Lecture 13

Historical Foundations and Human Factors Ergonomics

• Origins of Cognitive Psychology:

  • Emerged during and immediately following World War II ($\text{WWII}$).

  • Driven by the technological realization that human attentional capacity is strictly limited.

  • Military equipment grew increasingly complex, leading to catastrophic human errors (e.g., military aircraft falling out of the sky not due to mechanical failures like oil pressure or enemy fire, but due to attentional overload from excessive cockpit instrumentation).

• Physical Ergonomics:

  • The practice of designing physical objects to match the human body's anatomical shape and physical proportions.

  • Relies heavily on statistical population data to optimize comfort and usability for the majority of people (e.g., ergonomic seating).

  • Evolution of Video Game Controllers:

    • Atari Lynx: Early bulky, boxy handheld design.

    • PlayStation $3$ Banana Prototype: A silver, banana-shaped controller prototype that was quickly withdrawn after manufacturers recognized human hands are not shaped like bananas.

Cognitive Ergonomics and System Design

• Cognitive Ergonomics Defined:

  • Building systems, interfaces, and tools that match human mental capacities and cognitive processing limits rather than just physical shape.

• Avoiding Over-Complexity:

  • Designing systems with excessive controls overwhelms human cognitive capacity.

  • Example: A hypothetical automobile equipped with $2$ gear sticks and dozens of dashboard buttons creates overwhelming cognitive complexity.

• Avoiding Over-Simplicity:

  • Overly simplified interfaces require excessive steps to execute basic commands.

  • Example: A minimalist keyboard featuring a single letter-selection button and a "choose" button.

  • Example: Text messaging on legacy mobile phones via T$9$ numerical keypads. To type a name like "Zachary", a user had to press the "$9$" key $4$ consecutive times just to generate the single letter "$z$".

• Interface Design and Working Memory:

  • Early Apple mice featured only $1$ button, ignoring human manual dexterity and multi-finger control.

  • Modern tablet and smartphone interfaces (e.g., Apple iPad and iPhone) utilize a standardized $4 \times 5$ icon grid with a fixed dock at the bottom.

  • This layout is deliberately aligned with human working memory constraints, specifically the magical number seven plus or minus two ($7 \times 2$ or $7 \frac{+}{-} 2$ operational limits). Systems force users to navigate through submenus rather than overwhelming them with simultaneous choices.

Defining Attention and Visual Awareness

• Sensory Processing Pipeline:

  • Perception: The sensory entry point through which environmental energy enters through the eyes, ears, and other sense organs.

  • Attention: The formal beginning of cognitive processing. It functions as a selective gate mechanism that chooses specific information for further cognitive processing while actively rejecting non-selected information.

  • Attention acts as a decision point (locus) operating early or late in the processing sequence.

• Modes of Attention:

  • Focused Attention:

    • Occurs when an individual is required to focus on specific goal-relevant targets while ignoring distractors.

    • Selective Counting Paradigm: When instructed to count basketball passes made exclusively by individuals wearing white shirts, observers arrive at the correct total of $14$ passes.

    • Gorilla Inattentional Blindness: A person in a gorilla suit walking across the basketball court is routinely missed by observers engaged in focused counting. Detecting the gorilla during a focused counting task is not an indicator of superior capability, but rather demonstrates a failure to maintain strict attentional focus.

  • Diffuse Attention:

    • Occurs in the absence of a specific task goal, allowing an individual to broadly take in an entire environmental scene.

    • Without task-specific constraints, diffuse attention enables the detection of unexpected background details or anomalous scene elements.

Theoretical Models of Attentional Capacity

• The Cognitive Capacity Metaphor:

  • The dominant theoretical view models the brain as a computational system with finite processing capacity (analogous to hardware limits such as comparing a PlayStation $5$ or an RTX $5090$ GPU to higher processing demands).

  • Because cognitive resources are limited, unselected environmental stimuli are omitted from processing and vanish from conscious awareness.

  • Inattentional Blindness: The failure to perceive fully visible objects or events because attention is engaged elsewhere.

  • Inattentional Blindness Blindness: A metacognitive failure where individuals remain oblivious to their own susceptibility to inattentional blindness, creating severe safety risks (e.g., walking into traffic while operating a smartphone).

