My implementation of a fully flexible Physics 132

Showing my flexible learning setup.
A fully flexible setup.
Allowing participation in-person, synchronously remote, and asynchronous is quite complex!

As described in a previous post, the relaxation of the mask mandate at UMass Amherst around spring break motivated me to introduce a third way to attend Physics 132. For some time, this course had provided students the option to either attend in-person or engage asynchronously by watching the videos of lecture recorded by Echo360 later and completing in-class quizzes within the week via Moodle. However, the lifting of the mask mandate and a general push here at UMass promoting “flex learning” motivated me to add a synchronous remote option. For the second half of the semester, therefore, there were three different ways of engaging with the course: in-person, asynchronously, and the new synchronous remote.

Key features of the course that need to be present in all the different modalities

  1. The class is taught in a flipped style: students are responsible for engaging with the fundamentals of the content before we begin the unit via reading and homework problems. Students need the ability to get help with this material.
  2. To ensure adequate mastery of the preparation, there are (almost) daily one-question quizzes. Students need to be able to complete these quizzes regardless of their modality.
  3. Class time is spent in various ungraded activities including problem solving, conceptual questions. Students need to be encouraged to actively engage with the material as opposed to passively listening/watching.

How the different modalities achieve these goals


  1. Help with preparation: For those students who are physically present, we offer a slew of help hours in our help room. This room is open for significant parts of the day as shown in the schedule below. While the specific assignments for each TA are listed, any of the students in any of the service courses offered by our department can attend any hour.
  2. Engagement with quizzes: The quizzes are done via the iClicker system. This technology is the standard audience participation system at UMass and most of the students, being second and third year students, already have one. Therefore, these remotes do not imply an additional cost.
  3. Engagement with in-class activities: This goal is easiest to implement in-person. The social pressures of the room: giving dedicated time plus having myself and my TAs walk around, encourages most students to engage with the material.

Asynchronous attendance

  1. Help with preparation: In addition to the help room, we also organize Zoom-based help hours.
  2. Engagement with quizzes: In addition to the in-class quizzes, a related quiz is posted to Moodle each day. Students who do not take the quiz in class have one week to complete this quiz which has an 8min time limit.
  3. Engagement with in-class activities: Encouraging this behavior in students taking the course asynchronously is notoriously challenging. One way I tried to inspire students to go beyond just watching the videos is by segmenting them into smaller pieces and adding a card into the videos encouraging students to pause, try to solve the problem before moving on. To assist with this time consuming task, I hired a physics major with experience in video editing. This student’s physics knowledge was sufficient to figure out logical breaks.
A screenshot of the video uploaded to Echo360 system. Note the captions provided by PowerPoint, the camera of the room, and the instructions overlaid encouraging students to pause the video and solve it before moving on.
A screenshot of the video uploaded to Echo360 system. Note the captions provided by PowerPoint, the camera of the room, and the instructions overlaid encouraging students to pause the video and solve it before moving on.

Synchronous zoom-based attendance

  1. Help with preparation: the Zoom-based help hours are also useful for these students.
  2. Engagement with quizzes: UMass students who purchase an iClicker also gain free access to the cloud-based app which they can download to their mobile devices. Via the app, students can engage with the iClickers the same as students who are attending in person.
  3. Engagement with in-class activities: Students who choose this option attend class via Zoom. A single TA is assigned to manage the Zoom room: answering questions, facilitating conversation etc. The aim is for this guidance, plus the dedicated time during the class session, to promote students trying the problem as opposed to simply waiting for me to go over it.

Different ways students can move between the modalities

Primarily in-person

Some students will come to almost every class of their own volition. Others like to have some extrinsic motivator to encourage their attendance. For both of these groups, I offer the opportunity to join in-class teams which I organize via the CATME system which helps ensure equitable team formation. Students who elect to join such teams have an explicit attendance expectation enforced through the in-class quizzes; team members are limited in the number of times they can use the Moodle quizzes to four times or less during the semester. However, the various remote options are available for the occasional absence.

Significantly remote

This course is dominated by second and third year students as shown by the figure below. However, there are also a significant number of students in their final semester before graduation. Many of these students have a lot of external responsibilities: job interviews, touring graduate schools, etc. Other students have unanticipated life events that require them to be away from campus for extended periods, while still others find remote attendance to be more accessible. These students might complete a significant portion of the course remotely: either synchronously or asynchronously. These students are still encouraged to work together. However, there is no official recognition of their groups and, as such, no attendance requirement at all.

