Attention Flashcards
Meaning of attention according to William James
“Attention is…the taking into possession of the mind, in clear and vivid form, of one out of what seem several simultaneously possible objects or trains of thought. Focalisation, concentration of consciousness are of its essence”
General types of attention
Active attention - where attention is controlled ‘top-down’ i.e it is controlled and purposefully deployed due to some motivational reason for attending to a specific item
Passive attention - where attention is deployed in a ‘bottom-up’ fashion due to some property of external stimuli
What is focused attention?
Kind of thing we see when we present participants with two or more stimuli and ask them to attend (and/or respond) to only one of them
What is divided attention?
We can think of this like multi-tasking. We present participants with multiple stimuli and ask them to attend (and/or respond) to all of them
What is sensory buffering?
- Info that comes into thew senses if buffered
- Aim of the buffer is to hold all the incoming info until the attentional system can act on it
- This provides a valuable block to the incoming info long enough for the attentional system to select the important info
Types of sensory systems
2 types:
Vision –> Iconic
Auditory –> Echoic
Sperlings Iconic memory experiment
- Presented participants with a number of cards via a tachistoscope
- Each of the cards contained a number of letters, between 2 and 12, and these were presented to participants for 50 ms
- Participants were asked to report as many items as they could from the display –> whole report technique
- Brief display meant that we have to rely on the mental representation, not on the image
- Representation of the image is called the icon
What did sperlings iconic memory experiment find?
- Most participants could report back the complete display as long as there were fewer than 5 items
- If there was a greater number of stimuli onscreen than 5, participants only reported back between 4 & 5 items
- Suggests the capacity of sensory memory is 4.5 items
- Participants reported that they felt they had seen more, but couldn’t remember it
Sperlings cookie jar senario experiment
- Participants see the same display as they did previously
- Hear a tone after the display is removed telling them which line to report
- Partial report as they only report a part of the display
Results of Sperlings partial report
- By scaling up how many letters they reported against how many there were to possibly report that participants were able to report nearly the whole display
- Concluded there is no limit to sensory memory
- All of your perception is available to you
- Limitation is actually the speed at which you are able to report the required information
How is iconic and echoic memory limited
- Decay process
- Takes only 300 milliseconds for half the info to be lost
- After 1 second there was no advantage to partial report vs. whole report
- If the info is not moved onwards rapidly from sensory memory, then the icon stored will decay
- At some point it will be too poor to be recoverable
- It is attention that needs to act on the info stored to move it further in the info processing system
Cocktail party phenomenon
Cherry (1953) With several people in a room, there are physical qualities of the stimuli that we can take advantage of to differentiate people we are listening to: - Sex of the speaker - Voice intensity - Voice localisation
Dichotic listening task
- Present the same voice at the same intensity to both ears but with different messages
- Participant asked to ‘shadow’ one of the messages
- The question becomes, what do the participants notice about the message in the ear they are not attending
Results of dichotic listening task
- Message in non-shadowed ear is very poorly remembered
- If the speaker in thee unattended ear switched language, the participants generally failed to notice
- If they started speaking backwards it wasn’t noticed
- If a pure tone was played, or the pitch or gender was altered in the unattended ear, that was almost always detected
- Moray (1959) inserted a word 35 times into the message, yet still participants didn’t notice
Attentional selection theories
2 contrasting ideas:
- Early selection: We select out info at an early stage of processing
- Late selection: We perform detailed analysis of incoming stimuli before making a decision as to what is attended to.
