J Form Des Learn (2017) 1:31–44
              DOI10.1007/s41686-017-0005-1
              Developing an Integrative STEM Curriculum for Robotics
              Education Through Educational Design Research
              T. J. Kopcha1 & J. McGregor1 & S. Shin1 & Y. Qian1 & J. Choi1 & R. Hill1 & J. Mativo1 &
              I. Choi1
              Published online: 22 June 2017
              #Association for Educational Communications & Technology 2017
              Abstract This paper presents an integrative standards-based         toK-12STEMeducationarguethatteachingSTEMinamore
              STEMcurriculumthat uses robots to develop students’ com-            connected manner, especially in the context of real-world is-
              putational thinking. The need for the project is rooted in both     sues, can make the STEM subjects more relevant to students
              the overall lack of existing materials as well as the need for      and teachers^ (Honey et al. 2014, p.1). The need for stronger
              materials that directly address specific STEM standards in an       STEMeducationisdrivenbyseveralfactors.First,manypro-
              integrative fashion. The paper details the first mesocycle of an    fessions now demand a workforce that can engage in STEM
              educational design research project (EDR) in which a robust         thinking and skills such as creativity, innovation, critical
              theoretical frameworkwascreatedtosupportthedevelopment              thinking, problemsolving,andinformationandmedialiteracy
              of a 2-week series of robotics lessons. Analysis of evaluation      (Kennedy and Odell 2014). In addition, the demand for stu-
              data from 5 fifth-grade teachers and their students revealed        dents to engage in STEM thinking and skills is predicted to
              that the integrative curriculum supported student problem           increase as technology advances and becomes more readily
              solving and teacher practices that supported cognitive de-          available (Johnson et al. 2015; Gonzalez and Kuenzi 2012).
              mand. Implications for research, design, and instruction are           Robotics is promoted as an important piece of STEM edu-
              discussed.                                                          cationbecauseitintroducesstudentstocomplexmathematical
                                                                                  and scientific thinking. Bers (2010, p. 2) noted that Brobotics
              Keywords Roboticseducation .Educationaldesignresearch .             can be a gateway to learning applied mathematical concepts,
              Design-basedresearch .K-12STEMeducation .                           the scientific method of inquiry, and problem solving.^
              Computationalthinking                                               Robots can replicate the physical movements of humans and
                                                                                  allow students to develop mental representations of abstract
                                                                                  mathematical ideas (Han 2013; Kennedy et al. 2014).
              Introduction                                                        Additionally,theuseofrobotscanincreasestudentmotivation
                                                                                  andencouragepersistence when students encounter challeng-
              In the USA, Science, Technology, Engineering, and                   ing and complex learning scenarios (Kennedy et al. 2014;
              Mathematics (STEM) education is seen as an important step           McGill 2012; Perlman 1976;Dicketal.2005).
              toward ensuring a successful future for the country (National          The mathematical and scientific thinking associated with
              Academies 2007). BAdvocates of more integrated approaches           robotics education is commonly referred to as computational
                                                                                  thinking. Computational thinking is a process that involves
              ThisresearchwasconductedaspartoftheResearchfortheAdvancement        Bsolving problems, designing systems, and understanding hu-
              of Innovative Learning (http://rail.coe.uga.edu)                    man behavior, by drawing on the concepts fundamental to
                                                                                  computer science^ (Wing 2006, p. 33). Computational think-
              * T. J. Kopcha                                                      ing promotes activities that are central to expertise in mathe-
                  tjkopcha@uga.edu                                                matics and science, including abstraction, problem decompo-
                                                                                  sition, prediction, and iterative, recursive thinking and error
              1   Department of Career and Information Studies, University of     detection (Barr et al. 2011; Grover and Pea 2013;Sengupta
                  Georgia, 850 College Station Rd, Athens, GA 30605, USA          et al. 2013). Research indicates that computational thinking is
              32                                                                                                   J Form Des Learn (2017) 1:31–44
              fundamentally beneficial for students’ academic performance         integrative STEM curriculum (see Capraro and Han 2014;
              in the STEM classroom (Grover and Pea 2013;National                 Capraro et al. 2013;Morrisonetal.2015)—is more likely to
              Research Council 2011;Wilenskyetal.2014).                           be appealing to and consequently implemented by teachers.
