lWouldn’t be it faster if we store the fact that “canary can fly” rather than a broader category?
canary and animal lCollin and Quillian tested this prediction by measuring reaction time to these statements. They then showed the same idea i.e. participants are slower to respond to higher level of the network which is “canary is an animal” vs. “canary is a canary”. They also tested at property level and found similar results in sentence verification task lAnother property of this semantic network is the concept of spreading activation which activities spread along the link connected to the node e.g. to move through the network from robin to bird, it activates the node and also other linked concepts such as animal and ostrich etc. causing them to be primed easier to be retrieved lThis is determined by a lexical decision task where participants had to read many words with real and nonsense words. Some words are associated with particular concept e.g. bird while some aren’t associated at all shown in the graph lResearchers found that people respond to associated words faster than not associated words. Researchers suggest that associated words also are activated easier to be accessed and retrieved resulting a faster reaction time lCriticism of Collins and Quillian’s Hierarchical Model: -Many pointed out that it can’t explain the typicality effect which is the fact that typical words are responded faster than atypical ones e.g. “canary is a bird” is verified quicker than “ostrich is a bird” but they are equally far away from the “bird” node -This conflicting data led Collins and Quillian to propose a new model which is Collins and Loftus Model lResearchers proposed that concepts that are closely related are connected through shorter lines e.g. vehicle connected to fire engine as a