General Knowledge

Episodic vs. Semantic Memory (Tulving)

What is Tulving's distinction between episodic and semantic memory?

• Episodic memory: Memory for events that happened to you personally; emphasizes when, where, or how an event happened

• Semantic memory: Organized knowledge about the world, including knowledge about words and factual information; not tied to a specific time or place

Concepts and Categorization

What is a concept? Why are concepts important?

Concept: A mental representation of a category (a class of items that belong together)

Concepts allow us to:

• Organize knowledge efficiently

• Make predictions

• Interpret new information

Importance of concepts:

1. Storage efficiency 

   • We can combine a wide variety of similar objects using a single, one-word concept, greatly reducing storage space 

2. Concepts help inference making 

   • Ex. A young child knows a member of the category "fruit" has the attribute "you can eat it" 

   • When encountering a new fruit, she infers (usually correctly) that it's edible 

3. Situated cognition approach 

   • We tend to code a concept in terms of the context in which we learned it, making it difficult to transfer concepts from classroom to real-life situations 

Concepts and Categorization

What is categorization? What two processes work together?

Categorization: The process of grouping objects into categories (e.g., "fruit"). The physical category is stored as a mental representation distributed throughout the cerebral cortex.

Two essential processes:

1. Generalization

   1. Recognizing shared features among category members

   2. Applying category label to new instances (e.g., seeing a new dog breed as a "dog").

2. Discrimination

   1. noticing differences between conceptual categories (e.g., telling a dog apart from a wolf).

Classic Feature Theory

What is classic feature theory? What are its 3 characteristics?

Core idea: All members of a category have specific features in common.

3 characteristics:

1. All members of a category have all of the features 

   • Category membership requires possessing the appropriate necessary and sufficient features 

2. The distinction between different categories is clear and logical 

   • Categories have sharp, well-defined boundaries 

   • No question about whether something belongs or not 

3. All members of a category are created equal 

   • Every member that possesses the necessary features is equally representative 

   • A 2-year-old nephew and a 32-year-old unmarried man would both be equally valid examples of "bachelor" 

Classic Feature Theory

What is the problem with classic feature theory? Give the "bachelor" example.

Not all members of a category are created equal.

• Consider the category "bachelor": 

  • A 32-year-old unmarried male cousin seems like a much better example than a 2-year-old nephew or an elderly Catholic priest 

  • Yet all three individuals are indeed male and unmarried 

  • A "necessary and sufficient" model would need to conclude that all three deserve to be categorized as "bachelors" 

• Conclusion: Not all members of a category are created equal, challenging the classic feature theory 

The Prototype View (Rosch)

How did Rosch challenge classic feature theory (The prototype ciew of concepts)?

• Rosch disputed all 3 characteristics of classic feature theory, proposing that categories are organized around prototypes 

• Key challenges: 

  1. Not all members of a category have all of the features 

     • Members vary in which features they possess 

  2. Categories can have fuzzy rather than sharp boundaries 

     • No clear-cut distinction between categories 

  3. Some members are "better" than others 

     • Categories have graded structure 

Alternative: Categories are organized around prototypes.

The Prototype View (Rosch)

What is prototypicality? Give an example of typicality effects

Prototypicality: The degree to which an item is representative of its category.

• Ex. A robin and sparrow are prototypical birds (penguins and ostrich are not) 

Typicality effects (sentence verification):

• In sentence verification tasks, reaction times reveal prototypicality: 

• Example: 

  • “A robin is a bird” → faster 

  • “An ostrich is a bird” → slower 

• People judge typical items (prototypes) faster than nonprototypical items 

The Prototype View (Rosch)

What are two important characteristics of prototypes?

1. Graded structure of categories – categories are not all-or-none; members vary in representativeness.

2. Family resemblance – no single feature is shared by all members, but each example has at least one attribute in common with some other example of the concept.

Levels of Categorization (Rosch)

What are Rosch's three levels of categorization? Give examples.

1. Superordinate level (most general) – e.g., furniture, animal, tool

2. Basic level (moderately specific) – e.g., chair, dog, screwdriver (has special, privileged status)

3. Subordinate level (most specific) – e.g., desk chair, collie, Phillips screwdriver

Levels of Categorization (Rosch)

What experimental findings show the special status of basic-level categories?

1. Basic-level names are often used to identify objects:

   1. People typically use basic-level names to identify objects (e.g., "pen" not "writing instrument" or "Paper Mate Flair pen").

2. Recall: When shown superordinate or subordinate terms, people often recall the basic-level version later.

3. Semantic priming: Basic-level names (e.g., "apple") produce priming; superordinate names (e.g., "fruit") are too general and not helpful.

4. Brain activation: Superordinate terms activate prefrontal cortex; subordinate terms activate parietal region (visual search).

Feature Comparison Model

What is the feature comparison model? What are the two types of features?

Core idea: Concepts are stored in memory according to a list of features or attributes.

