Publications
Teacher dashboards in mathematics classrooms tend to provide teachers with information on student performance that are often linked to classroom management systems, online course systems, or peer-tutoring software. Teacher dashboards also tend to emphasize features that support teachers using a transition or direct instruction model. In our approach, we iteratively designed, developed, tested, and refined a teacher dashboard that is linked to a student digital collaborative environment with an embedded problem-based mathematics curriculum. In this study, we investigate teacher dashboard features that support teacher enactment of problem-based mathematics curriculum embedded in a digital collaborative platform. We report on design principles that guided the development of three teacher dashboard features: (1) monitoring evidence of student thinking in real-time or after class, (2) accessing workspace for whole-class discussions of the problem, and (3) creating and sending just-in-time supports. The pedagogical advantages and challenges teachers face throughout the iterative development process are also discussed. Evidence from observational data and teacher interviews suggests that the organic synergism generated from the student and teacher digital platform offers several ways that teachers are provided with new and timely information from teacher dashboards that supports problem-based mathematics teaching.
Game-based learning environments are designed to provide effective and engaging learning experiences for students. Predictive student models use trace data extracted from students' in-game learning behaviors to unobtrusively generate early assessments of student knowledge and skills, equipping game-based learning environments with the capacity to anticipate student outcomes and proactively deliver adaptive scaffolding or notify instructors. Reflection is a key component of self-regulated learning, and it is critical in effective learning. However, there is currently limited work exploring the utility of reflection for inducing accurate predictive student models. This article presents a predictive student modeling framework that leverages natural language responses to in-game reflection prompts to predict student learning outcomes in a game-based learning environment for middle school microbiology, CRYSTAL ISLAND. With data from a pair of classroom studies involving 118 middle school students, we investigate the accuracy of early prediction models that utilize features extracted from student trace data combined with word embedding-based representations (i.e., GloVe, ELMo) of student reflection responses. We evaluate the accuracy of the predictive models over time using data from incremental segments of each student's interaction with the game-based learning environment, and we compare against models that omit student reflection features. Results reveal that models encoding students' natural language reflections with ELMo word embeddings yield significantly improved accuracy compared to other representations, with the greatest accuracy demonstrated by an ensemble of predictive models. We discuss the implications of these results for the design of game-based learning environments.
Combinatorial proof is an important topic both for combinatorics education and proof education researchers, but relatively little has been studied about the teaching and learning of combinatorial proof. In this paper, we focus on one specific phenomenon that emerged during interviews with mathematicians and students who were experienced provers as they discussed and engaged in combinatorial proof. In particular, participants used a wide variety of cognitive models to interpret multiplication by a constant when reasoning about binomial identities, some of which seemed to be more (or less) effective in helping produce a combinatorial proof. We present these cognitive models and describe episodes that illustrate implications of these cognitive models for our participants' work on proving binomial identities. Our findings both inform research on combinatorial proof and highlight the importance of understanding subtleties of the familiar operation of multiplication.
This paper addresses two aspects of integrating mathematics education with engineering education that may address persistence of engineering majors (and STEM majors more broadly): an emphasis on modeling as a vehicle for more authentic learning activity (Niss et al. 2007), and the need for measures that can support academic units' efforts to collect local data about student attainment of program goals. In this paper, we contribute: (1) a measure for modeling self-efficacy and its corresponding design process; (2) a measure for modeling competency and its corresponding design process; (3) a preliminary analysis of the relationship between modeling competency and self-efficacy. We argue that such instruments address a genuine need of engineering departments (as well as STEM education researchers) to have a means for collecting local data on students' modeling self-efficacy and competency.
Computational thinking and activity are vital aspects of what it means to conduct scientific and mathematical work. In light of this, some propose that students' mathematical education should include an integration of computing into their mathematical experiences, giving students opportunities to engage with computational tools as they reason about mathematical concepts. In this commentary, we make a case that the international RUME community should focus on studying the integration of computing in research in undergraduate mathematics education. We situate this discussion within existing literature. Then, we suggest ways in which researchers can incorporate ideas related to computing, and we propose ideas for how investigations into computing might practically be incorporated into our already-existing research foci. Ultimately, we hope to motivate other members of the RUME community to join us in what we consider to be a timely and exciting endeavor.
