“One Student Does All the Work”: Rethinking Collaboration for Computational Thinking Settings

“One Student Does All the Work”: Rethinking Collaboration for Computational Thinking Settings

Collaboration can improve learning outcomes through several mechanisms, including opportunities for verbalizing and elaborating one’s ideas and resolving potential discrepancies with peers to facilitate critical thinking (Andrews and Rapp 2015). These benefits are particularly useful in STEM fields, which often involve individuals with diverse perspectives working together to solve complex problems. Many studies have demonstrated gains in student understanding of STEM content when allowed to engage in discussion with peers (Barron 2000; Smith et al. 2009;). Collaboration is defined as a “coordinated, synchronous activity that is the result of a continued attempt to construct and maintain a shared conception of a problem” (Roschelle and Teasley 1995, p. 70). There is a growing interest in the assessment of skills associated with collaboration, particularly in problem-solving situations, a critical component of STEM teaching, given the importance of cooperation in a variety of contexts, including the school, workplace, and military. Advantages of collaboration for problem-solving is that collaboration affords a more effective division of labor, the incorporation of solutions from group members with differing perspectives, knowledge, and experience, and enhanced solution quality by the ideas of other group members (Graesser et al. 2017). However, in K-12 settings, educators often report that productive collaboration is difficult to support and assess.
As a result, we developed a rubric called Co-ACT to assesses students’ collaborative problem-solving skills during CT at the individual student level. The goal of this workshop is to support teachers in how to determine which components of collaboration the students need support in and then to reflect on the ways their teaching can shift students toward more productive collaboration. This workshop will be useful whether your students are in face-to-face or online/remote learning environments. In this workshop, we will dig into the ins and outs of productive collaborative problem solving for computational thinking.
The rubric consists of five dimensions: peer interactions, positive communication, pattern generalization/decomposition, data analytics, iterative thinking. The first two dimensions are social interactions that are needed for any collaborative problem-solving activity. These are peer interactions and positive communication. The dimension of peer interactions within collaborative problem solving and computational thinking, students are expected to refer to the guidelines of the rubric with their peers to identify group goals and monitor progress towards completing the tasks. Group members then discuss how to divide tasks relying on one other’s expertise to equitably complete the work. Students rely on their group members to check for accuracy in the process (e.g., Does the way we are approaching the task make sense?) and the content (e.g., Is the content accurate?). Students provide one another with feedback to help them gauge how they are doing or redirect tasks. There are three attributes: monitors tasks and checks for understanding with peers, negotiates roles and divides work to complete tasks, and provides peer feedback, assistance, and/or redirection. The second dimension is positive communication. Similar to collaboration in other contexts, positive communication is essential to efficiently work towards solutions. Students participating in CPS and CT activities are expected to respect one another to foster productive contributions by all members. This dimension has three attributes: respects others’ ideas and compromises, uses socially appropriate language and behavior, and listens and takes turns.

The third dimension is pattern decomposition/generalization is a distinguishing characteristic of computational thinking is the ability to break down complex problems into more manageable parts. Usually, a model, representation or simulation is used to plan the approach and ultimately design an algorithm that can be generalized to other problems. When working on CT tasks together, we would expect to see students discussing ways to approach modeling, share tasks related to breaking down the problem, finding patterns and removing extraneous parts, and codesigning algorithms. This dimension consists of four attributes: discusses ways to break down a problem, discusses patterns in the problem and ways, negotiates methods to reduce problem complexity, co-designs algorithms/procedures.

The fourth dimensions is data analytics, which is one of the hallmarks of computational thinking is the ability to collect, analyze, and represent data. However, when assessing collaborative problem-solving in these contexts, we would expect students to work with group members to develop systems, making sense of the data together, and negotiate how the data is represented. This dimension contains three attributes: Discusses data collection approach, consults with peers about sense-making or finding patterns.

The fifth dimension is iterative thinking, which includes the process of prototyping to include testing and debugging, and iterated designs. When assessing collaborative problem solving during iterative thinking, we would expect students to engage with their peers using discussions, feedback, and critique. It includes two attributes: tests and debugs prototype or solutions with peers and iterates designs or prototypes with peers.

The last three dimensions (pattern decomposition/generalization, data analytics, and iterative thinking are unique to collaborative problem solving during computational thinking practices, which we have aligned to Weintrop et al. (2016)’s framework of CT-STEM Taxonomy. The term practices are used to represent the combination of both knowledge and skills. Researchers have proposed several definitions and frameworks for CT (Grover & Pea, 2013). The structure most closely aligned with this study is the CT-STEM taxonomy of practices that focuses on the applications of computational methods in STEM areas (Weintrop et al., 2016). The taxonomy was created from an analysis of interviews with STEM professionals, existing inventories and standards documents, and exemplary educational activities to identify existing real-world instantiations of computational thinking and practices. The taxonomy consists of four strands: data practices, modeling/simulation practices, computational problem-solving practices, and systems thinking practices. We aligned the Co-ACT rubric two of the strands: the data practices and computational problem-solving practices. Each consists of several sub-strands such as manipulating, visualizing, and analyzing data and choosing effective computational tools, troubleshooting and debugging, and assessing different computational solutions to a problem. The two other strands modeling and simulation (MS) and systems thinking (ST) practices are related to the CT collaboration activities in which the students’ participated, but there was not a clear alignment. For example, the students rarely compared models (assessing computational models a sub-strand of MS) or discussed how their findings could impact a more extensive system (communicating information about a system in ST). In this way, DP and CPSP are most appropriate for the scope of this project, given the focus on data science instruction and assessing gains in elementary students.

Workshop Activities:
We will begin the workshop by engaging the participants in an experiential collaborative activity and then reflecting on the experience while looking at the dimensions of collaboration in the rubric. Then, through the use of videos, collected from Grade 3-8, we will examine what these CT and CPS practices look like in classrooms. Then, we will use the rubric to determine the students’ skill proficiency of each skill using the rubric in order to teach the participants how to use the rubric from an instructor standpoint. At this point, we will examine how collaboration can both disrupt and contribute to status issues (e.g., race, gender, SES) in classrooms and how this rubric can provide a pathway for teachers to unearth some of those status issues and provide support for how to reduce some of those status issues in their classroom through productive collaboration. Then, we will brainstorm ways to shift instruction to better support students, and how it would look differently from a variety of contexts. We will also discuss how the rubric can be used in peer-assessment and self-assessment.

Teacher educators, in-service teachers, and curriculum coordinators would be interested in this workshop as it focuses on how to support teachers to promote productive collaboration. As well, we will examine which components of collaboration are most challenging and how to work around those issues.

The research team is a combination of teacher educators (Quigley and Herro), teachers (Lojek and Edkins), curriculum coordinator (Owens), and a graduate student (Plank). All have utilized the rubric and been a part of the development process. In this way, the team represents both the research and practitioner sides of the rubric implementation.

Learning objectives
1. Explore computational thinking and collaborative problem solving for K-12 settings.
2. Examine what productive CPS looks like in classroom contexts.
3. Determine challenges for CPS and explore workarounds for those challenges
4. Utilizing the rubric from an instructor standpoint
5. Discuss ways to implement the rubric from peer and self-assessment standpoint
6. Examine how this rubric can disrupt status issues (i.e., race, gender, SES) in the classroom

Long-term support
All participants will have digital access to the rubric as well as a subscription to the YouTube channel for videos aligned to the rubric so they can train teachers in their contexts. In this way, the participants will have full access to the rubric and training module to train their students, teachers, or researchers.