Wk 10 - Attention Flashcards
Describe attention (x3)
Concentration and focusing of mental effort
Selecting what is relevant from sensory input and processing it for appropriate action – the prioritising of cognitive operations, in order to allow response selection
Does not exist separately from cog processes – is an organiser of, in line with our goals
Describe early vs late selection issues in attention (x2 plus 2)
Broadbent’s filter theory: perceptual features, eg voice etc, used to filter out irrelevant message; a structural model – filter stops info flow through system No semantic processing of unattended info, but still bias outcomes, shown by: River bank/money bank experiment Late selection theories: that everything is processed semantically, and then selection occurs 2 responses (tap and shadow) can’t be organised to 2 inputs at same time, so difficult to critically test theory
Differentiate feature/conjunction search
Feature search used to distinguish target on 1 simple feature
Is pre-attentive: rapid, automatic, unconscious, parallel processing
Conjunction search used when object defined by number of features, eg Where’s Wally
Is focussed: slower, requires, attention, serial processing
Outline FIT (feature integration theory) (x3)
Human visual search is result of two distinct search processes:
Pre-attentive (single feature) or focussed (conjunctions)
Serial is slow, but efficient in terms of enabling more complex processing than could be otherwise achieved given neural constraints on capacity to form automatic detector for every possible combo of features
Describe evidence for endogenous control of visual attention in visual cuing tasks (x5)
Contingent capture Roger Remington
Not stimulus driven exogenous capture – what we see is what we’re set for
Spatial cue is given about location of upcoming target
If it’s defined as red, red cue captures attention even if not predictive of target position
Whereas a bright highlight (green) doesn’t capture
What is the rationale of RSVP? (x6)
To examine temporal rather than spatial attention
One location, time pressured
Letters etc displayed after one another in same location, about 100ms/item
Ps asked to look for certain targets, then asked questions about them at the end
Conceptual/semantic processing shown even at this rate by Mary Potter
Post-target intrusions common (confusing features of target)
Describe the AB (x5)
When the processing of task 1 causes interference with task 2 - a failure of awareness
Established through two-target RSVP
T1 and T2 will be flashed in amongst 13 other numbers/symbols
Find impairment at a few hundred ms
Out to about T1 + 6, but not seen at T1 + 1 – processed as one event
What are three popular theories of the AB?
- Chun and Potter 2 stage model:
- Boost and bounce theory, Olivers and Meeter 2008
- Wybel’s Episodic Simultaneous Type Serial Token – a computational model (mathematical expression)
What are three key concepts of attention?
Capturing attention: out of our control/bottom-up; guided by expectations/top-down
Selective attention: responding to relevant/ignoring irrelevant
Divided attention: among concurrent tasks
Describe Chun and Potter 2 stage model of the AB (x4)
Items matched with memory representation (conceptual processing)
Then consolidated in WM in order to report
AB due to limits on memory consolidation of T2 – as T1 not yet completed
Fits that more difficult T1 to process, harder T2 will be
Describe Olivers and Meeter 2008 Boost and bounce theory of the AB
Boost increases the activation
Bounce inhibits distractor activation
AB due to lag in rise of boost and bounce
Distractor immediate after T1 gets some of the boost/accesses working memory
Then selection of distractor causes strong bounce which affects T2
So inhibition deployed to help target selection causes the AB
Doesn’t explain AB for sequential targets: even when no distractor to cause bounce, still get an AB - more capacity limits than strategic effect
Describe Wybel’s Episodic Simultaneous Type Serial Token – a computational model (mathematical expression) of the AB (x4)
Type is representation used to ID something
Token is representation of particular occurrence of something
Combine to record stimulus occurrence in space/time – where/when
WM encoding suppresses attention to new distractors = AB
Three explanations of attentional limitations are…
Structures – often shown using box and arrow models: bottleneck/reaction time models strongly predict behaviours; shut the gate – to exclude rest of info; stores for info retention
Processes: resources with capacities are specific but open to manipulation depending on task demand; divide resources to limit attentional costs etc; improve resources through training (maybe)
Strategic: developed strategies used to manipulate attention; limitations = coordinating action
Describe the river bank/money bank experiment
Attended message had homograph (same written form but different meaning) – ‘they threw stones at the bank’
Unattended message had river or unrelated control word
Ps to choose sentences closest in meaning to the attended message
More likely to choose river meaning if ‘river’ occurred in unattended
Must have processed the meaning of river, although couldn’t say they’d heard it
Endogenous control of attention is… (x4)
Eg (x1)
Shifting attention voluntarily Directed by goals Internal Top-down Tune out conversation to listen to another
