United States

Logistic regression and random forest classification models were used to predict students at risk of receiving a failing grade (D or F) in an introductory mechanics class (N = 2061 students). The dataset consisted of both institutional and in-class variables gathered from the first two weeks of instruction, with the in-class variables (homework and clicker scores) found to be particularly important for the model’s classification accuracy. Additional methods such as model regularization and upsampling of the failing class through the synthetic minority oversampling technique (SMOTE) were found to slightly improve model performance. The best performing models achieved a balanced accuracy of around 80%. An apparent performance ceiling of around 87% was found by providing the model with the course test average, indicating that substantially increasing performance above 80% is unlikely with data available early in the semester.

AAPT SM 22

Using Machine Learning to Predict Student Performance in Introductory Mechanics

This meeting is a great opportunity to explore sessions and workshops designed with your professional development in mind. Present your research and exchange ideas on the latest in teaching and pedagogy. Participate in a wide range of events, activities, plenaries, and so much more, connecting you with peers and colleagues who share your commitment to excellence in physics education.

AAPT WM 22

technical paper

AAPT meetings are a great way to reflect on your teaching, recommit to best practices and re-energize your career with new research, products, and techniques. Listen to internationally known plenary speakers, attend sessions, workshops, and special activities. This meeting is an excellent opportunity to share your expertise and learn from and network with leaders in the field.



In this study we examine the impacts of elements of an advanced physics lab on a student’s self-efficacy. Self-efficacy is a person’s belief in their ability to perform a certain task. This has been studied in introductory physics courses and introductory physics labs, examining changes in a student’s self-efficacy based on aspects of course design and student interactions. The advanced physics lab offers a new arena in which to investigate these changes. While students are all majoring in physics, there is significant variation between students in their confidence in their abilities. In addition, their experimental design and execution skills are typically not developed at the start of the course. We will discuss how the use of particular course elements change a student’s self-efficacy and make recommendations based on our results.

Development of Self-Efficacy in an Advanced Physics Lab

Some university physics students volunteer in informal physics programs like youth camps, demo shows, and public lectures. These experiences help physics student volunteers to build connections with faculty and their peers, connect their formal learning to the real world, and support their physics identity. These experiences can also influence the career pathways of their participants. Here, we examine the career experiences of alumni who had volunteered in informal physics as students and have since gone on to become informal physics program leaders. By looking at this subgroup, we hypothesize that program leaders had some impactful informal physics experiences that led them to careers in informal education. We can use this as an initial step to understanding the commonalities in experiences for volunteers more broadly. The resulting themes could be of use to practitioners in better understanding and improving their own programs.

Becoming an Informal Physics Program Leader: Experiences and Choices

The Brief Electricity and Magnetism Assessment (BEMA) is a multiple-choice instrument measuring a student’s conceptual understanding of electricity and magnetism. This study applied modified module analysis—partial (MMA-P), a network analytic method that identifies groups of responses using partial correlations, correcting for overall instrument score, to post-test BEMA data from one institution over many years (N =12,214). In module analysis, groups of correlated responses are called “communities.” As in previous applications of MMA-P to mechanics conceptual inventories, a number of communities related to physics concepts and some communities related to the structure of blocked items in the inventory were identified. This work supports previous results of an analysis of the Conceptual Survey of Electricity and Magnetism that suggest the existence of a rich collection of misconceptions and naive conceptions about electricity and magnetism which are consistently applied by students after instruction in introductory physics.

Network Analysis of the BEMA with Modified Module Analysis-Partial

Challenges and opportunities emerge in the volatile landscape of higher education. These challenges and opportunities motivate physics departments to pursue change efforts in their undergraduate programs. APS initiated the Effective Practices for Physics Programs (EP3) project to help physics departments learn and engage in effective change. Part of this EP3 project is the Departmental Action Leadership Institute (DALI) which is a year-long institute for physics faculty pursuing departmental change. Through their participation in the DALI, physics faculty learn about change practices and are offered ongoing guidance on how to create, lead, and sustain local Departmental Action Teams (DATs) around change efforts. In this talk, I will introduce the goals and structure of the DALI. I will also describe the EP3 research goals that aim to document the role of the DALI in supporting broad stakeholder engagement, problem-scoping, and effective use of data within local departmental change efforts.

