This week, a lot of folks on twitter were discussing a paper in Physics Today by Natasha Holmes about the shortcomings of laboratory instruction. I’m not going to specifically get into the details of this paper here, but it was the stimulus for me writing this blog post. It’s a good read, and you should check it out if you haven’t. Anyway, based on my earliest experiences in Physics Education Research, I am not the least bit surprised that it is a typical outcome for most contexts of physics laboratory instruction to have little or no impact on student learning of physics content.
Why am I not surprised? (Spoiler: students can struggle to recall even the outcome of many lab activities, let alone any concepts or principles that relate to those activities.)
Here’s an example from my very first research project while I was a graduate student at Arizona State University. Students in an introductory college physics course would take a traditional laboratory at ASU that was separate from lecture. Each week students collected data (often using vernier equipment) about a different physics topic, plotted various data, and verified that results were consistent with theory. One week after students completed any given lab, I administered a post-test of some sort, related to the topic. What I learned fairly quickly is that you didn’t need to be very tricky with your questions. Often the best ones literally just asked students to state back what they had observed in the lab.
For example, for a lab on projectile motion, students use a photogate to measure the initial speed of a horizontal projectile (rolling off of a lab table), and also use a timing pad to on the floor measure time of flight. Students observed that the time of flight was independent of speed and that the horizontal distance was directly proportional to the initial speed. Students were asked to make plots of this data and also to answer questions
So the next week, the post-test survey I administered asked students to rank the time of flight for three experiments – ones that left the table with speeds v, 2v, and 3v. Some students got a slightly different version of the ranking tasks which was to compare the time of flight for landing at horizontal distances x, 2x, and 3x. The questions included drawings of the situation that basically mimicked their lab experiment. Usually about half of the students could correctly answer such questions after the lab. (Note: my experience with this data led me eventually to a more careful study of how students were answering these and other similar questions).
Conclusion: If one week later, many students cannot remember the results of these types of labs, we probably don’t stand much of a chance of them learning, remembering, and being able to apply any underlying principles that the lab result was intended to support. Right?
What’s wrong with labs? (Spoiler: It’s not labs per se that matter, but how and whether laboratory activities become relevant objects in broader discourse of students’ learning)
So, the truth is this. I actually do think that laboratory activities can play a supporting role in students’ understanding of concepts, including enhancing students’ ability to transfer those concepts to novel situations. That said, lab activities alone are pretty useless and irrelevant. What matters more is the broader context in which a lab resides, and even furthermore I would say its how students “frame” the role of lab in the broader context of their learning.
So how might I judge whether a laboratory activity I’ve planned is playing a meaningful role in students’ learning? The most crucial thing I look for is what happens in the coming weeks. If we as a class never explicitly mention a lab during ongoing discussions–what we saw, what we concluded, or what it meant– that lab was likely meaningless. That’s it. That’s really my only criteria. I don’t mean the lab was inherently meaningless. I mean that our subsequent activity rendered that lab activity meaningless, retrospectively. In trying to improve things for future courses, it could be that I need to ditch the lab, or it could be I need to better plan for how that lab will be continuously rendered meaningful through subsequent activity, or I need to help with reframing lab.
Lab activities I just do not think have any inherent value, just as activities of mining raw materials has no inherent value outside a broader activity. It’s what happens subsequently in a process that determines it’s value.
Brian, what the heck are you talking about? (Spoiler: here are some examples to help clarify what I mean)
Newton’s 2nd Law Lab:
On twitter recently, I tweeted about a Newton’s 2nd Law lab that I have adapted and refined over time, which is basically just a half-atwoods with force sensor on a cart and motion detector. In that lab, I deliberately have students in class take data using a 1.0 kg mass. A twitter friend of mine remarked that he might be worried that this could accidentally reinforce a misconception about how force and acceleration are related (i.e., F = a). I agreed that this seems like a possible outcome, but perhaps only if you are looking at this lab in isolation and expecting students to “discover”, “abstract”, or “generalize” the correct ideas from it. But that’s not how I look at this lab (or most labs), I look at this lab as something designed for talking about later, to uncover its meaning through subsequent activity. Specifically, I aim for students to start thinking of net force as meaning something like, “It’s the kind of (collective) forcing that would cause a 1.0 kg object to accelerate at the same value.” We expand from that specific lab by thinking about data / experiments for objects that are heavier and lighter with in-class clicker questions and post lab questions. But most importantly, it’s something I try to work into lessons during every single day in following weeks while we solve force problems. If I am doing an example problem, I won’t merely use the equation Fnet = ma. Instead, I will calculate Fnet and talk about what that means for a 1.0 kg object, what that should mean for our actual object, and figure out exact values using proportional reasoning. If I want to be really explicit, I might make an acceleration vs. force graph similar to lab. Similarly, when students are working problems at whiteboard, I expect students to always explicitly determine a net force value, and I ask questions about what that means, for a 1.0 kg object, for their actual object, and often other masses. I ask, how do they know? Why does it make sense? What observations have we made that support that way of thinking about how objects respond to forces?
