The data-driven approach of Complex Forma Mentis aims at gathering data about how high school students perceive and structure their knowledge of scientific concepts.

CFM adopts free association tasks, i.e. cognitive experiments where participants have to quickly produce target words when primed with a given cue (cf. Kenett et al. 2018a). In the last few years, networks of free associations have been shown to be powerful predictors of early word learning (Stella et al. 2017) and have been successfully used for characterising the creativity (Kenett et al. 2018b) and openness to new experiences in adults (Kenett et al. 2018c). These results indicate that the mental organisation of concepts captured by free associations has an influence on learning, creative thinking and positive/negative stances towards real-world settings. 

The above network reports free association data relative to “education” and “science” from the Edinburgh Associative Thesaurus, reporting associations of concepts provided by at least two different subjects in the experiment. Notice that concepts such as school and progress act as important bridges for interconnecting science and education.

Building a forma mentis networks of free associations is useful for addressing important issues of STEM education. For instance, do students generally perceive STEM subjects as positive concepts? Do students tend to associate negative positive concepts together? What are the concepts they associate with “chemical reaction” or “equations” or “simulation”? Are high school students able to relate “math” to their everyday experience? Do students structure their knowledge in different ways when compared against science teachers or scientists?


Example of a network of free associations for English speaking toddlers, based on the University of South Florida association norms. Word associations give rise to an association network where nodes represent concepts. Notice that associations relate with the meaning of concepts and can give rise to communities of words representing meaning contexts. In the above example concepts relative to bath furniture (purple nodes) cluster together away from concepts relative to containers or supports (red nodes).