• Resource Allocation Models:

  • Single-Resource Pool vs. Multiple-Resource Pools:

    • Multiple-pool theories propose independent resource pools for distinct sensory modalities (e.g., visual processing resources vs. auditory processing resources).

    • Example: A person can drive a vehicle while listening to a spoken radio broadcast, but cannot listen to $2$ simultaneous radio streams or drive $2$ vehicles at once.

    • Exception: Complex or high-risk driving maneuvers (e.g., executing a difficult turn) force drivers to turn down the radio, demonstrating that highly demanding tasks draw from a unified cognitive capacity pool.

• Selection for Action Theory:

  • Proposed by Alan Allport in the late $1990\text{s}$ as an evolutionary alternative to capacity-limit models.

  • Posits that attentional limits exist because motor action capabilities are physically restricted (humans possess only $2$ hands and can execute a limited number of physical actions at any given moment).

  • Processing all environmental inputs simultaneously would trigger conflicting action tendencies; attention evolved to select only the specific inputs required to execute immediate, coherent physical actions.

Locus of Selection: Early vs. Late Models

• Concept of Locus:

  • Refers to the precise stage within the cognitive processing stream where attentional filtering occurs.

• Early Selection Models:

  • Filtering occurs early in the processing chain, prior to semantic (meaning) analysis.

  • Selection is based purely on physical characteristics (e.g., voice pitch, acoustic volume, sound location, or distinguishing human speech from background noise).

  • Unattended inputs are discarded before their meaning can be extracted.

• Late Selection Models:

  • Filtering occurs late in the processing chain, after semantic recognition and full meaning processing have taken place.

• Experimental Evidence:

  • Dichotic Listening and Shadowing Paradigms:

    • Participants wear stereo headphones receiving distinct auditory messages in each ear and are instructed to shadow (repeat aloud continuously) the message presented to $1$ target ear.

    • Early Selection Support: Initial experiments showed participants could report physical traits of the unattended channel (e.g., speech vs. noise, male vs. female voice) but could not report the language, individual words, or semantic content.

    • Late Selection Support (Anne Treisman): When a coherent narrative was split across ears mid-sentence ("The trees on the left side of the road seem to be filled…"), participants automatically switched channels to follow the semantic meaning, proving that meaning in the unattended ear was being processed prior to filtering.

  • The Cocktail Party Effect:

    • The ability to detect one's own name spoken in an unattended background conversation within a noisy environment.

    • Controlled laboratory testing by Wood and Cowan demonstrated that only $35 mechanism percentage$ ($35ND$ or $35$ out of $100$, expressed as $35\text{\%}$) of individuals actually hear their name in an unattended auditory channel.

    • The widespread belief that individuals always hear their name (often predicted at an $8\text{\%}$ baseline expectancy by participants) is driven by confirmation bias: people remember instances where they heard their name, but remain unaware of the majority of instances where their name was spoken and missed.

    • The $35\text{\%}$ detection rate is frequently interpreted as temporary "attentional slippage" or "leakage" rather than evidence for continuous late selection across all inputs.

• Flexible Locus of Selection (Nilli Lavie's Load Theory):

  • The locus of attentional selection is dynamic and shifts depending on the perceptual and cognitive load of the primary task.

  • Low Perceptual Load: The primary task leaves spare cognitive capacity. Unconsumed capacity automatically spills over to process background distractors semantically (resulting in Late Selection).

  • High Perceptual Load: The primary task consumes all available processing capacity. No spare capacity remains to process background distractors, causing them to be filtered out purely by physical traits (resulting in Early Selection).

The Role of Attention in Visual Search and Feature Integration

• Preattentive vs. Serial Visual Search:

  • Preattentive Search (Pop-Out Effect): Visual search for a target defined by a single unique feature (e.g., locating a red circle among green circles) occurs effortlessly and in parallel across the entire visual field simultaneously.

  • Serial Search: Visual search for a target defined by a conjunction of multiple features (e.g., unique combinations of color and shape) requires item-by-item serial scanning.

  • Example: Where's Wally / Waldo illustrations require serial visual search because the target is defined by a specific combination of features (red-and-white striped shirt, blue jeans, glasses) shared by surrounding distractors.

