Guided by theories of socialization and possible selves, this study examines how STEM doctoral students perceive their academic and professional trajectories. More specifically, we rely on four years of interview data from 66 doctoral students in the biological sciences to explore students’ perceived trajectories, focusing on the salient identities and experiences that shape the way students identify and describe their graduate experiences over time. Findings reveal wide variation in terms of how students described their trajectories, with some students describing linear trajectories and/or unchanging career interests, while others described their developmental trajectories as highly turbulent and non-linear. These perceived trajectories were largely shaped by student-advisor interactions, the value students placed on becoming “independent” scientists, and the privilege students brought with them to their graduate programs.

Theories of embodied cognition argue that human processes of thinking and reasoning are deeply connected with the actions and perceptions of the body. Recent research suggests that these theories can be successfully applied to the design of learning environments, and new technologies enable multimodal platforms that respond to students’ natural physical activity such as their gestures. This study examines how students engaged with an embodied mixed-reality science learning simulation using advanced gesture recognition techniques to support full-body interaction. The simulation environment acts as a communication platform for students to articulate their understanding of non-linear growth within different science contexts. In particular, this study investigates the different multimodal interaction metrics that were generated as students attempted to make sense of cross-cutting science concepts through using a personalized gesture scheme. Starting with video recordings of students’ full-body gestures, we examined the relationship between these embodied expressions and their subsequent success reasoning about non-linear growth. We report the patterns that we identified, and explicate our findings by detailing a few insightful cases of student interactions. Implications for the design of multimodal interaction technologies and the metrics that were used to investigate different types of students’ interactions while learning are discussed.