Published: June 15, 2013

How can I teach undergraduate chemistry students to build pictures in their minds, and consider what might happen if something changes about that system?  By giving them an example system that they can watch and manipulate.  And how can I get my 600 students to explore such systems with guidance from an expert and discuss their findings?  Ask them to bring their laptops to class!

The real wonders of chemistry occur not at the macroscopic level where we experience everyday life, but at the microscopic level where protons, neutrons, and electrons exist as atoms, atoms interact to form molecules, and molecules interact with each other do something cool.  In chemistry (and especially in General Chemistry), we want our students to be able to conceive of matter at this fine level, but they can’t observe these minute interactions directly, even in the lab.  Really getting to the guts of chemistry can require a lot of imagination, especially since these systems are complex with lots of parts, and dynamic so that all of those parts move around at the same time.

So how can I teach undergraduate chemistry students to build these pictures in their minds, and consider what might happen if something changes about that system?  By giving them an example system that they can watch and manipulate.  And how can I get my 600 students to explore such systems with guidance from an expert and discuss their findings?  Ask them to bring their laptops to class!

Let me talk about the simulations first.  The PhET group at CU has created dozens of simulations (sims) for chemistry (and other fields related to math, science, and engineering) so that students can observe and explore phenomena at the microscopic scale.  Even better than many lab experiments, the PhET sims let the user control and manipulate the system, sometimes changing variables that would be hard to change in real life.  Some sims even show data that would be obtained for models that we’ve since proven incorrect, like the “plum pudding” and orbit models of the atom.  These simulations are really built so that students can explore ideas about each system, pose their own questions (“What happens if...”) and answer them, and construct their knowledge about the system.

Since the PhET sims have so many options for exploration, you can’t just tell a student, “Here’s a sim.  Learn everything you can about this topic.”  Very few students would be able to construct good knowledge.  Students need more guidance about what they can explore, especially for sims that show complex phenomena.  A great way to use the sims with students is with guided inquiry -- asking a question where students can predict an outcome, then explore their prediction with the sim.

I’ve used sims in my course in several ways -- in lecture, where I show the simulation on my computer and ask students to make predictions, then show the changes; in recitation (20 students to 1 TA), with a worksheet-style activity; and as homework, with a short set of questions.  But I really think that students’ learning is best when they can explore with an expert, manipulate the variables themselves, ask for clarification, and receive quick correction if their thinking goes astray.  If we use the simulations in lecture, I’m also (more) sure that students are doing the activity, all students do it the same day (instead of across a week’s worth of recitations), and I can be flexible during the lecture with the direction of exploration.  So in my General Chemistry course (2 lectures with 200 and 400 students, respectively), I invited my students to bring their laptops to lecture so that we could work with several of the simulations together.

Of course, there are some challenges.  The expert to learner ratio is 1:200 or 1:400.  I’ve just invited hundreds of distracting devices into my classroom.  And now I’ll need to keep all of those people engaged and on task.  But I’m up for the challenge!  In upcoming posts, I’ll report about using three simulations during four class lectures (two different weeks) in these large lecture sections, what worked and didn’t in this course, and whether I’m crazy enough to do this again in the future.

During two weeks of the Spring 2013 semester I invited students in my very large lectures (200 and 400 students, respectively) to bring their laptops to class so that we could work with simulations from the PhET group together in class.  The planning for these lectures had two main components -- getting students prepared, and having the lecture prepared.

Preparing the students was easy -- they just needed to download the simulations and bring their laptops.  We worked with three main simulations on student computers during the lectures:  Models of the Hydrogen Atom (two lectures during the third week of the course), Molecule Shapes, and Molecule Polarity (one lecture each during the eighth week of the course).  Students were instructed to download the simulations ahead of time and test them out on their computer, with reminders in the previous lecture and by email.

Preparing the lectures was more challenging.  My co-instructor the course, a PhET simulation expert, and I wanted students to use the sims in several different ways, including exploration, answering open questions, making and testing predictions, and as a resource to check their answers to other questions.  So we planned our lectures to allow all of these practices.

My co-instructor has done research about the level of student exploration of sim features and the level of guided direction.  We wanted to set aside a few minutes for students to do some broad exploration of each sim, so we prepared a few broad questions in one slide and asked students to come up with some answers.  After a few minutes we regrouped for whole-class discussion and filled in the answers on the tablet PC.

