3.4 Selective & Divided Attention

3.4 Selective & Divided Attention

Attention is the set of cognitive mechanisms by which the brain allocates its limited processing resources to prioritize specific environmental inputs, thoughts, or actions while suppressing competing distractors. Because human cognitive capacity is strictly finite, processing every sensory input simultaneously is impossible. Cognitive psychologists categorize attention into selective attention (focusing on one stimulus while filtering out others) and divided attention (attempting to perform multiple tasks concurrently). This section examines classic filter models of selective attention, experimental paradigms like dichotic listening, controlled versus automatic processing, and attentional limits.

The Cognitive Bottleneck & Selective Attention Models

Selective attention functions as a cognitive filter, determining which sensory inputs receive higher-level semantic processing and enter conscious awareness. Over several decades, cognitive psychologists proposed competing theoretical models to explain where in the human information-processing stream this selection filter operates.

Sensory Input ──> Sensory Register ──>[ Filter Location? ]──> Semantic Processing ──> Conscious Memory

1. Broadbent's Early Selection Filter Model (1958)

Donald Broadbent proposed that information processing is constrained by a rigid physical bottleneck.

Sensory Input ──> Sensory Register ──> Selective Filter ──> Bottleneck ──> Perceptual Processing ──> Short-Term Memory
                                      (Physical traits:
                                       pitch, volume)

• Mechanism: All sensory stimuli enter an initial short-term sensory register. A selective filter then screens incoming messages based strictly on low-level physical characteristics (e.g., pitch, loudness, spatial origin, voice gender).
• All-or-Nothing Filter: Unattended messages are completely blocked before any perceptual or semantic (meaning-based) processing can occur.
• Critical Flaw: Broadbent's model cannot explain the Cocktail Party Effect—the empirical observation that an individual engaged in a deep conversation can instantly notice their own name being spoken softly in an unattended conversation across a noisy room. If unattended inputs were filtered out prior to semantic processing, one's name could never be recognized.

2. Treisman's Attenuation Theory (1964)

To resolve the limitations of Broadbent's model, Anne Treisman modified the early selection framework by replacing the rigid filter with an Attenuator.

Sensory Input ──> Sensory Register ──> Attenuator ──> Mental Lexicon ──> Semantic Processing ──> Conscious Memory
                                    ("Volume knob"    (Activation
                                      weakens input)   Thresholds)

• Mechanism: The attenuator does not block unattended stimuli; instead, it acts like a "volume knob," weakening (attenuating) their signal intensity while allowing the attended signal to pass at full strength.
• The Mental Lexicon & Thresholds: Attenuated stimuli pass to the mental lexicon—a storehouse of words and concepts, each possessing a specific activation threshold:
  • Most words require strong, un-attenuated signal strength to reach their activation threshold.
  • High-priority concepts (such as one's own name, "Fire!", or words relevant to current context) possess extremely low activation thresholds. Even an attenuated, weakened signal is sufficient to cross these low thresholds and trigger conscious awareness.

3. Deutsch & Deutsch Late Selection Model (1963)

J. Anthony Deutsch and Diana Deutsch proposed a radical alternative in which all incoming sensory stimuli are processed semantically before selection occurs.

Sensory Input ──> Sensory Register ──> Full Semantic Processing ──> Late Selection Filter ──> Conscious Awareness / Response

• Mechanism: The selection filter is located after perceptual and semantic analysis. All inputs are analyzed for meaning, but only information that is semantically relevant to the observer's current goals is selected for conscious memory and behavioral response.
• Trade-Off: Late selection models account for complex priming and unattended semantic processing, but require immense cognitive effort to analyze every distractor semantically.

4. Kahneman's Capacity Model of Attention

Daniel Kahneman shifted focus away from structural bottlenecks, modeling attention as a single pool of flexible cognitive capacity or effort. The amount of available cognitive capacity varies with physiological arousal (Yerkes-Dodson Law) and is allocated based on task difficulty, enduring dispositions (e.g., involuntary response to sudden loud noises), and momentary intentions.

