Time-on-task has been shown to predict student performance in Computer Science courses [3], making it a useful tool for teachers to identify which students need extra support. Previous work has found that fine-grained metrics (e.g. keystroke-derived) for time-on-task produce stronger predictions of performance when compared to coarse-grained metrics (e.g. submission-based). This poster starts by replicating previous findings at scale, specifically that keystroke-derived time-on-task metrics predict grades at the assignment level.

Small previous studies exploring student programming process have demonstrated behavioral metrics' (i.e., EQ, Watwin, RED) ability to predict grades. However, this relationship deteriorates when scaled across contexts [1]. This paper replicates Price et al's finding on anonymized data collected from thousands of learners across the globe that use the Codio platform and discusses the context-dependent nature of these behavioral metrics and grades noted in previous work as potential reasons for this finding.

Parsons problems ask students to unscramble lines of code by dragging and dropping them. Du, Luxton-Reilly, and Denny (2020) summarized Parsons problems benefits as being able to (1) "test programming concepts, be it syntax or logic, separately," (2) provide immediate feedback, (3) improve student engagement, and (4) reduce cognitive load. However, one major challenge of implementing Parsons problems is the need to code and host each question. While previous tools have attempted more user-friendly, graphical Parsons creators and even host the resulting problem sets (such as the Parsonizer), they do not leverage the full power of the most ubiquitous Parsons implementation (http://js-parsons.github.io/documentation/) which allows teachers to check resulting variable values and run unit tests. This demo provides an in-depth tutorial on using Codio's open-source graphical interface to generate Parsons problems (https://codio.github.io/parsons-puzzle-ui/) and shows teachers how to host the resulting problem sets on Github for free (https://github.com/codio-content/hosting-parsons-on-github-template). We are actively seeking feedback on how to make this project more usable for educators.

There is a lack of clarity on how to increase the adoption of research-based pedagogy in CS education. Past research indicates that evidence from education studies was not an influential factor, while educator excitement was a significant one. Based on an educator survey, we found that peer-reviewed citations correlated with educator familiarity and intended adoption. Interestingly, perceived benefits and challenges, which we used as a proxy for excitement, had a statistically significant yet smaller effect size. These results suggest that despite previous findings, evidence from education research might influence adoption. This work also offers a quantitative metric for educator excitement.

Many students are introduced to computing through its infusion into other school subjects. Advocates argue this approach can deepen learning and broaden who is exposed to computing. In many cases, such interdisciplinary activities are student-driven and collaborative. This requires students to balance multiple learning goals and leverage knowledge across subjects. When working in groups, students must also negotiate this balance with peers based on their collective expertise. Balance and negotiation, however, are not always easy. This paper presents data from a project to infuse computing into high school statistics using the R programming language. We analyze multiple episodes of video data from two pairs of students as they negotiated (1) the statistics and computing goals of an activity, (2) the knowledge needed to meet those goals, and (3) whose expertise can help achieve those goals. One pair consistently reached agreement along these dimensions, and engaged productively with both subject matter and computing. The other pair did not reach agreement, and struggled to accomplish their tasks. This work provides examples of productive and unproductive interdisciplinary computing collaborations, and contributes tools to study them.

Research on students' learning in computing typically investigates how to enable individuals to develop concepts and skills, yet many forms of computing education, from peer instruction to robotics competitions, involve group work in which understanding may not be entirely locatable within individuals' minds. We need theories and methods that allow us to understand learning in cognitive systems: culturally and historically situated groups of students, teachers, and tools. Accordingly, we draw on Hutchins' Distributed Cognition [16] theory to present a qualitative case study analysis of interaction and learning within a small group of middle school students programming computer music. Our analysis shows how a system of students, teachers, and tools, working in a music classroom, is able to accomplish conceptually demanding computer music programming. We show how the system does this by 1) collectively drawing on individuals' knowledge, 2) using the physical and virtual affordances of different tools to organize work, externalize knowledge, and create new demands for problem solving, and 3) reconfiguring relationships between individuals and tools over time as the focus of problem solving changes. We discuss the implications of this perspective for research on teaching, learning and assessment in computing.