Education Heartbreakers (and hiatus)

I seem to be taking an unintentional hiatus thanks to my real work and my personal projects. I might as well make it official – after this post, I’m on hiatus until further notice.

I’m going to use an analogy from gaming to talk about things happening in online education right now. The initial example is from tabletop roleplaying games, but I’ll add a few examples from other areas as well.

The “fantasy heartbreaker” is a concept in RPG design circles. This phrase comes from a post by Ron Edwards over a decade ago, with more detail in a follow-up post and many years of refinement. You need to know a little context to unpack the whole meaning, so let me give you the short version.

Dungeons and Dragons is the great grandaddy of fantasy RPGs. It is well-known and well-beloved, having been many people’s first game. Not everyone enjoys it, but everyone who plays RPGs knows it. Most of the folks who’ve played it for many years know that it has a few idiosyncrasies, oddities, and failings. Many of us create “house rules” that change the game in one way or another.

A “fantasy heartbreaker” is a case where someone takes their house-ruled version of D&D and goes through the difficult work of publishing it, usually with cover text that extolls its virtues in comparison to D&D. The problem with this is that we gamers already have our own house rules that work for our group. This person’s “new” game offers one or two tweaks to a game that we’d then have to re-learn. There’s little incentive to pick it up, regardless of how much “better” it is or how much “simpler” or “more realistic” it is. There are dozens of fantasy heartbreakers littering the gutters of game design. Most of them have one or two great ideas… but one or two great ideas do not make a good game, and they especially don’t make a game that can take down the D&D juggernaut.

To give a few other examples: Imagine that someone who comes up with a motorized bicycle, without realizing that motorcycles, motor-scooters, or vespas existed. Imagine someone who comes up with a “phenomenal, new” seven-card version of five-card-draw poker. Imagine someone who creates a smartphone, but this time, with a physical keyboard. Imagine further that these folks think their inventions would revolutionize the field. They’re earnest. They’re hard-working. They really believe. They really don’t know. It’s heartbreaking.

Let’s coin a related phrase for a few things I’ve seen in the past year: the Coursera Heartbreaker.

A Coursera heartbreaker is an online course delivery platform that purports to “fix the problem with online education” in much the same way that fantasy heartbreakers “fix the problems with D&D.” It does so by recreating an existing course delivery platform to about 90% accuracy, and then adding The One Feature That Will Fix Everything.

(Actually, that’s not entirely accurate. Coding is not fast or easy. These systems recreate the most basic features of an existing system, meaning that for the first few years of their life they will lack the majority of interesting and/or difficult features. They will lack polish, depth, and reliability.)

want to see things that revolutionize education. So do the people who create these heartbreakers. I’m worried that these hard-working folks were inspired by people who knew nothing about the field they were entering and changed it completely by creating something amazing and unexpected. Such things absolutely do happen in this world. Unfortunately, those are one-in-a-thousand events. Not knowing what else already exists is more often a handicap than an advantage. In the case of both fantasy heartbreakers and Coursera heartbreakers, people re-tread the same ground that dozens of others have walked, wasting time losing out because of it. Skipping your research is not a laudable act.

These pieces of software are heartbreaking not because they’re bad, or because they’re unpolished, but because their creators are earnestly putting them in competition against juggernauts without doing their research. There’s nothing wrong with creating them, but there is something unfortunate about parading them around, and at this point I just sort of sigh when I see one.


Side note: Interestingly, there is a phenomenon where some games that might once have been heartbreakers take off and survive. There’s this “old-school revolution” within game design, with games that are throwbacks to older versions of D&D. They’re not huge (well, with one exception that most people don’t consider as old-school), and they’re a very splintered market, but they do fairly well and are well-respected. They’re a lot more like expansions and supplements for older games, intended to evoke a particular feel, rather than games designed to fix a particular problem with the old game.

