lecture 2 - attention and distraction Flashcards
the automatic and deliberate control of behaviour
- Not everything we see or hear enters ‘consciousness’ ie looking at something isn’t the same as ‘noticing’ it
Can we decide what to pay attention to? Ie select one ‘message’ and suppress others
the stroop task - J Ridley Stroop 1935
- Congruent list - where words of colour are that colour eg red is red
- Control - xs of a colour
Incongruent - words of colour are not that colour eg word blue is coloured green
- Control - xs of a colour
the task has many variations
results In notes - ptps have a slower reaction time on the incongruent condition and the fasted reaction Time on the congruent condition. we are slower in incongruent as have to suppress the word meaning
selective attention
- Process that controls our awareness of particular categories of events in our environment
- Allows us to select one ‘message’ and suppress others
Why do we need selective attention?
- Allows us to select one ‘message’ and suppress others
Broadbent 1958
brain has limited capacity for conscious processing - can’t process everything that enters the brain - we would be overwhelmed. limited capacity and only hold certain amount of info in consciousness at a time.
both examples of selective attention
- We appear not to process the background details while focusing on the card trick
- But
We don’t seem to be able to ignore the words when we’re trying to focus on the colour (the ‘stroop effect”
- But
The stroop effect
An experimental paradigm that allows us to study selective attention mechanisms
Stroop incongruent condition creates a situation where two attention processes are in conflict with each other (and hence allow us to measure their effect on each other)
Attention can be controlled via automatic versus controlled processing
- Controlled (deliberate) processing
Involves ‘mental effort’
Limited (Broadbent, 1958)
Subject to distraction- Automatic (obligatory) processing
Happens independently of ‘effort’
Causes distraction when incongruent with the focal task (and facilitation when congruent)
- Automatic (obligatory) processing
psychology of attention
- We can study attention processes in tasks where they are in conflict with each other (Stroop incongruency c/w congruency or neutral trails)
And in tasks where they help each other (Stroop congruency)
Automatic processing for interacting with the environment
Task = is the object the right way up?
Yes/no response on keyboard with either right or left hand.
object action compatibility effect, in response latency, on the hand of response when classifying objects (such as in the cup example) that are orientated so as to have optimal hand of grasp (right hand in cup example). if your right handed your more likely to place a cup don with handle on right and vice versa so would look wrong to you the other way. usually hold with index finger.
picture in notes
Why do we need attention?
- Enhanced processing of info falling within the ‘focus’ of attention (poster et al 1980)
Focus of attention on context/ goal relevant information - So that we don’t waste limited resources processing irrelevant information
- We can ‘focus’ our attention to enhance processing of relevant information (attention as a ‘spotlight’)
If attention enhances processing of the thing we’re attending to, what happens to the things we’re not attending to?
- We can ‘focus’ our attention to enhance processing of relevant information (attention as a ‘spotlight’)
‘Posner’ cueing paradigm
look at centre of screen then an arrow then centre plus. have to press left or right button on where target stimulus (digit) is on the screen.
- Congruent ‘endogenous’ cueing
Interval between -> and appearance and disappearance of 5 too short to allow for eye movement. cues attention to the right so faster to detect 5 on right side. digit is in same direction as arrows.
(‘Attending” not the same as ‘looking at”) (endogenous = voluntary attention direction)- Incongruent endogenous cueing
- Congruent c/w incongruent ‘endogenous’ cueing ->3 <- 3
- Response times (RTs) congruent < RTs incongruent
flash of light and then target appears. congruent target in same area as light. incongruent target is other side to flash of light. external cues.
- Congruent ‘exogenous’ cueing - exogenous = reflective automatic attention. Attention is drawn automatically towards stiumulus by external stimulus eg flashing light, noise etc. outside affects attention. - Incongruent ‘exogeneous’ cueing - Congruent c/w incongruent ‘exogenous’ cueing - +3 +3 Response times (RTs) congruent < RTs incongruent
people are faster in cued condition compared to non cued. shift attention to target location faster as faster processing there.
brief recap
- Attention can be controlled via automatic and deliberate controlled processes
- We can study these processes in task conditions where they are in conflict (Stroop incongruency) or help each other (object-action compatibility)
Both automatic and controlled processes can be biased by cues (Posner paradigm: exogenous/endogenous cueing effects)
models of selective attention
- Early selection models: Items that are not attended to will not get selected for perceptual processing
Late selection models: All information is attended to and gets selected lateron in processing chain
‘early’ selection - Broadbent 1958
diagram in notes
limitations on processing
bottleneck is between selective filter and perceptual system. there is a limited capacity. selective filter is third step.
linear model but may not happen in parallel in brain.
