Ch 9 Semantic Organization

Front 1

Hierarchical Network Model

Back 1

Proposes that items are categorized by using the hierarchical relations specified in a semantic network
-Features stored at highest level - all levels below that are true
2 Assumptions
1)Takes time to move from 1 level to another
2)Takes additional time to retrieve features stored at one of the levels

Front 2

Typicality Effect

Back 2

Finding that the more typical members of a category are classified more quickly than the less typical category members
(eg)easier to verify that a canary is a bird than to verify an ostrich is a bird

Front 3

Feature Comparison Model

Back 3

Items are categorized by matching the items features to category features
-Meaning of words stored in memory as a list of features
-When concepts have too many or very few features in common a fast yes or no decision can be made
-When concepts have some, but not a lot of features in common, then an additional check of their defining features is made--> Slower Response
-Accounts for Typicality Effects & Category-Size Effect

Front 4

Feature Comparison Model Limitations

Back 4

1)Relies on ratings to make most of its predictions
2)Proposal that all comparisons require computations: We use the features of concepts to compute their degree of similarity
3)Argument against necessary or defining features
-Characteristic features are more salient and directly observable than defining features

Front 5

Defining Feature

Back 5

Feature that is necessary to be a member of that category - Possessed by all members of a category
(eg)Concept bird: Defining Feature = has wings, has feathers

Front 6

Characteristic Features

Back 6

Usually present in members of that category, but not necessary - Shared by some but not all category members
(eg)Concept robin: can fly, is small

Front 7

Category-Size Effect

Back 7

Finding that members of smaller categories are classified more quickly then members of larger categories
(eg)Verifying that a collie is a dog is quicker than verifying that collie is an animal
-Smaller Category(dog) requires fewer inferences than the larger category(animal)
-Smaller category is part of larger category and appears lower in hierarchy so its reached sooner

Front 8

Spreading Activation Model

Back 8

Accounts for response times by formulating assumptions about how activation spreads in a semantic network
-Concepts joined together by links that show relationships
-Length of link represents the degree of semantic relatedness between 2 concepts
-When concept is processed, activation spreads outward along paths, but loses strength over distance
-Activation of 2nd concept decreases amount of activation of 1st concept
-Semantic Priming

Front 9

Spreading Activation Model - Limiations

Back 9

Includes assumptions of both hierarchical net models and feature comparison models
1. Too many assumptions - can explain anything
2. Too few clear cut predictions - unless specific predictions can be tested, can't determine whether models are any good