Lecture 6 - Semantic Memory Flashcards
In a sentence verification task, what is the the subject and the predicate? E.g. a canary is a bird
subject - canary, predicate - bird
In a sentence verification task, what is a set inclusion?
What set the subject belongs to e.g. a canary is a bird ‘isa’
In a sentence verification task, what is a property-attribute?
A property of the object - e.g. a canary has feathers - ‘has’
What are concepts represented by?
nodes
Relationships between concepts are represented by…?
links
Describe 2 aspects of the Hierarchical network model
- concepts are organised in a hierarchy
2. cognitive economy - property attribute is stored at the highest level. ANIMAL>BIRD>CANARY
What does the idea of Hierarchal network model suggest?
sentence verification RT is a function of levels that you have to travel through in the hierarchy
What is the main problem with the hierarchal network model?
Conrad (1972) RT is better explained in terms of frequency of co-occurence of concept and property rather than levels - associative strength
What are three more problems with the hierarchical network model?
- RTs did not always mirror hierarchal relationship
- within-category typicality effects
- faster negative judgements for closer concepts
What are three characteristics of the spreading activation model? (Collins & Loftus, 1975)
- concepts are organised non-hierachically
- links vary in associative strength/accessibility
- activation of a concept spreads to other concepts link to it
What is the semantic priming effect?
Response to a word is faster following a semantically related word e.g. car + truck
What is the semantic priming effect used to study? 2 things
- organisation of semantic memory
2. automaticity
What is the feature comparison model? (Smith, Shoben & Rips, 1974)
It assumes that knowledge is represented as distributed features in semantic space
What do network models assume?
knowledge (e.g. canary has wings)
What is multidimensional scaling?
Subjects rate how similar a pair of concepts are e.g. sheep-goat. Data are represented in semantic space on a graph
e.g. domesticity (wild-domesticated), size (small-large) etc.