Old-school education (of many different varieties) is still very much in vogue, so if the RPG market is an indicator, these might have a chance of surviving in a niche market. Most Coursera heartbreakers, though, are labors of love that are doomed by their very design.

This Is Not Science

Today is more about science than education. Here are two links to recent stories of bad science. The second one is rather larger, but the first one reminded me of it.

I’m not going to go off on some kind of “science is bad and the scientific method has has failed us” rant here, not because I have “faith in science”, but because the failures that those articles are describing are not science.

Lack of randomized studies, lack of control groups, pay-to-publish, prestige publishing, these aren’t science. They’re trying to wear science’s clothing like it’s some kind of game. It’s not a game. People use these results to create, to build, to teach, to treat and heal. When someone uses a medical study to treat patients and then finds out that the study was a fraud, the cost for that is measured in corpses.

I’m against people pretending that their research is solid when it turns out to be primarily opinion. I couldn’t care where they’re faking it. Weakening one area of scientific endeavor weakens the whole, even if only in the realm of public perception.

There’s a comment from the second article about the need for “naming and shaming”, and a suggestion for a Consumer Reports-style journal that reviews other journals. In the interest of that, here’s a link to Jeff Beall’s list of predatory journal publishers.


The two boarding schools I worked at, Northfield Mt. Hermon and Hyde School, both lost some great people in recent weeks. NMH lost David Demaine, and Hyde lost Paul Hurd.

Paul died in a car crash – he had apparently had a heart attack while driving. Luckily, no one else was injured. I barely met Paul, but I knew him by reputation. He was one of the Old Guard at Hyde, a staunch advocate of character education and someone who really knew how to jump in with both feet.

David died due to complications from cancer surgery. His wife Gail died in January of last year. I knew David tangentially from my time as a student – Gail was my Sophomore english teacher, and their son David and I took Russian together – but also from my time as a faculty member. David and Gail both were very warm and helpful to me in my year at NMH. They always had time to listen, and were there with a hand on the shoulder when I needed encouragement.

Boarding schools see their share of grief. Tragedies can take children and adults of all ages. As both a student and teacher, I appreciated that these diverse but tight-knit communities would come together in the face of loss, to remember the good that people brought to the world. Some people had their church or family or neighborhood to turn to. We had our teachers and classmates, students and colleagues, and in that time they were family.

I think you never really know someone until you see them through someone else’s eyes – it is a shame that we so often wait until people are gone to share what we see in them. Hyde and NMH are both poorer for the loss of these teachers, and richer for the things they left behind.

David’s obituary

Joe Gauld’s goodbye to Paul


“Lots of students” is not a course format.

As many of you know, I’ve worked on a mid-sized online course. Mid-sized means about 15,000 registrants – we’re much larger than most distance-education courses, but smaller than the really huge courses with 50,000-100,000 registrants. It’s interesting to note that in terms of students who pass the course, we’re on par with some of the bigger courses at ASU or Texas A&M. Massive indeed.

There are a lot of articles about MOOCs these days, including those that say (already, not two years into the process) that their days are numbered. There seems to be a fair amount of confusion about what makes something a “MOOC” and what that means about the course itself. Let me break down the acronym.

M – Massive. Large numbers of students.

O – Open. No prerequisites, no cost, preferably reusable materials.

O – Online. Fairly obvious.

C – Course. Intended for education.

Nothing in that acronym says anything about how the course is to be run. There are some substantial differences between how, say, 8.01X and PoetryX are delivered. Both follow the typical lines for on-campus courses for those two departments, but 8.01 is a physics course that involves a lot of lectures, homework, and mathematics, while poetry is taught in a more conversational manner, focusing on discussions and writing assignments. The online versions of these courses follow the on-campus versions fairly closely.

