lecture 1- models of attention Flashcards
What is meant by selective attention?
“The ability to focus on that which is important to the task at hand, while ignoring or suppressing task-irrelevant information”
what are the answers to WHEN and HOW we attend to information?
WHEN = early or late selection models (pre or post-processing)
HOW = structure or capacity models (does system have structural limitations or limited capacity)
explain Broadbent’s filter model
- information not attended to is filtered out before any processing such as identification (and therefore NOT processed)
- filter that focuses attention- structural limitation that stops us from attending to info not supposed to
what is the evidence for the early-selection model?
- Cherry 1953: dichotic listening
- what can be reported of unattented speech = BASIC PHYSICAL CHARACTERISTICS (voice/noise, gender, if becomes same in both ears (channels))
- what can’t be reported of unattended = SEMANTIC MEANING (content of message, normal/reversed speech, language/language changes)
- finding: only material (and therefore semantic meaning) in selected channel processed -> early
what did Broadbent and Cherry conclude?
- channel selection takes place (input from one channel processed at a time)
- there is a cost of switching (switching between channels takes time- 300-500ms)
what are the elements of Broadbent’s model?
- sensory store (store for physical characteristics)
- filter (selection of what is to be attended to)
- perceptual/meaning analysis (processing)
- STM (where encoding happens)
what does the early-selection model predict?
- when filter is focused on one channel, there should be no real processing beyond basic physical elements
- dual tasking is actually doing two things one after the other (involves channel switching)
what is evidence against early selection and structural limitations?
- evidence of processing in the unattended channel (still semantic info being encoded)
- 30% (Moray 1959, Wood & Cowan 1995)
- physiological response to words in non-attended channel that were pre-associated with shocks (even synonyms- Corteen & Dunn 1974, Von Wright et al. 1975)
- following coherent message when switched to ‘unattended’ channel (Treisman 1960)
what is Treisman’s modified model?
- ‘attenuated not eradicated’ model
- unattended info not filtered out, but attenuated (turned down) prior to processing
- attenuator instead of filter at same point as early
- signal from unattended is considerably weaker
- BUT some stimuli are ‘louder’/’stronger’ so are processed when unattended
- e.g primed stimuli (name)
explain the late-selection model
- all incoming info processed automatically to highest level
- all stimuli’s physical characteristics are processed, named and categorised in parallel
- filtering happens after meaning processing (at level of response)
- bottleneck is prior to response (still structural limitation)
what is the evidence for late selection?
- strong evidence of processing of unattended messages
- participants use material from unattended channel to resolve ambiguous sentences (Lackner & Garrett 1972)
- stroop (both elements compete for response, causing interference)
why is unattended information remembered?
- filter is leaky (unattended info is identified/processed, then selection happens after)?
OR - the filter is slipping (attention system slips and attends and processes to material it shouldn’t)?
(2 = early selection -> still focusing on things, just wrong things)
what are meant by capacity models?
- no structure bottleneck
- cognitive system has limited amount of processing capacity
- allocates bandwidth to tasks
- limitation is how much resource you have to spend
- any task demands a processing capacity, and performance is dependent on conscious allocation of capacity to task
what does a tasks processing demand depend on?
- difficulty (more difficult = less bandwidth (??))
- individual expertise
what does conscious allocation of capacity to a task depend on?
- enduring dispositions (habits and preferences)
- momentary need
- evaluation of demands on capacity
what is the role of arousal on conscious allocation of capacity?
- too much arousal = reduced focus, more irrelevant details noticed
- too little = no real motivation
- increased levels can counteract capacity limits (determines bandwidth we have to spend)
what is evidence for capacity models?
- dual-tasking
- do 2 tasks separately and then together
- peformance drops (accuracy/RT) when do them together -> dual task decrement
what are two findings about dual-tasking?
- reading prose while writing to diction - 6 weeks of practive led to competence (Spelke, Hirst & Neisser 1976)
- two visual vs two auditory had larger dual task decrement than doing one of each (Treisman & Davies 1973)
explain Lavie’s perceptual load theory
- resolves early/late selection debate
- combines idea of attentional capacity (early) and that attention is allocated automatically (late)
- system is limited, but will process everything until it runs out of capacity
- selective attention determined by perceptual load of task
- low demand (simple stimuli) = more info is processed (increased chance of non-relevant info being processed as attentional capacity spills over from relevant) (LATE)
- high demand (complex stimuli) = only focus on certain aspects as attentional capacity used up (non-relevant filtered out) (EARLY)
describe the setup for the original experiment for Lavie’s perceptual load theory
- single letter OR line of different letters (low vs. high PL)
- say whether z or x is present
- compatible/incompatible/neutral distractor capital letter present in periphery
- incompatible distractors supposed to reduce performance (as Z is also processed as well as x)
- as interference of distractor causes response competition
- incompatible RT longer than compatible RT: COMPATIBILITY EFFECT
explain the results of the original experiment for Lavie’s perceptual load theory
- prediction: increasing PL should reduce compatibility effect (as distractors not processed under high PL, leading to less interference)
- low PL: strong compatibility effect (large RT difference between incompatible and compatible distractor)
- high PL: no sig compatibility effect (similar RTs) -> attention captured by other letters, reduces interference from distractor
what did Cartwright-Finch & Lavie 2006 find about inattentional blindness?
- inattentional blindness: not noticing task-irrelevant elements in the visual scene
- did same experiment as original but low PL = saying colour of line and high PL = saying which line was longer
- high PL had less participants noticing black square in corner
- high PL reduces compatibility effect and increases inattentional blindness
what is the role of WM in load theory?
- WM controls attention (picking up on and responding to the right things)
- there are general cognitive control processes beyond the perceptual system
- even when distractors aren’t processed, they gain control of behaviour
- cog control processes maintains a clear distinction between relevant and irrelevant info
what experiment was done to investigate WM loads role in interference?
- dual-task experiment
1. given set of digits to remember during task
2. task: categorising faces as congruent/incongruent (distractor task while holding numbers in memory) - low WM = 0, 1, 2, 3, 4 seq given
- high WM = different sequence each time
3. after categorising, asked to report seq
what did the experiment investigating role of WM on interference find?
- high WM load increases impact of incongruent pictures/names (longer RT)
- indicates increased interference under high WM load
what are the two systems that determine selection attention and how does overloading each of them affect it?
- cognitive load (active system- controlled processing): overloading increases interference
- perceptual load (passive system- automatic processing): overloading reduces interference