- Their primary feature is that if items are not attended to they are not selected for perceptual analysis and so play no further part in information processing.
- However late selection models argue that all info is attended to and is only selected later in the info processing chain, after perceptual analysis of the stimuli.
- Most of the influential ESMs were developed in the 1950s and 1960s when organisational psychology was making large inroads into workforce behaviour. Psychologists such as Donald Broadbent were interested in how psychological principles could be applied to understanding real-life problems, such as operating air traffic control systems or navigating a plane, both of which require extraordinary attention and selective attention.
- Broadbent’s (1958) model of attention proposed a filter theory of attention which, suggested that processing information was like the operation of a filtering system, at the time it was popular as it was testable and falsifiable however evidence has shown that features of the model were incorrect.
- Filter theory = a channel of communication would process information and transmit this information to other cognitive systems for analysis.
- Broadbent suggested that the filter system initially processes information from a ‘sensory store’ and transfers it to other cognitive systems. This was an all-or-nothing model where only selected material would pass through the filter system. This selected material would then make its way to a limited capacity P(erceptual) system which would identify the material.
- The all-or-nothing feature of the model can explain why material presented to the unattended ear in dichotic listening experiments is not processed.
- Moray’s (1959) experiments suggested that the basic feature of Broadbent’s model was wrong.
- Moray found that when participants were instructed to switch attention from one ear to another during the experiment, they were able to do this when the instruction was along the lines of, ‘Robyn, switch ears.’ According to the model, this channel should have been blocked and remained unattended to: the participant should have been attending exclusively to another channel.
Another set of experiments also demonstrated that listeners could follow messages that were switched from one ear to the other. For example, a narrative would begin in one ear and be switched to the previously unattended ear (Treisman, 1960, 1964). Participants, contrary to the filter model, would switch attention to the unattended ear to follow the narrative.
‘late’ selection (deutsch and deutsch 1963)
limitations on responding.
attended and unattended info.
bottleneck is between selective filter and response as selective filter is second to last step.
dichotic listening meaning
The first experiments to investigate the nature of attention scientifically took advantage of the fact that we have two ears.
- Selective attention is not achieved by simply closing a sensory channel. Some information can break through into consciousness due to its nature eg Moray (1959) found that people can hear and remember their name presented to the unattended ear or Nielsen and Sarason (1981) found that people tend to immediately notice sexually explicit words presented to the unattended ear.
- The fact that some types of information presented to the unattended ear can grab our attention indicates that even unattended information undergoes some verbal analysis.
- If the unattended information is ‘filtered out’ at some level, this filtration must not occur until after the sounds are identified as words.
- Several studies have shown that information presented to the unattended ear can affect our behaviour even if we do not become conscious of the information. the information can produce implicit memories, of which we are unaware (Cleermans, 1993)
Von Wright et al (1975) showed that words previously presented along with an unpleasant electrical shock would produce an emotional reaction when the words were presented to the unattended ear. Even when the participant was not consciously attending to the voice, the information produced a non-verbal response – a classically conditioned emotional reaction. Thus, the unattended information could trigger the recall of an implicit memory.
dichotic listening (shadowing) experiment
devised by cherry 1953 - used to settle debate between early and late selection. you ignore inputs into right ear = unattended info. you attend the information into your left ear and say it aloud.
what happens to the unattended info and how would we know?
listeners can tell if unattended message is voice or noise, male or female ie relatively superficial, physical aspects of input. but can’t report anything that was said - even single words repeated many times, or even if it was in a foreign language.
McKay (1973) showed that information presented to the unattended ear can influence verbal processing even when the listener is not conscious of this information.
in a later recall test - asked did you hear the sentence “ they threw stones at the side of the river?’ or “they threw stones at the building society?”.
they answer the river which means the meaning of non-attended message was processed as
they received the ignored inputs of ‘river, flood, swim’ and attended input ‘they threw stones towards the bank’. so their answer is congruent with meaning of ambiguous word
however the answer you get depends on how you measure your dependent variable
attenuation model - treisman 1964
the model is a compromise - diagram in notes. weakened processing of unattended information.
bottleneck is between attenuating filter and selective filter.
unattended information is weakened. physical features are filtered out in attenuating filter. in perceptual system a dictionary unit - meaning extraction - processing of unattended input depends on threshold.
attention system more flexible and dynamic
- Treisman (1960) showed that people can follow a message that is being shadowed even if it switches from one ear to the other. Suppose a person is shadowing a message presented to the left ear, while the message to the right ear is unshadowed. Apparently, the switch occurs when the message begins to make no sense.
