selective attention

Question 1

what is the cocktail party problem?

what sort of listening is this similar to in a lab?

Answer 1

trying to divide attention between voices

dichotic listening - play 2 different streams of spoken words in earphones in each ear

Question 2

what did Cherry find?

Answer 2

we can’t understand/remember the contents of 2 concurrent spoken messages

best we can do is alternate between attending selectively to the speakers

Question 3

when is shadowing messages successful and not successful?

and what is it?

Answer 3

• repeating aloud one of the messages whilst ignoring the other
  successful: if messages differ in physical properties e.g voice

not successful: if differ only in semantic or lexical content

Question 4

what 2 changes do people notice/not notice in unattended messages?

Answer 4

1. physical changes - e.g location and voice

2. NOT semantic changes e.g if changed from meaningful to meaningless

Question 5

who focused on shadowing and unattended messages?

Answer 5

Treisman and Moray

Question 6

why don’t we process meaning if unattended words?

Answer 6

unattended words are filtered out early on basis of physical attributes before we access meaning

so aware of physical changes but don’t think to process meaning

Question 7

how to extract meaning from unattended messages?

describe this process?

Answer 7

have to switch attention filter between the unattended and attended messages

slow and effortful process

Question 8

what is Broadbent’ filter model of human information processing?

Answer 8

selective filter (all-or-none and can’t be bypassed) directed to only one source at a time and is done early

• after sensory analysis but before semantic analysis and conscious memory

only info that passes through filter is recognised etc.

Question 9

why may filtering not be all-or-none and what does this mean?

Answer 9

all or none - info either gets through the filter or not at all

evidence that unattended speech could activate meanings - partial breakthrough:

1) own name in unattended speech
2) interpretation of lexically ambiguous word in attended influenced by meaning of words in unattended

3) conditioned word with electric shock
skin response evoked by the word in unattended message even though not ‘noticed’ that word was there
also happened for similiar words so meaning activated

Question 10

describe late selection theories?

Answer 10

both attended and unattended words processed up to and including identification and meaning activation

relevant meanings picked out if salient or currently relevant

Question 11

what don’t late selection theories explain?

Answer 11

skin response to unattended probe word weaker than to attended (should be the same)

selection on the basis of sensory attributes more efficient than selection on basis of meaning

Question 12

what is Treisman’s filter-attenuation theory?

Answer 12

early filter but later selection

not all-or-none (block) but attenuates info from unattended sources (weakens force - so unattended words can still activate meanings if salient)

early filtering is an optional strategy (not bottleneck - meaning necessary)

Question 13

how did a task show that early selection is optional not a structural bottleneck?

Answer 13

played sequence of digits to right ear
had to be shadowed (spoken aloud) and left ear unshadowed

found more letter detected in shadowed (right ear) than left ear

Question 14

what did Daniel Simon find?

what does this suggest?

Answer 14

that many missed highly salient events in unattended stream when focusing on another stream

events of unattended stream even though in visual field not processed to level of meaning

suggests that there’s an early filter for visual info

Question 15

what is the attentional spotlight of covert attention?

what is meant by covert attention?

Answer 15

covert attention - attending to parts of visual field away from fixation

attentional spotlight-
endogenous cueing (arrow suggesting which side image would appear)
made people respond faster in all tasks
so faster in expected and slower in unexpected side

Question 16

describe exogenous cues?

compared to endogenous cues?

Answer 16

reaction time faster after change in stimulus when box flashed (exogenous)

exogenous attraction of spotlight is faster than endogenous movement

Question 17

what did Mangun find?

Answer 17

P1 component larger in activity when stimulus appeared in attended location than unattened

filtering happening very early

Question 18

findings suggesting endogenous selection occurs very early in processing?

Answer 18

fMRI BOLD response reacted to a greater degree when attention directed to that side (endogenous selection)

Question 19

how does the efficiency of early selection depend on processing load?

evidence to support this?

what is this theory called?

Answer 19

perceptual load theory

selection good when input takes up all processing capacity as no space for extraneous distractors

when proessing load low, spare capacity to process extraneous distractors and influence later dealt with by selection process

Question 20

what are the 3 functions of selective attention?

Answer 20

1. defensive filtering - protecting high level limited capacity systems from overload
2. positive selection-for-action - prioritising one object (most beneficial) to aid intended actions or further processing
3. Treisman’s feature-integration theory - integrates features in visual field to form object representations (bottleneck between sensory processes and representation)

Question 21

what is the difference between what early and later researchers used to measure how early selective attention takes place?

Answer 21

early - listening tasks

later - visual reaction time tasks

Question 22

describe how we attend to something using vision?

Answer 22

look at object which we are attending to bring retinal image onto the fovea

Question 23

evidence in support of perceptual load theory?

Answer 23

if processing load low, then incongruent distractor slows response relative to congruent (if one letter different to the other as opposed to the same)

if processing load increased then distractor didn’t slow down response if incongruent as distractor not even processed

Question 24

what does Treisman’s feature-integration theory predict?

Answer 24

1. unattended features won’t combine properly

2. unattended objects won’t result in stable object representations