                 The inclusion of a robot in a lesson, however, does not             The purpose of this paper is therefore to describe an inte-
              guaranteethatstudentswillengageincomputationalthinking.             grative standards-based STEM curriculum that uses robots to
              Students are most likely to engage in computational thinking        develop students’ computational thinking. The need for the
              when robotics tasks are presented in a functional environ-          project is rooted in both the overall lack of existing materials
              ment—that is, an environment that combines problem-                 as well as a need for materials that directly address specific
              solving scenarios and explorations that make authentic use          STEMstandards in an integrative fashion. Using educational
              of STEM skills (Pea 1987). In a functional environment, stu-        design research (EDR) to guide our process, we first devel-
              dents are challenged to identify and solve complex problems         oped and operationalized a conceptual framework in curricu-
              bytestingandretestingtheirprogrammingoftherobot.While               lar materials that met specific math, science, and engineering
              the robot serves as a concrete external embodied representa-        standards. The curricular materials were then implemented in
              tion of students’ thinking (Han 2013), students must also en-       5 fifth-grade classrooms over a 2-week period as part of stu-
              gage in abstraction as they think through and program the           dents’regularclassroomlessons.Evaluativedatawerecollect-
              physical movement of the robot. Such activities are important       ed from teachers and students to demonstrate the efficacy of
              for STEM education because they help students develop the           the framework in practice.
              confidence and persistence needed to deal with ambiguous               TheremainderofthepaperdescribeseachphaseoftheEDR
              problems and collaborate and engage in rigorous academic            project as well as the results associated with the evaluative data.
              discussions with peers (Barr et al. 2011).                          Providing a detailed account of our design and overall curricu-
                 Despite the noted benefits of robotics education, teachers       lumaddresses the pressing need in the robotics education liter-
              struggle to integrate it into the mainstream classroom. This        ature and offers insight into the design of integrative STEM
              struggle stems from an overall lack of curricular resources for     materials for other curriculum developers and STEM scholars.
              teaching robotics in an age-appropriate manner (Barr et al. 2011;
              Khanlari2016).Inaddition,existingmaterialstypicallyfocuson
              programming the robot rather than presenting an environment         Materials and Methods
              for applying subject-specific knowledge. When working with
              five elementary classrooms, Chang et al. (2010) found that a        McKenneyandReeves(2012)describedhowEDRfocuseson
              focus on programming made it more difficult for teachers to         an educational innovation that is developed, tested, and
              see how integrating robots could help students learn subject-       researched over multiple iterations in an applied context. A
              specific standards. In a recent study of 11 elementary and          key characteristic of EDR is that the innovation is grounded
              secondary teachers, Khanlari (2016) similarly found that            in theory yetdevelopedinclosecoordinationwiththeintended
              teachers viewed robotics activity as an additional burden that      audience,whoprovideregularandfrequentinputandfeedback
              took time away from preparing students for testing in core sub-     about the innovation. Thus, EDR projects attempt to inform
              jects. Without clear connections to subject-specific learning,      theory from a scholarly perspective while also resulting in an
              teachers are more likely to see robotics activities as something    innovationthatisvaluedandgroundedinareal-worldcontext.
              that fits outside the regular curriculum rather than a tool for        ThegenericmodelforEDRconsistsofthreephases:analysis
              meeting subject-specific standards (Karim et al. 2015).             and exploration, design and construction, and evaluation and
                 Aviable alternative to focusing on programming would be          reflection (see McKenney and Reeves 2012). Analysis and ex-
              to create curricula for robotics education that present a func-     ploration focuses on defining a problem and potential solutions
              tional environment. In a functional environment, content from       that are grounded in theory and existing research as well as
              multiple STEMareas(e.g., math and science) is contextualized        practical considerations about the problem itself. Design and
              andappliedthroughtherobottosolveanauthenticproblemand               construction focuses on building an educational innovation in
              engage students in computational thinking. It is the integrative    a way that embodies the theories and practical considerations
              nature of this approach that has the potential to improve robot-    from the previous stage. Data collected on the innovation are
              ics education. Teachers are more likely to integrate robotics       then evaluated and reflected upon to improve the innovation in
              activities when the curricular materials clearly and directly meet  the future, as well as generate implications for theory and design.
              specific content standards (Chang et al. 2010; Karim et al.            McKenney and Reeves (2012) described how these three
              2015). In addition, curricular materials developed in an integra-   phases are situated in a Bflexible, iterative process^ (p. 77) in
              tive fashion have greater potential to develop students’scientif-   which an intervention is refined and theoretical knowledge is
              ic thinking and expertise (Sengupta et al. 2013). Thus, a curric-   generated through ongoing activity. Each individual phase
              ulum that uses robots to teach and apply concepts by drawing        represents a microcycle of activity that has its own distinct
              on multiple STEM subjects rather a single one—that is, an           qualities and purpose. Two or more phases/microcycles are
              J Form Des Learn (2017) 1:31–44                                                                                                    33
              often combined for reporting purposes or to aid in decision-         authentic, and enable collaboration in a safe environment for
              making before engaging in another iteration of the project.          exploration (Kapur and Bielaczyc 2011).
