Abstract
Educational robotics engage students in an integrated STEM approach that helps students understand STEM concepts as well as increase positive perceptions of STEM subjects from an early age. When the COVID-19 pandemic broke out, physical robots in a face-to-face classroom became an impossibility. A virtual robot program was quickly developed to function with a familiar coding platform to provide students and teachers with an alternative robotic solution that could be used from anywhere. In this paper, the usage data from over a million students globally will be interpreted alongside two teacher case studies. This combination of data provided insights into the virtual robot as a learning tool, as well as a teaching resource. The teacher case studies also revealed a set of critical needs that facilitated teaching in such unpredictable circumstances. Finally, this data indicates that the virtual robot learning environment could be used as a symbiotic compliment to a physical robot to help students gain confidence with iterative programming, increase excitement for educational robotics, and provide teachers with a highly flexible teaching option moving forward.
Keywords
Virtual robot, educational robotics, teaching robotics, COVID-19 solutions, STEM education, computer science, programming
Introduction
Robotics and computer science have become increasingly integrated into primary and secondary school (kindergarten through 12th grade) in the United States in recent years, spurred by national reports and policies. In 2015, the National Science Foundation stated that the acquisition of science, technology, engineering, and math (STEM) knowledge and skills is increasingly vital to Americans to fully engage in a technology-intensive global economy, that it is critical for everyone to have access to high quality education in STEM topics. The National Science and Technology Council’s Committee on STEM Education put forth a report in 2018 to outline a federal strategy for STEM education. This report notes that, “The character of STEM education itself has been evolving from a set of overlapping disciplines into a more integrated and interdisciplinary approach to learning and skill development. This new approach includes the teaching of academic concepts through real-world applications and combines formal and informal learning in schools, the community, and the workplace. It seeks to impart skills such as critical thinking and problem solving along with soft skills such as cooperation and adaptability.” This national focus on STEM learning has been accompanied with increased research and innovation in educational settings on how to better incorporate technology into the classroom for STEM topics.
Robotics provides a hands-on way for students to explore STEM concepts. Basic STEM topics are important topics in primary and secondary education, as they are essential prerequisites for advanced college and graduate study as well as increasing technical skills in the workforce1. A metanalysis2 revealed that generally, educational robotics increased learning for specific STEM concepts. Studies across many age groups revealed that robotics increases student interest and positive perceptions of STEM subjects3, 4, 5, which in turn increases school achievement and furthers science degree achievement6, 7, 8. For high school students, robotics has been used to support college preparedness and technical career skills9, 10, 11, while robotics has been introduced to elementary school students to develop inquiry and problem-solving skill, and foster positive perceptions of STEM topics12, 13. Introducing educational robotics has been especially beneficial to young students, who can begin to form negative attitudes toward STEM subjects as early as 4th grade14. Young students benefit from an integrated learning context and develop more positive attitudes toward STEM subjects with early experiences of success15.
Research has also shown that introducing robotics during teacher pre-service education increased teacher self-efficacy, content knowledge, and computational thinking skills16. While logical that the benefits of robotics would be found in teachers as well as in students, the introduction of robotics in formal teacher education is still limited. In many countries, traditional teacher education focuses on discipline-based topics in science and math, leaving most teachers underprepared in engineering and technology17 and less confident teaching STEM topics not covered in formal teacher training or making connections across STEM disciplines18, 19. Bybee20 noted that this limitation of STEM topics in teacher education leads to an underrepresentation of engineering and technology, particularly in K–8 education. While the benefits of including robotics in teacher education is clear16, an alternative could be achieved through continuing professional development and informal learning through communities of practice. Bandura21 expressed the critical aspect of social learning contexts, and from that concept Lave and Wenger22 outlined the concept of communities of practice (CoP). For a CoP, members gather around a shared interest in a domain, develop a community, and share research and insights to further skills and knowledge—developing a practice22. In lieu of robotics in formal teacher education, informal learning and CoPs could provide similar benefits to teachers, and furthermore, to students.
