In a lexical decision task, participants see a prime word and then must decide if a target is a real word. 'Doctor' primes 'nurse' — responses to 'nurse' are faster after seeing 'doctor' than after seeing 'lamp.' Spreading activation theory predicts that 'hospital' would prime 'nurse'...
AMore than 'doctor,' since hospitals are more directly associated with nurses in a professional sense
BEqually to 'doctor,' since both are semantically medical and priming is categorical
CSomewhat, but likely less than 'doctor' if the 'hospital'–'nurse' associative link is weaker than the 'doctor'–'nurse' link
DNot at all, because spreading activation only travels one link at a time and 'hospital' is not directly linked to 'nurse'
Spreading activation predicts priming proportional to associative link strength, not just category membership. If 'doctor'–'nurse' is a stronger or more direct link than 'hospital'–'nurse,' then 'doctor' will prime 'nurse' more. Both primes should produce some facilitation since both are semantically related to 'nurse,' but the degree of priming varies with link strength. This is precisely why reaction time experiments can map the fine-grained structure of semantic memory.
Question 2 Multiple Choice
Collins and Quillian's (1969) hierarchical network model predicted equal verification times for 'A robin is a bird' and 'A penguin is a bird,' since both require one taxonomic step. The observed typicality effect — robins are verified faster — is best explained by which account?
APenguins are less common words, so lexical access is slower regardless of the semantic relationship
BThe hierarchical model is correct; the typicality effect is an artifact of participants being more familiar with robins
CTypical members like robins have stronger or more numerous associative links to 'bird,' receiving more spreading activation than atypical members like penguins
DPeople store typical and atypical category members in separate memory systems with different access speeds
Spreading activation networks explain typicality effects naturally: a typical bird like a robin has many strong connections to bird features (flies, has wings, sings, builds nests) and to the 'bird' node itself, so it receives activation from many directions simultaneously. An atypical member like a penguin has fewer or weaker links — it shares fewer features with the bird prototype. The hierarchical model's assumption of equal one-step distances for all category members is the flaw that typicality effects expose.
Question 3 True / False
In spreading activation theory, semantic priming occurs because seeing a prime word causes activation to spread through the network before the target appears, so related concepts are already partially active when the target word must be processed.
TTrue
FFalse
Answer: True
This is the core mechanistic account of priming in spreading activation models. The prime activates its node, activation spreads along associative links to neighboring nodes with strength proportional to link weight, and nodes that are pre-activated require less additional activation to reach threshold for recognition. The result is faster reaction times for targets semantically related to the prime. This mechanism makes the model empirically testable: stronger associative links should produce larger priming effects.
Question 4 True / False
Patients with semantic dementia lose knowledge of typical category members (like 'dog') before atypical ones (like 'hyena'), because typical items are encountered more frequently in daily life and are therefore more vulnerable to degradation.
TTrue
FFalse
Answer: False
The opposite is true: atypical members are lost before typical ones. The explanation is not frequency of encounter but network structure. Typical members like 'dog' have many strong connections to the 'animal' node, to shared features, and to other typical members — they receive activation from many directions and are robust to partial damage. Atypical members like 'hyena' have fewer, weaker links and depend on a smaller number of connections that degrade early. This gradient of loss from periphery to center is precisely what spreading activation networks predict.
Question 5 Short Answer
Why does the spreading activation network model predict that semantic dementia patients will lose knowledge of atypical category members before typical ones? What does this pattern reveal about how knowledge is organized in semantic memory?
Think about your answer, then reveal below.
Model answer: Typical category members (like 'robin' for birds) are densely connected — they share many features with the category prototype, have strong links to the category node, and receive spreading activation from many directions. Even as some connections degrade, activation still reaches them along multiple pathways. Atypical members (like 'penguin') have fewer strong connections and depend on a small number of links that are more vulnerable to damage. When anterior temporal lobe atrophy reduces network connectivity, peripheral nodes lose their activation supply first while densely connected typical members remain accessible longer. This reveals that semantic memory is not a flat list but a weighted network where knowledge is represented through the density and strength of relational connections.