              The combined microcycles are referred to as a mesocycle,                Problems that are presented in a functional environment
              which typically represent one iteration of an educational de-        require students to exhibit high levels of cognitive demand.
              sign research project.                                               Cognitive demand is Bthe kind and level of thinking required
                 This paper presents the first mesocycle, or iteration, of         of students in order to successfully engage with and solve the
              EDR on our robotics curriculum. The cycle began with a               task^ (Stein et al. 2000, p. 11). Problems that support high
              thorough analysis and exploration of the literature. Next, a         levels of cognitive demand provide learners with an opportu-
              conceptual framework was designed and operationalized in             nity to form conjectures, justify strategies, interpret and refine
              the development of the curriculum. Finally, the curriculum           solutions, and make connections between concepts (Boston
              wastestedinlocalschoolsandevaluativedatawerecollected,               and Smith 2009; Tekkumru-Kisa et al. 2015). While the use
              analyzed, and reflected upon. A description of each phase in         of authentic, ill-structured problems in a functional environ-
              the context of this study is provided below.                         ment improves the likelihood that students will encounter a
                                                                                   high level of cognitive demand, it is no guarantee. Learners
              EDRPhaseI:AnalysisandExploration                                     mustbepressedtojustifyandexplaintheirthinking,aswellas
                                                                                   makeconnections to prior knowledge and experiences.
              Areview of the literature on STEM education and computa-                Another characteristic of a functional environment is col-
              tional thinking yielded a specific set of three comprehensive        laboration. When students engage in cognitively demanding
              design principles to guide our curriculum development: (a)           tasks within a functional environment, collaboration affords
              create a functional environment, (b) embed opportunities for         students the opportunity to engage in deeper levels of dis-
              embodiedlearning,and(c)integratemultipleSTEMstandards                course. Engaging in deeper discourse may enhance learning
              into the curriculum.                                                 through students explaining and justifying their thinking with
                                                                                   peers and allowing students to discuss multiple ideas and/or
              Create a Functional Environment                                      waysofsolvingproblems(LinandAnderson2008;Nussbaum
                                                                                   2008). Researchers have found that students perform better
              Functional environments are a key characteristic of a learning       when working with their peers to coordinate their efforts to
              environment that supports computational thinking. As Pea             solve problems (Lanzonder 2005; Witney and Smallbone
              (1987) noted, functional environments Bhelp motivate stu-            2011). Through collaboration, students can critically observe
              dents to think mathematically by providing mathematics ac-           andmonitoroneanotherwhichmayleadtotheearlydetection
              tivities whose purpose goes beyond learning math^ (p. 103).          of errors (Lanzonder 2005; Witney and Smallbone 2011)and
              According to Pea, mathematics becomes functional because             contribute to each other’s knowledge, in turn enhancing un-
              technologies prompt the development of mathematical think-           derstanding and filling knowledge gaps (Manloveetal. 2006).
              ingasameansofsolvingproblemsratherthananendinitself,
              andstudents interpret the world mathematically in a problem-         EmbedOpportunities for Embodiment
              solving context. Pea suggested that these functional environ-
              ments also have a social aspect in which students are collab-        Embodimentisamethodthatallowsstudentstointeractwith,
              orating and engaginginacademicdiscourseinordertosolvea               experience, and learn from authentic situations (Stoltz 2015).
              problem and to motivate mathematical thinking.                       Dr
                                                                                      awing on theories of embodied cognition, the embodiment
                 Functional environments typically involve an authentic            method emphasizes creating perceptual and bodily experi-
              problem in which learners contextualize learning around a            ences for students to learn abstract concepts by experiencing
              complex, open-ended problem. These types of problems are             those concepts in context and interacting with the real world
              often design-focused and are typically seen in settings such as      aroundthem(Han2013).Itusesperception,action,andphys-
              engineering (Kapur 2008). Such problems create an opportu-           ical movement to externalize and make visible mathematical
              nity for learners to engage in productive failure. Productive        thinking and problem-solving processes (Alibali and Nathan
              failure is the knowledge that forms when a learner attempts a        2011; Daily et al. 2015;Han2013).
              problem, fails, and has to construct a new potential solution           In the case of robotics education, both the student and the
              based on the failed results (Tawfik et al. 2015). The failure is     robot can serve as an embodied agent. As students program,
              productive in that the thinking and problem-solving processes        test, and retest their solution strategies using the robot, they
              developed by learners during this experience become poten-           canseeaconcreterepresentationoftheirthinking(Han2013).