Unfortunately, the COVID-19 pandemic caused widespread global disruption to in-person learning, affecting nearly all students worldwide23. Hands-on learning experiences were suspended, which was a foundational portion of most robotic STEM curriculum, including the robotic curriculum used by the VEX educational robotics line. Remote learning solutions were needed to quickly provide a virtual learning environment that could still help students engage with STEM topics in an authentic, meaningful way. VEX Robotics quickly created VEXcode VR (hereafter simply referred to as “VR”), a platform with a virtual robot that could be used in similar ways as a physical robot.
This paper will review the usage data collected by the VR platform to gain insights into how this virtual substitute was during this global disruption. Two case studies will also be presented which provide context for how teachers implemented VR in their remote learning environments. The two primary research questions for this paper are as follows:
- What insights can usage data and teacher case studies reveal about student learning with VR following the COVID-19 outbreak?
- What insights can teachers provide on the implementation of VR into the classroom?
The chaos sewn by COVID-19 was particularly felt by educators. Decades of experience and lessons designed for in-person learning was instantaneously upended, yet this disruption also encouraged educators to experiment with new tools and teaching methods. Understanding the decisions made and outcomes achieved from the perspective of the educators who led through innovative solutions can provide insights into how to incorporate new technology to strengthen student learning in robotics and STEM subjects moving forward.
Methods
VEXcode VR. When schools in the United States closed in March of 2020, a solution was needed that could keep students engaged with robotics and STEM topics while working remotely. VR was developed and launched on April 2nd of 2020, mere weeks after most schools went to a virtual format. VR Activities were created to be consistent with the other robotic curriculums with interdisciplinary lessons aligned to content standards. The VEXcode VR coding platform is the same as the coding environment students would normally use with physical robots with the addition of the virtual interface, as seen in Figure 1. In lieu of a physical robot, students create projects to control a virtual robot in a thematic “playground” that changes based on the activity. Beginning coding students use blocks-based programming, and advanced students use text based on Python.
Figure 1. The VEXcode VR platform interface for the Coral Reef Cleanup Activity.
VR activities were created to be interdisciplinary, combining the computer science skills that are foundational to controlling a virtual robot with topics from science or math. Over the course of these VR activities, students not only learn about programming, but also scientific inquiry, mathematical thinking, and technical literacy—all components of an integrated STEM framework19. The unique circumstances brought about by COVID-19 required that students be able to work through lessons independently in blended, synchronous, or asynchronous settings. To accomplish this, students are introduced to learning objectives and the goal for the activity. Direct instruction is then used to provide step-by-step instruction and intentional scaffolding to sequence learning for understanding24, 25. Students then receive targeted scaffolding leading up to solving the final coding challenge26. Students learn that robotics and coding are used to solve practical, interdisciplinary problems. For example, in the Coral Reef Cleanup Activity, students are challenged to navigate their robot around a coral reef to collect as much trash as possible before their solar-charged battery dies. Pollution is a global problem that will be solved by tomorrow’s students, and engaging in these authentic, scenario-based projects help students apply computer science skills across disciplines.
Figure 2. The mission context for the Coral Reef Cleanup Activity.
Considering that students are separated from their instructors, the virtual environment needed to be as seamless as possible to reduce split attention and cognitive load27, 28. Students can drag and drop commands into their project, and watch their robot navigate the VR playground in the same window. Students can add any number of blocks at a time, running the project after each addition, to see how their robot moves in the playground. This provides students with immediate feedback and early feelings of success.
Additionally, remote learning created practical hurdles that VR needed to overcome. School computers often have restrictions for downloading applications, causing the addition of a program to be a hurdle in the most normal of circumstances, let alone when students are remote with school computers. But students may not even have access to school computers to do their work. To maximize the access to VR, the program was built to be entirely web based (no downloading or plugins required) and to run on many different types of devices to increase the likelihood that students would be able to use it.
Results
Usage Data. The data presented is provided by Google Analytics. As VEXcode VR is entirely browser-based, there are a number of different metrics which provide insight into how this virtual robot environment has been used globally. Since its launch in April 2020, there has been an increase in VR users monthly, that have combined to over 1.45 million users in more than 150 countries.