We also wrote clicker questions based on the simulation and its images.

Preparing for this type of “lecture” is really a challenge. The hardest part is developing a clear set of learning goals for the students to help guide them to construct new knowledge and build on their existing knowledge.  Knowing how long each sim-related activity will take is a bit of an art, so you have to keep your eye on the clock during the lecture.  It also helps to be flexible during the lecture to handle questions from students and decide how to tackle any technical challenges.

So that was the plan.  What could go wrong??

Up next in my series: Was this experiment a success?

When your lecture seems like chaos, how can you tell whether it “works”?

A class session in which a third of the students have laptops out and most students are engaged in exploring a simulation of the microscopic world looks (and sounds!) like chaos.  How can anyone tell whether these students are just messing around with the computer or if they’re actually learning something?  In this post I’m going to discuss some ideal -- and some more realistic -- indicators of success.

First, I should say what “success” would look like.  My goal for using simulations during class is to help all of my students gain better or deeper understanding of these topics than I think we could achieve with just the textbook and some lecture slides.  Ideally, students will be able to clearly and correctly explain each topic, draw pictures of these phenomena, and predict what might happen if you alter the system.  That’s a lot to ask for, and a lot of learning to measure.

Ideally, I’d try to measure each of these items compared to students from a previous semester.  Unfortunately, I didn’t have this opportunity because this is the first time I’ve taught this particular course, and I didn’t have comparison data from other students.  So I don’t have the “hard data” that, as a scientist, I’d use to draw a strong conclusion.

However, I surveyed my students after each of these sims-in-class weeks to see what they liked and disliked about using the sims in class, to what degree they thought using the sims in class improved their learning, and for their open feedback on this topic.  I also asked exam questions based on the simulation topics (sometimes using screenshots from the actual sims) to gauge their comprehension.  Finally, I was able to observe students during the class sessions and see their level of engagement with the activity, including discussion with their peers about open-ended and clicker questions that I used during each session.  Engagement may not be the same as learning during these sessions, but I suspect that they are often related.

All of these metrics of success focus on the student side of learning, but there are also ways to assess potential growth in my teaching, too.  “Success” for myself in this exercise includes whether I can achieve the learning goals that I have set out in each session, whether the questions I’ve posed to students are clear and solicit the responses I’m hoping for, and whether, if it didn’t quite work out during the first week of sim use, I was able to make an improvement for the next time.

In my final post on this in-class experiment, I’ll reflect on any successes for my students’ learning and my teaching when using sims on student computers in large lectures.

During the Spring 2013 semester I invited my students to bring laptops to class to run simulations during two weeks of the course.  These class sessions were a little bit crazy, with about ⅓ of students bringing computers to the very large lectures (enrollment of 200 and 400 students).  After each set of in-class sim use, I surveyed students about their opinions of using sims in class.  Overall I think this experiment was a success, both for my students’ learning and for my growth as an instructor.  In this post I’ll discuss what worked, what didn’t work, and what I’d try to change in the future.

What Worked

I organized these class sessions to allow students a few minutes to explore the sim and answer a few broad questions, and later in each class session students used the sims to answer several clicker questions.  During these classes, I wandered the aisles of the classroom during open-answer exploration and clicker questions to observe the students.  Students that were using the sim on their computer or sitting next to those students were engaged in deep, active discussion about the activity.  Although I can’t be sure that they were learning everything I wanted them to, it was clear that they were participating and thinking.   Students nearby, in the same row or in the row behind the computer, watched but rarely participated in a discussion.  Students who weren’t able to see a peer’s computer reported that they mainly watched the sim projected on the instructor’s screen.  I think these activities were quite beneficial for students that were able to actively participate.

I asked students whether they felt that using the simulations in class on their own computers was useful for their learning.  Most students reported that the simulations helped their learning “a little”, “somewhat”, or “a lot”, with higher ratings after the second set of simulations.

Finally, there was some positive student feedback about using the sims in class.  Many students noted that it really helped them to visualize some of the topics that are hard to draw on paper (the energy levels for the electronic structure of the hydrogen atom, the three-dimensional shapes of molecules); however, many requested more guidance and clarity during the class session to make the simulation use more productive.

