This project explored how rehearsal of pedagogical skills in a mixed-reality classroom simulator supported STEM GTAs’ teaching as well as learning and affective outcomes for undergraduate students taught by the trained GTAs. For this project, a new learning environment was developed for the mixed-reality classroom simulator TeachLivE^TM to resemble a physical science laboratory environment, as shown in Figures 1 and 2.

We created and published a model for creating professional development modules to be implemented in a mixed-reality classroom simulator. In this project, GTAs rehearsed four pedagogical skills across four modules. Module 1 featured cold-calling (i.e., calling on non-volunteering students by name) paired with error framing (i.e., acknowledging that a student shared an incorrect idea in a way that frames errors as natural and/or beneficial). Module 2 featured questioning strategies, and Module 3 featured group management strategies. Finally, Module 4 allowed GTAS to integrate all four skills or to focus on a single challenging skill. We created three discipline-specific lesson scenarios for introductory physics, chemistry, and calculus. During the projects, over 100 GTAs received pedagogical training, impacting at least 6,000 undergraduate students. Consent rates for research were low among the calculus GTA cohort, we can not meaningfully analyze some data for this cohort, such as classroom observations.

Analyzing observations of both the physics and chemistry GTAs, we identified distinct instructional styles than had previously been described in the literature. For chemistry GTAs, these instructional styles included: Responders; Active Lecturers; and Initiators. For physics GTAs, the styles included: Group-work Facilitators; Waiters; and Whole-class Facilitators. In contrast to prior work, our findings suggest a relationship between GTAs’ instructional styles and student behaviors, with more interactive instructional styles correlating with higher student engagement. This analysis suggested that both the nature of the laboratory activities and the GTAs’ amount of prior experience impacted GTA instructional style, with less experienced GTAs tending to be more interactive than experienced GTAs. Relatedly, analysis of physics GTA teaching behaviors showed that GTAs were less interactive when the classroom in which they were implementing the same active learning curriculum changed from one with tables that supported multiple students sitting together to one with individual desks.

We conducted observations of the physics and chemistry GTAs in semesters with various amounts of simulator rehearsal of the target skills (i.e., no training, one session of training, and four sessions of training). We found that both new and experienced GTAs who completed all four simulator modules in one semester demonstrated a shift towards actively implementing whole-class facilitation strategies rather than passively waiting for their students to ask for help. We also observed that these GTAs implemented questioning and cold calling techniques. Specifically, we found that simulator training increased the number of high-quality questions GTAs asked during a class session.

Additionally, we analyzed a particularly tricky aspect of active learning instruction– what an instructor should say when a student makes a mistake in front of other students. Overall, we found error framing is difficult for GTAs to implement. We analyzed the error framing statements GTAs made in the mixed-reality classroom simulator and found that GTAs used both explicit and implicit forms of error identification and a variety of error framing techniques, including framing errors as natural, framing errors as beneficial, and positively acknowledging the error. From interviews with undergraduate students about some of these exemplar GTA error framing statements, we found that the specific framing that a GTA implemented as well as the GTA’s tone impacted student’s comfort. Some GTAs and some undergraduate students suggested that GTAs should not explicitly identify student mistakes. We analyzed GTAs use of error framing statements in their actual classrooms and found that in the rare cases that GTAs made error framing statements, they tended to use implicit, indirect strategies, which aligned with those that would support student comfort.

Overall, this project demonstrated that rehearsal in a mixed-reality classroom simulator supported GTAs’ use of teaching skills that are associated with active learning curricula, which have become common in introductory STEM courses.

Last Modified: 02/28/2025
Modified by: Jacquelyn Chini