Friction Lab: I have also developed some friction stations that I think are a pretty decent introduction to the topic, but again, the question isn’t if a lab is good or bad by itself. One of things I expect out of this lab is for students to spontaneously bring up stations from that lab when we get stuck on a problem or a clicker question about friction. One clicker question in particular where I see this happening is about pulling objects at angles along rough surfaces. The particular clicker asks students how the angle of the pull might affect how easily the object budges. In this clicker question, there are certain lab activities I am looking for students to reference during discussion, including observations about normal force,observations about the effect of normal force on friction, and observations about angled vectors and components.
How do you get students to treat labs as activities relevant to future discussion? (Spoiler: treat them as if they are actually relevant to future discussions and plan for later activities that are designed to make use of that relevance).
So, I don’t think there is one answer to this, but you certainly don’t just hope it happens. Here are three concrete things you can do to help support it, but there are many more.
- Make sure that laboratory activities are meaningful to prior discussions. Many times students don’t know why they are doing a particular lab, or how it’s connected to their prior learning. By making sure students have a sense of purpose and connectedness with lab before hand, this makes them more likely to extend that connectedness in the future. Some of the ways I do this are following: using clicker questions to raise an issue (scroll to end of this post to see an example from forces), using demos to establish a tentative result and present a challenge (Newton’s 3rd Law), or sequencing questions, activities, and discussions that help us refine an understanding through observation and testing (circular motion). It’s a good sign when students can say why they are doing something and/or what they are trying to figure out.
- Ask questions, use prompts, and phrases that imply connectedness and expansiveness. (What will you want to remember about this lab tomorrow?, How is your answer to this question supported by what we observed earlier today? How is this result similar or different than what we saw before? How does this observation change your thinking about? How does this outcome influence our answer to the question of ___? What ideas from today’s lab will be helpful when you go to ___ ? ). If you are interested in this more, see this paper about expansive framing and transfer.
- Design (writing) activities that require students make use of observations from lab activities as evidence in making claims about ideas. Here are some examples of students referencing lab activities about light in a course of I taught:
This is a great post, and it’s quite timely as I’m about to go give a talk about the activities I plan for non-majors. Your criterion for what makes a good lab really resonates with me. Certainly in hind-sight I would say that the activities (labs, demos, whatever) that are mentioned later in class to tie concepts together or to get out of a nasty snag with the material are the ones I think were most successful.
I wonder if the reverse is true. In my department the vast majority of the time the lab on some topic comes after the classroom time devoted to it. Would monitoring the comments in those labs help one understand which classroom activities were best? I guess I’m wondering if there’s a fundamental difference between hands-on inquiry experiences and classroom lecture-like experiences in that regard.
I think, the general principle applies… if I lecture on something, it’s learning value is largely determined in future activity. Like I’d want to know how students use what took place during my lecture to enhance their learning activities—to what extent, in what breadth of contexts, to what end? Lectures are mostly bad for same reasons labs are bad, neither are inherently bad, they are just often rendered useless through inadequate design, and a lot of design is about how it you plan for just connects to other times and places. Lab and lecture both benefit from being treated like rehearsal spaces for what we hope the might do in the future.
I liked your blog post Brian and the observations you made about the physics labs.
Have you ever explained the lab procedures to the students in class and have them predict what will happen? Then confirm or reject their predictions by the actual lab results. I think this would ultimately be what you’re doing now as a follow-up but reversing the order such that the follow-up is actually the lab itself (with perhaps a large group discussion to share in the following class). Curious what the results would be.
Thanks for a thought provoking read!
Thank you Michael. We do a fair amount of prediction-observation type activities, be it demos, labs, or other observations. One of my favorite ways to do quantitative prediction labs is the following: https://teachbrianteach.wordpress.com/2016/03/01/converging-lenses/