Exogenous control is… (x2)
Eg (x1)
Automatic response to important stimulus
Botton-up
Attention captured by hearing friend’s name in conversation
Inattention blindnes is... (x1) Eg experiment (x1)
A phenomena of endogenous control of attention
Gorilla experiment – looking for the gorilla makes you miss the curtains changing colour, and the team-member who leaves part way through
Change blindness is... (x2) Eg experiments (x2)
A phenomena of endogenous control of attention - related to inattention blindness
Failure to notice change of stimuli
Eg objects on a table (Nevin and Simons 1997), the interviewer’s appearance/gender after builders walk through with cladding sheet (Simons and Levin 1998
Unilateral neglect is… (x1)
And results form… (x1)
Providing neurophysiological evidence for (x1)
Inattention to other side of space
Not just visual, but attentional deficit - joining contralateral objects with lines eliminates problem
Separate what vs where and when processing
Limitations of the ‘spotlight’ concept of attention are highlighted by… (x3)
ie (x1)
Inattention and change blindness
Unilateral neglect
It’s not just where you look
Debate over whether attention and consciousness are the same thing continues… (x2)
They’re separate: things get to awareness sometimes without attention; argument for evolved capacity to let important stuff break through
They’re involved, attention is gateway to consciousness, but no solid examples of things you don’t need to attend to in order for it to enter consciousness
Dichotic listening tasks are used to identify… (x1)
And involve… (x6)
Results led to formation of… (x1)
The fate of unattended to info
Ps hear a different message in each ear concurrently, through headphones
Attend to 1 message, and shadow it (repeat aloud as heard)
Shadow ear – the one that’s attended to
Look at how the two messages interact, what they noticed: the physical features – speech vs music, and gender, pitch tone of voice
But not meaning – considered to be the end goal of processing
Can sometimes hear own names, but not accents or meaning of other messages
Early selection models of attention, eg Broadbent’s filter theory
Nellie Lavie’s Load Theory of atterntion selection holds that is… (x1)
Depends on the cognitive load of task involved
Debate over attentional control continues over whether it’s … (x2)
Evidence points at… (x1)
Resources (which could be changed) vs
Capacity (bottleneck supported by evidence)
Kahneman, 1973 proposed attentional control as processing capacity based on processes rather than structures… (x3)
Attention as process of allocating resources to inputs
Concurrent tasks possible contingent on difficulty
Resources are increased under arousal (motivation)
Capacity theory of attention control proposes… (x1)
And is tested through… (x2)
Limited general resources
Divided attention attention tasks
Concurrent task limits set by task difficulty
Visual attention is assessed through… (x1)
Involving… (x4)
Visual search tasks
Identify the object among the distractors
Measure reaction time or accuracy – speed/accuracy trade-off
Response latency (RT) is primary DV: increases as identification, decision making or response selection get difficult
Accuracy a better DV for difficult/time-pressured tasks
Two limitations of FIT…
Features don’t always pop-out – such as in change-blindness
Configural patterns can over-ride basic features
Heterogeneity (difference) or homogeneity (similarity) of distractors – much harder if all distractors are different
Contingent capture is…(x1)
And involves… (x3)
That counters which theories? (x1)
A form of top-down control in visual search
What we see is what we’re set for
Spatial cue is given about location of upcoming target
If its defined as red, then red cue captures attention even if not predictive of target position, whereas a bright highlight (green) doesn’t capture
Exogenous/FIT theories
Masking in vision is… (x1)
And pattern masks… (x2)
When two stimuli shown close together, one stimulus masks another
Interfere with target processing
Compete with targets to engage perceptual processing – they ensure brief stimuli don’t reach awareness
Discriminating targets from distractors involves… (x2)
Attentional set: preparedness to select target features/reject distractors
Attentional engagement: locking attention onto targets
Feature detectors are… (x1)
Which fit into which theory of visual attention?
Modules/processors each of which detects a colour, or shape (e.g., angle/line) whenever it appears in the processors location
Treisman’s FIT
RSVP is… (x4)
Rapid Serial Visual Presentation - visual attention task, finds the AB
Stream of about 12 items – letters, digits, words, pictures – presented in a fixed location
Items occur one at a time, at about 100 ms per item
Position of T2 relative to T1 is varied – the lag
Ps report items at the end of each stream, need to ignore distractors
The lag-dependent deficit in T2 accuracy in RSVP results is… (x3)
The AB
The fewer items between T1 and T2, the lower is T2 accuracy
Accuracy recovers by about lag 8 (T2 is the 8th item after T1)
Lag-1 sparing in RSVP results is…
(about half the time)
T2 accuracy is unaffected at lag 1 (when immediately preceded by T1)
And then the AB deficit appears at lag 2
What are the two main classes of theories regarding the AB?
Capacity: the brain is full
Attentional control is lost
