Lecture 7 - Selective Attention

Define cognitive psychology?

Study of mental operations that mediate our experience and behaviour

Describe the cognitive model of processing information?

- Stimulus

- Information processing:

Perceptual processes = perception + attention

Storage + manipulation processes = memory, learning & experience, thinking & reasoning

Response related processes = preparation of motor responses e.g. language, speech, writing, grasping

- Behaviour

What did William James (1890) say that attention is?

"The taking into possession of the mind, in clear and vivid form, of one out of what seem several simultaneously possible objects or trains of thought"

Define attention?

Cognitive process = selects a small amount of incoming information from our environment to be processed, while excluding other concurrently occurring information

Why are attentional processes used?

Too much information coming into the senses = used to limit amount of information

Define selective (focused) attention?

Form of attention = kind of thing we see when we are presented with two or more stimuli = attend to/respond to only one of them

Define divided attention?

Think of this as "multi-tasking" when we are presented with multiple stimuli + attend to/respond to all of them

Define active/'top-down' attention?

Purposeful, deployment of attention e.g. paying attention in lectures, revision = focus on particular task/event = actively engage + complete task

Define passive/'bottom-down' attention?

Attention is deployed because of external stimuli e.g. loud noise, flashing light, sudden movement = switch to attend to this

Define sensory buffer/memory?

Holds information coming from the senses

What happens in the sensory buffer after information comes in (Sperling, 1960)?

Sensory buffer has a large capacity but decays fast = in 300 msecs most of the information is lost

What does the sensory buffer block?

Blocks incoming information long enough for attention to select the important/relevant information

What can be seen in patients with right parietal lobe damage?

Unilateral neglect = inability to perceive half of their visual field/other info coming in

What does the selectivity of attention say what we perceive depends on?

What we attend to = can change if we have micro switch in attention

Describe the experiment of selective auditory attention?

Words in right ear + different words in left ear

3 examples of low level information?

- Location

- Pitch

- Gender

3 examples of higher level information?

- Language

- Meaning

What task was used to study selective auditory attention + describe it?

Dichotic listening task = participants presented with an auditory message in each ear with headphones

- Participants hears the same voice at the same intensity (pitch) to both ears but with different messages

- Participant asked to 'shadow' (i.e., repeat) one of the messages

- Question is 'What do the participants notice about the message in the 'unshadowed' ear'?

Why is shadowing used?

Ensures participants conform to instruction and repeat/shadow the message from one ear + ignore the message in the other ear

Describe Cherry (1953) experiment?

- Used dichotic listening task to examine how much information is retained from the unattended ear

- Participants shadowed a message from one ear = attended (shadowed) message

- Different message was played in the other (unattended) ear

- Asked questions about the messages played both in the attended and in the unattended ear

Describe Cherry (1953) results?

- Participants recalled information in shadowed ear very well

- Information in the unattended message was very poorly remembered = only basic physical characteristics were encoded + remembered

What did participants in Cherry (1953) experiment notice in the unattended message?

- Switch from voice to tone

- Switch from male to female

What did participants in Cherry (1953) experiment miss in the unattended message?

- Switch to new language

- Forwards vs backwards speech

What did Moray (1959) find?

Presented in the unattended ear the same word 35 times = still participants failed to notice it

What did Deutch (1986) find?

Even something as low-level as shadowing a melody in the attended ear = participants struggled to have any knowledge of a melody played in the unattended ear

Why can we focus on conversation with one person in a cocktail party?

Can follow one conversation in a cocktail party situation as long as we pay attention to the person speaking = attention to our conversant allows us to block all other conversations out and have little knowledge of them = attention acts like a 'filter' or 'bottleneck'

Who was Donald Broadbent (1926-1993) + what did he do?

- 1st to use computer analogies to make a considerable contribution to the study of human cognition

- Realised importance of psychological factors + especially attention, during flying, whilst working in the American Air Force during and after WWII

- Research into attention and memory started shortly after the war and has influenced scientists ever since

Describe Broadbent (1958) method?

- Condition 1 = report the numbers in the order they were heard - sequentially (84, 59, 26)

- Condition 2 = report the numbers by ear (852, 496)

Describe Broadbent (1958) results?

- Condition 1 = when participants were asked to report the numbers sequentially, they were very poor (20% correct)

- Condition 2 = asked to report by ear they were very accurate overall (65% correct)

Describe Broadbent's Filter Theory: Sensory Register?

Incoming information from multiple channels e.g. 2 ears = initially processed in parallel e.g. pre-attentive analysis + placed in sensory buffer

Describe Broadbent's Filter Theory: Selective Filter?

