attention Flashcards
Covert Paradigm - Posner (1980)
Subjects have to detect or identify targets that occur at one of two locations
Must keep eyes still
People are faster if target appears at cued location than non-cued location
Debate on if we have two systems or one system with two controls
Exogenous cues
Appear near target
Do not predict where target will appear
Automatic
Attention is automatic and involuntary
Endogenous cues
Appear at centre
Predict where target will appear
Wilful
Have to deliberately move your attention
Nakayama & Mackaben (1989) - performance decrease/increase based on cue time
Participants had to search for an odd target and report it
Target location cued (from 0 to 500ms) just before the display
Display very brief and then masked
Performance improves as cue target increases over about 100ms but then increasing cue to target makes things worse
Nakayama & Mackaben (1989) - Cue to target meanings
If cue happens immediately before target then it’s kinda shit
Cue comes a little bit (20-30ms) before target then performance increases - Cue attracts attention and gives us a head start
Peaks at 100ms
But with more time then performance decreases to like if cue wasn’t there
Exogenous attention can’t be sustained
Spotlight metaphor
Attention is a spotlight, highlighting one thing whilst the rest of the world is dark
Two methods of controlling the spotlight
Automatically moving to something interesting
Searching and being deliberately moved, constantly for something interesting
Attentional network
Posterior parietal lobe
Superior colliculus
Pulvinar
PET and fMRI has confirmed the role of parietal cortex in attention shifting in control subjects
Card trick
Only process the card we are focusing on
Ignore that every card that we could choose from is gone
Shows that we have change blindness
Change blindness
If they are presented with no interval, then the difference is easy to see
If they are shown with just 0.1s interval then the difference is hard to see
Have to use endogenous attention
Neisser and Becklen (1975)
Two superimposed videos
Subject press button whenever hand clapped or ball passed
Could do either task alone but not together
Head up displays
Having flashy spedometer? Not good.
Just looking at something doesn’t ensure attention
Brawn and Snowden (2000)
Attention must be object based to an extent, but space is still important
Cues either exogenous (lines flash) or endogenous (subject told which more likely)
Subject faster on the cued object
Visual search
Have to use endogenous attention in order to search for something like a character in Where’s Wally
Pop out effect
Easier to find red circle in left panel
Uniqueness attracts exogenous attention
Have to use endogenous attention on the right to search all cases
Feature search
Search for things defined by a single feature seems easy and effortless
Reaction time does not increase with increasing number of things to search
Item seems to pop out
Conjunction search
Searching for things defined with a conjunction seems hard
Reaction time increases with items
Increases twice as much for absent than present trials
If you have to search each item in turn until you find it then you will find it when you are half way through
Feature Integration Theory - Treisman’s
Each feature in the image is registered in feature maps
Maps used without any attention
To join these features requires attention - attention is the glue
Can do feature search without attention as no gluing is required
For conjunction you have to move attention from location to location gluing things together
Treisman and Souther (1985)
According to FIT it is easier to search for a lollipop in a field of circles than the other way round
Problems with FIT
Things that would not normally be regarded as features can also show pop-out
Other issues with FIT
Higher order things such as facial expressions also pop out
Nakayama and Silverman (1986) - Issues with FIT
Conjunction of depth and colour pops-out
Conjunction of movement and colour pops-out
Guided search (Wolfe)
You can use some knowledge (target is red) to restrict your search to these items
Model uses the idea of a salience map
Visual search in eye movements
Strong relationship between number of saccades and time taken to search
In a feature search, eyes quickly go to target
In a conjunction search, eyes flit all over
Number of fixations made is predictive of how long it takes for someone to find a target - more fixations = longer
Movement of attention - Superior colliculus
Has a direct connection from the LGN
Important in control of eye movements and attention
Has direct connection to extastriate cortex
Has multisensory input
Movement of attention - Muller et al (2005)
Stimulation of the superior colliculus produces eye movements
Pre-cuing paradigm
Cued the location by doing a tiny pulse that would not trigger eye movements
Targets towards the cued location were detected better than any other area
Shows close link between attention and eye movements
Right parietal lesion patient
Can identify shapes
Global perception must be ok
Crosses out only right elements of each object not the space
Visual stimuli themselves pull attention to the right (exogenous)
Visual neglect - visual search
Hold up fingers on both hands and wiggle one hand’s fingers
If they say no fingers wiggle then they have visual neglect
