ch8-Human memory & cognition
Terms
undefined, object
copy deck
- Concept Identification
- Task that requires deciding whether an item is an example of a concept, where concepts are typically defined by logical rules
- Conjunctive Rule
-
Uses logical relations AND to relate stimulus attributes (such as small & square)
-Both attributes have to be present to be a conjunctive rule - Disnjunctive Rule
-
Uses logical relations OR to relate stimulus attributes
-A pattern that has either of these attributes - Attribute Learning
- Concept identification task in which people are told the logical rule (conjunctive...) but hvae to discover the relevant attributes
- Criticisms of Concept Identification Paradigm
-
1)Real-word categories are unlike the categories studied in the lab-Highly artificial & unrelated to the categorization tasks we usually encounter in the real world
2)It assumed that all members of a concept are equally good members
- Rules fail to predict typicality ratings
-Even if something can be defined on the basis of rules it may contain examples that differ in the typicality effect - Natural Categories
-
Real-World Categories
-Hierarchical-Each level contains many objects, but the variety decreases as the category becomes smaller
-Some members seem to be better representatives of the category than others - Superordinate Level Category
-
Top of hierarchy - Largest category-Members share few attributes
-(eg)Furniture, tools, vehicles - Basic-Level Category
-
Intermediate category in middle of hierarchy (eg)Table, saw, truck
- Most important because they're the most differentiated from one another-1st category we learn & most important in language
-Avoids: members sharing too few or too many attributes
-Members share many attributes but also have attributes that differ
-Categorization is fastest at basic-level - Subordinate-Level Category
-
Bottom of hierarchy-Smallest caegory (eg)Table-lamp, jigsaw, pickup truck
-Members share many attributes - Prototype
-
An item that typifies the members in a category & is used to represent the category
-Prototype of category is usually the "average"-Represents central tendency
-Think of objects from same basic-level (not avg shape of furniture, but avg shape of chair)
-Average shape is impossible at superordinate level - Family Resemblance
-
A measure of how frequently the attributes of a category member are shared by other members of the category
-(eg)A car has wheels as its attribute, so we'd count the vehicles that also have them
-Good representatives of a category have high family resemblance - Stereotype
-
An attribute value believed to be representative of social categories
-Exaggerating within group similarity - Prototype Model
-
Classification strategy that selects the category whose prototype is the most similar to the classified item
-Doesn't require many comparisons to classify a pattern
-Compares novel pattern with single pattern in each category
-Person creates prototype to represent each category & classifies a novel pattern by comparing it with the category prototypes - Feature Frequency Model
-
Classification strategy that selects the category having the most feature matches with the classified item - Matching Features
-Looks at features of the pattern & compares how many times they exactly match features of the category patterns - Exemplar Model
-
Proposes that patterns are categorized by comparing their similarity to category examples
-Base decisions on examples in the categories
-Nearest-neighbor & average distance