One decade ago, NSF funded two efforts to create a computing education research (CER) community in the U.S. and Europe. Nicknamed "Bootstrappers" and "Scaffolders,"a these two projects helped create a CER community among existing computing faculty who had an interest in computing education. More than 40 computer science faculty became active in CER. Many publications have resulted, the ICER conference has been created and established, and many of those faculty members now hold leadership positions within the computing education community. What the Bootstrappers and Scaffolders did not focus on was growing the number of Ph.D. students in CER, and trying to establish the legitimacy of CER as a research discipline within computing departments at research-oriented schools. (It is difficult to place CER within schools of education, as computing is not a required component of primary or secondary schools.) With the exploding interest in computing education and technology, with the awareness of the challenges in computing education at the K–12 level, with NSF's recent push toward disciplinary-based education research in STEM fields,3 and with a huge growing interest in offering computer science courses via distance, NSF funded another pair of workshops. The purpose of the first summit was to allow the CER community to identify the most important research questions facing the discipline. The second summit focused on identifying and developing strategies to overcome structural impediments to supporting CER at the university level.
This column focuses on the first summit: the most important research questions facing CER. Interested researchers were invited to submit a white paper addressing key research questions likely to be faced by CER over the next five years. The accepted white papers (available from http://purl.stanford.edu/mn485tg1952) asked research questions primarily in five areas: broadening participation, computing in K–12 (primary and secondary), computing in STEM education, students and learning issues, and tools. A few papers were written at a meta-level, looking at the direction of CER and its impact, and looking more broadly at challenges facing the teaching of computing at all levels. Here, we summarize across the five research areas while highlighting the two most compelling agendas that currently look to be in contention in terms of garnering researchers' attention.
A critical issue for CER is "computing for all." Not unlike other STEM disciplines, computing suffers from under-representation of women and minority groups. Understanding why there is such little diversity in computing is a critical area of research in CER, but even more critical is research into interventions that make computing available for females, underrepresented minorities, low-income students, students in community colleges, as well as students with disabilities. The focus on increasing access to computing moves a lot of attention in CER into primary and secondary school education. It is the level of schooling where we can reach "all."
In computing education, we tend to focus on coming to understanding algorithms and information representations. The summit focused at a different level. Computing education researchers study ways to better understand how conflicting cultural values and self-belief that impact motivation, and tackling these through design practices and tools that foster meaningful learning. Because we deal with such a flexible medium that can easily shift modalities, there is no reason why computing tools and programming environments are not more accessible to those with disabilities. Because computing is typically an elective subject in primary and secondary school, computing education researchers look for any way they can to slide computing into schools, including integrating through humanities (including creative writing, theatre, and media production), through science classes, and through mathematics classes.
The focus in CER on broadening participation makes CER different from other discipline-based education research, like physics education. Our first keynote at the summit was from Nobel laureate Carl Wieman, who argued for rigorous research in physics (and in computing) education. Wieman noted that physics education research has largely stayed away from research in K–12, due to the difficulties working in this space. But the motivation to diversify physics education is not as strong as it is in CER. Computing is even less diverse than physics. The economic value for computing is even greater than physics, so the lack of access to computing education is an even greater cost to underrepresented groups than is a lack of access to physics education.
Lance Pérez offered a keynote on the history of engineering education research, and lessons CER could learn. In particular, Pérez noted that engineering education has largely advocated for separate engineering education departments, leading to the conflicting needs to both be strong teachers (of introductory engineering curricula) as well as effective engineering education researchers. He advised computing education researchers to develop a model that avoids the problems that engineering education has faced.
Research in computing education must necessarily incorporate research into gender studies and other non-majority perspectives. Perspectives and experiences are shaped by the constant navigation of multiple shifting identities of gender, race, class, religion, and ability. Computing education researchers were urged to disaggregate empirical data in CER based on race, ethnicity, gender, and socioeconomic status. Studies at the intersection of race, ethnicity, and gender are needed, to better understand why certain groups choose to enter and persist in computing while others do not. We don't come to understand others' perspectives by counting them, so CER necessitates the inclusion of qualitative research methodologies (such as interviews and case studies), so the richness of personal narratives may inform action.