There are two reasons that most MOOCs follow similar formats, with some combination of lecture videos, texts, discussion boards, and computer-graded questions. One is that some of the professors who want to move their courses online are most interested in duplicating the on-campus experience as closely as possible. The other is that the professors who want to do something unique and different find themselves stymied by the available tools. edX is one of the more flexible systems out there, and still I complain a lot about its lack of flexibility. I’m working on a few more courses right now, mostly in advanced areas in the sciences. The professors I work with want to create something exceptional and useful and powerful and – frankly – not that difficult to do, and the system isn’t up to it yet. (Yet.) The more powerful the underlying systems become, the more unique and interesting things we’ll be able to do. We just need to keep pushing for more development and more features.

Don’t confuse “class with lots of students” with “class involving lecture video and computer-graded questions.” We do little courses too (SPOCs – Small, Private Online Courses) and we’re limited by the same tools. Within these tools we have some flexibility, but we’re stretching and pushing for more as hard as we can. If we manage to keep the same enrollment numbers, the definition of MOOC will spread out fantastically as we develop more and better tools.

If the numbers drop, people might claim that “the MOOCs failed,” but there’s nothing in that acronym that’s a process – there’s nothing there to fail.

College Acceptance Required?

Corbett High School in Oregon requires students to get accepted to college in order to graduate from high school:

What a bizarre requirement. They don’t have to go, they don’t even have to want to go. They can still choose to go into the military, travel the world, get a job, what-have-you. They just have to fill out at least one college application and get accepted. (Oregon community colleges apparently accept everyone who applies.)

The phrasing emphasizes the words “every” and “all” to the point where I almost cut-and-pasted the whole thing into something where I could downcase all of the letters. The word “choice” appears continuously, in what have to I assume is unrecognized irony. Yes, yes, I get the idea that removing students’ choice of whether or not to apply to a safety school opens up a choice for them later on, but I doubt that any of them were unaware of that particular choice.

This school, in particular, is likely to have a ~96% college entry rate to begin with. I’m not sure why this wasn’t just handled as an internal college counseling requirement, but perhaps that wasn’t easy to do with their current setup.

It’s fascinating to compare this to events in Texas, where Algebra II just came off the graduation requirements.

Wonderful quote from that one:

A 2003 Stanford University study (PDF link) of six states found that less than 12 percent of high school students were aware of course requirements for their local universities. In fact, simply mailing high-achieving low-income students more college-enrollment information increased the number of applications those students sent to selective colleges, researchers at Stanford and the University of Virginia recently found.

I added the PDF link; I’m fairly sure it’s the one they’re talking about.

I would say that I have nothing against every student applying to college, but I know how overloaded college admissions officers are. Applying for a dozen schools is not unusual these days. Even the college board now says “five to eight is usually enough.” Even if things stabilized at that number (and the average number has been climbing), if every high-schooler in the US were required to apply to college, the number of applications would be essentially impossible to handle well. From everything I’ve heard from admissions folks, the applications they receive now are not handled well, due primarily to a lack of manpower.

This post intentionally left without a concrete conclusion.

Post-Turkey-Day Coma

No edublog post today. Will return on Monday.

Wolfram Problem Generator

There’s a joke in here somewhere along the lines of the Java ProblemFactory joke, but I digress.

Wolfram, the folks who make MathematicaWolfram Alpha, and various oddities, have come up with a Problem Generator for certain branches of mathematics. If you don’t want to sign up for a trial of Alpha Pro to try it out, I’m going to give you a quick overview here.

The short version is that they’ve created a random problem generator that takes a template, fills it with random numbers, and asks a straightforward procedural problem similar to back-of-the-textbook practice problems. No word problems, just “find where this is discontinuous” or “find the median of this list of numbers” or “factor this expression.” You can ask it for easy, medium, or advanced problems. Advanced problems involve doing a little more of the work yourself (unfactored expressions in rational functions, for example) and more intimidating numbers with square roots and such. They provide step-by-step solutions that use the exact random values given to you.

In terms of drilling exercises, this is fairly handy. Some of it is higher-level material, so that’s nice. There’s arithmetic, number theory (very basic – factor this, LCM of that), high-school algebra, single-variable calc, linalg (add, subtract, cross, determinant), and super-basic statistics (avg, mean, mode, and range).