A growing body of work has shown that two specific study strategies help explain differences in learning and achievement in gateway courses: spacing (breaking up study sessions across multiple days) and self-testing (actively recalling information from memory). However, it is still unclear whether the benefits of these strategies are applicable in more advanced biology courses, and whether promoting effective study practices in these courses (spacing and self-testing) is related to increased use of these practices and greater learning outcomes. We studied two senior-level microbiology courses that were taught by the same instructor. Using a quasi-experimental design, one course additionally received a light-touch study skills intervention, where the instructor introduced the concepts of spacing and self-testing while also providing reminders to students about utilizing these strategies. We found that, while the intervention was not related to increased use of spacing and self-testing, both strategies were positively related to learning, as measured by the final course grade. Results from multiple regression analyses revealed that engaging in spacing throughout the course was the most consistent predictor of final course grade, even after accounting for other study strategies, demographic characteristics, and prior academic achievement. Our results add to the literature emphasizing the importance of spacing in increasing students’ achievement in STEM courses. © Springer Nature Switzerland AG 2021.
EQUIP is a free, customizable observation protocol for tracking patterns of student participation in STEM classrooms (https://www.equip.ninja). EQUIP generates data analytics that are disaggregated by student social markers (e.g., race, gender), which makes it a useful tool for tracking patterns of inequity in student participation. However, prior studies have not yet established how many observations are needed to create a representative picture of instruction. In this study, we use g-theory and simulations with Cramer’s V to analyze observations from 20 undergraduate mathematics instructors to determine how many classroom observations are needed, and how this differs by individual codes. We found that Gender could achieve stability in just a few observations, whereas codes such as Instructor Response, Instructor Solicitation Type, and Instructor Solicitation Method required nearly 20 observations. Thus, we recommend that users account for their specific context and needs with EQUIP when determining the ideal number of observations to conduct, using this research as a baseline. We also compare the g-study and simulations approaches, bringing up new methodological questions for the field. © The Author(s), under exclusive licence to Springer Nature Switzerland AG 2021.
This quantitative study examined student participation in an introductory project-based engineering course offered in fully face-to-face and hybrid course modes (N = 160). This course attempted to counteract trends of decreased student motivation and high attrition rates among engineering majors. Mixed-design analysis of variance examined differences in motivational constructs including student self-efficacy, effort regulation, and interest in engineering, as well as engineering skills throughout the course and across instructional modes. None of the motivational constructs were associated with significant decreases throughout the course nor with differences across instructional modes. However, students’ engineering skills increased throughout the course with no significant differences across course modalities. Furthermore, interest in engineering and effort regulation were positively associated with course performance. The instructional modality was not significantly associated with course performance. Overall, this study provides an example of a project-based introductory engineering course which may help maintain student motivation and foster student success in engineering. © The Author(s) 2021.
Beliefs play a central role in human development. For instance, a growth mindset-a belief about the malleability of intelligence-can shape how adolescents interpret and respond to academic difficulties and how they subsequently navigate the educational system. But do usually-adaptive beliefs have the same effects for adolescents regardless of the contexts they are in? Answering this question can reveal new insights into classic developmental questions about continuity and change. Here we present the Mindset x Context framework and we apply this model to the instructive case of growth mindset interventions. We show that teaching students a growth mindset is most effective in educational contexts that provide affordances for a growth mindset; that is, contexts that permit and encourage students to view ability as developable and to act on that belief. This evidence contradicts the beliefs alone hypothesis, which holds that teaching adolescents a growth mindset is enough and that students can profit from these beliefs in almost any context, even unsupportive ones. The Mindset x Context framework leads to the realization that in order to produce more widespread and lasting change, we must complement the belief-changing interventions that have been aimed at students with new interventions that guide teachers toward classroom policies and practices that allow students' growth mindset beliefs to take root and yield benefits.
Introduction: Experiencing low support from teachers can be associated with low academic achievement. Nonetheless, individual- (i.e., ability self-concepts) and contextual-level (i.e., parental support) protective factors may help adolescents to display academic resilience. This study examined whether high school students? math ability self-concepts and parental support can mitigate the possible negative association between perceived low math teacher support and their math achievement. Method: Correlational data were drawn from the High School Longitudinal Study (N = 14,580, Mage =17.42 in 11th grade, 51% female), a nationally representative study of high school students in the U.S. The measures of protective factors (i.e., math ability self-concepts and parental support) were obtained from the surveys administered to students and parents in 9th grade. Students? perceived teacher support and their math achievement score were measured in 11th grade. A series of linear regression analyses were estimated to test our hypotheses. Results & Conclusions: Perceived low teacher support was negatively associated with adolescents? math achievement. Adolescents? math ability self-concepts were directly and positively associated with their math achievement. The interaction between perceived low teacher support and ability self-concepts in predicting adolescents? achievement varied by parental support. The association between perceived low teacher support and adolescents? math achievement was not statistically significant when adolescents were high on one of the protective factors. That is, high parental support may be protective for adolescents with low math ability self-concepts. This study highlights the interaction between adolescents? academic motivation and parental support in demonstrating resilience to perceived low teacher support.