Figure 3. The countries with VR users globally.
Given the timeline of COVID-19 and the VR release, we also reviewed usage over time. As shown in Figure 4, the numbers of users rose quickly shortly after the release, then decreased during the summer months when students were out of school. The typical return to school months (August/September) saw a significant increase which persisted the rest of the school year. The periodic drops in the number of users indicate less usage on the weekends and over holiday periods.
Figure 4. The number of users over time since the launch of VR.
A project is a program that students create for a lesson or challenge. Projects do not have to be saved in order to run, but a saved project is downloaded for a user to come back to at a later time. There were over 2.52 million saved programs. However, a project does not have to be saved in order to be run. Because VR is entirely browser based, editing a project and testing it happens immediately by selecting “START.” There have been more than 84 million project runs in the software, indicating that students tested their projects at frequent intervals. Due to this immediate feedback loop, students had the opportunity to experiment and iterate at a much faster pace compared to working with a physical robot. This iterative process is a good indication for student learning as multiple iterations have been shown to maintain student engagement and interest29.
| VEXcode VR Data | |
|---|---|
| Users | 1,457,248 |
| Saved Projects | 2,529,049 |
| Run Projects | 84,096,608 |
| Countries | 151 |
Table 1. All VEXcode VR usage data from April 2020 to April 2021.
Certification Data. In addition to the VR program itself and the curriculum to accompany it, VR includes a free teacher training called CS with VEXcode VR Educator Certification Course. Since its launch in June of 2020, over 550 educators have completed the certification, which contains over 17 hours of curriculum and support, to become a VEX Certified Educator. The certification course contains 10 units of material aimed at preparing teachers who may have no experience with computer science or robotics. The content spans topics such as the basics of programming, how to code the VR robot, how to teach with the VR activities, and how to implement VR in a classroom. Figure 5 shows both the number of certified educators monthly and cumulatively from June of 2020 until March of 2021. Trends in the data show an increased number of certified educators around back to school time, which includes August and September and into October of 2020.
Case Study 1
Aimee DeFoe is the principal at the Kentucky Avenue School, a small private school in Pittsburgh, USA, that combines traditional and innovative teaching and learning methods. Like most schools, the Kentucky Avenue School was disrupted by COVID-19 and had to identify alternative plans for the start of the Fall 2020 school year, not knowing how circumstances would change. The first six weeks of the year were taught entirely virtual, and the remaining year was spent in a hybrid format with student cohorts alternating days of in-person and remote instruction. Even when students were learning at home, it was crucial that students continue to engage in the same problem-solving and critical thinking activities as in the classroom setting.
Aimee chose to use VR with her sixth and seventh grade students for several reasons. As VR was an entirely virtual learning environment, students would be able to switch between home and school without changes in policy impacting their learning activities. The block-based coding environment would not be intimidating for students new to coding and there were activities designed for differing experience levels. She also believed that students would find the VR robots exciting and motivating—which she found to be true. When reflecting on what she hoped students would get from VR, Aimee stated:
I was hoping that using VR would be every bit as rigorous, challenging, and exciting as using physical robots, and that my students would not feel like they were missing out on an experience, but rather gaining a new kind of coding experience that was just as exciting. I wanted them to feel the same kind of accomplishment they would have felt in the classroom when they have to reiterate and persist through challenges and then finally achieve success.
As the only robotics teacher, Aimee taught 23 students once a week between the start of school and winter break, for a total of 15 lessons. Students began with the “Computer Science Level One - Blocks” course. Aimee worked through the first unit with students as a group, but for the remaining lessons let students work at their own pace and acted as a facilitator. Most students finished between seven and nine units, with the additional ocean cleanup activity.
Aimee found that students were very motivated by the challenges in the lessons; so much so that it was sometimes difficult getting them to work through the lesson systematically. Some students who struggled with attention or reading needed additional support, and the greater than/less than and Boolean concepts to be challenging. However, most students had the right amount of challenge, struggle, and success. Students were excited at the idea of working with physical robots when returning to class. After working with VR, Aimee noted, “Everyone left the class as a more confident coder, without a doubt.”
