United States

Discipline based education research often compares student outcomes (e.g., performance, affect) across treatments. These quasi-random experimental designs can entail students in section A experiencing a treatment and students in section B experiencing a control. However, if the differences between students are not accounted for, student-non-equivalence can confound results or mask true treatment effects. 
Linear regression is one way to control for student non-equivalences that arise from quasi-random designs. However, linear regression only works for continuous outcome data that meet certain assumptions. As DBER research questions have become more sophisticated and researchers begin measuring more nuanced student outcomes, statistical analyses must extend beyond linear regression. Here I will review extensions of the linear regression model – generalized linear models (glms) – and compare seven glm types. By the end of the session, participants will be able to articulate why controlling for confounding variables is critical and identify when to apply different glm methods. 


AAPT SM 21

Beyond linear regression: Analyzing common data types in DBER

data analysis

statistics

regression

Join AAPT and Physics Educators worldwide in developing and strengthening your professional connections by attending the 2021 AAPT Virtual Summer Meeting. 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. Attending the virtual summer meeting is an excellent opportunity to share your expertise and learn from and network with leaders in the field.

Please log in with your AAPT account or register for the event if you haven't already.

If you are registered please note that you need to be logged in with your AAPT email / credentials in order to get access.

technical paper

Computer simulations can provide an efficient and low-stakes way for students to discover and apply scientific principles. However, until recently it has been difficult to customize simulation parameters for particular students so that each student is exploring a virtual space that is unique to them. Such customization can make the exploration more authentic and may discourage direct copying of other students' work. The arrival of PhET-iO allows not only this type of customization, but also allows autograded questions based on each set of unique parameters.

Customizing student experiences by pushing randomized values to simulations

Due to COVID-19, remote physics instruction has become ubiquitous. This research examines which practices are highly impactful for creating social connections between teaching assistants and students in synchronous remote settings. We recorded interactions between students and teaching assistants in introductory physics active learning problem solving sessions and coded teaching assistant behaviors for types of social practices (e.g., addressing students by name or expressing empathy about course workload). By administering a validated survey of social alignment to students at the end of their recorded sessions, we were able to determine which practices were most impactful to students’ feelings of social connection with their instructors. The results of this study have implications for training teaching assistants in facilitating remote problem-solving sessions.

Teaching Assistants’ High Impact Social Practices in Remote Physics Recitations

A classic lab activity for electrostatics involves mapping out 2-D equipotentials on semi-conductive paper, allowing students to probe the shapes of electric fields and potentials. But what can be done for distance students or students who cannot attend a hands-on lab? To address this, we developed a graph-matching challenge in which students must lay out charges to produce a desired field and/or potential. The simulation activity runs entirely in Excel and features 3-D surface plots that are visually more stunning than the flat line drawings proposed in most textbooks, and arguably more evocative than the deliverables produced in the traditional lab. We discuss technical details and student reaction.

Equipotentials-matching challenge: an alternative to the classic plotting lab

This is a follow up on a paper presented during the summer 2020 virtual meeting. We have now completed 20 problems integrated into the expert TA system as the Physics Video Analysis Collection. Students in my introductory classes in fall 2020 solved these problems as part of their homework assignments. I also had a student from that class look at all the problems and we have and new hints and feedback, common wrong answers identified as suggested from the student perspective have been added. These changes have improved on what was already an excellent pedagogical experience. I will show them in the expert TA system, focusing on pedagogy not available in more traditional problems and I will also report on student experiences in learning from them.

Video-based graphical analysis homework problems

The electric field is one of the more abstract concepts taught in introductory physics. Introductory E&M courses often start by introducing Coulomb’s Law—an expression for the force between two point charges—before eventually defining the electric field as a ratio of the electric force to the charge on a “test charge”. This definition can result in some confusion among students; for example, students often have difficulty with the idea that the electric field is independent of the “test charge”. We decided to switch the order in which these topics are taught in a calculus-based introductory physics course: first introducing the electric field as the cause of electric forces, and later introducing the electric field caused by a point charge. We will share some pedagogical affordances of beginning E&M in this fashion as well as some observations from our initial implementation.

Teaching Electric Field before Coulomb’s Law

Kinematic graphs is a popular topic for algebra-based physics, but rarely are the graphs used to solve problems. This talk will communicate experience I have gained in how to teach students to solve kinematics problems directly from the graphs.

Solving Kinematics Problems Using Kinematic Graphs

An Apparatus for the Lab of Uniform Circular Motion
Richard Guarino, Jingbo Ye
Department of Physics
Southern Methodist University

Uniform Circular Motion (UCM) is discussed in college physics. A lab that is commonly used to demonstrate the relationship between the centripetal force and the angular velocity usually involves a string. The string that connects the rotating object to a weight or a force sensor is turned 90 degrees by a pulley. This scheme has the problem that the string will twist and the pulley has friction. Utilizing the force and gyro sensors in a Pasco Force Sensor and with a custom rotor, we constructed a UCM lab with precision measurement results. A photogate is used to double check the angular velocity measured by the gyro sensor. This double measurement of the same quantity provides students with an example and training of calibration and cross-check in scientific work. We will present the design and performance of this lab, and share the lab manual with interested fellow lab instructors. 


An Apparatus for the Lab of Uniform Circular Motion

Many newly reformed introductory physics laboratory courses prioritize student engagement with scientific practices, in line with a significant body of PER research that suggests these courses often do not adequately enhance students’ conceptual knowledge. However, is it possible for introductory physics labs to focus on student engagement with scientific practices AND generation of conceptual knowledge simultaneously? In this presentation, I discuss how our Introductory Physics for Life Sciences (IPLS) lab instructional team refined existing evidence-based instructional strategies and curricular activities to elicit students’ simultaneous engagement in scientific practices and the generation and utilization of new conceptual knowledge during investigations. We show how these refinements resulted in students collaboratively generating conceptual mechanistic knowledge of complex biophysical systems and using this knowledge in open-ended experiments to build scientific models of related scientific phenomena. These examples provide evidence that physics lab courses can engage students with scientific practices and conceptual learning simultaneously.

Practices and Concepts in Physics Labs: Can We Do Both?

The acceleration of a falling coffee filter released from rest is always much less than the free-fall acceleration of gravity, g. This is true even near the initial release, when the part of the drag force proportional to the velocity is negligible. This small acceleration occurs because there is a part of the drag force that is proportional to the acceleration. This part is called the added mass. The added mass is a large effect for falling coffee filters. Nesting the coffee filters does not change the added mass. The measured value of the added mass from the initial acceleration agrees with the theoretical value calculated for an ideal fluid. In experiments using a motion sensor or video analysis, it is easy to measure both the added mass and the steady-state drag force.

Measuring the Added Mass of a Falling Coffee Filter

Integrating of engineering design activities in physics has been shown to be effective in improving student achievement, attitudes, and learning in K-12 settings. Prior research supports the idea that engineering design activities have the potential to bring value to a physics course and to facilitate transfer between engineering courses and physics. During the Spring 2021 semester we introduced an engineering design challenge, to design a simplified Martian lander, in the physics labs midway through the semester. Using survey methods, we examined student interest, metacognition, and transfer. Our results found that most students made connections to other classes and their majors. Students also reported using metacognitive monitoring and control strategies during the design project. However, student perceptions of the workload limited student interest. Implications for implementation of engineering design activities will be discussed.

Premium content

Downloads

Next from AAPT SM 21

Customizing student experiences by pushing randomized values to simulations

Similar lecture

How STEM faculty enter Discipline Based Education Research