Research methods in cognitive science - Modeling, Observation & Evolutionary approach

Information processing uses a computer metaphor to simulate cognitive processes: input processing and output generation.
Also called Boxes-and-arrows models (similar to flow charts).
Information flow can be bottom-up, top-down, or interactive.

Visual input drives processing upward through intermediate representations (features: colors, shapes, sizes, etc.).
Pathway illustrative labels (from slide):
- Bottom-up processing leads to activation of intermediate level features and interacts with long-term memory.
- Visual input (blurry face) → intermediate features (colors, shapes, sizes, etc.) → long-term memory activation.
Key idea: Perception starts with sensory data, which is progressively transformed into higher-level representations without prior knowledge shaping the initial interpretation.

Prior exposure or knowledge about the input or its context shapes perception and interpretation.
The interaction is between long-term memory and current visual input, with expectations guiding interpretation of ambiguous stimuli (e.g., a blurry face).
Significance: Explains how expectations and context influence what we perceive, not just what is sensed.

Each node simulates a neuron; nodes can connect to multiple other nodes.
Activation can spread between nodes (spread activation).
Illustrative statement: If A is a dog, the network can propagate related features or concepts through connected nodes.
Significance: Provides a model of learning and pattern recognition that emphasizes distributed representation and parallel processing.

Machine learning: Learn what?
- Train the computer to classify a huge data distribution.
- Develop an algorithm from the training data.
- Largely relies on statistics and programming.
- Also used a lot in neuroimaging data analysis.
Robotics: Video example and prompt
- Link: https://www.youtube.com/watch?v =Qh2yT-AL1V8
- Prompt: From the video above, how do robots communicate with each other?
Significance: Demonstrates how AI/ML and robotics study cognition through computation, data-driven learning, and interaction among agents.
Reference note: Baddeley (2010) is cited in the interim summary as a key source on modelling approaches.

Box-and-arrow (computer metaphor):
- Works like flow charts (e.g., working memory model).
- Describes processing routes in perception and recognition: bottom-up and top-down/interactive.
Connectionism (artificial neural networks):
- Leads to machine learning, deep learning, and AI.
- Emphasizes distributed representations and learning through connection weights.
Reference: Baddeley (2010).

Derived from the ecological approach.
Core distinction: What is observable versus what is not observable (internal cognitive processes are not directly observable).

Non-naturalistic; a hybrid of experiment and observation.
Example topic: Environmental factors affecting students’ attention in a large lecture hall.
Typical manipulation: Environmental factors (e.g., lighting, room temperature) to observe impacts on attention or attendance.
Source of ideas (example and topic): https://tzuhui99.pixnet.net/blog/post/44604306

Clinical interviews: begin with open-ended questions, then follow up with questions depending on responses.
Questions may be designed beforehand; many clinical questionnaires are standardized.

Observing natural behaviors to enhance ecological validity.
Pros: More natural than lab experiments.
Cons: Lacks experimental control; can take longer to observe target phenomena (e.g., infant development).

Research approaches: Lab experiment and Naturalistic observation.
Both methods can inform about human cognitive evolution.
Slide shows numeric bullet: 7 9 62 4 38 (likely slide counts or data cues from the session).

Natural selection – “Survival of the fittest” leads to evolutionary pressures.
Example: Enhanced reasoning skills may be advantageous when cheating or solving problems in social contexts.
Goal: Understand cognition by finding evidence from evolution.
Evidence sources: Fossil records and evidence from non-human primates.
Big question: How did humans evolve to the current level of intelligence?

What fossil records tell us about cognitive evolution:
- Development of technology
- Anatomical changes
- Cultural and societal traits
- Emergence of language (question marks indicate uncertainty in this area)
Reference: https://images.app.goo.gl/pGrBT9TTZY98FYhWA

Experimental methods:
- Neuroimaging: fMRI – Answers the spatial “Where” question.
- EEG – Answers the temporal “When” question.
- Behavioral measures: Button-press responses (reaction time, accuracy); How priming works and time course of processing.
- Eye-tracking.
Observation methods:
- Naturalistic observation
- Non-naturalistic observation (including controlled observation and clinical interviews)
Modeling:
- Box-and-arrow model
- Connectionism (artificial neural networks)

Structuralism: Focuses on the "what" by studying mental elements through introspection.
Functionalism: Focuses on the functions of mental processes (the "why" we do X).
Empiricism: Knowledge built by accumulating experiences.
Behaviorism: Opposes introspection; focuses on stimulus–response; behavior explained by reinforcement and conditioning; distinguishes Classical vs. Operant conditioning.
Nativism: Pre-wired biological functions; opposes strict behaviorist views.
Ecological approach: Derived from functionalism; focuses on real-world settings.
Evolutionary approach: Looks for origin and evolution of human intelligence.

Is it possible that a study could fit in more than one theory/approach? YES!
Example: A behavioral study examining why people categorize objects. What kind of study is it?

Is it possible that a study could fit in more than one theory/approach? YES!!
Example: To understand the evolution of human foraging behavior, a behavioral experiment was conducted to compare how chimps and humans learned from their mistakes when searching for natural resources for survival, and why both chimps and humans showed a certain pattern in their behavior. What kind of study is it?