The demographic breakdown of Physics 132.

Exams and labs were still in person

As described in a previous post, the exams for this course were hosted on Moodle. Even so, given the challenges of remote proctoring, all students, including those who completed most of the rest of the course remotely were required to come to exams in person except in exceptional circumstances such as COVID-19 isolation. Moreover, remote labs are always a unique challenge. Remote labs for electronics and optics doubly so due to the equipment requirements. Therefore, all students were also required to attend lab in-person.

How it went


In terms of the technical aspects of the fully flexible course, I think it went pretty well. Students seemed to like the flexibility and things went rather smoothly. A full picture of the setup is visible at the top of this post. In general, however, the Zoom room was run by one of my TAs who was also the facilitator of collaboration. A key for this TA was to provide a separate webcam for them to use as the camera angle of most built-in laptop webcams is not very good. The sound was also passed through this laptop as sound was the “straw that broke the camel’s back” for my surface’s hardware and bandwidth. We connected a Jabber microphone to the TA’s laptop. Sound, therefore, ultimately went from my mouth through the room microphone (which recorded the sound for the Echo360 viewing later), into the Jabber, and then to Zoom. Meanwhile, my surface hosted the slides with closed captioning, a Zoom chat window (for questions during the lecture), a webcam, and a document camera for demos.


This was a bit more mixed. I had a statement of expectations (reproduced below) which students had to read and earn a 100% quiz on (unlimited attempts) in order to get the Zoom link. However, the level of engagement on Zoom was still not the same as in class. Moreover, I had almost no students turning on their cameras, an issue with which most faculty who have taught remotely will be very familiar. Finally, I am concerned that some of the students who elected to engage via Zoom were simply doing it so as to not need to come to the classroom at 9:05 or 10:10 in the morning. I suspect that many of these students may have benefited more by being present. I will look into the data from the semester and see what it says.

Continue reading “My implementation of a fully flexible Physics 132”

Reflections on the use of Moodle-based exams in a large enrollment intro physics course

This past year, I have experimented with Moodle-based exams. This is a holdover from the remote instruction that I, along with basically everyone else, was forced to do during the COVID-19 pandemic. Personally, I am not a fan of traditional scantron-based multiple choice exams, preferring long answer. My motivation for this is based on the feeling that multiple choice exams test students’ ability to recognize a correct answer and/or use process of elimination to narrow down the choices to those that are most probable. Long answer exams, in contrast, require students to write detailed solutions and explain/justify their work in words – tasks which engage some of the “higher-order” domains of Bloom’s Taxonomy1.

Bloom's taxonomy
Bloom’s Taxonomy: labelled as creative commons licensed from

Despite my preference for long-answer exams, their implementation in courses of several hundred is difficult as the grading quickly becomes intractable. While some physics departments include the needed grading resources to grade several hundred long-answer exams into the TA workload allotment, UMass Amherst’s Physics Department does not by custom. My solution in the past has been to do a hybrid exam: 10ish multiple choice questions and then two long answer. Typically, one of these long answer questions would be a traditional “solve the algebra” while the other would require students to explain a physics phenomenon in words. While this worked in terms of respecting the TA workload, there were still challenges. The primary issue was turn around time. I felt that giving the TAs two weekends to complete the grading was the minimum. As such, a minimum of two weeks would pass before students would get their full exams back. Two weeks in my courses is typically a full unit out of the five in the course. After such a long turn around time, students had moved on and did not get as much learning out of the exam feedback as the would have if the exams had been returned more promptly.

Moodle-based exams, by contrast, allow for grades to be released immediately if the instructor so desires (I typically take a few days for review the scores to look for problematic questions etc.). In my mind, the pedagogical benefit of this immediate feedback compensates for the losses arising from a lack of long-answer questions particularly when the variety of auto-graded question types provided by Moodle is considered.

In contrast to paper scantrons, which are limited to five-choice multiple choice questions, Moodle allows for: unlimited numbers of choices, multiple select, entering numbers, fill in the blank (via drop down or string matching), and placing markers on images. In the case of physics, these question types can be quite useful. Questions which require students to enter a number are, of course, obviously applicable. The unlimited choices are also useful as all the possible options, for say the direction of a vector, can be available. With a little creativity, the other question types can also be employed. For example, questions which require students to place a marker on an image can be useful for ray diagrams; the configuration of lenses can be superimposed on a grid which is printed and given to students which they can use to solve the problem. Once students have figured out the location of all intermediate images and the final image, they can use the grid to place markers in the correct location on the screen.

