research on C'est la Z

CS Research by HS Students

Tue, 28 May 2024 14:02:01 -0400

Last week Devorah and I attended the Stuyvesant High School Alumni Association's annual fundraising gala. We were there as guests of the Alumni Association since, as retired teachers, gala tickets are generally out of our budget range.

We had a great time. I spent time with a bunch of former students spanning the years, some current teachers I hadn't seen since I left Stuy, as well as new Stuy alums and current students. I had a great conversation with one particular young man, a graduating Stuy senior, who's working on an interesting app. He asked if he could contact me to talk about it. I told him, I'd love to so I hope he reaches out.

At one point, I was talking to a current Stuy teaching involved in their research program - that is, the STEM research program, primarily focused around the Regeneron Science Talent Search - Westinghouse to us old timers. She lamented to me that she wished she could get one of the CS teachers into the fold - to basically run a CS version of the other research classes that already exist for Biology, Chemistry, Physics, and Mathematics. Each of those classes target students in their Junior years, hook them up with a researcher, and shepherd them through the process of ultimately submitting a project to the big competition.

I had to explain that I actually did run a CS "research" course back in the day but it was actually a bogus cover course since I didn't think that high school students, even with Stuyvesant's multi-year program had sufficient CS and Math to by and large do good research.

The class was created primarily because, at the time, I had set up and was running Stuy's network. I got neither pay nor time compensation for this. I also typically worked with a handful of students. They learned a lot about Linux and networking and helped keep things running. They also got nothing for their troubles - no grade, no pay.

The deal was that we'd create the class, they'd learn the networking stuff and we'd all have some time to keep things running. The kids would also get a nice high grade rewarding their work. They also had to cobble together a "research" project that they could enter into Westinghouse/Intel/Siemans/Regeneron.

Over the years, the class had some students who wanted to do the research thing and were not so much there for the networking side. Those kids had to do a more legit research project. Now, occasionally I had a kid who was able to do an excellent project. Spike's Sporknet project which was a design for a scalable, anonymous network. Really good work and ahead of his time. Brad's work on Captcha was also really great. The thing is, though, that most kids just didn't have the background to do solid CS research.

Even though Stuy's program was and is arguably the most rigorous HS CS program in the country, by second term Junior year, they're just finishing up CS1 (APCS-A). They'd be studying data structures during the second half of 11th grade. Since the research projects were always due in the early November, senior year courses don't have much affect.

On the math side, some might be taking calculus during their junior year but many will be in precalc. Few will have taken statistics and none will have taken discrete math or linear algebra.

These kids were and are tremendously talented but they just don't have the knowledge base yet. On top of that, on the programming side, they haven't yet gotten to the point where they can write large complex programs reasonably quickly. That doesn't typically happen until some time senior year, or the semester after data structures.

Now, in the sciences, the kids were always hooked up with a researcher in a lab. In the best case, they can do legit research. At Stuy, in all areas, including CS, there are usually a couple of outlier kids who had exposure early on and come in with a wealth of experience in their resaerch field. In the worse case though, these kids become glorified lab assistants. The thing is, though, this can be a really good experience. They are there for the entire process and learn a lot. In CS, the equivalent is basically having the kids implement an already designed algorithm or create part of a software system that isn't really showing them the research process. From what I saw, it just wasn't as rich of an experience.

After a few years, I dropped the network - too many headaches and soon after, got rid of the "research" course. I think it was the right choice. The research course could have continued and been a good experience for the couple of outlier kids each year who came in super strong but that wouldn't be a sensible use of resources with each class having over 1000 students.

So, unfortunately, I wasn't able to help this Stuy research teachers case for CS research in high school. I did point out that the kids might be able to leverage their CS and coding skills in their research in other areas. I had one student who ultimately did Physics research who made extensive use of his CS skills for data analysis and visualization. I believe he was a semi-finalist. I've also had former students do Astronomy research in college and grad school using CS skills they got at Stuy.