Computing education is a newcomer to the ranks of discipline-based education research. Mathematics education researchers know a lot about how students come to understand numbers and sets. Physics education researchers have detailed research into the misconceptions that students develop about the physical world from living in it without measuring it. But we know relatively little about the foundations of computing knowledge, and these are more critical issues as we move computing into primary and secondary schools. Foundational questions include "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?" were listed among the many key questions the learning sciences can help address. Even though these topics have been studied extensively since the 1980s, many questions remain unanswered and would benefit from contemporary research in the learning sciences in socio-cultural and situated learning, distributed and embodied cognition, as well as activity, interaction and discourse analyses.2
Researchers are working to define and validate learning progressions for K–12 computing. Some researchers are studying how people "become computer scientists" and to investigate the broader meanings of the computing discipline. The use of design-based research and design-based implementation research is common in the community to support research and assessment in naturalistic settings.
A more critical issue for CER today is how to develop enough teachers to support computing education in primary and secondary schools. We need to support the professional growth of computing teachers, and in parallel, build a robust community of computing teachers. How, for example, will the U.S. meet the CS10K challenge1 by 2016? A major focus for research in computing education is, "What do we teach computing teachers?" What do computing teachers need to know that's different from other STEM teachers? That discipline-specific knowledge of teachers is called Pedagogical Content Knowledge.4 Defining PCK for computing and figuring out how to teach it are major thrusts in CER today. Computing education researchers are developing just-in-time teacher professional development to help teachers communicate and learn best practices, robust assessments, and the latest introductory programming.
Sally Fincher's keynote focused on the needs of teachers in order to improve computing education. How do we make our computing education research matter and influence practice? She pointed out that many of the innovations in teaching reading failed because they changed reading in ways that was dissimilar to practice. We need to ensure that we in computing education are focusing on what the students actually need to know, or we risk becoming ineffective and useless.
Like other research areas, the CER community is asking itself if "big data" can help with our issues too. Some researchers are calling for data and evidence-driven analyses of student learning as part of adding rigor to drive practice in CER, including the use of big data to develop scalable methods for automated assessments (along the lines of emergent forms of assessment in MOOCs). The hope is to leverage big data to understand better the nature of learning programming and computational problem solving, the problems learners encounter, and the pedagogies and strategies that can be used to address them. Researchers are using evidence-centered design and data mining to study both cognitive and non-cognitive aspects of computational thinking.5
The summit exposed some of the tensions between research groups in CER about what are the most important research areas to be addressing. A large contingent whose research focused on issues of equity argued that all research in CER, whether content pedagogy, culture, and so forth needed to be viewed through an equity lens. Others argued for CER to focus on more foundational issues, such as learning progressions and integration of computing content and computational thinking into math, science, and non-STEM classes.
Tensions in CER around research methods mirror what is going on in other STEM education research areas. Traditionally, much of CER focused on social science methodologies. New researchers to CER are using big data analytic techniques, and borrowing approaches from other computing disciplines, most notably machine learning. How can these "newer" techniques (for example, looking at click-through rates, intermediate forms of student artifacts, and other forms of "big data") be used to answer our questions? Do they substitute for our other research approaches, or do we use them in addition to more traditional approaches?
Computing education research is unique among the STEM education research fields. The economic value of knowing computing is greater than any other STEM field, but computing is the least diverse of all the STEM fields. Thus, too many people are losing out on the advantages of computing. We have to figure out how to fix that, while still answering the fundamental questions of our discipline. How do people come to understanding computing, and how can we make it better?
The Digital Library is published by the Association for Computing Machinery. Copyright © 2014 ACM, Inc.
No entries found