- Information in one of those channels is selected to be allowed to pass through a filter based on its physical characteristics e.g. location (right/left ear as per instruction)

- While the selected information is passed on for further processing = message in the unattended channel is blocked by the filter + remains in the sensory buffer, during which time it is decaying

What does Broadbent's model assume?

Only simplest forms of analysis occur prior to the bottleneck = early-selection account

Why was the ear-by-ear recall better in Broadbent's experiment?

- Participants filtered/attended to the information based on ear = low-level characteristic processing

- When they were told to repeat the letters in the order they were presented = accuracy dropped = supports Broadbent's filter theory because switching from ear to ear is difficult and goes against the filter

Broadbent's experiment: What % of participants noticed 'you may stop now'?

6%

Broadbent's experiment: What % of participants noticed ' [name] you may stop now'?

33%

Broadbent's experiment: What % of participants noticed ' [name]'?

80%

Problem with the theory of Broadbent's experiment?

Even though the end of the meaningful message continued in the unattended ear = most participants ignored the instruction and switched the 'attended ear' = how did they know the message was continuing in the unattended ear unless they were processing at a semantic level?

What were Underwood's (1974) findings?

Naïve participants only detected 8% of the digits presented to the unshadowed ear, but an experienced shadower was able to detect 67%

What doesn't Broadbent's theory allow?

Unattended information to 'leak through'

What did Moray (1959) and Loftus (1974) show?

Some participants noticed their own name on the unattended ear

What did Treisman (1960) show?

Participants switch ears to follow the story from the attended to the unattended ear

What else became apparent in Broadbent's theory?

Access to information presented in the unattended ear was analysed beyond the purely physical (low-level)

What is Treisman's (1964) Attenuation Theory?

Early-selection account = filter is more flexible than Broadbent suggested

How are all stimuli processed in Treisman's (1964) Attenuation Theory?

All stimuli e.g. words, are processed at the low-level (pitch, gender etc) = then if there is enough capacity in the processing system, the stimuli are processed at the level of meaning, grammaticality

Describe Gray & Weddeburn (1960) support to Treisman's (1964) Attenuation Theory?

- Participants shadow the left ear

- Participants should report "Dear 7 Jane"

- Participants actually reported "Dear Aunt Jane'

- Suggests that participants were aware of some information from the unattended ear

How does Treisman's theory account for some of the problematic findings in Broadbent's theory?

When we hear our own name (Moray, 1959) = in Treisman's model we'll process that to a certain level = more likely to be perceived

4 similarities between Treisman's and Broadbent's theories?

- Attention acts as a filter

- This filter prevents the unattended input interfering with the attended input channel

- Which information is selected depends on physical characteristics

- Treisman's Attenuation theory = early selection account

2 differences between Treisman's and Broadbent's theories?

- Unattended channel is attenuated

- Top-down processes are important = sometimes listeners report a word from the unattended ear = 'breakthroughs' occur when the word is expected in the context of the attended message

What does Deutsch & Deutsch (1963): Response Selection Theory say about filter operation?

Operates after extensive perceptual and semantic analysis

What isn't the role of attention according to Deutsch & Deutsch (1963): Response Selection Theory?

Is not to aid perception e.g. object recognition, but to aid action by selecting certain stimuli to respond to over others

What type of model is the Deutsch & Deutsch (1963): Response Selection Theory?

Late selection model

Summary of Broadbent's theory?

Only some information is selected and that the rest is not processed

Summary of Treisman's theory?

All information processed pre-attentively + those inputs that best match the filter (our goals, pre-dispositions, context) continue on for further processing

Summary of Deutch & Deutch's theory?

Almost all incoming information from the senses are sent on for further processing = with final selection happening only just before a response

What did Corteen & Wood (1972) find to support the Response Selection Theory?

- Presented participants with lists of words = some of which were paired with a mild electric shock

- Then asked participants to shadow a message played on one ear = whilst in the other ear words from the previous lists (see point above) were presented

- During shadowing the attended ear = they recorded participants' galvanic skin responses (GSRs)

- There was a GSR on 38% of the words associated with the shock, despite participants denying having heard the words in the unattended ear

Evidence against late selection?

One such artifact was that there was some attention switching going on = meant that the supposed 'unattended' ear received some attention occasionally

What did Dawson & Schell (1982) study find?

Replicated Cotreen & Wood's (1972) study = increased GSRs in the unattended ear were found only when shadowing performance for the attended ear dropped = suggested that participants were switching ears as opposed to processing the information in parallel (as the Deutch & Deutch late selection model suggested)