The use of text mining to detect key shifts in Japanese first-year medical student professional identity formation through early exposure to non-healthcare hospital staff

Abstract

Background: Professional identity formation is nurtured through socialization, driven by interaction with role models, and supported through early clinical exposure (ECE) programmes. Non-healthcare professionals form part of the hospital community but are external to the culture of medicine, with their potential as role models unexplored. We employed text mining of student reflective assignments to explore the impact of socialization with non-healthcare professionals during ECE.

Methods: Assignments from 259 first-year medical students at Fukushima Medical University, Japan, underwent hierarchical cluster analysis. Interrelationships between the most-frequently-occurring words were analysed to create coding rules, which were applied to elucidate underlying themes.

Results: A shift in terms describing professional characteristics was detected, from "knowledge/skill" towards "pride [in one's work]" and "responsibility". Seven themes emerged: contribution of non-healthcare professionals, diversity of occupation, pride, responsibility, teamwork, patient care and gratitude. Students mentioning 'contribution of non-healthcare professionals' spoke of altruistic dedication and strong sense of purpose. These students expressed gratitude towards non-healthcare professionals for supporting clinical work, from a doctor's perspective.

Conclusion: Socialization with non-healthcare professionals provides important insights into the hospital working environment and cultural working norms. Through role modelling altruism and responsibility, non-healthcare professionals positively influenced student professional identity formation, promoting self-conceptualisation as a doctor.

Keywords: Altruism; Professional identity formation; Role modelling; Socialization; Text mining.

Fig. 1

A comparison of the frequencies of the four most-commonly-used words describing professional characteristics across all three time points. The numbers in the brackets are chi-square values. ^**p < 0.01

Fig. 2

A breakdown of how the 50 most-commonly-occurring words extracted from the reflective essays distribute across the three emergent networks leading to the extracted themes

Fig. 3

a Comparison of the frequencies of the seven extracted themes across the reflective essays, categorised by student self-reported impact. The black and white bars represent group A (most influenced by non-healthcare professionals) and group B (most influenced by clinical or allied health professionals), respectively. A chi- squared test was used to explore differences between the expression of themes depending on which department had the greatest self-reported impact on students. ^*p < 0.05. b Diagrammatic representation of the significant interrelationships among the themes based on the results of Fisher’s exact test. ^*p < 0.05, ^**p < 0.01, ^***p < 0.001