Case Study 2
Mark Johnston teaches seventh and eighth graders at Bel Air Middle School in El Paso, USA. For his STEM 1 course, Mark teaches Project Lead the Way Gateway courses on Automation and Robotics, and Design and Modeling to approximately 100 students. The STEM 1 course incorporated the VEX IQ robot to teach basic mechanics and foundational coding with VEXcode IQ (a plastic robot kit for younger students). This course is taught in the fall semester, so the initial COVID-19 disruption did not impact his robotics in the spring. However, in April of 2020 Mark saw the VEX VR robot and began to work with it. “When I saw that VR was using the same setup (ie. VEXcode), I was super excited because I saw the potential—like a puzzle piece I KNEW would fit perfectly with what I was already doing. When VR was updated to include Python, I was even more excited.” Mark created tutorial videos for other teachers, gathering a large following on social media platforms. Through his own nonprofit educational company, Mark offered a free summer camp for students on VR, in addition to teacher training in preparation for the 2020/21 school year.
Uncertain teaching circumstances make it difficult to plan. “When I realized the distance learning would continue into the 2020/21 school year, I decided to teach design first and then robotics… but so many things were up in the air, it was hard to plan anything. I didn’t know if we would be back in person or continue online—very little information was clear at the time. I ended up just mixing robotics and design together and just planned one or two days in advance.” Mark began using VR at the start of the school year (which would remain 100% remote until 2021) by picking different activities from the site, which worked well because there were different experience levels and editable instructions. When the Computer Science Level 1 - Blocks course was released, he walked students through it in its entirety, though noted that next time he’d distill lessons into shorter lectures. Using VR was inherently different than the in-person robotics lessons, but there were still a set of key goals Mark had for these lessons:
- Get students familiar with VEXcode
- Build confidence in programming (self-efficacy)
- Introduce programming ideas/vocabulary in a non-threatening way
- “Trick” them into using math without realizing it ;)
- Ask students to solve well-defined problems given constraints
- Introduce ill-defined problems
- Encourage a “fail and try again” attitude
- Keep problem solving fun
While a virtual experience was different, Mark found distinct advantages to using VR. Students were much less afraid of experimenting using VR versus RobotC (another coding language used with other robots). Mark also uses a measurement of how long it takes students to get a “win” to determine how good a STEM activity is, noting that, “if it takes too long for the student to get a positive result, it’s much harder to keep them engaged.”
There was an immediacy to VR that encouraged exploration and active engagement. Mark describes this type of “win” with an example of introducing VR to students:
Me: “Everyone open a new tab and go to vr.vex.com. Everyone see the site? Good. Now make the robot drive forward.”
Student: “How?”
Me: “See if you can figure…”
Student: “I figured it out!”
And then they are hooked! By that time, many of them are asking me how to do all kinds of different things. They are literally asking me to teach them!
Results and Discussion
VR as a Learning Tool. The usage data and case studies both provide insights into the first research question on how VR worked as a learning tool during the COVID-19 pandemic. The simplest takeaway is from the sheer volume of usage; the VR platform was used more than a million students around the globe, indicating the virtual robotic environment functioned well as a substitute for in-person learning during a crisis circumstance. The number of run projects (84+ million) was also a surprising finding when considering the number of individual users. On average, users were completing 57 project runs, showing a high degree of testing and iteration. This is a very promising result given the importance of developing a “try and try again” attitude in students. There are multiple possible ways to solve the VR activities, which is a critical lesson for students to learn. When students understand there are multiple solutions to a problem, there can be an increased likelihood that students will request feedback from teachers and also that they have higher comprehension of what they are learning30.
From the case studies, there is also confirmation that VR works as a low-stakes learning environment. Aimee noted that her students were more confident coders and were looking forward to working with the physical robots. Mark noticed that students were less afraid of experimenting as they coded in VEXcode VR and there was an immediacy to their sense of a “win” in this environment. When we consider these teacher observations in conjunction with the raw usage data, it seems to confirm that a virtual robot environment makes students feel freer to experiment and iterate during their learning process, and increases positive perceptions of robotics in general.
