What Is Attention & Why Do We Need It?

• Definition
  • William James: ability to focus on one object or thought while ignoring others.
  • Modern: process of focusing conscious awareness on specific experiences.
• Flexibility
  • Can be directed to:
    • External objects/locations (visual/auditory stimuli)
    • Internal thoughts (mind-wandering, planning, imagery)
• Why It’s Necessary
  • Environment provides far more information than we can process.
  • Attention selects a subset of relevant information for the current task.
  • Hal Pashler: our awareness is tiny compared to all available stimuli.

General Model of Attention

Input → Sensory Memory Store → Selector Stage → Working Memory

• Sensory Memory Store
  • All sensory input briefly stored.
  • Pre-attentive processing:
    • Rapid, automatic, unconscious
    • Parallel
• Selector Stage
  • Attention chooses which information progresses to working memory.
  • Processing becomes:
    • Conscious
    • Limited capacity
    • Serial/effortful
• Attention as Gatekeeper
  • Acts like a bottleneck:
    • Decides what sensory information reaches awareness.
    • Prevents overload.

Classic Attention Paradigms

Dichotic Listening Task

• Method:
  • Different messages to each ear.
  • Participant shadows (repeats) one message; ignores the other.
• Findings:
  • Good recall for attended channel.
  • Poor awareness of content in unattended channel.
  • BUT participants noticed sensory changes:
    • e.g. male → female voice.
  • Shows:
    • Unattended info still processed at sensory/feature level.

Cocktail Party Effect

• Ability to detect personally relevant info (name, own language, emotional content) in an unattended channel.
• Suggests:
  • Some semantic processing of unattended input.

Theories of Attention

Early Selection — Broadbent’s Filter Model

• Key idea:
  • Attention selects information early, based on physical properties:
    • Pitch, loudness, location.
• Process:
  • All info enters sensory store.
  • Filter selects one channel → working memory.
  • Unselected info = blocked from further processing.
• Supports:
  • Dichotic listening behaviour.
• Problems:
  • Cocktail party effect → meaning can break through.
  • Participants may follow semantic meaning across channels → suggests meaning is processed pre-filter.

Late Selection Theories

• Oppose early-only filtering.
• Deutsch & Deutsch model (extreme late selection):
  • All stimuli fully processed for meaning.
  • Selection happens just before response.
• Evidence:
  • Semantic content in unattended channel can influence behaviour.

Treisman’s Attenuation Theory (Middle Ground)

• Attention = attenuator, not all-or-nothing filter.
• Unattended input is weakened (attenuated), not destroyed.
• Signals passed to a hierarchy of analysers (features → words → meaning).
• Top-down factors (goals, expectations, relevance) influence which attenuated signals still reach awareness.
• Personally important words (e.g. name) can cross threshold even if unattended.

Capacity Theory (Kahneman)

• Attention = limited mental resource, not just a filter.
• Key points:
  • Difficult tasks require more attentional resources.
  • Easy or overlearned tasks become automatic (require less attention).
  • Performance depends on:
    • Task difficulty
    • Effort allocation
    • Arousal

Biased Competition Model

• Multiple stimuli compete for neural representation.
• Attention biases competition toward:
  • Relevant stimuli (top-down goals).
  • Salient stimuli (bottom-up features).
• Selection emerges from neural competition.

Pre-Attentive vs Attentive Processing

• Pre-attentive Processing
  • Automatic
  • Unconscious
  • Parallel
  • Processes basic features (colour, orientation, size).
• Attentive Processing
  • Conscious
  • Limited capacity
  • Serial / effortful
  • Needed to bind features into coherent objects.

Visual Search

Pop-Out (Feature Search)

• Target defined by single distinct feature (e.g. red among green).
• Fast, efficient; reaction time independent of set size.
• Handled by pre-attentive processing.

Conjunction Search

• Target defined by combination of features (e.g. red + vertical).
• Requires serial, attentive search.
• Reaction time increases with number of items.

