  another conversation says your name, you are very
  likely to notice
• shouldnt be able to recognize your name (or
  anything else), because this requires analysis at
  the level of meaning (i.e., analysis at a later
  stage of processing)

28
Dichotic Listening

• Treisman (1960)
• dichotic listening in which participants heard a
  different message in each ear, and were required
  to shadow one of them
• occasionally, the meaning of the shadowed message
  switched ears
• based on early selection, participants should not
  follow the meaning of the story when it switched
  ears
• mistakes like this were common

29
Early-Selection Theory

• first phase participants were asked to listen to
  a series of words
• each list contained three city names (e.g.,
  Minneapolis, Lafayette, Cincinnati) that were
  each followed by a slight electric shock
• the shock produced an autonomic nervous system
  response called a galvanic skin response
  (GSR)--slight sweating in the fingertips
• second phase participants performed dichotic
  listening and shadowing task
• the unattended message was a list of words which
  consisted of 12 critical words
• 6 words presented in phase 1 (the 3
  shock-associated city words plus 3 nouns not
  associated with shock) and 6 words not presented
  in phase 1 (3 city words, 3 nouns)
• some of the participants reported that they had
  heard words in the unattended ear, but they were
  unaware that these words were city
  namesseemingly, the words had not been
  identified
• results
• during the dichotic listening task, the three
  shock-associated city words from phase 1 produced
  a higher galvanic skin response relative to that
  produced by the 3 non-shock associated nouns from
  phase 1 (37.7 vs. 12.3)
• presumably they had to be meaningfully identified
  in order to produce a GSR
• the three city words not presented in phase 1
  produced a more intense GSR than the three nouns
  not presented in phase 1 (22.8 vs. 8.7)
• the GSR generalized to other items from the same
  semantic category indicating the original shock
  related words had been meaningfully identified

30
Attenuation Theory

• unattended information is not completely blocked
  from further analysis beyond sensory memory
  rather, it is attenuated or turned down
• early filtering of messages is partial, not
  complete
• unattended information (albeit weak) can make it
  through to working memory
• some words in our mental dictionary are
  permanently more available than others because of
  personal importance (e.g., your name) or are
  temporarily more available due to present
  circumstances
• e.g., the context of the sentence in Treismans
  ear-switching study or words that might indicate
  the arrival of shock in the Corteen Wood study
• so even the small trickle of information that
  makes it through the attenuating filter might
  well be enough to trigger recognition
• these stimuli require less information from
  bottom-up processing to trigger recognition
  because of the facilitatory effect of top-down
  processing (e.g., personal importance or current
  circumstances)

31
Late Selection Theories

• all incoming information (attended and
  unattended) makes it past the point of sensory
  memory and enters working memory to be identified
• after the information is identified, only the
  selected one enjoys further cognitive processing
  (conscious responding or entry into long term
  memory) non-selected information is quickly
  forgotten due to the limited capacity of working
  memory
• distinguishing attenuation theory and late
  selection theories
• attenuation theory proposes that identification
  of meaning in the unattended ear is the exception
  rather than the rule
• whether information is identified depends on the
  context or the exact nature of the information
• late selection theory argues that the
  identification of meaning is the rule rather than
  the exception
• identification of meaning is necessary all
  information (attended and unattended) is
  identified
• neuroscientific research is more consistent with
  an attenuation-type account
• attention is influenced by the type of top-down
  processing (e.g., context, current task demands,
  goals of the attender, etc.) proposed by the
  attenuation theory

32
Attention as Capacity (or Capacities?)

• the pools of resources are differentiated
  according to whether the input modality is visual
  or auditory and whether the response required is
  vocal or manual
• tasks interfere to the degree that they tap into
  the same pool of resources
• tasks requiring a vocal and a manual response
  will interfere less with each other than two
  visual tasks or two tasks requiring a manual
  response

• What are Mental Capacities/Resources?
• term is just a vague concept that's used as an
  explanation of attention, but it doesn't really
  explain anything it simply re-describes it
• definition is circular
• the notion of limited resources is used to
  explain why sharing two tasks is difficult but
  the fact that two tasks are difficult to share is
  used as evidence that there are limited resources

33
Cell Phones and Driving a bad combo

• Strayer and Johnston (2001) tested attention
  while driving and using cell phone
• participants performed a simulated driving task
  (moving a joy stick to keep a computer cursor
  aligned with a moving target)
• performed this task alone (single task condition)
  or in conjunction with a cell-phone conversation
  using either a hand-held or hands-free device
  (dual task conditions)
• control group listened to the radio instead of
  carrying on a cell phone conversation.
• periodically a red or green light would appear
  and participants were to respond appropriately on
  the joystick task (stopping or continuing
  depending on colour)
• the number of signals missed, and reaction time
  to the signal were compared in the single- and
  dual-task conditions
• participants carrying on a conversation were
  twice as likely to miss a signal in the dual task
  than single task condition (7 to 3)
• responses to the traffic light in the dual task
  conditions were 50 ms longer than responses in
  the single task condition
• when the second task was simply listening to the
  radio, there were no performance deficit
• using a hands-free device was no better than
  using the traditional hand-held version

34
Multimode Theory of Attention

• attention is flexible people can shift from
  early modes (processing only the physical
  characteristics of incoming stimuli) to late
  modes (processing the meaning of incoming
  stimuli)
• attending in each of these modes has an
  associated set of costs
• as selection proceeds to a later point in the
  information processing system, it takes more
  mental capacity
• Johnston and Heinz (1978) used a divided
  attention task--dichotic listening and simple
  light detection
• participants were required to shadow one of the
  presented messages while ignoring the other
• during the presentation of these messages, they
  also had to watch for a light signal, and were to
  press a key as quickly as possible when they saw
  it (included to assess the mental effort or
  capacity expended on the listening task)
• if a great deal of capacity was being used in the
  listening task, this would leave less capacity to
  perform the light detection task and slow down
  reaction time

35
Multimode Theory of Attention, cont.

• Johnston and Heinz varied the mode of attention
  (early vs. late) by manipulating the nature of
  the two presented messages, and hence the
  difficulty of the listening task
• early selection mode the two presented messages
  were different physically one was read in a
  female voice, the other in a male voice
• cheap in terms of attentional capacity, leaving
  more to perform the light detection task and
  reaction time should decrease
• late selection mode the two presented messages
  had the same physical characteristics (i.e., read
  by speakers of the same gender), but differed
  semantically (the meaning of the information
  presented was different)
• expensive in terms of attentional capacity
  leaving little capacity to perform the light
  detection task and reaction time should increase
• the difference in capacity required by the
  different listening tasks was assessed by
  examining how it affected response time in the
  light detection task

36
Multimodal Theory Results

• results
• the cost in RT (compared to performing the
  light detection task in isolation) was greater in
  the late selection mode than the early selection
  mode, supporting the multimode theory