Diffusion is a principle in Physics, Chemistry, and Biology. The rate of diffusion is affected by temperature, particle size, concentration, and material type. Students can model the rate of diffusion based on particle size by contrasting blue and yellow dyes. Two petri dishes containing agar-agar receive a drop of dye at the center. The radius of expansion is recorded over time. The variance of the distribution grows as sigma²=4𝐷𝑡, where sigma² is the variance, D is the diffusion constant, and t is time. Graphing variance versus time gives a slope of 4D. Diffusion constants vary by particle size, allowing for a size ratio comparison between blue and yellow dyes. Relating this to cells, students predict that smaller molecules diffuse into living cells, whereas larger molecules need some assistance from protein channels as in facilitated diffusion. In Physics and Chemistry, the data can be related to Kinetic Energy. 

I will present my observations of student machine learning projects from three different general education classes.  As part of a class on Galaxies and Cosmology, I guided the students through an application of neural network classification of galaxy images.  I also taught two three-week, intensive, project-based courses on Machine Learning.  Students learned tools to classify, predict, and generate data, and then they developed their own projects.  Students in the cosmology class were much less comfortable with quantitative analysis than those who specifically chose a machine learning class.  For their projects, students in the Machine Learning course tended to pick either image classification projects or predictive projects, usually relating to some kind of competition.  I will share my recommendations for how to make Machine Learning less intimidating to beginning students and what approaches to avoid.

Problem Roulette is an online study service offering low-stress practice to students preparing for examinations in introductory STEM subjects at University of Michigan. Using four years of service data, involving millions of questions attempted by thousands of students, we quantify the benefits of increased practice study volume in introductory physics.  Relative to a mean grade conditioned on students’ ACT or SAT math score, we find that grade point earned rises roughly quadratically with the logarithm of N_Q, the total number of questions attempted over the term, with an overall gain of 0.75 ± 0.15 points between 0<N_Q<1000. Using both test score and study volume to predict final grade, we measure demographic deviations, noting significantly lower scores among students whose parents never earned a college degree. Our findings can motivate low-volume students to study more and help teachers identify which types of students especially need such encouragement. 

One of the valid arguments against overuse of the multiple-choice questions for instructions and evaluation is that they do not allow the students to explicitly formulate their own ideas. Mini-case studies can provide a viable alternative allowing students to explore more open-ended scenarios with a potential to reduce the reliance on multiple-choice format questions. Course materials, in the form of case studies were created for use in the large-enrollment introductory physics classes for science programs. They were designed to target the most fundamental concepts of the first-year physics curriculum. These case studies are being used for collaborative group in-class activities in the partially flipped classroom. Examples of the activities will be demonstrated and discussed.

This talk describes several single-concept simulators I created after attending an AAPT HTML5 workshop. While these simulators have been specifically developed to aid instruction for a Next Gen Physics and Everyday Thinking (Next Gen PET) guided-inquiry course [1], they have also been helpful for instruction in other introductory courses. These available sims [2] can be used as dynamic diagrams to complement student discussions during in-class activities relating to the concepts of: initial models ("Mystery Tube"), Newton's Second Law, Centripetal Force, Drag Force, Reflection and Refraction. The "Mystery Tube" sim was developed so that remote and on-line students could participate in an activity that would have otherwise only had an apparatus available in-class.

There is a nice collection of HTML5 physics simulations available, at https://physics.bu.edu/~duffy/sims.html

We have been starting to collect worksheets to go with these simulations, to make the simulations more useful for teaching and learning. Some of these worksheets have been developed by us at Boston University, and others have been written by participants in AAPT workshops. We'll use this talk to give some examples of the worksheets.

At St. Catherine University, I used the idea of an app to re-invent the traditional final in the calculus based physics course. Rather than offer a paper exam, I had the students instead create an app, or the idea of what one would look like as thier final project. I will show how I constructed the project as well as will show some of the student examples garnered by using this as a way to have a final in the course. The results were very interesting and allowed students a fair amount of creativity and additional learning in the course. 

Science does not happen in a vacuum. Women make up less than thirty percent of the workforce in STEM fields in the US and despite increases in the number of women earning degrees in physics, the proportion of women in this field averages 20% (the lowest of all the physical sciences). Through research, interviews, and conversations students in my introductory physics classes (algebra-based & calculus-based) have analyzed experiences females (all over the world) in STEM fields encounter(ed). The interviewees ranged between four continents and three generations. The students got a sense of the treatment and struggles these women faced; got a sense of how similar and how different these experiences between continents have been; got a sense of how similar and how different those experiences might be between generations. I will report on the student learning outcomes, the students’ reactions to this project, the challenges faced, and the knowledge gained.

Since freshmen have not yet undergone any course of differential equations, it is common to give the solution of an equation without a proof. I think that by doing so, one misses an important point. The matter of fact is that these equations are very few, the first and simplest of them can be solved very easily, whereas the others can be derived from it step by step. In addition to reasoning the solution, one achives by this two goals: (a) These equations are useful for a large variety of cases. In a course of electricity and Magnetism, all the equations needed are those known from mechanics and no new equations are involved. (b) It turns out to be very useful for the mathematical course in differential equation, taught commonly at a later stage, since it gives differential equations the proper context. We demonstrate how to do it.

When lab equipment is operated via a computer, an offsite user remotely logged on to that computer can control the experiment.  We have been operating two such experiments at Kwantlen Polytechnic University (KPU) since 2017 as part of an online lab section for first year physics.  Referred to as CloudLab, students connect to the equipment, instructor and other students using Zoom.  Originally developed as RWSL and NANSLO in grant funded consortia, the current implementation at KPU includes a “cart on an inclined track” and a “electron charge to mass ratio” (or “e/m”) experiments.  We have also worked successfully to provide access to classes at other institutions. Hardware, software, student and instructor experiences, as well as future expansion plans are described.