Operationalizing agentic engagement in medical education PBL groups

Introduction
Student engagement is a useful perspective for educators seeking to understand how and why students react and learn in diverse learning activities. Until recently, the student engagement construct was conceptualized as being three-dimensional, containing a behavioural (implying students’ effort and attention), a cognitive (students’ sophisticated learning strategies) and an emotional (presence of interest/enthusiasm) dimension. This tri-fold conceptualization failed to represent the extent to which students contribute to, modify and enrich their learning. Such disparity between reactive and proactive learning prompted the addition of agentic engagement as a fourth component of student engagement (Reeve & Tseng, 2011). Agentic engagement is defined as the proactive, constructive and intentional contributions students make to the flow of instruction they receive in order to optimize their learning. Although agentic engagement was initially developed in secondary education, it also seems to be a relevant practice for benefitting optimally from problem-based learning (PBL) in higher education. In this study, we aimed to operationalize agentic engagement in higher medical education small PBL groups.

Methods
First, we developed a draft list composed agentic engagement acts in literature used to operationalize agentic engagement in different contexts. We complemented this list with potential relevant acts that we sourced from proactive behaviour characteristics, PBL practices and adult learning literature. We then performed exploratory observations to adapt the draft list of potential agentic engagement acts to our second-year pre-clinical medical student study group at the University of Gronignen. To facilitate the operationalization process, we built an observation following the guidelines of an existing observation reporting tool.

In the pre-clinical phase of the medical education programme, students meet twice a week in tutor group meetings to discuss patient-related complaints related to the week's theme. Using video recordings of seven different tutor groups (7 first and 7 second meetings in total), we analyzed which acts from the initial scheme were observed in our context and which additional, unique acts of agentic engagement could be identified in our higher education PBL context.

Results
We initially identified 10 distinct agentic engagement items that can be used to measure agentic engagement in small-group PBL teaching sessions. When tailored to our study population, only 8 acts were observed in our pre-clinical medical tutor groups. These were when a student: asks/answers clarifying questions, expresses expectations in line with the learning material, tells the teacher what they (dis)like, makes learning as interactive as possible, suggests co-operation with another student, creates alternative ways of going through the learning material, corrects the given content and enriches others’ insights.

Further analysis showed that agentic engagement behaviours covered three main themes: (1) choosing the content to be studied (34.9%, 15 contributions), (2) complementary additions (46.5%, 20 contributions), and (3) correcting content shared by peers (18.6%, 8 contributions). In addition to the three main themes, we identified agentic engagement that involved both individual students (unilateral contributions, 48,.8%) and multiple students (transactional contributions, 51.2%). Apart from student-initiated agentic engagement (65.2% of all contributions), we also observed student agentic engagement as a result of tutors’ encouragement to change the flow of the instruction (34.8% of all contributions).

Discussion (and conclusion)
When observing behaviour in higher education PBL groups, we identified unique agentic engagement behaviours that have not previous been described in higher education literature and are distinctive to our medical education population (i.e., students correcting their peers, students suggesting co-operation and students creating alternative ways of going through the learning material). Such student-initiated behaviour supports the conceptualization of agentic engagement in PBL pre-clinical medical education being one where students create the content and steer the flow of instruction (most commonly without the help of their tutor); rather than merely contributing to the instruction they receive.

Our study adds to the literature by operationalizing agentic engagement in a higher education PBL context. Gaining insight into how students exhibit agentic engagement during the learning process is important for teachers to adjust their instructional mode and support student agentic engagement more effectively.