Future Directions in Computing Education Summit Part One: Important Computing Education Research Questions

Abstract

The National Science Foundation (NSF) funded a summit in January of 2014 where researchers gathered to discuss future research directions for Computer Science (CS) education. The participants addressed the question: "What do you consider an important research question, or series of related questions, in CS education research?" This report summarizes the discussions from that summit. -- The most significant question raised in the summit was "How do we involve everyone in computing?" How do we provide computing education for all? Our "everyone" includes females, underrepresented minorities, low-income students in community colleges, as well as students with disabilities. When we talk about computing in compulsory education ages, we also have to consider how we provide computing education to people who may not be interested in computing or in STEM. We also have to think about computing for a broad range of abilities. -- The challenge of preparing enough teachers to meet our school-age needs in computing education is so large that we might consider an alternative approach. The summit participants explored the possibility of introducing computing through existing mathematics and science classes, using STEM-prepared teachers to learn enough computer science to teach introductory units. A growing body of research shows how programming activities can enhance learning in science and mathematics. -- We know too little about how learning occurs in computer science. Foundational questions to be answered includes "How do students learn to program and what does that development look like?", "What are successful and unsuccessful mental models of challenging computing concepts?", "How do we support successful transfer from beginner programming environments to real-world ones?", and "What are common challenges in conceptual understanding in computing course?" -- Computing and STEM share a symbiotic relationship. STEM can enrich computational learning while also providing valuable opportunities to embed CT in established and accessible (as well as required) STEM courses. The CT-STEM research agenda would thus need to seek answers to several questions, including, "What CT skills are most important for STEM?", "How can these skills best be taught through curricular units in STEM coursework?", "What STEM content can be reinforced naturally through CT?", and "How can we get STEM teachers to adopt and practice computational methods and approaches in their curricula?" These questions are also worth asking for particular STEM disciplines. What do Engineering faculty (for example) need to teach about computing, and how can computing help with engineering education? -- Most STEM discipline-based education research (DBER) areas have been around longer than computing education research. We need to learn lessons from these other DBER efforts to grow computing education and avoid the mistakes that others have faced. We have to identify the barriers to greater access and success in computing education in order to address those. For example, do we have requirements for mathematics that are greater than are necessary for success in computing? -- We know too little about how to assess learning computer science. We need to improve assessment if we're going to answer the foundational questions about how knowledge in computing develops. Summit participants are exploring the possibility of using "Big Data" to improve our understanding of computing knowledge development. -- The greatest need in providing computing education in schools is to provide more teachers. Many of the participants are exploring how to provide professional development and how to motivate teachers to learn computing. But some of the participants are exploring how we can help non-CS teachers to teach CS, the role of information education options (non-profit organizations, start-up companies, MOOCS, and on-line video sites). -- There has been a recent significant interest in learning analytics, tools, and grading, seen especially with the recent creation of the Learning@Scale conference. While Learning@Scale is not limited to studying CS, that seems to be the focus of early research. There is an opportunity to synergistically incorporate researchers and research from this growing Learning@Scale community into the more traditional computing education research community. -- Beyond getting students and teachers interested in computing, we need to understand how to improve the quality of graduates. We need to understand how to prepare computing graduates to use the most effective development processes, including team behaviors and parallel programming abilities. We need to prepare all STEM students to have the kind of fluency that they need with computing.