A Leadership Institute’s Role in Supporting Team-based Departmental Change

The Effective Programs for Physics Programs (EP3) Guide is an online, living document, that serves as a resource for departments seeking to improve or make changes to their undergraduate physics programs. The guide is based on past successful practices, identified by practitioners and researchers in physics education. Understanding how departments interact with the guide is of major interest, both to the EP3 project and other initiatives that aim to create resources for physics programs. In this presentation, we describe and compare how two different physics programs structure team-based approaches to using the EP3 Guide. Through data gathered during team meeting observations and individual team member interviews, we showcase how different stakeholders, within each respective team, develop understandings of the guide and connect it to their local contexts. Our findings shed light on how these resources become embedded (or not) in departmental joint work and become meaningful to specific departmental actors.

Team-based approaches to programmatic resources: Who the Guide really guides

The impact of active learning courses on students greatly depends on how students work in groups. It takes more than placing students in groups for group work to be effective and inclusive. Instructors' knowledge, attitudes, and beliefs about group work has been shown to have a tremendous impact on students' experiences and their performance. As part of a larger project aimed at supporting faculty to foster effective and inclusive group work, we interviewed studio physics instructors about how they use group work in their courses. We also surveyed students in these instructors’ courses about their attitudes and beliefs about group work. We analyzed the relationship between the instructional practices used by instructors and the impact of these practices on students' learning and attitudes about the course. In this presentation, we will report preliminary findings comparing the interview and survey data with implications for instructors and researchers.

Supporting Physics Instructors to Facilitate Effective and Inclusive Group Work

Student-centered introductory laboratory and recitation sessions are frequently led by graduate teaching assistants (GTAs) who have varied teaching experiences. GTAs from chemistry and physics participated in a professional development intervention designed to prepare graduate students for active learning. The project included GTAs rehearsing pedagogical skills during four training sessions in a mixed-reality teaching simulator. In each session, groups of two or three GTAs taught a lesson to five avatar-students, observed the other GTAs in their session rehearse, and participated in group feedback discussions. We compare GTAs’ use of “Stretch-It”, a questioning technique, in disciplinarily-homogenous (Chemistry or Physics GTAs only) groups to disciplinarily-mixed (Chemistry and Physics GTAs) groups by comparing how GTAs used questioning within and across the simulator sessions. Then, we analyze the feedback discussions for two groups, one physics-only group and one disciplinarily-mixed group. Finally, we discuss benefits and drawbacks of this type of interdisciplinary collaboration.

Investigating the Effect of Discipline in Mixed-reality STEM GTA Training

Jupyter has become a widely used tool in recent years especially for computational physics. Nbgrader is an extension for Jupyter that assists in building, managing, and grading assignments using Jupyter. This talk will discuss specific ways I use nbgrader to improve assignments in computational and theoretical physics.

Improving Assignments using the Nbgrader Extension for Jupyter

Three interactive MATLAB Live Script tutorials in computational magnetism for first year university students are described. These tutorials explore magnetic fields of finite solenoidal and toroidal current-carrying helical coils, permanent magnets, and a complex analytic expression for the field of a cylindrical current sheet. The tutorials supplement laboratory experiences in electromagnet and permanent magnet field mapping and in eddy currents and induction.

Live Script Tutorials in Computational Magnetism

Because electric fields cannot be touched or seen, simulations are often utilized to enhance students' understanding of them by providing them with a visual representation of electric fields and the motion of test charges within them. We built and tested an electric field "sandbox" simulation that dynamically represents the electric field lines, field vectors, equipotential lines, and the voltage of the charges that the user places anywhere on the screen. We then gamified the simulation with the intent of further improving motivation and engagement in the material. Everything we created was built in JavaScript, so they will run on most browsers on a computer or mobile device. Results and student feedback from a subsequent controlled study of the efficacy of these instructional tools will be discussed. The results indicated that no individual tool was significantly more effective than the others, but pre- to post-diagnostics were significant across all three groups.

Premium content

Downloads

Next from AAPT SM 22

Development of Self-Efficacy in an Advanced Physics Lab

Similar lecture

Predicting At-Risk Students in Introductory Physics using Machine Learning