5 JANUARY/FEBRUARY 2003 The authors argue that design-based research, which blends empir- ical educational research with the theory-driven design of learning environments, is an important methodology for understanding how, when, and why educational innovations work in practice. Design- based researchers��� innovations embody specific theoretical claims about teaching and learning, and help us understand the relationships among educational theory, designed artifact, and practice. Design is central in efforts to foster learning, create usable knowledge, and ad- vance theories of learning and teaching in complex settings. Design- based research also may contribute to the growth of human capacity for subsequent educational reform. Eproblems ducational researchers, policymakers, and practitioners agree that educational research is often divorced from the and issues of everyday practice���a split that cre- ates a need for new research approaches that speak directly to prob- lems of practice (National Research Council [NRC], 2002) and that lead to the development of ���usable knowledge��� (Lagemann, 2002). Design-based research (Brown, 1992 Collins, 1992) is an emerging paradigm for the study of learning in context through the systematic design and study of instructional strategies and tools. We argue that design-based research can help create and extend knowledge about developing, enacting, and sustaining in- novative learning environments. Definitions of design experiments abound (see Bell, 2002a). We use the phrase design-based research methods deliberately (after Hoadley, 2002) to avoid invoking mistaken identification with ex- perimental design, with studies of designers, or with trial teaching methods. We propose that good design-based research exhibits the following five characteristics: First, the central goals of designing learning environments and developing theories or ���prototheories��� of learning are intertwined. Second, development and research take place through continuous cycles of design, enactment, analy- sis, and redesign (Cobb, 2001 Collins, 1992). Third, research on designs must lead to sharable theories that help communicate rel- evant implications to practitioners and other educational design- ers (cf. Brophy, 2002). Fourth, research must account for how designs function in authentic settings. It must not only document success or failure but also focus on interactions that refine our un- derstanding of the learning issues involved. Fifth, the development of such accounts relies on methods that can document and con- nect processes of enactment to outcomes of interest. Design-Based Research: An Emerging Paradigm for Educational Inquiry by The Design-Based Research Collective Educational Researcher, Vol. 32, No. 1, pp. 5���8 Why We Need Design-Based Research: Challenges of Context, Design, and Enactment The last few years have seen a renewed effort to close the ���credi- bility gap��� (Levin & O���Donnell, 1999) in educational research. Some see this gap as arising from unscientific research approaches (e.g., NRC, 2002), while others point to the detachment of re- search from practice (Lagemann & Shulman, 1999). Educa- tional research that is detached from practice may not account for the influence of contexts, the emergent and complex nature of outcomes, and the incompleteness of knowledge about which factors are relevant for prediction (Robinson, 1998). Claiming success for an educational intervention is a tricky business. If suc- cess means being certain that an intervention caused learning, then we need to look carefully at the intervention in a particular setting. However, research in this model would be difficult to generalize to other settings. On the other hand, if success means being able to claim that an intervention could be effective in any setting, then we should study effects across a variety of settings in order to generalize. However, this kind of research leaves many questions unanswered about how any observed learning was caused by interactions between intervention and setting. To ad- dress these problems, we view educational interventions holisti- cally���we see interventions as enacted through the interactions between materials, teachers, and learners. Because the interven- tion as enacted is a product of the context in which it is imple- mented, the intervention is the outcome (or at least an outcome) in an important sense. In addition, the design of innovations enables us to create learning conditions that learning theory suggests are productive, but that are not commonly practiced or are not well understood. For example, the Jasper Series (Cognition and Technology Group at Vanderbilt, 1997) was one of the first learning environments that presented students with an opportunity to develop compu- tational skills by grappling with real-world scenarios. This inno- vation allowed the testing of the tenets of ���anchored instruction,��� which included the belief that learning should be contextualized, and of ideas that mathematics learning should be more closely tied to students��� experience. Successive classroom trials with dif- ferent versions of these series contributed to an understanding of the characteristics of an effective ���anchor,��� clarified issues for the theoretical refinement of transfer, and showed how social inter- actions play a role in metacognition. Illustrations of Design-Based Research Design-based research methods focus on designing and exploring the whole range of designed innovations: artifacts as well as less