---

Experimental Paradigms in Attention Research

Cognitive psychologists utilize precise behavioral paradigms to test attentional mechanics:

                      ┌─────────────────────────────────┐
                      │  Attentional Research Paradigms │
                      └────────────────┬────────────────┘
                                       │
         ┌─────────────────────────────┼─────────────────────────────┐
         ▼                             ▼                             ▼
┌─────────────────┐           ┌─────────────────┐           ┌─────────────────┐
│ Dichotic        │           │ Inattentional   │           │ Change          │
│ Listening Task  │           │ Blindness       │           │ Blindness       │
├─────────────────┤           ├─────────────────┤           ├─────────────────┤
│ Headphones play │           │ Failure to see  │           │ Failure to      │
│ different audio │           │ fully visible   │           │ detect visual   │
│ to left & right │           │ unexpected item │           │ changes across  │
│ ears. Shadowing.│           │ when focused.   │           │ scene breaks.   │
└─────────────────┘           └─────────────────┘           └─────────────────┘

Dichotic Listening & Shadowing Tasks

In a dichotic listening task, participants wear headphones that play two distinct auditory messages simultaneously—one to the left ear and one to the right ear. In a shadowing task, the participant is instructed to repeat aloud the message presented to the attended ear word-for-word as it is being spoken.

• Findings: Participants readily shadow the attended ear. When asked about the unattended ear, they notice physical changes (e.g., tone changing to a buzz, speaker switching from male to female), but fail to report semantic contents or language changes (e.g., switching from English to German), supporting attenuation and early selection dynamics.

Visual Attentional Failures

• Inattentional Blindness: The failure to perceive a fully visible, unexpected object in the visual field because attention is intensely focused on another task or object (e.g., viewers counting basketball passes failing to notice a person in a gorilla suit walking through the game).
• Change Blindness: The failure to detect noticeable alterations in a visual scene following a brief visual disruption, flicker, or saccadic eye movement (e.g., failing to notice that an interviewer was swapped for a different person behind a moving door).

---

Divided Attention & Multitasking

Divided attention occurs when an individual attempts to attend to or perform two or more tasks simultaneously. Human capacity for multitasking is heavily restricted, with dual-task performance typically showing significant impairment (dual-task interference).

Key Determinants of Dual-Task Performance

1. Task Similarity: Tasks that rely on the same sensory modality or neural processing infrastructure suffer severe mutual interference. For example, reading a textbook while listening to a talk radio show both compete for verbal processing structures (Wernicke's and Broca's areas). Conversely, combining a visual task (driving) with an auditory task (listening to music) causes less structural interference.
2. Task Difficulty: Highly complex tasks demand a greater proportion of the available pool of cognitive effort, leaving insufficient resources for secondary tasks.
3. Practice and Automaticity: Extensive practice transforms task execution from controlled processing to automatic processing.

---

Controlled versus Automatic Processing

The shift from novice performance to expert mastery is defined by the transition from controlled to automatic processing:

Treisman's Feature Integration Theory

Feature Integration Theory explains how attention binds visual traits together:

• Pre-attentive Stage: Visual features (color, shape, motion) are extracted automatically, effortlessly, and in parallel across the visual scene.
• Attentive Stage: Focused attention acts as "cognitive glue," binding individual features located at a specific spatial position into a unified object.
  • Feature Search: Target differs from distractors by a single trait (e.g., red circle among blue circles). Target "pops out" instantly regardless of distractor count (parallel search).
  • Conjunction Search: Target shares traits with distractors (e.g., red circle among red squares and blue circles). Requires serial search, scanning items one by one.

---

High-Yield MCAT Strategy & Common Traps

• Broadbent vs Treisman: Remember that Broadbent's model uses a hard filter based only on physical traits (cannot explain the Cocktail Party Effect). Treisman uses an attenuator ("volume knob") and activation thresholds in the mental lexicon.
• Inattentional vs Change Blindness: Inattentional blindness involves an unexpected new object entering the scene while focused. Change blindness involves comparing two states of a scene across a disruption and missing a modified detail.