Dr. Richard Olexa responded via twitter to my Orlando Sentinel op-ed on the perils of the State University System’s emphasis on online instruction. He said this: “As an alum and someone who took Physics at FSU, I say this very respectfully: I have a lot of students who wouldn’t even be able to give it the old college try without online (blended) courses….IMHO, we need to do our part to ensure that the online experience and rigor mirrors the traditional course. I think it can be done.” He also said that he had taken a traditional lecture physics course at FSU and had not taken a studio physics course, which I described in the op-ed.
Dr. Olexa teaches at the State College of Florida Manatee/Sarasota and St. Petersburg College. He earned a bachelor’s degree in Psychology at FSU in 2005, a master’s degree in biomedical sciences at the Philadelphia College of Osteopathic Medicine in 2007, and an M.D. at Ross University School of Medicine in 2010.
To start with: If you are saying that we can teach a fully online physics course that is at least as effective for promoting student learning as a traditional lecture class, then you are 100% correct.
The graph below shows “normalized learning gains” in Newtonian mechanics from three types of introductory physics courses, as reported by a team from MIT (Colvin et al.) in the September 2014 issue of The International Review of Research in Open and Distributed Learning. The normalized learning gains are measured by pre-testing and post-testing using the Force Concept Inventory. The bar on the left side of the graph was measured by Hake et al. (1998) in traditional lecture classes. The center bar is from the online course developed at MIT and is reported by Colvin et al. Not only is the online course as good as the traditional lecture course you took at FSU, it is better.
But the traditional lecture course wasn’t what I was defending in my Orlando Sentinel op-ed. I was defending the type of course that is shown on the right side of the graph – the interactive engagement course. While interactive engagement courses come in various shapes and sizes, the model that first UCF and then FSU (which was inspired by UCF) implemented is called SCALE-UP. SCALE-UP was developed at North Carolina State University by a team led by Robert Beichner and has been implemented at more than 250 institutions, including MIT (the New York Times article on the MIT implementation is here). Our first SCALE-UP classroom (we call it Studio Physics at FSU), built into FSU’s large classroom building in 2008, is shown below. Two more have been renovated in the Carothers Building since then.
Here are some questions you might have about SCALE-UP/Studio Physics:
Is it more expensive to run than a traditional lecture class? No. The normal staffing model in our 60-80 student Studio Physics classes is one professor, two graduate TA’s and one undergraduate “learning assistant”. It turns out that the personnel cost per student is about the same as it is for a big lecture class, which includes not just lectures but also recitations and labs – all of which need to be staffed.
Does it require more physical space than a traditional lecture class? No. Yes, the SCALE-UP/Studio room you are looking at above requires more square feet per student than a lecture hall. But a lecture class also requires a recitation space and a lab facility. Add those up, and once again the SCALE-UP/Studio room is equivalent.
Does anybody actually achieve those learning gains you show in the graph? Well, we do. The graph below shows our Force Concept Inventory normalized learning gains in our calculus-based first-semester introductory classes (PHY 2048C) from 2015 to spring 2017 (you probably don’t want to see all the data since we started in 2008). We are consistently in the interactive engagement sweet spot. If the Board of Governors was in a good mood, they’d congratulate us for measuring student learning every semester. But they haven’t noticed.
Now the really big question you raised:
Can you deliver a SCALE-UP/Studio learning environment online? Not yet.
There are two key elements in a SCALE-UP/Studio environment. One is face-to-face conversations among students. In our classroom, we set students in collaborative learning groups of three, and seat three such groups at a round table of nine. Our three SCALE-UP/Studio rooms seat 63, 72 and 81 students. Another popular format for smaller classes is D-shaped tables each seating two groups of three students each. A 30-seat classroom at FSU’s lab school is shown below.
The other key element is face-to-face conversations between students and instructors.
One leader of the Physics Education Research community told me during my two years as chair of the American Physical Society’s Committee on Education (2013-14) that physical face-to-face conversations are “high bandwidth” because of all the non-verbal channels that are in play during such conversations. If we could figure out a way to completely reproduce that same bandwidth with a remote conversation, then we could stop making Florida’s legislators all travel to Tallahassee – they could have their sessions remotely. Such an arrangement would eliminate an entire class of problems.
So what’s the best we can do to serve students remotely instead? That was the subject of a master’s thesis written (or rather, designed) by a Harvard architecture student – who happens to be my daughter Allison. She is employed by Overland Partners in San Antonio. Here was her idea: Instead of having all of the tables at which students are seated at a central location, put one (or two) tables at each of several remote locations. If each of the tables had six or nine students seated, you’d still have the face-to-face bandwidth among student peers in the room. And if you wired the room in which the table is located with lots of audio and video pickups and gave the students in the room video/audio access to other student table locations and the instructor location as shown below, you’d still have some of the contact that characterizes a physical SCALE-UP/Studio room.
So in principle, this could be done. But it’s not cheap, and it would require staff at the remote locations to maintain both lab equipment and the audio/visual connections.
And that, Richard, is the best response I can come up with.