Feature Integration Theory (Treisman)

• Basic features processed pre-attentively in parallel.
• Binding features into objects requires:
  • Focused attention
  • Sequential, effortful processing.

Top-Down vs Bottom-Up Attention

• Bottom-Up (Stimulus-Driven)
  • Attention captured by salient properties:
    • Brightness, motion, sudden onset.
  • Automatic, exogenous.
• Top-Down (Goal-Directed)
  • Attention guided by goals, expectations, knowledge.
  • Voluntary, endogenous.
• Spatial-Based Attention
  • Selecting locations in space.
• Feature-Based Attention
  • Selecting specific features (e.g. red items, moving objects).

Classic Attention Phenomena

Stroop Effect

• Task:
  • Naming the ink colour of colour words.
• Interference:
  • Slower + more errors when word meaning conflicts with ink colour.
• Explanation:
  • Reading is highly automatic.
  • Automatic processing interferes with controlled colour naming.

Inattentional Blindness

• Failure to notice unexpected objects when attention is engaged elsewhere.
• Stronger when unexpected object:
  • Differs in features from task items.
  • Is unrelated to the attended task.

Mobile Phone Study

• Visual tracking task + unexpected red cross.
• Tracking accuracy:
  • Similar with vs without phone (~77–78%).
• Noticing unexpected object:
  • 10% noticed on phone vs 70% without.
• Conclusion:
  • Phone use drastically reduces awareness of unexpected events, even if primary tracking task seems intact.

Heads-Up Displays (HUDs) & Pilots

• Haynes (1991):
  • ~25% of pilots in simulator didn’t see another plane crossing during landing.
  • Eyes looking outside, but attention on HUD.
• Implication:
  • Overlayed displays (HUD/AR) can monopolise attention and cause serious inattentional blindness.

Change Blindness

• Failure to notice changes between consecutive scenes.
• Often occurs with:
  • Visual interruptions (flicker, cuts).
  • Distractions or divided attention.
• Example:
  • Simons’ door study: people fail to notice that the person they’re talking to is swapped mid-conversation.

Eye Movements & Covert Attention

• Overt attention: where eyes are looking.
• Covert attention: where mind is focused.
• Key point:
  • Eye position ≠ guaranteed indicator of attentional focus.
  • You can attend without looking and look without attending.

Clinical & Extreme Attention Phenomena

Spatial Neglect

• Usually due to right parietal lobe damage.
• Patient ignores left side of space (or vice versa).
• Behaviour:
  • Eats food only from right side of plate.
  • Dresses only right side of body.
  • Reads only right side of page.
• Assessment Tasks
  • Line cancellation → marks mainly on right.
  • Line bisection → midpoint shifted to right.
  • Copying drawings → omits left-side features.
• Case: Peggy Palmer
  • Stroke → right parietal damage.
  • Omits or distorts left side even in mental imagery.
  • Shows neglect affects both perception & imagination.
• Insight:
  • Parietal lobes critical for:
    • Spatial awareness
    • Integrating perception + memory of space.

Simultanagnosia

• Inability to perceive more than one object at a time.
• Symptoms:
  • Bumps into objects not currently attended.
  • Can describe single object but not full scene.
• Cause:
  • Lesions at parietal–occipital junction.
  • Disrupted attentional networks.

Attention & Magic / Misdirection

Card Trick Example

• Audience asked to choose one card from six.
• Magician then shows new set; all six cards are different.
• Audience believes only their chosen card is gone.
• Why?
  • They encode only the attended card, not the others.
  • Non-attended cards never reach strong memory.

Misdirection (Apollo Robbins)

• Exploits:
  • Narrow, selective nature of attention.
• Core idea:
  • If attention is captured in one place, changes elsewhere go unnoticed.
• Related to:
  • Inattentional blindness
  • Simultanagnosia-like constraints in normal brains.