EDUCATIONAL RESEARCHER 6 concrete aspects such as activity structures, institutions, scaffolds, and curricula. Importantly, design-based research goes beyond merely designing and testing particular interventions. Interven- tions embody specific theoretical claims about teaching and learning, and reflect a commitment to understanding the rela- tionships among theory, designed artifacts, and practice. At the same time, research on specific interventions can contribute to theories of learning and teaching. One example is the BGuILE project (Reiser et al., 2001), a program of research to support inquiry learning in biology. The underlying innovative approach was the use of discipline-specific scaffolds for data analysis and explanation construction. Initial design efforts focused on developing software and activity struc- tures to support student investigations of episodes of natural se- lection, followed by cycles of design, enactment, and analysis to successively refine the relationships between technological scaf- folds for inquiry and social scaffolds to support scientific dis- course. Besides improving the curriculum design, this project yielded concomitant findings concerning psychological aspects of students��� understanding of scientific explanation and argu- mentation (Sandoval, 2003), as well as the role that artifacts and the social interactions around them play in such processes (Tabak & Reiser, 1997). Such findings provide insights into the complexity involved in de- veloping knowledge and skills, and they help us understand the role that teachers play in capitalizing on the affordances of learning materials, but they could easily have gone unno- ticed had the research focused solely on the summative effects of the intervention. Design-based research methods respond to emergent features of the setting. For example, the Passion School Project (Joseph, 2002) articulated theoretical principles that should drive an interest-based curriculum, including the principle that students develop competencies in adult-defined learning objectives through engagement in authentic work in their area of interest. Micro-analyses of student interactions with activities based on that principle enabled redesign and refinement of activities, and ultimately refinement of the underlying interest-driven learning framework. Thus, emergent behaviors of students in response to activities drove development of the intervention and develop- ment of theory. These developments would have been unimag- inable in the absence of real student choices. Finally, in design-based research, practitioners and research- ers work together to produce meaningful change in contexts of practice (e.g., classrooms, after-school programs, teacher on-line communities). Such collaboration means that goals and design constraints are drawn from the local context as well as the re- searcher���s agenda, addressing one concern of many reform efforts (Robinson, 1998). Engaging such partnerships across multiple settings can uncover relationships between the numerous vari- ables that come into play in classroom contexts and help refine the key components of an intervention. In particular, these part- nerships can help us distinguish between a ���lethal mutation��� (Brown & Campione, 1996)���a reinterpretation that no longer captures the pedagogical essence of the innovation���from a pro- ductive adaptation���a reinterpretation that preserves this essence, but tailors the activity to the needs and characteristics of partic- ular classrooms. For example, Baumgartner (1999) described how several teachers adopted different strategies to manage the tension between performance goals (e.g., building effective fish- ing rods) and explanatory goals (e.g., understanding why the rod works the way it does) in a science and engineering curriculum. The role that local interpretation plays in successful implemen- tation became salient by examining the cases in which different teachers��� strategies achieved similar instructional goals. Indeed, such reinterpretation is inevitable and necessary. Sustainable in- novation requires understanding how and why an innovation works within a setting over time and across settings (Brown & Campione, 1996), and generating heuristics for those interested in enacting innovations in their own local contexts. Relationships Between Design-Based Research and Other Methodologies Design-based research has re- cently been described as a po- tentially fruitful methodology for generating causal accounts of learning and instruction that could form the basis for system- atic, randomized clinical trials (Levin & O���Donnell, 1999 NRC, 2002). We see design- based research as raising impor- tant questions for research applied to practice and for research methods, generally. However, randomized trials are not necessarily the appropriate end goal of our research approach we do not understand issues of context well enough yet to guarantee that randomized trials are the best means to answer the questions we care about. The use of ran- domized trials may hinder innovation studies by prematurely judging the efficacy of an intervention. Additionally, random- ized trials may systematically fail to account for phenomena that violate this method���s basic assumptions���that is, phenomena that are contextually dependent or those that result from the in- teraction of dozens, if not hundreds, of factors. Indeed, such phe- nomena are precisely what educational research most needs to account for in order to have application to educational practice. We would suggest, however, that design-based research can gen- erate plausible causal accounts because of its focus on linking processes to outcomes in particular settings, and can productively be linked with controlled laboratory experiments or randomized clinical trials (cf. Brown, 1992) by assisting in the identification of relevant contextual factors, aiding in identification of mechanisms (not just relationships), and enriching our understanding of the nature of the intervention itself. For example, the CoMPASS Importantly, design-based research goes beyond merely designing and testing particular interventions.