Lessons from Teachers. When we consider the second research question on what insights teachers can provide on the implementation of VR into the classroom, we can identify several commonalities from the case studies. Both case studies revealed information on how teachers made decisions and implemented solutions during COVID-19, but also on what was needed in order to provide an effective learning solution for students in a virtual and hybrid environment. These themes include flexible solutions, continuity, and curriculum and support. These findings should be considered as requirements for all technology solutions, as supporting teachers supports students.
Given the uncertainties around teaching conditions, both Mark and Aimee noted that they needed flexible solutions. Remote learning could change to face-to-face learning, or some form in between. VR could continue to be used in any setting, but also offered flexibility in its approach. Students could be engaged in structured teacher-led lessons as Mark used with the activities and course, or student-led learning at their own pace as Aimee described. Teachers also needed flexibility in the experience level, both in terms of activities and the type of programming languages offered to meet the needs of all students.
Continuity of learning was indicated as important in both case studies. Aimee noted that after working in VR, students were excited about getting to work with the VEX V5 robots that were waiting when in-person learning resumed. VR was serving as a steppingstone to working with physical robots, and increasing student excitement and positive perceptions. Mark also noted that the continuity of VEXcode from VR to IQ was very important to him: “I cannot tell you how awesome it is that VEX has a very simple to follow progression from 3rd grade to college all using VEXcode! And with VR, they can start learning it from home!”
Curriculum and support were clearly critical for the success of VR in this evolving teaching in learning situation. The VR units provided all the content for students to learn as well as the material needed to teach the lessons. Not all teachers have a background in computer science and coding. Aimee noted that the block-based program was also not intimidating for her in addition to her students. Mark also said that he was not accustomed to teaching as much computer science, and had to learn the lessons himself before teaching. However, Mark acknowledged, “If things were to go back to “normal” tomorrow, I will now be able to teach the programming portions of my class with more confidence.” Teacher support for the curriculum and programming of VR are vital for the implementation of VR in the classroom.
Digital learning is not just for students; teachers are also reaching out to learn about teaching practices and resources through technology and social media. Teachers in nearly 50 countries completed the VR certification. A global community of practice is forming around VR. Mark began posting videos on VR on social media and quickly had more than a thousand followers; through his work with VR, he made friends with teachers in Slovenia and Taiwan. As teachers share their experience and practice, students ultimately benefit from these informal teacher support groups. Communities of practice could provide a bridge between the current availability of educational robotics and the inclusion of this technology in formal teacher education. As more teachers become familiar with educational robotics through professional development, such as the 550+ teachers who completed the certification course, or through informal learning communities, more students will be introduced to integrated STEM learning.
Conclusion
VEXcode VR was created in a time of great uncertainty and great need for immediate solutions. Innovative solutions can come about from urgent situations. VR has touched more than 1.45 million users who saved more than 2.52 million projects and ran more than 84 million projects—in more than 150 countries. Even though the pandemic has impacted students and teachers around the world, VR has enabled students and teachers to engage with robotics and computer science concepts regardless of physical barriers. From the teacher case studies, themes of flexibility, continuity, curriculum, and support were identified as important to teaching with technology in such uncertain and challenging circumstances.
Moving forward from this unprecedented time, the lessons learned from the creation and implementation of VR indicate avenues for its use in the future. The usage data combined with the teacher case studies show that students felt less inhibited to iterate while coding in the virtual environment. This suggests that VR may be a valuable scaffolding tool that could be used in conjunction with physical robots. This is also supported by the need for flexibility; using VR as a learning tool in combination with a physical robot could provide an optimal, flexible robotic learning environment where an easy, at-home option supplements the in-person physical robotics curriculum. We look forward to future research to investigate how teachers could combine virtual and physical robotics in a post-pandemic world.
Acknowledgements
We gratefully acknowledge Aimee DeFoe and Mark Johnston for sharing their teaching experiences and valuable insights.