A well-developed mental model is crucial for effectively studying physiology core concepts. However, mental models can be difficult for students to represent and for instructors to evaluate and correct. Systems modeling as a visualization cognitive tool may facilitate mental model development. On the other hand, evidence of mental model development may also be represented verbally, in writing, and therefore, be evaluated. In this study, analysis of writing prompt completions illustrated progress in physician assistant student mental model formation of physiology core concepts, such as homeostasis and cell-cell communication, over time. Two cohorts of physician assistant students were invited to voluntarily submit completions of writing prompts five times over 16 months. Sessions included submissions pre- and post-small group systems modeling participation. Word frequency and word association cluster dendrogram analyses were conducted on submissions using the tm text mining package in R to provide insight into progressive changes in core concepts of word use and associations. Students demonstrated expanded core concepts systems thinking over time. This was apparent through the increased use of systems process terms, such as homeostasis, in submissions immediately following systems modeling activities. Students also increasingly included terms and associations emphasizing cell-cell communication and systems integration. The inclusion of these concepts within student mental models was demonstrably enhanced by participation in systems modeling activities.NEW & NOTEWORTHY This study applies text mining, an artificial intelligence form of natural language processing, to evaluate a series of physiology student-written prompt completions. Text mining of student writing in physiology has not yet been reported in the literature. Through the application of this technique, longitudinal trends in student development of mental models of core concepts were identified and visualized through word frequency distributions and cluster dendrograms.
Keywords: core concepts; physician assistant; physiology; systems modeling; text mining.