This research paper discusses pedagogical and instructional approaches for K-12 education. As demand for engineers and computer scientists continues to increase, it is important to nurture a passion for STEM in high school students. However, a significant obstacle to contend with is low student motivation in rural and economically disadvantaged areas, which in turn may cause a lack of student success. In this paper, we examine students' motivation by using a STEM curriculum in which students from rural and economically disadvantaged schools are taught remotely using Zoom and a telepresence robot. Over the course of 11 weeks-approximately one semester-a curriculum that included a mixture of programming, engineering, and manufacturing was taught to 11th and 12th grade students in three different classes. Remote teaching was done by college students with a background in engineering (e.g., student instructors). These schools were used because of their remote location and underrepresentation in research. To measure student motivation in a teaching environment that uses telepresence robotics and Zoom, researchers collected the following data: a daily class progress document filled out by student instructors; weekly discussions between student instructors and researchers; and the submissions of student projects. Data on student instructor motivation was collected with surveys that had a mix of Likert scale and free response questions. The ratio of phrases about amotivation to phrases about motivation of the students was collected from surveys, discussions, and progress documents and analyzed. We found that the three classes exhibited varying levels of interest in the program, and that differences in class motivation fluctuated over the course of the semester. Classroom motivation may be influenced by the curriculum or technical issues unique to a remote course with co-located students. Finally, we found that the traditional relationship between student motivation and instructor motivation is not disturbed even though the students were co-located and the student instructor was remote.

This research-to-practice paper investigates how undergraduates organically transition into researchers through the lens of Legitimate Peripheral Participation, Legitimate Peripheral Participation is a contextual learning phenomenon through which new members become core members of a community. These peripheral members begin with the execution of peripheral low-risk and simple activities, gradually accruing knowledge and experience as they take on more complex and essential activities. They gain legitimacy and access to more complex involvement through interaction with, and mentorship by, senior members of the community. Our research examines how seven undergraduate students, who are involved with a research lab, transition from peripheral roles to core membership through observation and peripheral engagement, skill development, community integration, increasing responsibility, and collaborative projects. These undergraduates interact with and are mentored by both graduate students and faculty members in an organic fashion, with no assigned mentorship roles existing between the undergraduate students and graduate students or faculty. The focus of our research is not on how one recruits undergraduates to work in a research lab, but on how the identity of an undergraduate as a researcher, both in their own mind and in those of their coworkers, is nurtured and developed. Through the process of surveys and interviews we extract the core elements of the undergraduates' experience and explore factors that might contribute to or hinder their transition into being more active researchers. The degree of involvement and identification as a researcher varied significantly between the undergraduate students. The key elements we discover at social engagement, physical presence, and near-peer interaction.

Instructional discourse around physically-based material typically involves the use of gestures to focus attention in the joint visual space of the instructor and the student. Such gestures are performed in conjunction with speech, thus situating the discourse in the space of what the student is learning or doing. This kind of situated discourse is missing in distance learning. The lack of situatedness between conversants contributes to communication inefficiencies, as was recently highlighted when classes began using Zoom during COVID. For instruction that involves hands-on learning, the inability for instructors to index the student's space and vice versa is a more significant problem. The use of telepresence technologies allows this spatial gap to be bridged but can often leave a communication gap. Little evaluation has been made into how these mechanisms alter gesture and speech. This paper explores how giving the instructor situated gestural access to the student's workspace in distant learning settings alters gesture and speech, with the goal of guiding design of future systems that are mindful of how these systems work, not simply whether they work. Our research is based on a system that supports situated instructional discourse through augmented reality gesturing. The instructor is able to see the student's workspace and perform gestures that are superposed into the student's workspace. This shared gestural space is used in tandem with a traditional video chat. Two spatially-focused tasks were examined: a two-dimension problem that required varying levels of assistance, and a three-dimension problem with complete guidance by the instructor needed. We evaluated how gestures were used to augment verbal communication and the types of gestures and verbal utterances used. We found that static pointing G-hand deictic gestures were more common in the augmented reality mode than in the co-situated mode and complex spatial structuring gestures (e.g. orientation gestures) occurred more often in the co-situated mode than the augmented reality mode. Additionally, students provided a range of body postures and working orientations to encourage or discourage invasion of their personal workspace in co-situated settings, and several subjects indicated that the augmented reality made them feel more comfortable with sharing their gestural space. These results provide a framework to guide the design of future improvements of telepresence systems. Subjective statements of comfort by subjects also suggest this may be a profitable avenue for exploration with students who are reluctant to engage physically with instructors.

“What are we learning this for?” is a common refrain heard in education. Showing how engineering practices are relevant to everyday needs like building science models in K-12 classes may lead to greater motivation for students in STEM learning. This paper investigates how horizontal integration of different streams of learning may enable students to see why and how each element of knowledge is relevant to the whole. By combining microcontroller-based electronics, programming, and 3D design and fabrication to build science models, students may develop a stronger appreciation of how the component technologies interrelate (internal contextualization) and how these integrated computational models contribute to the enterprise of science (external contextualization). Furthermore, students in science classrooms see how computation can make physical models dynamic to represent changes and relationships more effectively. The implicit representational science models and the incorporation of technology and computation into a regular science curriculum make the knowledge transferable beyond an individual lesson. Students learn the component technologies cumulatively through a sustained process of engagement across multiple science modeling projects. The major challenges of this approach are how the science modeling projects may be designed and what factors have to be considered in the designs to realize this learning progression. To answer these questions, this paper describes a rigorous design process for the Physical Computational Models (PCMs) that incorporates science model correctness, progression of learning across all streams, pedagogical efficacy, classroom dynamics and manufacturability. In addition, we implemented a year-long project using our Horizontal Learning (HL) approach at two local public schools. This paper reports in detail two PCMs we developed for authentic 5thand 6thgrade classrooms to illustrate the integration of the learning streams in science, computing, digital electronics, and 3D design and fabrication. By analyzing three lessons we developed, the result demonstrated that our approach did produce the incremental learning progressions characterized by two primary principles in these streams across multiple HL interventions.

Multimedia approaches retain a powerful ability to convey numerous streams of information simultaneously. When used appropriately, the uptake and comprehension of these multiple streams can be accomplished with relative ease, despite multimedia approaches typically conveying more information than their single medium counterparts. Given their power, this paper investigates how multimedia approaches can be leveraged to convey spatially arranged text. We do this with a special eye towards Individuals with Severe Low Vision, a group often forgotten in discussions of multimedia and who have a much-reduced capacity to receive information via the visual medium. We accomplish our investigation by comparing two tools for reading mathematical expressions, a form of spatially arranged text. The first is MathSTAAR, a novel system which treats mathematical text as a form of interactive multimedia. The second is JAWS, a screen reading tool which provides access to mathematical text via audio only. Researchers found that the multimedia approach reduced the amount of time it took for participants to relocate mathematical content while also reducing Frustration and Effort.