Spotlight of attention Flashcards
How does the organisation of vision contradict the idea of a limited capacity?
The brain, particularly the visual system is organised into parallel processing streams/ pathways with simultaneous processing.
Therefore cannot localise the processing limitation in attention capacity.
May be a problem of integrating the information?
What is the binding problem?
how different features, that are processed across functionally specialised brain regions, are “bound” together to produce coherent representations of whole objects.
What is FIT?
Feature integration theory
Anne Treisman
proposed that attention is the brain’s solution to the binding problem
different attributes of a visual scene are processed in functionally different specialised areas (feature maps)
attention acts as a spatial spotlight that binds together features of the attended object to create an object file
without attention only basic features are processed
there no fundamental limit to the amount of information we can process rather we need attention to coordinate across different functionally specialised systems or brain regions
What is change blindness
Rensink et al. (1997)
despite the subjective feeling that we process a lot of information from a scene, demonstrated that this is actually limited
Flashing image - each time there is a reasonably substantial change but this is only noticed a few seconds later when told to look for it.
Only at that point do we bind the relative features together to be able to compare from one flash to the next and realise that’s where the change is.
Consistent with Treisman’s theory of Feature Integration
Treisman and Gelade’s (1980) visual search experiments
Task 1: feature search
Is there a red ‘O’?
Pop out search - target is defined by a unique feature that’s not shared by any of the distractors
only need to look in colour maps and as there is only one red thing, immediately go to this location.
Search time is consistently very quick irrespective of set size.
If not present RT increases with set size as need to scan more locations.
Task 2: conjunction search
Slow serial search
no single unique feature defining the target (e.g red Xs and blue Os)
unable to look for single feature map
must search through different locations, attention causes feature to be combined can then compare to target.
strong relationship such that subjects are slower, the more items in the display.
Slope on ‘no’ trials is double that of the slope on ‘yes’ trials
As have to search through all the items as opposed to on average about half
Only conjunction needs binding and therefore its only conjunction that needs spatial attention. Attention must be spatial if features need identifying.
For spatial: Brefczynski & DeYoe (1998)
evidence suggest that attention acts spatially, like an internal ‘spotlight’ that highlights particular areas of the visual world
used fMRI to show that attention activates retinotopically mapped areas of visual cortex.
showed people stimuli in different locations in the visual field
attending to different parts of the visual field elicited activity that matched the retinotopic mapping corresponding to the different areas in the retina
manipulated so that visual stimulus was fixed and participants need to attend to different locations while keeping gaze fixed
activity in cortex corresponds to attended area as if actually occurring in those areas
Evidence for attention being spatial from neglect
Patients ignore the LHS of the visual world thus attention has important spatial properties
McLeod et al. (1988)
strong evidence that attention is directed towards coherent objects rather than just locations in space
shown that target objects defined as conjunctions of features can be found very quickly in visual search experiments
Participants were asked to find as quickly as possible a conjunction of form and motion
E,g moving x
Found that participants could find these conjunctions very quickly, independent of set size.
Suggests participants attention can be drawn to a whole meaningful object, a conjunction of features
Movement is inherently linked with space so can still ignore this and result in pop out search.
Chechlacz et al. (2010)
Visual neglect can show ‘object-based’ effects.
studied patients who ignored the left-hand side of objects regardless of whether the object itself appeared in the left or right visual field.
Apples task
Piece of paper with apples
Patients asked to draw a line through all of the apples that haven’t had a bite taken out
Only put line through those on the right hand side of the page
But some patients will put a line through apples/ objects regardless of where they are on the page but will put a line through some apples they shouldn’t where the bit has been taken out on the left hand side
As if ignoring the left hand side of the object regardless of where it is in space
Simultaneously measures the spatial and object neglect that is ignoring complete apple on the left hand side versus the left hand side of objects.
Can show a mix of these depending on where the damage is in the parietal lobe.
O’Craven et al. (1999)
showed that when subjects pay attention to one feature of an object (e.g., its motion), they automatically seem to pay attention to other features of that object (e.g., whether it is a face or a house), as reflected in fMRI activation of areas of visual cortex that are specialized for those features.
If attention is object based the idea is that if i’m paying attention to an object, i will process all of the different features of that object
Showed subjects stimuli where there is a photo of a face and house overlaid, one of which was moving.
Motion (V5)
Faces (FFA)
Places (parahippocampal place area PPA)
Direct subjects attention to different attributes for example movement.
Look at activity in other regions to see the degree to which attention is inherently paid to other aspects of the moving object.
E.g paying attention to movement and its the face that moving; therefore should expect to see increased activity in the face area despite all of the information occurring within the spatial spotlight of attention
Results were consistent with prediction
When you process one part of an object you inherently process all of the features of that object in terms of this object based attention
Biased competition theory
Desimone and Duncan’s (1995)
objects compete for limited processing capacity and the control of behaviour. Attention operates to bias this competition by selecting the most relevant object(s). This selection can occur based on any of the object’s features, including—but not limited to—its location. Selection of an object boosts its processing (at the expense of competing objects), leading to greater activity in neural representations of all features of that object.
