Tema: Beregninger og aktiv læring
Program:
- 13.45-14.00 Kaffe/te og pepperkaker
- 14.00-14.05 Velkommen
- 14.05-14.50 Doing physics means feeling confusion - Professor David Hammer (Tufts University School of Arts and Sciences)
Abstract: The core work of scientists is to engage with what they do not know: Doing science means finding gaps or inconsistencies to fill or reconcile. The experience is both intellectual and emotional, and the same is true for students.
For more than a decade, my advisees have been guiding me toward research on how students’ feelings entangle with what they think of as knowledge and learning, that is with how affect entangles with epistemology. This research has an impact on my teaching, ~100 college students in introductory physics, and it raises questions for possible roles of artificial intelligence.
14.50-15.35 An American in Oslo (with apologies to Gene Kelly and Leslie Caron) - Professor Gerald Feldman (George Washington University)
Abstract: The FYS 1100 introductory course at UiO has been undergoing a restructuring process over the past year, with the aim of incorporating more active learning elements and promoting student engagement. I have been spending the Fall 2023 semester here at UiO as a co-instructor in the course to assist in the implementation of pedagogical reforms (such as collaborative group learning and hands-on activities) that move the course in this direction.
I will give an overview of the current course structure, and I will describe some of the innovations that we have added this semester in this ongoing process. Finally, I will propose some recommendations for upcoming semesters that will continue to make the course more student centered and enhance student learning.
- 15.35.15.50 Pause, gløgg og bakevarer
- 15.50-16.20 Insights into Assessment Practices in the United States: Unraveling 'Gentle Assessment' in Computational Physics Instruction - Researcher Hannah Christine Sabo, Fysisk institutt UiO
Abstract: Although computation has long been integrated into the physics curriculum here at the University of Oslo, its prevalence at the undergraduate level has been increasing worldwide. Research on computational physics has also grown, yet there remains a lack of understanding regarding the most effective assessment methods for computational physics. To address this gap, we interviewed fourteen university instructors teaching physics courses that incorporate computational elements, aiming to explore their learning goals, support techniques, and assessment approaches.
Through thematic analysis of the interviews, we identified a prevalent grading approach termed 'gentle assessment,’ a grading rationale often employed in computational physics settings (usually introductory) designed to reduce students’ negative affect and (feelings of) risk around the introduction of computation into physics courses. In this presentation, I describe gentle assessment and its common characteristics.
- 16.20-16.50 Teaching Computing into Introductory STEM Courses at Scale - Professor Danny Caballero, Michigan State University / Fysisk institutt UiO
At Michigan State University (MSU), we have developed several introductory STEM courses in which students learn to use computing. While each course covers different disciplinary content (e.g., introductory mechanics, elementary statistics), each aims for students to better understand core disciplinary ideas while using computing as a tool to do so. Students learn how to write programs and create visualizations that model particular systems, that explain certain trends in data, or that determine the significance of some effect.
In this talk, I will discuss 3 specific courses. In introductory physics, students work in groups of 4 to solve open-ended problems and develop Python code collaboratively. In elementary statistics, students work in groups of 6 to construct visualizations and develop models of real world data using R.
Finally, in introductory computational science, students work in groups of 4-6 to make models (including machine learning models) of phenomenon as well as of data using Jupyter notebooks. In each of these courses, we are teaching hundreds of students per semester with growth likely to continue.
I will present the goals for these courses, how they are structured to run at scale, and the importance of different structural and institutional supports for their continued success. This work has been generously supported by the National Science Foundation as well as Michigan State University's Lappan Philips Foundation and Colleges of Natural Science and Engineering.
- 16.50-17.20 Many dimensions - key trends. Using transformer-based text embeddings to perform literary analysis. Markus Fleten Kreutzer, Jonas Timmann Mjaaland, Halvor Tyseng - students at UiO.
Transformer based language models have revolutionized the generative artificial intelligence space over the past twelve months.
Empowered by the ada-002 model from OpenAI, we suggest a novel way of performing literary analysis of large texts corpora. Our research points to the possibility of a quantitative, deductive, and painless solution to a traditionally demanding problem.
- 17.20-17.30 Short yearly review and closing remarks - Professor and Director of the center, Anders Malthe-Sørenssen
-
17.30 Aperitif, CCSE
-
18.00 Middag med juletapas/Tapas dinner, CCSE