Cognition: Chapter 5 - Attention and Performance

What is the definition of attention according to William James (1890)?

Attention is taking possession of the mind, focusing on one thing vividly while ignoring others.

What is the purpose of attention?

Prevents sensory overload, aids survival (spot danger, find food), and supports everyday tasks (finding keys, holding conversations).

What is active (top-down) attention?

Goal-directed attention guided by expectations (e.g., searching for a red car in a car park).

What is passive (bottom-up) attention?

Attention triggered by external stimuli (e.g., turning at a sudden bang).

What is focused attention?

Concentrating on one stream of information (e.g., listening to one friend in a noisy bar).

What is divided attention?

Attending to two or more streams simultaneously (e.g., walking while texting).

What is external attention?

Attention directed at sensory input from the environment.

What is internal attention?

Attention directed at mental processes like planning, memory, or imagination.

What is the cocktail party problem (Cherry, 1953)?

The difficulty of following one voice among many in a noisy environment.

What does dichotic listening show?

People can shadow one ear but recall very little from the unattended ear; detect pitch/tone changes but not language changes → strong attentional filtering.

What did Broadbent (1958) propose?

Early selection model: filter info by physical features; only attended input reaches meaning/STM. Strength = explains poor recall; Weakness = can’t explain hearing your own name.

What did Treisman (1964) propose?

Attenuation model: unattended info not fully blocked, just weakened; important info (like your name) can still break through.

What did Deutsch & Deutsch (1963) propose?

Late selection model: all input processed for meaning, bottleneck only at response. Weakness = unrealistic capacity demands.

What do ERP studies show about unattended input?

Unattended info gets less processing but is not fully ignored; own-name effect shows semantic-level processing.

What did Mesgarani & Chang (2012) find?

Auditory cortex encodes attended speech more strongly than unattended.

What did Olguin et al. (2018) find?

Both attended and unattended streams are processed, but comprehension is better for attended.

What is Posner’s (1980) cueing paradigm?

Task with valid, invalid, or neutral cues. Valid cues = fastest reaction, neutral = medium, invalid = slowest. Shows covert attention shifts without eye movement.

What are Corbetta & Shulman’s (2002) two networks?

Dorsal network = goal-directed/top-down; Ventral network = stimulus-driven/bottom-up “circuit breaker.”

What evidence supports Corbetta & Shulman’s model?

Brain scans: dorsal for expectation, ventral for unexpected detection. TMS/disruption impairs respective functions. Lesions = poor goal-driven or poor detection.

What other attention networks exist?

Cingulo-opercular (sustained alertness); Default mode network (mind-wandering); Fronto-parietal (task switching/flexible control).

What is neglect?

A disorder, usually after right-hemisphere stroke, where patients ignore the left side of space or objects. Egocentric = left side of whole scene; Object-centred = left side of each object.

What is extinction in attention disorders?

Failure to notice a stimulus on the neglected side only when something is also present on the non-neglected side.

What is pseudoneglect?

Healthy people show a slight leftward bias.

What is the neural basis of neglect?

Right temporal, parietal, frontal cortex + insula implicated.

What happens in unconscious processing in neglect?

Patients may deny left-side stimuli consciously, but the brain still processes them; amygdala responds to emotional stimuli they “don’t see.”

What is the airport example in visual search?

Ultra-rare items (e.g., guns in luggage) often missed; training and feedback help improve detection.

What is Treisman & Gelade’s (1980) Feature Integration Theory?

Two-stage search model: parallel stage (automatic detection of simple features), serial stage (attention combines features). Errors = illusory conjunctions.

What is Wolfe’s (2011) dual-path model of search?

Selective pathway = bottleneck, detailed analysis; Non-selective = gist of scene. Real search uses both.

What is the texture tiling model?

Peripheral vision encodes more than assumed; explains crowding and expert performance in real-world vision tasks.

What is cross-modal attention?

Attention is not isolated to one sense; senses interact.

What is the ventriloquism effect?

Vision biases perceived sound location (dummy’s mouth “talking”).

What is temporal ventriloquism?

Sounds shift perceived timing of visual events.

What are real-world applications of multisensory attention?

Driving signals (lights + sounds) reduce braking times; vibrating car seats improve reaction; multisensory warnings save lives.

What is covert attention?

Attention shift without eye movement.

What is egocentric neglect?

Missing the left side of a whole scene.

What is object-centred neglect?

Missing the left side of each object.

What is an illusory conjunction?

Incorrectly combining features from different objects.

What is the ventriloquism effect (in perception)?

Visual cues bias perceived sound location.

What did Cherry (1953) show with dichotic listening?

Little processing of unattended ear, except for gross physical changes.

How do Broadbent, Treisman, and Deutsch & Deutsch differ?

They differ in bottleneck location: early (Broadbent), attenuated (Treisman), late (Deutsch & Deutsch).

What did Mesgarani & Chang (2012) show with electrodes?

Auditory cortex encodes attended speech more strongly than unattended.

What did Posner (1980) prove with cueing experiments?

Covert attention speeds responses to valid cues and slows responses to invalid ones.

What did Corbetta & Shulman (2002) propose?

Two networks: dorsal (goal-directed) and ventral (stimulus-driven).

What are illusory conjunctions (Treisman & Gelade, 1980)?

Errors where features from different objects are incorrectly combined.

What did Wolfe (2011) propose about visual search?

Two pathways: selective (detail) and non-selective (gist).