The lack of educational sophistication, however, is what makes this less than impressive to me. It’s more of an aid to textbook manufacturers than it is to students or teachers. You can practice doing integrals until your hands fall off, and lord knows people need practice with their skills, but as we all know at this point, Students do not overcome conceptual difficulties after solving 1000 traditional problems. (Yes, that’s in italics because it’s the title of a paper.) You won’t know what an integral is any better afterwards than you did before.

It’s also not extensible. Where’s the tool that allows us to make our own problems? How do we create our own templates? Will this list ever get expanded, or is this just something that someone came up with on their 10% time?

This might be a good product for textbook manufacturers, so they can create books faster and more accurately. It might be good for math teachers as a way to get quick problems for their tests, things they can wrap words around to make real problems and not just exercises. It’s good for drill-and-kill practice, to the extent that such things are necessary. If you’re saying “But we already have all of that stuff in thousands of books and practice sheets…” then you’re hitting on the exact reason I don’t think this is very useful.

We should be working to figure out good ways to use tools like Alpha in the classroom. Right now I just have this bizarre image of someone just sitting there with the Problem Generator open on one page, and Wolfram Alpha open on another page, plugging questions from one into the other and returning the answers.


For those who aren’t yet familiar with him, the rarely-updated blog of Edward “Joe” Redish:

Redish is a very big name in physics education research. My favorite post so far is “The World is an Ill-posed Problem”.

Teaching as Acting

Teaching in the classroom requires a fairly broad set of skills. As an absolute minimum, it requires two: the ability to relate to your students, and knowledge of the subject material you’re teaching. We can break these down in all sorts of ways (classroom management and lecture delivery are both part of relating to your students), and add many other skills (lesson planning and knowledge of misconceptions, for example), but relatability and subject knowledge will do as a minimum.

When you’re in an online environment, putting together a course on your own brings in other skills. Not many people have all of them, so online courses are typically designed by a team. You need one person to handle technical matters, one to write, one to supply subject knowledge, one to interact with the students, and (assuming you’re using video) someone to edit the videos and someone to appear in them.

It’s not unreasonable to replace that last person with someone who has no content knowledge. After all, in a molecular biology course, your techie doesn’t need to know how to… uh… yeah, I know nothing about molecular biology, so I can’t even come up with a reasonable example. Your video editor doesn’t even need to know what protein biosynthesis is, let alone how it works. (Thanks for the example, Wikipedia!) So why should the person delivering the speech need to know anything? They just need to stand and deliver, and do so in an entertaining and captivating manner. Your writer and subject knowledge expert can handle the scripting.

If this seems strange, it’s because we expect a lot from classroom teachers, and we consider the production of an online course to be something quite like teaching. In fact, it’s something quite different. So if Matt Damon wants to do a guest lecture for PoetryX, or Tilda Swanson wants to put a few hours on film for Fundamentals of Neuroscience, why would we say no?

I understand that there’s a concern about this appearing… less than genuine, I suppose. If the content is accurate and the delivery more stirring, I say go for it anyway.

An issue of honesty

You might have heard about a month ago that George Washington University has been lying about its admissions process.

I have difficulty writing about this while keeping my temper, so I’ll try to be brief. This is not an issue of one person’s definition versus another’s. This is not an “everybody does it” issue or a “internal process, not a public matter” issue. This is a school that advertised one thing, told it to even its own workers, and then went and did something else.

There’s nothing wrong with considering a student’s ability to pay when you take them in. It sucks, but you have to pay the rent. (And the salary, and the overhead, and benefits, and and and…) But you can’t say “I’m covering my eyes” and then peek through your fingers.

Now, to compound their… it’s not a mistake, it’s an outright lie… they’ve decided to double down on the “Oh, this is just a matter of definition” defense.

No. It’s a lie. Own up, get honest.