• Feature Integration Theory (Anne Treisman):

  • Posits that the fundamental role of attention is feature binding—the process of combining separate visual features (e.g., color, shape, orientation, motion) into unified object representations.

  • Features are processed independently and in parallel across specialized preattentive feature maps.

  • The spatial spotlight of attention is required to bind features present at the same spatial location.

  • Illusory Conjunctions: When attentional resources are brief, distracted, or overloaded, features drift and bind incorrectly (e.g., rapidly presenting a blue triangle and a red square causes observers to report seeing a red triangle).

Change Blindness and Perceptual Illusions

• Mechanism of Change Blindness (Rensink and O'Regan):

  • Demonstrates that the human perception of complete, detailed visual awareness is an illusion.

  • Flicker Paradigm: Alternating $2$ images with a brief blank flash ($180\text{\,ms}$ interval) disrupts the automatic visual motion signals that normally capture attention.

  • Without motion cues, detecting substantial structural changes (e.g., a disappearing tree branch, shifting sky boundaries, or a head surround changing from black to white) requires an effortful serial search, often taking upwards of $30\text{\,s}$ or failing entirely.

• Perceptual vs. Cognitive Illusions:

  • Perceptual Illusions: Persist continuously even after the underlying structural mechanism is fully understood.

  • Cognitive Change Blindness: Disappears permanently once attention is directed to the changing element. Once identified, the change becomes immediately obvious and impossible to ignore.

• Film Anachronisms and Media Continuity Errors:

  • Gladiator ($2000$, set in $50\text{\,AD}$): An extra wearing modern blue jeans walks backward in the background while Russell Crowe pats a horse; wristwatches appear on extras in the amphitheater. These errors are routinely missed due to object-based attention focusing on foreground actors.

  • Raiders of the Lost Ark ($1980$, set in the $1930\text{s}$): An unregistered bystander in modern attire walks through an open bazaar scene on the far left of the frame.

  • Media Aesthetics and Production Details:

    • K-Pop Demon Hunters: Features a crow background scene that functions as a sponsored advertisement for a Samsung mobile device.

    • The Matrix: Filmed at Martin Place fountain in Sydney. Producers cast real-world twins and triplets as background extras to create the illusion of a computer-generated simulation. The film uses specific color grading: a heavy green tint represents the simulated virtual world, whereas a refreshing blue tint represents the real world.

    • Synthetic Image Anomalies: AI-generated images frequently contain contextual visual errors that bypass casual observation (e.g., multiple individuals sharing identical faces, missing microwave hardware, un-cut cakes with sliced portions distributed, or missing refrigerator handles).

Real-World Applications and Attentional Distraction

• Real-World Person Swap Paradigms:

  • Desk Counter Swap: A desk clerk ducks down behind a counter to retrieve paperwork and is replaced by a completely different individual (featuring different hair color, such as black to blonde, and differing height). A high proportion of participants fail to detect the substitution.

  • Door / Direction Swap: An experimenter asks a pedestrian for directions on a public street. Two workers carrying a door pass between them, during which the experimenter is swapped for a different person. Pedestrians routinely continue providing directions without noticing that the individual has changed.

• Mobile Device Interference and Attentional Control:

  • Auditory/Haptic Notifications: Receiving a silent phone vibration while performing a task causes substantial cognitive disruption and performance collapse. The distraction occurs because cognitive resources automatically shift to off-task thoughts regarding the sender or message content.

  • Media Multitasking: Research demonstrates that self-described media multitaskers do not possess superior parallel processing capacity. Instead, they suffer from reduced attentional control and an inability to suppress irrelevant internal and external distractors.

• Endogenous vs. Exogenous Attentional Capture:

  • Endogenous Attention: Voluntary, top-down, goal-directed attentional focus generated internally by the individual.

  • Exogenous Attention: Involuntary, bottom-up, stimulus-driven attentional capture triggered automatically by salient physical events in the environment.

  • Visual vs. Auditory Capture: Text displayed on visual presentation slides involuntarily captures visual attention over spoken auditory information, causing individuals to read slide text rather than process verbal explanations.

• Administrative Tracking:

  • Engagement tracking verification code: "BBForBanana".