Two types of features:

1. Defining features – essential to meaning; necessary for category membership (e.g., for "bachelor": male, unmarried).

2. Characteristic features – descriptive but not essential; common but not required (e.g., for "bachelor": lives alone, dates).

Feature Comparison Model

How does the feature comparison model explain the typicality effect?

The model accounts for why typical examples of a category (e.g., "robin" for bird) are verified faster than atypical ones (e.g., "penguin"). 

• Ex. When verifying a statement (e.g., "A penguin is a bird"): 

  1. The model compares the feature lists of the concept ("penguin") and the category ("bird"). 

  2. Typical members share many characteristic features with the category, leading to a high overall feature overlap and faster "yes" responses. 

  3. Atypical members share primarily defining features but few characteristic features, resulting in lower overall overlap and slower verification. 

Network Models – Hierarchical Network Model

What is the core concept of network models of semantic memory?

Core concept: Semantic memory (your knowledge about facts, concepts, and meanings) is organized like a web or network in your brain. 

• Nodes: Represent concepts (e.g. bird, animal, robin 

• Nodes are connected by links: Links represent relationships between concepts 

Network Models – Hierarchical Network Model

Describe the hierarchical network model (Collins, Quillian, & Loftus). What are its two key features?

This model explains how we quickly verify statements like "A robin is a bird" or "A robin is an animal." 

• Key features 1: Hierarchical structure 

  • Concepts are arranged in levels from general to specific (e.g., Animal → Bird → Robin). 

    • Properties stored at the highest level are more general 

      • Example: "Can breathe" is stored at the Animal level, not repeated for Bird and Robin. 

    • Properties stored at the lower levels are more specific 

      • Example: "Can fly" is stored at the Bird level 

     

• Key feature 2: Spreading activation 

  • Activating one node triggers activation that spreads along links to related nodes 

    • Cognitive economy: The closer two concepts are in the hierarchy, the faster the activation and the quicker the verification. 

      • Example: Thinking of "robin" activates "bird" (fast), which then activates "animal" (slower). 

        • Response time increases with distance between nodes in the network. 

Parallel Distributed Processing (PDP) Approach

What is the PDP approach to semantic memory? How does it contrast with localist representations?

• Each concept represented by a distinct pattern of activation across a large set of nodes (distributed representations) 

  • e.g. It’s a pattern across many nodes (fur, bark, animal, pet, etc.) 

• Contrasts with localist representations (single node per concept) 

• PDP simulates human performance 

  • Three Key Abilities:  

    1. Spontaneous generalizations – conclusion about a general category 

       • You meet several engineering students who are conservative → 
         Your brain detects a pattern → 
         👉 You generalize: “Engineering students tend to be conservative” 

        

    2. Default assignment – Applying a general belief to a specific new case 

       • Example: Assuming a new engineering student is conservative based on others 

        

    3. Pattern completion – Recognizing something even when input is incomplete or wrong 

       • Example: Identifying a word even with missing or misleading information 

Models of Semantic Priming

What is semantic priming? What are two models that explain it?

Semantic priming: Identification of a word (CAT) is faster and more accurate when preceded by a related word (DOG) vs. an unrelated word (SUN).

Two models:

1. Morton's logogen model (localist representation)

2. Masson's PDP model (distributed representation)

Models of Semantic Priming

How does Morton's logogen model explain semantic priming?

• Each concept represented by a single node/cell (localist representations) 

• Source of semantic priming: Spreading activation from prime node to target node in the semantic (meaning) system  

  • Processing "DOG" activates related nodes like "CAT" 

  • "CAT" is partially pre-activated, so it's recognized faster 

   

• Priming happens because of connections BETWEEN separate nodes 

Models of Semantic Priming

How does Masson's PDP model explain semantic priming?

• Masson's PDP model (distributed representations) 

  • Each concept represented by a distinct pattern of activation across a large set of nodes/cells (distributed representations) 

  • 2 types of nodes: 

    • Perceptual ("P")/Spelling Nodes – how the word looks (“C-A-T”) 

    • Conceptual ("C")/Meaning Nodes – how the word looks (“C-A-T”) 

  • Source of semantic priming: Related items represented by similar patterns of activation across all Conceptual/Meaning Nodes 

    • "DOG" and "CAT" have similar activation patterns 

    • Less adjustment needed to recognize related words 

Semantic Memory, Schemas, and Scripts – Definitions

What is semantic memory? Give examples.

General knowledge about the world, including concepts, facts, and meanings of words.

Organized structurally to allow efficient retrieval and interpretation.

Helps us understand new information, make predictions, and fill in missing details.

Semantic Memory, Schemas, and Scripts – Definitions

What is a schema? How are schemas built?

A schema is generalized knowledge about a situation, event, object, or person.

Represents organized knowledge in semantic memory.

Built from past experiences, cultural knowledge, and repeated exposure

Semantic Memory, Schemas, and Scripts – Definitions

What is a script? Give an example.

