Using Discussion Board Data to Hire Teaching Assistants Arnaud Deza University of Toronto St. George Toronto, Canada arnaud.deza@mail.utoronto.ca Haocheng Hu University of Toronto St. George Toronto, Canada hc.hu@mail.utoronto.ca Vaishvik Maisuria University of Toronto Mississauga Mississauga, Canada vaishvik.maisuria@mail.utoronto.ca Michael Liut * University of Toronto Mississauga Mississauga, Canada michael.liut@utoronto.ca Andrew Petersen * University of Toronto Mississauga Mississauga, Canada andrew.petersen@utoronto.ca Bogdan Simion * University of Toronto Mississauga Mississauga, Canada bogdan.simion@utoronto.ca ABSTRACT Teaching Assistants (TAs) fill critical roles in instructional teams, spending one-on-one time with students, providing feedback, and serving a critical mentoring role. As a result, the process of selecting TAs is important, yet instructors of- ten have little information when making hiring decisions. We used data collected from a course discussion board to identify statistics that are indicative of participation and that might be useful for hiring TAs. We then analyzed opast TA hires and found that there are opportunities to use discussion board metrics to increase the number of potentially qualified appli- cants for TA roles and to better inform instructors about the engagement of past students as they select candidates. Author Keywords Educational data mining; learning analytics; discussion board; discussion board analytics CCS Concepts •Social and professional topics → Computing educa- tion; Computer science education; •Applied computing → Learning management systems; Interactive learning envi- ronments; Collaborative learning; INTRODUCTION Large courses require Teaching Assistants (TAs) to operate, and in larger institutions, many of those TAs are former stu- dents or current undergraduate students. TAs are important members of the instructional team, with many spending sig- nificant time mentoring and providing feedback. As a result, the process of selecting TAs is an important, if under-studied, element of course preparation. The TA hiring process varies by institution, but in many cases, the hiring instructor has relatively little information about ap- plicants who have not held a TA position before. Most appli- cants will not have relevant experience, making performance * corresponding authors Sixth SPLICE Workshop at L@S 2020 “Building an Infrastructure for Computer Science Education Research and Practice at Scale”, SPLICE’20, August 12, 2020, Virtual Event in related courses the most commonly available piece of evi- dence. Instructors faced with such hiring decisions often have no information about students beyond these grades. In partic- ular, instructors may not have any information about an ap- plicant’s level of engagement with their peers or about their communication skills, yet we know that communication skills are necessary for effective teaching behaviours [3]. Data from discussion boards can play an effective role in bridging this knowledge gap in contexts in which former students (in par- ticular, current undergraduate students) might be considered for TA positions. In our particular context, undergraduate TAs play a crucial part by filling peer-mentoring roles in first year courses while they are senior undergraduates. In this study, we use data collected on the Piazza discussion board [10] and analyze which participation statistics (e.g., posts viewed, number of days active, number of questions asked and answered, etc.) can be utilized to identify addi- tional potential candidates for TA positions. These indicators are mapped against each respective student’s course grade to provide a better overview of a student. Additionally, by cross- referencing with hiring data from a subsequent year: 1. We find that while several applicants with high grades and high active participation/basic engagement in the prior year do get hired, there are some who are not even considered. While interviewing hiring instructors it was found that this was due to a lack of information about those candidates. We expect that these analytics may benefit instructors to make more informed decisions in selecting and interview- ing TAs who have the right attributes. 2. We find that quite a few students with high grades, who were active participants on the discussion board in the prior year, simply do not apply to become a TA at all. This sug- gests that better communication about what it takes to be a TA and how to apply might attract more applicants and increase the pool of good candidates. In the next section, we look at some of the learning theory that suggests that facilitating community is an important as-