CH3 Flashcards
function of attention for perception:
attention is necessary because the perceptual system has limited processing capacity so attention selects a few items to receive perceptual processing. It basically prevents overload. Also, the binding problem can be solved with attention by selecting a limited spatial area for processing, so that the features within that area can be combined to reconstruct the objects.
function of attention for awareness:
attention is needed to bring perceptual information to consciousness. Unattended objects are present on the retina, and they may be processed a little, but attention is needed to register the results of this processing
function of attention for action
attention is needed to constrain possible actions; to make sure there’s no behavioural chaos
Binding problem
how is it all combined into one conscious experience?
early-selection view by Broadbent
the human information processing system is
like an information channel with limited capacity. He thought that attention works to
select information at an early level based on attributes like the location or pitch. His
filter theory says that there is a filter that protects the information-processing model
from being overloaded. It assumes that information from the senses first enters a
buffer where it can be held for a short time. Information that fits the filter is then
passed along to the limited capacity channel, where it can be identified. The results
of this analysis are then sent to a response system and may be used to update
expectations about what is likely to occur in the given situation. However, sometimes
potentially relevant information gets past the selective filter, like hearing your name.
This is why Treisman proposed the filter attenuation theory: the selective filter does
not completely block out unwanted information, but only attenuates or reduces the
strength of unattended stimuli.
Treisman proposed the filter attenuation theory:
the selective filter does
not completely block out unwanted information, but only attenuates or reduces the
strength of unattended stimuli.
According to
Treisman, early selection is necessary when perceptual load is high in order to prevent
interference within the various perceptual analyzers.
late-selection view by Deutch and Deutch & Norman:
attention is not needed to
perceptually process and identify items, but is needed to create a more durable
representation of the information. All information is processed to a semantic level;
this can explain why relevant information sometimes is identified.
Shiffrin, Diller and
Cohen argue that
basic visual information is processed in parallel across the visual field, and
attention is applied to a location only when a response must be selected.
Focusing the attentional spotlight:
it can be adjusted to focus on a small or larger
region. This can also be manipulated.
Moving the attentional spotlight:
the time needed to focus attention at a location
increased as the eccentricity of the target increased, so attention probably travels
through space at a constant velocity. However, this has been criticised because it can
also be interpreted differently: in terms of the time needed to identify peripheral
targets. It seems like attention jumps from one position to another such that
resources are allocated to a new location as they are released at the old location.
LaBerge: one letter or whole word. Zoom lens for which
there is inverse relation between illuminated area and the concentration of attentional resources.
The resolution of the attentional spotlight
not everything within the focus of the
attentional spotlight may be attended. However, detection of a probe is faster when
attention is directed at the probe’s location than when it is not, and attention operates
differently on objects than on empty space. We can conclude that attention moves
from item to item, but that the items might sometimes be perceptual groups.
A better metaphor of selective visual attention is
that of a gradient of resources that is
allocated to a region in space rather than a spotlight. The gradient may vary in size, and
resources are assumed to fall off from the center of the gradient to the edges. The gradient
contains not only the current focus of attention, but also the results of previous attentional
allocation. So, activation can build up over time and across more than one attentional
fixation.
Directing attention to an object or region…
increases the efficiency of processing. The frontal
cortex may exercise attentional control by inhibiting areas or objects that should not be
attended. Paying attention to one thing can imply decreasing the weights of unwanted inputs
rather than simply increasing the weights of desired inputs.
Overt orienting
changing the positioning of the senses to improve
perception.
Covert orienting
attention is directed to a location other than the focus of the eyes. Covert orienting does not affect which information is registered by the senses, but it
may affect the output of perceptual processes by directing attention to specific locations or
items. You can study covert orienting with “filtering” tasks, in which participants are shown a
number of stimuli and asked to attend to just one of them.
Exogenous orienting:
reflexive orienting, driven by stimuli, also called peripheral cues. Most
often happen for stimuli that appear suddenly.
Exogenous orienting can be prevented by top-down attentional control.
Endogenous orienting:
controlled orienting,
done by the observer. They need to be processed by the observer and intentionally acted
upon in order to have an effect.
Cuing benefits (or costs)
are computed by comparing the
reaction time and accuracy on valid trials with that of invalid trials. Valid peripheral cues
benefit performance, especially when the interval between the cue and target is only about
100 ms. A reversal of this benefit happens when it’s >300 ms. This is called inhibition of return. Not only the stimulus onset but also the stimulus offsets can attract attention. The
attention-attracting effect of abrupt onsets or offsets may be due to the changes in
luminance that they create.
Effects on the P1 component are seen for both nonpredictive and predictive exogenous
cues.
inhibition of return
Valid peripheral cues
benefit performance, especially when the interval between the cue and target is only about
100 ms. A reversal of this benefit happens when it’s >300 ms, and is called inhibition of return
Is space special?
Space is more important for selection than is form, but it is possible
that selection can be based on nonspatial features. Attending to an object of a
relevant colour does not facilitate the identification of objects with the same colour,
but objects that were near the location of the attended object. Attention seems to
operate on a spatial representation of visual space, so you could argue that space is
special.
According to purely space-based models of selection
attention directly operates on a spatial representation of visual space. Eg, in the
feature integration theory, attention is allocated to regions of space and acts to bind
features in those regions to form objects. So, objects are identified only after
attention has been directed to regions of space. Other models assume that attention
selects objects rather than regions of space. They say that attention selects only
regions of space that are occupied by objects. Space-based theories assume that
attention can select regions of space whether they are occupied by objects or not.
Duncan proposed that
we attend to objects and not to
space, and that when judgements are made about two objects, attention must
be switched from one object to the other, which takes time. Space is not the dominant factor of attention, but attention operates on objects formed on the
basis of earlier perceptual processes.
Moore, Yantis, and Vaughan:
if attention is truly object-based, it should
operate not only on object representations that are explicit but also on
perceptions that are the result of earlier perceptual completion processes,
e.g. Gestalt principles. They had an experiment where two vertical rectangles
were presented that were either occluded by a third horizontal rectangle or
not. They found that selection can occur on object-based representations that
are formed by perceptual completion processes.
Evidence suggests that attention operates on…
objects, even when these objects are only virtually present. However, spatial location and distance also play important roles in selection. Maybe attention is either space-based or object-based depending on task demands.