A script is a type of schema that describes a sequence of actions and their order (a schema for an event).

Key characteristics: time-ordered structure; includes typical roles, actions, outcomes; automatically activated in familiar situations.

• Example (restaurant script): Enter → wait to be seated → look at menu → order → eat → pay bill → leave.

Core Functions of Schemas and Scripts

What are the 4 core functions of schemas and scripts?

1. Organizing knowledge

2. Guiding attention

3. Enabling inference

4. Influencing memory

Research on Scripts – Inferences (Graesser et al., 1979)

Describe Graesser et al.'s (1979) study on script inferences and false memory.

Method: Participants read "John went to a restaurant. He ordered chicken. He left a large tip."

Inferred information: Readers automatically inferred actions like "He sat down," "He ate the chicken," "He paid the check."

Participants then were given a Recognition test: Participants confidently but incorrectly stated that "He ate the chicken" was in the original story.

Conclusion: Scripts fill in missing details, increase understanding, but create false memories.

Script Identification (Trafimow & Wyer, 1993)

What did Trafimow & Wyer (1993) find about the timing of script activation?

The Research Question: Does the timing of script activation matter for memory? 

Key finding: Memory for script-related events is significantly better if a script-identifying event is presented early in a sequence.

• Ex. Early activation (e.g., knowing it's a restaurant) primes the script, guiding encoding and organization of subsequent information.
  

Conclusion: Early schema activation improves understanding and memory for related information

Memory for Schema-Consistent & Inconsistent Material

What did Brewer & Treyens (1981) find about memory for schema-consistent items?

Method: Participants waited in an office, then surprised with a memory test for the room's contents.
Findings:

• Schema-consistent items (desk, chair) → good recall (hits).

• False memories: Many falsely recalled schema-consistent but absent items (e.g., books on shelves).
  Conclusion: Schemas increase both hits (correct recall) and false alarms (incorrect recall).

Memory for Schema-Consistent & Inconsistent Material

What did Davidson (1994) find about memory for schema-inconsistent material?

Information that violates schemas can be highly memorable when it is vivid, distinctive, and disruptive to expectations.

Example: Seeing a firefighter cooking a gourmet meal is more memorable than seeing a firefighter holding a hose.

Conclusion: Unexpected information can stand out and be encoded strongly, even though schemas usually guide memory.

Boundary Extension (Intraub & Berkowits, 1996)

What is boundary extension? Give an example.

Definition: The common tendency to remember having seen a wider angle of a scene than was actually presented.

Example: Viewing a photo of a kitchen; later recall includes areas not actually shown.

Conclusion: Perception and memory are constructive – we create a mental model extending beyond sensory data using visual schemas.

Inferences and Spatial Schemas (Bransford et al., 1972)

Describe Bransford et al.'s (1972) study on inference and spatial schemas.

Method: Two groups read slightly different sentences:

• Group A: "Turtles rested beside a log"

• Group B: "Turtles rested on a log"
  

• Test: Both groups asked if they had heard "A fish swam beneath the floating log."
  

• Result: Group B (turtles on the log) often falsely believed it was originally presented.
  

• Explanation: The word "on" activates a spatial schema where the log is above water, supporting the inference that a fish could be beneath it.
  

• Conclusion: Schemas shape interpretation and alter memory through inference

Gender Stereotypes as Schemas – ERP Study

How do gender stereotypes function as schemas? What did Osterhout et al. (1997) find using ERP?

Gender stereotypes are organized knowledge structures about traits, behaviors, and roles associated with males/females in a culture.

Osterhout et al. (1997) – ERP study:

• Method: Measured brain activity while participants read stereotype-consistent vs. inconsistent sentences (e.g., "The nurse himself...").

• Finding: The brain shows a distinct, rapid electrical response to stereotype-inconsistent information, demonstrating that stereotypes are automatically activated and violations detected immediately.

Implicit Association Test (IAT) – Greenwald, McGhee & Schwartz

What is the Implicit Association Test (IAT)? How does it measure gender stereotypes?

The IAT measures the strength of automatic, often unconscious, associations between concepts and attributes.

Method:

• Participants categorize stimuli using two response keys.

• Key pairing alternates: e.g., Male/Career vs. Female/Family, then switched to Male/Family vs. Female/Career.

• Measure: Response time – faster responses indicate stronger associations.

Key finding: Participants are faster when stereotypically congruent pairs share a key (Male+Career, Female+Family) and slower when incongruent (Male+Family, Female+Career).

This reveals automatic associations outside conscious awareness.

Priming Effects on Gender Stereotypes (Blair & Banaji, 1996)

What did Blair & Banaji (1996) find using priming with gender-related words?

Method: Priming participants with a gender-related word influences response to a subsequent related word.

Finding: Faster reaction times for stereotype-consistent word pairs, confirming that these associations are deeply ingrained and automatically activated.

Indicates: Schemas operate automatically.