This is a short list of “successes” for the class, but I think they’re enough to get me to try this again in the future, with several modifications for both my part and the students’.

What Didn’t Work, and How to Fix It

First, a lot of students were resistant to using simulations during “lecture” time and felt that they were not being “taught” the material.  Most students think that “learning” happens when I talk and they write, but a couple decades of science education research have shown that’s not the case!  Most CU students are used to clickers and are willing to go along with clicker questions as “learning”, but using computer simulations in class was too far out of the box for many.  I think this resistance signifies that I haven’t done a good enough job of explaining to my students how learning happens, and why doing constructivist activities in class is learning. In future courses where I want students to use sims on their computers, I know I need to start this discussion at the beginning of the course and return to it often so that students can better appreciate active learning techniques.

Second, we had several challenges with actually using the sims during class.  Many students were resistant or unable to bring a laptop to class.  Some students commented that they don’t have a laptop, and some said that they don’t want to carry one all day.  Nevertheless, about a third of students did bring laptops.  It seems like that should make a nice arrangement for groups of three students per computer.  But the computers were not distributed evenly around the lecture hall, so some students were in a laptop-less desert.  Despite encouragement from the instructors, students would not move, even just one seat to be near someone with a computer.  Perhaps this is a consequence of the fixed-seating arrangement of lecture halls, but many students had the opportunity to move over a few seats and would not go.  I think one solution to both of these problems might be to form small groups (3-4) of students early in the course so that they get used to working together.  One group task would be to make sure that a group member brings a laptop on days when we use simulations in class.  Setting up groups in a class of 200 will be a challenge, but it might be a way to create a little bit of community and address this problem.

We also faced technical challenges to get the sims running on student laptops during class.  We instructed students to download and run the sim on their laptop before coming to class, but only a few students did so.  So when we asked students to start up the sims during class, many needed to download the sim and update relevant software.  This caused two problems.  First, in the larger lecture, we saturated the wireless network, and second, it wasted class time when I wanted students to be working with the sims.  The network issues were an easy fix after I told OIT about the problem.  Getting students to prepare for class is a little more challenging.  I think the key might be to give an pre-assignment worth a few points that requires students to explore the sims and therefore download and make them run on their computers.

I could also improve my in-class plans for sim days.  I wanted to give students time to explore the simulation, answer overarching questions about each topic, and create clicker questions to chain together the learning goals.  During the first week, when we used the PheT Models of the Hydrogen Atom sim, the overarching questions were quite broad, and students reported on class surveys that they wanted more guidance during this activity.  I think their comments were warranted; this sim shows several models of a complex system, with multiple representations of the models.  That’s a lot to process on the fly during class!   (Again, an advantage of a pre-class sim assignment.)  During the second week of in-class sim use we worked with the Molecule Shapes and Molecule Polarity sims.  Both of these are less complex than the Hydrogen Atom sim, which made them a little easier for the students.   I also did a little more demonstration of some of the sims’ features and really worked to focus and clarify the open-answer, exploratory questions.  After this round of sims, students reported that these sim days were more helpful for their learning than the first round.  So I made some progress, but further improving the guidance and outline for these classes will be an added improvement.

Finally, I also used the simulations to write exam questions to connect to the images that students had already seen in class.  The question below was on an exam a few days after the first week of sim use, but before students had performed a lab experiment on the same topic.  Unfortunately, students did not perform well on this question (45% answered correctly).  I hadn’t asked a quantitative question about the relationship between wavelength and energy, so I’m not sure how to interpret the low scores.  Discussions with students after the exam showed that some were confused because two of the lines weren’t the full height of the graph.  Students had observed these short lines when using the sim in class, and no one who proofed the exam (my co-instructor and three chemistry graduate students) realized that the different heights of the lines would add an unintended challenge for students.  Overall, I think using the sims to create exam questions is a good practice, and I intend to continue and reuse some of these in the future.

The Outlook

I think this experiment to have students bring their laptops to class to explore simulations of microscopic phenomena was a success.  Despite the challenge, I think that the sims that show more complex phenomena are better for in-class use because students can benefit more from expert guidance.  Making sims-in-class work better  will require a lot of improvements.  I’ll certainly be attempting this again in my classes, implementing the changes I outlined above.  I think I’m on the right track.