The network approach to assess the structure of knowledge: Storage, distribution and retrieval as three measures in analysing concept maps

Abstract

AbstractWe present three new standardised network concept map (CM) measures that can provide unique information about learning‐related progress, which cannot be determined from previously known measures. Grounded in cognitive development theory on the one hand, and network theory on the other hand, our measures reveal how knowledge is stored, distributed and retrieved. We validated the new measures by testing their ability to discriminate between CMs of respondents with different levels of competency in statistics (students before and after taking an introductory statistics course and experts in the field of statistics). We also validated our measures against the most commonly used traditional and network measures. Based on a small sample of respondents, we show that two of the newly proposed compound measures reveal significant differences between experts and novices in the field, with higher values for experts, showing that expert knowledge is better distributed, more connected and balanced. More importantly, our measures were sensitive enough to show learning‐related progress for students, albeit statistically non‐significant, while common indicators from network theory did not demonstrate these small shifts. The validity of our new measures can be inferred from the consistency of the results from different sets of measures. Practitioner NotesWhat is already known about this topic A concept map (CM) is a knowledge assessment tool that can be used to evaluate student understanding of a topic and learning‐related progress. However, there has been a lack of standardised, complex and domain‐general CM metrics. Network theory has been shown to be useful in developing a standardised approach to CM evaluation. What this paper adds Based on the theory of cognitive development, we derived three standardised network measures of a CM: knowledge storage capacity, knowledge distributivity and knowledge retrieval. The new measures were empirically found to discriminate between groups of students with different levels of competency. Implications for practice and/or policy The new measures provide unique information about learning‐related progress. The newly developed measures enable educational researchers to make more objective evaluations of individual map elements, while simultaneously allowing for a holistic view of the quality of the knowledge structure of students.