However, by the time the person realises that ‘crept out of flowers’ makes no sense, the rest of the message, ‘the swamp’, has already been presented to the right ear. Because the person is able to continue the message without missing any words, he or she must be able to retrieve some words from memory. Thus, even though an unshadowed message cannot be remembered later, it produces some trace that can be retrieved if attention is directed to it soon after the words are presented.
How do we know that unattended information is processed?
- Dichotic listening studies have shown that un-
attended channel can influence responses
(auditory modality)
Priming studies have shown that unattended information can influence responses (visual modality)
priming
- Exposure to one stimulus influences the processing of a subsequent target stimulus without conscious guidance or intention
- Positive priming: Enhances processing of target (faster naming times, less errors)
Negative priming: Inhibits processing of target (slower naming times, more errors)
negative priming
task in notes
name black word (target) and ignore grey word (unattended distractor)
there is also a separate study of target facilitation and distractor suppression processes - task is to read red letter and ignore green letter. distractor inhibition - baseline - target facilitation. measure response times/ errors
Metzler and Parkin 2000 - frontal lobe patients show positive instead of negative priming - deficits in distractor suppression not target facilitation.
Negative priming
* Unattended information influences subsequent responding i.e. must have been processed to a certain degree
* Treisman (1964): Attenuation model:
Unattended material is processed in a weakened, attenuated form
adaptive significance and the control of behaviour
the ‘cocktail party effect’ - cherry 1958
sometimes people hear their names being mentioned .
we want a system that responds to significant events even when were not ‘paying attention’ to them
- Selective attention to auditory messages has practical significance beyond the laboratory For example, sometimes we have to sort out one message from several others without the benefit of such a distinct cue; we seldom hear one voice in one ear and another voice in the other. We might be trying to converse with one person while we are in a room with several other people who are carrying on their own conversations. We can usually sort out one voice from another eg the cocktail-party phenomenon.
- In this case, we are trying to listen to the person opposite us and to ignore the cross-conversation of the people to our left and right. Our ears receive a jumble of sounds, but we are able to pick out the ones we want, stringing them together into a meaningful message and ignoring the rest. This task takes some effort; following one person’s conversation in such circumstances is more difficult when what they are saying is not very interesting. If we overhear a few words of another conversation that seems more interesting, it is hard to strain out the cross-conversation.
- The cocktail-party effect study was reported in 1959 by Moray.
- The laboratory finding seemed to mirror that of hearing one’s name mentioned in someone else’s conversation at a party even though you were not attending to that conversation
- Only 33 per cent of participants responded to the cocktail-party effect in Moray’s study.
- Is there a psychological difference between those who attend and those who do not and would such a difference reflect different means of processing auditory or cognitive information?
- Working memory could be the difference between those who attend and those who do not. Conway et al hypothesized that the least capable participants identified their names because they failed to demonstrate the working memory facility which would allow them to attend to the channel they were meant to be attending to and ignore the channel they were not meant to attend to.
Participants in the experiment were asked to complete a selective listening exercise similar to that of Moray. Participants listened to messages through headphones but were told only to attend to one channel; the participant’s name would occur in the unattended channel. Participants also completed a working memory exercise that involved reading a simple mathematical equation followed by a word (e.g. ‘Is (6 + 4) / 2 = 5? DOG’), solving the equations and, at the end of all trials, writing down as many of the presented words as they could remember. The number of equation and word displays in each trial varied between two and six. Those who excelled at this task were significantly better at ignoring their name than were those who performed less well.
recap what happens to unattended information
- Early versus Late Selection Models
- Dichotic listening and negative priming experiments to study processing of unattended information
- Attenuation model as compromise
Selective attention as a balancing act
* No all-or-nothing system; not as simple as early versus late attention -> Treisman’s Attentuation model
* We need to focus on the context-relevant information to enhance processing
* But we also need to be flexible, if new, significant (dangerous) information comes along
* Distraction may be a ‘good’ thing in some types of setting
e.g., does the Stroop Effect reveal a flaw or a strength in our processing system?
key issues
- Attention not the same as seeing, hearing, etc.
- Automatic and controlled processing
- We can create laboratory tasks where these processes oppose or help each other to study them
- We’re limited
- ‘Focussing’ attention enhances processing, but it’s not impervious to distraction
- What happens to unattended information (and, importantly, how would we know?)
- Where do the limitations on processing arise?
- i.e., at what point does selection take place?
What is attention for (what does it enable/prevent)?
- i.e., at what point does selection take place?