Abstract:

Problem-solving is an essential skill for aspiring scientists to develop. Yet, the kinds of problems we ask students to solve in most physics courses are devoid of real-world context, artificially constrained, or remove many of the opportunities for students to make their own decisions (e.g., what kinds of assumptions to make). Though this makes grading easier, it does not adequately prepare students to solve the kinds of problems they will encounter as working scientists or engineers. Recent research has investigated how expert scientists solve problems they routinely encounter in their work, and has characterized the problem-solving process as 29 different decisions to be made. The way to become an expert physicist, then, is to practice making each of these decisions with timely and specific feedback from an instructor. In this talk, I will show examples of how this decision-making can be assessed in detail. I will also discuss how we have adapted an introductory physics 1 course to focus on teaching these kinds of problem-solving skills. Preliminary results show that it is possible to teach real-world problem-solving skills while covering all of the same content as you would otherwise. The data also show that students who receive problem-solving instruction perform better on a common final exam, and receive higher grades in physics 2 compared with students who received traditional instruction. Finally, problem-solving instruction has proved to be more equitable than traditional instruction: there is no correlation between exam performance and high school physics preparation. This means that students who have never taken a physics class before college are able to succeed just as well as students with one or two years of high school physics instruction. We hypothesize that the primary driver of these results is giving students opportunities to practice solving more realistic physics problems, meaning that any instructor should be able to achieve these results if they change the way they think about writing homework and exam problems.

Bio:

Dr. Eric Burkholder is an assistant professor in the department of physics and department of chemical engineering at Auburn University. He has a broad range of research interests in physics and engineering education including problem-solving, sensemaking, equity, and what factors contribute to student success at the graduate and undergraduate levels. Before transitioning to educational research in 2018, Eric studied the physics of self-propelled colloidal particles like bacteria and nanorobots.

Abstract:

Abstract:

Abstract:

This contribution is devoted to presenting a central issue in the field of complex systems: given a set of experimental data representing the state of a physical variable fluctuating in space and/or time, how can we infer the main characteristics of the fluctuations that generate the experimental traces from the data? How can we build a general, reliable and efficient model for them? In past years, many efforts have been made in this field for different classes of stochastic processes, including, for instance, generalized Langevin equations and jump-diffusion processes. Complex system study in terms of time series analysis constitutes an intriguing and challenging branch of statistical physics, embracing diverse fields such as medical sciences, finance, economy, surface science, turbulence, and so forth. In the case of turbulence, for instance, the energy transfer across different scales can be recast in the framework of a generalized Langevin equation, where velocity fluctuations constitute the dynamical state of the system. The main focus of this presentation is to show an application of a general framework based on stochastic differential equations in the context of space plasma turbulence. In light of the latest space mission, in fact, observations of solar wind provided with an increasingly higher resolution allows us to investigate peculiar properties of turbulent fluctuations transitioning from large scales, at which the plasma behaves like a single fluid, to small scales, below which the ion motion decouples from the electronic fluid. In turbulence, it is common practice to study the statistical properties of the energy redistribution among structures of different sizes in terms of scaling laws of the fluctuation statistics. In the case of space plasmas we prove that: i) the correct scaling laws can be obtained through a stochastic model of the Langevin type, ii) alongside usual statistical properties, the stochastic modeling allows us to derive the asymptotic limit of individual sets of fluctuations, thus giving a precise predictions of the dynamics on a local level. The methodology illustrated in this presentation is general and therefore suitable for application in very different scientific fields in a multidisciplinary perspective.