<<Back to Cognition Class Home Page
---------------------------------------------------------------------------------------
MODELS OF ATTENTIONSTRUCTURAL MODELS
(Broadbent, Norman, Treisman)
The metaphor of these model is that of a BOTTLENECK. A lot of sensory information reaches our processing system, but at some point part of this information is filtered out and only a small portion is further processes and influences behavior.
What is the type of experimental paradigm used to test these models? The paradigm has to test the process of SELECTION (something will pass through the bottleneck and something else will be filtered out). A very important paradigm in the auditory domain is the dichotic listening task (for example, shadowing).
If we choose this metaphor, what is the relevant question that we want to ask?
One basic question suggested by this model (or metaphor) is WHERE the bottleneck is, that is, at what point of the process the information is blocked and not further processed. There has been an endless (and maybe not so useful) debate on the locus of the bottleneck.
EARLY SELECTION THEORIES
Some people (for example, Broadbent) proposed that the bottleneck was at the beginning of the information processing flow. All the stimuli were analyzed just for the simples physical features. After this stage only the relevant stimuli were further processed, and the other were filtered out. The main point of these models is that there is a CENTRAL PROCESSOR that has limited capacity. To avoid overload, only a limited amount of information can be allowed to the central processor in each moment.
LATE SELECTION THEORIES
Other researchers (for example, Norman), did not think that there was a central processor of limited capacity. They proposed that the real problem was response selection, NOT INFORMATION PROCESSING. According to Norman, information is completely processed up to semantic level and the bottleneck is at the level of response selection.
A second problem is: HOW does the bottleneck work?
For example, Broadbent proposed that the bottleneck worked as a filter. External stimulation was analyzed on the basis of the most elementary features (physical characteristics) and then filtered. Non relevant information (for example, the message from the unattended ear in the dichotic listening paradigm) was filtered out and not further analyzed.
Anne Treisman proposed that the bottleneck was an ATTENUATOR rather then a filter. Nonrelevant information was not filtered out but just attenuated, so that particularly salient information (such as the participant's name) could still pass through the bottleneck and be analyized up to the semantic level.
CAPACITY MODELS
(Kahneman, Navon & Gopher, Johnston & Heinz)
The metaphor in these models is that of RESOURCES. According to this metaphor, our processing system has a limited amount of processing resources. When all the resources are used, further information need to "wait" to be processed until part of the resource become available again.
What is the type of experimental paradigm used to test these models? The paradigm has to test the use of resources when different tasks need to be executed. A very important paradigm is the DUAL TASK paradigm. In this paradigm, the participant is asked to perform two tasks simultaneously and the effect of executing one task is measured as cost in executing the second task.
If we choose this metaphor, what is the relevant question that we want to ask?
One basic question suggested this model (or metaphor) is HOW MANY type of resources we have.
SINGLE CAPACITY MODELS
Early capacity models such as Kahneman's model, proposed that there is a single pool of resources that different information processing need to use. This means that whatever tasks we choose, performing one task will always produce a cost on the performance of a second task.
MULTIPLE CAPACITY MODELS
Several empirical data suggested that performing a task does not always produce a cost in the execution of a second task. It depends on how similar the two tasks are (in particular, how similar are teh input and the output modality of the two tasks). These results led to the proposal of multiple capacity models, such as Navon and Gopher's model or Johnston and Heinz' model.
Do you have some suggestions on how these models can be related to how the brain works?
------------------------------------------------------------------
ATTENTION AND INHIBITION
In the last 15 years several researchers have started to emphasize the importance inhibition mechanisms in cognitive processes. Our brain has a very large number of inhibitory connections which are fundamental for its functioning. It we had only facilitatory connections, any small sensory stimulation would produce a huge wave of activation, similar to an epilectic seizure.
However, inhibitory mechanisms seems to be important not only for physiology, but also for cognition. In the next weeks we will see examples of the importance of inhibitory processes in memory and language. The idea is that inhibition mechanisms are essential for flexibility and efficiency in information processing.