              tially transferable to novel situations in the future (Kapur         Students can also act out the movement of a robot, relating
              2008). Theproblemsandinstructional structures that promote           aspects of computational thinking and programming to their
              productive failure are complex, utilize prior mathematical un-       ownphysicalbody.Boncoddoetal.(2010)foundthatlearners
              derstandings, develop new mathematical understandings, are           were better able to comprehend STEM concepts when they
              34                                                                                                    J Form Des Learn (2017) 1:31–44
              imagined themselves as robots and moved their own bodies             positively impact students’ motivation and attitudes to learn
              according to the given instructions; this eventually helped          STEM-relatedcontent(Stohlmannetal.2012;Meyrick2012).
              them understand how the robot moved when programmed.                 To bring the benefits of an integrative STEM curriculum to
                                                                                   students, however, teachers need a Bcoherent curricula in
                                                                                   which computational thinking, programming, and modeling
              Integrate Multiple STEM Standards                                    arenottaughtasseparatetopics,butareinterwovenwithlearn-
                                                                                   ing in the science domains^ (Sengupta et al. 2013,p.353).
              STEM education is increasingly seen as an integrated ap-
              proach to teaching and learning in which science, technology,
              engineering, and mathematics disciplines come together to            EDRPhase2:DesignandConstruction
              address content standards in a unified and cohesive manner
              (Brownetal.2011).Inthisintegratedapproach,educators are              In the second phase of our EDR mesocycle, a curriculum
              expected to teach across the four subjects rather than teach         called, Danger Zone: A STEM-integrated Robotics Unit,was
              them separately (Capraro and Han 2014; Capraro et al.                developed using the conceptual framework from phase 1. The
              2013;Morrisonetal.2015).AnintegrativeSTEMcurriculum                  curriculumwascomprisedofaseriesofsixlessonsdeveloped
              would therefore make connections between the STEM sub-               for the fifth-grade classroom. These lessons met a variety of
              jects; it would teach students to apply their integrative knowl-     objectives that spanned engineering, math, and science stan-
              edge to solve a real-world problem in an authentic situation         dards;lessonsweredevelopedtotaketen50-minclassperiods
              using hands-on, technological tools, equipment, and proce-           to complete. Table 1 contains the complete objectives by les-
              dures in innovative ways (Wang et al. 2011).                         son. Several objectives aligned with multiple STEM disci-
                 Research has shown that using an integrative approach to          plinesandspannedmultiplelessons,includingapplyingmath-
              making connections between STEM subjects offers several              ematical concepts (i.e., decimals, coordinate algebra) to the
              benefits for students. First, an integrative approach enables        programmingoftherobot,engagingintheengineeringdesign
              learners to gain authentic problem-solving experience                cycle (i.e., identify constraints, generate potential solutions,
              (Furner and Kumar 2007). Through this authentic problem-             test and revise solution based on results), and drawing on
              solving approach, students are able to improve higher order          science knowledge to understand the problem and solution.
              thinking skills and problem-solving abilities (Stohlmann et al.         Thesixlessonswereorganizedsuchthatstudentswerefirst
              2012). A true integrative STEM curriculum can help students          introduced to an authentic science context for solving a rich,
              become better Bproblem solvers, innovators, inventors, self-         open-ended problem using the robot (lesson 1). The lessons
              reliant, logical thinkers, and technologically literate^             thenledstudentsthroughbuildingandprogrammingtherobot
              (Stohlmann et al. 2012, p 29). Another benefit of integrated         (lessons2and3),exploringpotentialsolutionstotheproblem,
              STEM is that teaching science and math together can                  and finally using mathematics to generate and present a final
              Table 1   Lesson objectives for danger zone curriculum
              Lesson                       Objectives
              1: Danger zone               ●Identify problem goal, constraints, and possible solutions
                                           ●Explore the science content of the task
                                           ●Explain the steps in the engineering design process
              2: Build-a-bot               ●Construct a robot for the given task
                                           ●Identify the mechanical components of the robot
                                           ●Define the role of the central processing unit (CPU)
                                           ●Explain the difference between input and output devices
              3: Primary programming       ●Actoutthebasic programming commands
                                           ●Programtheir robot to follow basic commands
                                           ●Applythemathematical concepts of fractions and decimals to their programming
              4: Purposeful programming    ●Further examine science content that will impact programming (i.e., specifics types of volcanic terrain)
                                           ●Applythemathematical concepts of decimals, measurement, and coordinate algebra to their programming
                                           ●Engageintheengineering design process to program and navigate their robot (e.g., plan, test, evaluate, and revise)
              5: Prime optimization        ●Usemathematics (decimals, measurement, and coordinate algebra) to optimize programming and planning
                                           ●Engageintheengineering design process to program and navigate their robot (e.g., plan, test, evaluate, and revise)
                                           ●Determine their best problem solution
              6: Share                     ●Share their results with peers
                                           ●Explain and justify their approach to solving the problem
                                           ●Engageinacademicdiscussions around programming challenges