Broadbent’s theory of attention
- Keen on investigating the nature of the bottleneck
- Used the dichotic listening task to investigate how people would manage to recall information as you overloaded them
Results of Broadbent’s theory of attention
- Incoming info from multiple channels is initially processed in parallel and placed in a sensory buffer
- One of these channels is selected to be allowed to pass through a filter based on its physical characteristics
- While the selected message is passed on for further processing, along a limited capacity channel, the message blocked by the filter remains in the sensory buffer, during which time it is decaying
- Although messages are short-lived in the buffer, we can switch between the input channels (in this case echoic memory), as long as too much decay hasn’t occurred, we can pass them through one after the other
Problems with Broadbent model
- Doesn’t allow for any of the unattended info to ‘leak through’
- Primarily used naive participants
- Underwood (1974) found that naive participants only detected 8% of digits presented to a non-shadowed ear, but an experienced shadower was able to detect 67%
- Access to info presented in unattended ear was analysed beyond the purely physical
- Moray (1959) found that although most of what was presented in the unattended ear went by unnoticed, participant were much likely to detect their own names in the unshadowed ear
Triesman’s (1964) attenuation theory
- Believed there was a filter but disagreed with Broadbent about its nature
- Incoming stimuli proceed through a hierarchy of processing levels (physical, syllable pattern to meaning0
- If insufficient processing capacity test towards the top of the hierarchy are omitted
Similarities and differences between Treisman’s and Broadbent’s models
Similarities:
- Assumption of a filter
- Filter prevents the non-shadowed input from interfering with the selected input channel
- Physical differences between the sources are a part of the selection of one channel over the other
- Both are early-selection accounts
Differences:
- Triesman’s model doesn’t assume a complete phasing out of the non-shadowed message, rather it is turned down/dampened
- Pre-attentive analysis is more complex, it’s not just on physical differences and cam be as complex ads semantic analysis
Support for the Treisman model
- More easily accounts for some of the findings
- E.g. when we hear our own name (Moray, 1959)
- In the Treisman model, we will process that to a certain level meaning that it is more likely to be perceived
- The reporting of different sets of numbers to each ear being serially reproduced is easier to explain if we first assume they all have a level of processing, with priority for one being the physical channels
Problems with the Treisman model
- Complicated
- We have a pre-attentive analysis that extends as far as semantic processing
- This is so far down the processing stream we wonder what is left for attentional processing
Late selection theories
- All info makes it to memory and that selection of information to act upon is done at that level
- Shadowing tasks are not ideal for testing: By asking someone to shadow one ear you are essentially selecting out the information in that one ear to save prior to any detailed analysis
- There are some clever ways of examining the viability of a late selection model, primarily through subliminal priming
Late selection theories: Deutsch & Deutsch (1963)
- Late selection model
- Broadbent supposed that only some information is selected and that the rest is not processed
- Treisman suggested that it is all processed pre-attentively, pre-working memory, with those inputs that best match the filter continuing on for further processing
- Deutsch & Deutsch assume that almost all incoming stimuli are sent on for further processing, with final selection for the very topmost level occurring at the level of working memory
Support for Deutsch & Deutsch
Lewis (1970)
- Dichotic listening task, with shadowing of one ear
- Words in the unshadowed ear could be semantically unrelated or synonyms of words in attended ear
- Vocal response to shadowed words were slower if they occurred with a semantically related word in unshadowed ear
What is the perceptual load model of attention
- Does not assume a specific point at which items are selected
- Dependent on the difficulty of the task
- When load is high, selection is made early
- When load is low, selection is made late
What is orienting
Our response to an immediate change in thee environment
What is overt orienting
The alignment of sensory receptors to the location of an external stimulus
What is covert orienting
The alignment of mental resources to an external stimulus which can be involuntary or voluntary
Unilateral neglect symptoms
- Damage to thee right parietal lobe
- Patient can’t orient (move attention) to one half of the visual field
- Visual, auditory, and tactile neglect have been documented
- No sensory damage
- Bumping into objects and people on their left side
- May fail to groom the left side of their body or eat the left side of what is on the plate
- Voluntary attention and movement to the ‘neglected’ side is possible but doesn’t always occur naturally
What can we tell about attention from neglect?
- Informs us about how selective attention is allocated
- SPace is important in visual input
- Selective attention selects from that (space-based) representation - or organisation - of the visual field
Spatial cueing task
- Known also as Posner spatial cueing or spatial orienting task
- Simple methodology and simple stimuli
- Used with patients and animals
- Reveals mechanisms of orienting: how selective attention is allocated
- Reveals the units of orienting: what is selected
Exogenous cueing
Valid trials - the event influencing attentional deployment to a spatial position facilitates responding to the target
Invalid trials - the event influencing attentional deployment to a spatial position hinders responding to the target
Neutral trials - neither hinder nor facilitate responding to the target
Endogenous cueing
Valid trials - if the symbolic information presented at fixation influences attentional deployment to a spatial position should facilitate responding to the target
Invalid trials - if the symbolic information presented at fixation influences attentional deployment to a spatial position should hinder responding to the target
Neutral trials - should neither hinder nor facilitate responding to the target
Differences between exogenous and endogenous
- Outcome is very similar
- There is deployment of attention to a region of space, but they act differently
How does attention act like a spotlight
- Attention starts at our fixation point
- Arrow appears within it
- Spotlight moves in the direction of the arrow