An example usage of the "place marker on image" question type applied to ray diagrams.
An example usage of the place marker on image question type applied to ray diagrams. Students place the markers for “Intermediate image” and “Final image” in the correct spots. The grid helps students precisely transfer their work from a paper version of the setup to the screen.

Instructors can also include videos or animations in the exam questions. This can be useful to clarify questions that are difficult to word in an unambiguous way. Videos can also provide a connection between the exam and very real phenomena. A question can include a video of, for example, a demonstration which the students then need to explain as part of answering the question.

Beyond the student experience of the exam, Moodle provides other advantages. Students who need to be away on exam day due to minor illness or participation in a University sanctioned event, can take the exam remotely. The experience that almost all instructors have been forced to develop over the course of the COVID-19 pandemic makes this relatively straight forward, particularly considering the small fraction of the students who will need such accommodations. Also, Moodle makes it easy to, over time, build up a nice question bank of problems. Tags can be used to keep track of the content of each problem as well as to record the last time a problem was used.

Of course, exam integrity is always a key concern when considering computer-based exams. The proctored environment is one key. Another important step to maintaining exam integrity is to remove any need for students to switch windows: provide printed equation sheets and require external calculators, for example. There are also several other features one can build into the exam which help promote exam integrity which, for obvious reasons, I do not want to share publicly. If you are an instructor and would like to hear some of these tips, please do not hesitate to reach out to me.

In summary, Moodle (and other LMS systems) provide a nice format for hosting exams in large enrollment courses. Such exams are, in my opinion, still inferior to full long-answer paper-based exams. However, such exams can be difficult to implement for large enrollment courses and generally result in very delayed feedback which we know to limit the educational value of examinations. Moodle provides a nice middle-ground between sophisticated questions and rapid feedback. In addition, such systems provide the instructor additional benefits reducing the need for makeups and maintaining lists of questions.

  1. I acknowledge the limitations of Bloom’s Taxonomy particularly in terms of thinking about levels, but it is a useful framework for conversation.

Full hybrid here we come!

As mentioned in a previous post, UMass recently dropped its mask mandate. My goal is, as always, to have as many students feel comfortable in my class as possible. In this case, that means offering a way for those who may not wish (for whatever reason) to be in a large class of unmasked folks to participate. As such, I have finally taken the plunge to offer a fully hybrid option for Physics 132 for the second half of the semester. Students will now be able to participate in one of three ways:

  1. Attend class in the “old-school” way.
  2. Engage asynchronously via watching recordings of class which have been chunked into smaller pieces by my undergraduate video editor.
  3. Attend class remotely and synchronously via Zoom.

We will see how this goes!

Embedding Discord Into Moodle

Prior to the Spring 2020 semester, I used to use the Piazza platform for asynchronous communication among my students. However, they have since moved to a different economic model. They used to be free, but now there is a cost to either the institution or the instructor. The other alternative is that the students are frequently reminded to contribute. On top of this annoyance, the student-contribution model lacks certain key features such as the ability to lock the platform during an exam.

Thus, starting in Spring 2020, I began to seek alternatives. My first thought was the professional communication tool Slack which I tried during the Fall 2021 semester. However, this simply did not take off. I did not get the same level of engagement that I was used to with Piazza. When I polled the students, they mentioned that “it was just one more platform to keep an eye on.”

With this feedback in mind, I wanted to find something that would integrate with Moodle, the UMass Learning Management System. The “Enhanced Moodle Forum” (then in beta) seemed to meet the need, and I tried that during Spring 2021. However, participation was still not what I expected/hoped it would be.

In both of these fall semesters, when my official efforts at a asynchronous communication platform, were falling flat, my students independently setup a Discord server. In both cases, the Discord was more active than my official platforms. Thus, for the Fall 2021 semester, I decided to try Discord. Most of my students were seemingly already on it. As I investigated further, I discovered WidgetBot ( This bot allows for a Discord server to be embedded within another webpage. Folks can even comment directly from the webpage without a Discord account. Discord + WidgetBot, therefore, seemed to meet all my needs: an asynchronous platform that students will actually use and can be embedded within Moodle thereby reducing the number of platforms students must check. The result, screenshot below, has been quite successful.

With widgetbot, a Discord server can be embedded within a Moodle page.