So, I guess that's it. Am I wrong on this? I certainly could be. Maybe I wasn't back then but times have changed. Sure, if a school had unlimited resources for unlimited classes, having a CS research class would be great even if it only had a handful of students. With budget and space constraints though, I think a full class of another advanced elective is more valuable.

All this said, I'd love to hear differing opinions.

How my views on education research were shaped

Wed, 07 Mar 2018 00:00:00 +0000

After reading a couple of comments on my last post where I talked a bit about practitioners vs researchers I thought I'd expand and expound a bit.

While there are education researchers that I very much respect, overall, I'm skeptical of education research. Note that I'm not talking specifically about CS Ed research but rather education research in general.

Let's go back to the beginning. I entered teaching from industry. Goldman Sachs to be specific. I started in a shortage area - math and barely had the credentials to do so. The transcript evaluator at the DOE counted my Algorithms and Numerical Methods courses as Math courses to get me over the credit hump.

I had zero ed courses under my belt and was thrown into the fire. Fortunately, I landed in a school, Seward Park High School, with a very supportive faculty. I wouldn't have survived my first year without Mike G, Bruce B, Jonathan G, and many others.

Then, I had to start taking education credits to make my license permanent. Some classes were taught by professors some by adjuncts. The adjuncts were invariably experienced teachers and the professors were research faculty. In many classes, I had to read the latest research.

I found that what "the research" told us to do was contrary to what some of the best teachers did in practice. I kept hearing about the "right" way to teach but it frequently didn't jive with reality. When the true way worked, circumstances were right and the technique aligned with the teacher's natural tendencies. When it didn't, why not? Many reasons.

lack of prep time.
too many students (per class, per day, per term)
different populations than the one researched
factors the researcher didn't include (and there are tons of them)
teachers with different personalities and skill sets

and each of the above points opens up to scores of specifics.

This theme continued throughout my career. I'd ask teachers in other subject areas about the current state of the art research. What was the current trends in their fields. Generally the best teachers would summarize what was going on and then either say that it didn't work for them or that if it did, they didn't take it seriously because in five or ten years "the research" would tell them to do something different and then they would all of a sudden be bad teachers.

This was another point. The right way to do it has changed over and over during my career.

Then there's the reproducibility problem. Apparently it's pretty bad in education research. I mentioned in my earlier post that when a researched does it once on a small group it's research but when a teacher did it over and over it's an anecdote.

When I was young I would listen to a master teacher tell me about how he set up some experience - the one I'm remembering now is an experiential / discovery lesson on triangle geometry. He'd talk about all the hard work that went into prepping his class for this type of lesson as well as the prep time for the lesson itself. He told me how he did it four years ago and it was great but then the year after it fell flat and how he changed it for the next year. How it worked in third period but not seventh and why he thought it played out the way it did. That's a lot more reproduction than most 'research' actually gets and this is what a master teacher does all the time.

Finally, I got tired of hearing how to do it by people who don't actually have to. If you're going to talk the talk, you should walk the walk. As a side, this problem is really bad with "thought leaders."

All this sounds like I'm really down on education research or at least the overarching system. I probably am but there are many education researchers doing great work and I respect them immensely.

They know they don't have a magic bullet:

This is what I tried with this population at this time and this is what happened. You might want to try it or a variation on it.

That's a far cry from "here's how to teach"

The researchers who acknowledge that their experience as college researchers or even teaching faculty are different from a k12 teacher and that no matter how hard they try to work out all the variables they can only ever account for a tiny fraction of them.

The key is that they're adding another drop into the bucket of knowledge. Over time it adds up. I have great respect for everyone doing this.

These researchers have to keep prodding us teachers to re-evaluate our practice and then either take or incorporate the research or reject it after determining if it would be good or not for our population. At the same time, it's up to practitioners to keep pushing back against magic bullet research and the one true way.

We all have a role to play in this game.