March 8, 2025

Cognitive scientists have a long-standing interest in mapping out how information in our brains is organised and retrieved. Semantic memory is the part of long-term memory that stores facts and information about the world, and is conceptualised as a network of concepts that are connected based on associations and relationships. In order to study the large-scale structure of semantic memory, cognitive scientists may use computational network approaches to analyse behavioural data retrieved from classical psychological tests such as semantic fluency tasks. In a semantic fluency task, individuals are asked to list as many examples as they can from a specific category within a time limit, for example, naming as many animals as possible within a minute. 

In “Investigating the network structure of domain-specific knowledge using the semantic fluency task” (Memory & Cognition, 2023), Assistant Professor Cynthia Siew (NUS Psychology) and Anutra Guru (University of Oxford) explored differences in knowledge structures estimated from undergraduate students’ (experts) and high school students’ (novices) responses on domain-general and domain-specific semantic fluency tasks. Domain-general networks refer to broad and general categories like animals and fruits while domain-specific networks refer to academic disciplines like biology and physics. The researchers used computational methods from network science to quantify underlying knowledge representations of the academic disciplines used in the domain-specific semantic fluency tasks. 

Siew and Guru found quantitative differences in the network structures of undergraduate and high school students’ knowledge representations. Specifically, undergraduate students had more efficiently connected and less compartmentalized knowledge networks across both general and domain-specific cue words. This supports the hypothesis that additional years of education, especially at the university level, help develop more integrated and efficiently organized knowledge structures. It also corroborates with previous research showing that experts possess a more principled and holistic understanding of problem domains. Compared to novices, experts focus on deeper conceptual aspects of problems rather than superficial features. 

The findings of this study have significant implications for education and cognitive science. They illustrate how additional years of education lead to more integrated and efficiently connected knowledge networks in students. Such insights can influence educational practices by informing curriculum design and teaching strategies to better enhance students’ learning experiences. By examining differences between experts and novices, researchers can refine theories of expertise development and deepen the understanding of cognitive processes. 

Read the full article here.