2024 ASEE Annual Conference & Exposition

Large universities can often have introductory computing courses with hundreds of students, dozens of TAs, and multiple TAs on duty at the same time. Universities have responded to this increase in scale with two main solutions: 1) increasing staff sizes by hiring more Teaching Assistants (TA) and, 2) using automated feedback systems to take the place of instructional staff. In this work in progress paper we investigate what occurs during office hour interactions, especially those in an intermediate data structures course at scale. We hope to use this understanding to develop a conceptual model of what gaps exist in current TA training methodologies, investigate the “black box” of office hours, and explore a potential taxonomy of office hour interactions. To collect this data, we are using a self-hosted version of MyDigitalHand, an automated office hours queueing system. We have modified our instance of MyDigitalHand to collect additional data to answer our research questions.

This study analyzes student-TA interactions in a third year Data Structures and Algorithms (DSA) course at a large R1 public university in the US during the Fall 2023 semester. This course was chosen in part for the difficulty of the average interaction. We believe that many questions brought to TAs in this course are driven by student misconceptions. We also think that much of the load on TAs should be pointing the students away from a potential code error message and towards a misunderstanding of the project rather than debugging student code. The preliminary data contains the average interaction time (for students as well as TAs), total questions asked and answered, and responses to feedback questions on the topic of the interaction and whether the individual thought the interaction solved the student’s issue. One question asked of teaching staff also investigated whether students had misconceptions about what their “real” problem was. This data has allowed us to target interviews and observations towards specific interactions to develop a more complete view of office hour interactions at this level.