We will talk about two examples:
(1) inhibitory processes in selective attention and the negative priming paradigm
(2) inhibitory processes in executive functions.
SELECTIVE ATTENTION AND INHIBITION
If we don't believe the strongest version of the early selection models (in which the nonrelevant information is filtered out completely and "disappear" from the system), we need to ask: What does it happen to the ignored information after selection?
According to early theories of selective attention, ignored information simply decays (as it happens, for example, to information from the sensory stores if it's not transferred to a more long term memory store). However, some results from the selective attention literature suggest otherwise.
Consider a situation like this:
Which kind of paradigm does this paradigm remind you? Which kind of results do you expect, for example if you compare reaction time to these stimuli with reaction times to a display in which a single letter is presented?
What is the prediction for the "passive decay" models in the second condition?
The results of this type of paradigm however, is that the response to the second stimulus in the second condition is always slower than the response to he second stimulus in the first condition. Can you find an explanation for this result?
One explanation is that when we select the red A in the first display, we also inhibit the blue B. This inhibition will prevent the B from interfere too much with our response to the red A. What does it happen if the inhibition lasts for a certain period? When we are responding to the second display and a B occurs again (this time as the target letter), we may have some troubles responding to it, because of the inhibition that was previosly associated with the B. The increase in reaction times that is observed when a previously ignored stimulus is re-presented as target is called NEGATIVE PRIMING effect.
This inhibition model also predict that inhibition during selection is an important process and that less efficient inhibition mechanisms will be associated with less efficient selection of nonrelevant information. Several results support this proposal. For example, patients with schizophrenia do not show negative priming effect AND show abnormalities of the frontal lobe circuit AND show deficits that may be related to impaired inhibition.
For example, schizophrenic patients are slower in the Stroop task than normals and have problems with other tasks that seem related to frontal lobe functions such as the Wisconsin Card Sorting Task.
COGNITIVE DEVELOPMENT AND INHIBITION
During the period from 6 to 12 months of age, infants acquire the ability to dissociate reaching from the line of sight. Adele Diamond studied this phenomenon by placing a transparent plastic bos containing a toy in front of an infant. The open portion of the box is placed in the front, but the infant views the toy through the box's transparent top. Infants between 6 and 8 months can retrieve the toy only when the box is positioned so that they can look through the open side. If they are looking through the top of the box, they unsuccessfully try to reach the toy from the top. Only by 9 months, infants can retrieve the toy from the side opening while looking at the top of the box.
This happens because in this situation conflictual information is present. Which conflictual information?
The infant cannot solve the conflict between a strong perceptual cue (the toy visible in the box) and the information that it is not possible to reach the toy from the the top of the box.
Difficulty in dealing with conflictual information is also associated with higher cognitive functions. For example, 3-year-old children don't seem to have the capacity for deception, at least when perceptual cues guide them to give the correct version. In a pretty cruel experiment, Russel and colleagues, asked 3-year-old children to lie to another children pointing an empty locations instead of a location where chocolate was hidden. The most interesting result is that children seemed not be able to learn to lie. They continued to point to the correct location even after 20 trials! Because children usually learn very rapidly in absence of conflict, this performance suggests that the problem in this situation lie in the difficulty of overcoming the conflict between what they know to be true and what they have to communicate.
Another task that is difficult for schizophrenic and people h frontal lesions is the Tower of Hanoi task.
TOWER OF HANOI TASK (GROUP)
1) Which kind of operations do you have to perform in order to solve the problem?
2) What do you think may be the role of inhibition in this task?
3) Which kind of errors would you expect in this task in frontal lobe patients?
COGNITIVE FAILURES QUESTIONNAIRE
Now, let's give a look to the cognitive failure questionnaire that I ask you to fill out. In a study, Tipper and Baylis found that the presence of negative priming in a group of normal subjects was correlated with the score in this questionnaire.
How are this cognitive failures associated to inhibition? In particular, in this questionnaire there are a lot of questions associated to memory and to planning. Can you find some example that can be explained with a lack of inhibition? What is the relationship between distractability and inhibition?
---------------------------------------------------------------------------------------