I'm Kieran Mathieson, geek, educator, and innovator. This site is about my work on learning.Kieran

We know a lot about learning. Researchers have studied feedback, motivation, deep learning, lots of other stuff. We can help students learn about technology, social issues, critical thinking, and working in teams.

Universities don't use this research. Learning is less effective and efficient than it should be.

I want to:

  • Build open source software, that implements ideas from learning research. Authors use it to make Web sites that replace textbooks.
  • Help authors learn about learning research, so they can help students learn more effectively and efficiently.
  • Help authors make a living from their work. If it's a full-time gig, they'll have the time to do quality work.
  • Help professors improve course outcomes, without being learning research experts. Keeping up with their own field is hard enough!

My main project is CyberCourse (http://cybercour.se). It's about half complete. It extends the venerable CoreDogs. CoreDogs is a complete (almost) intro course on Web tech. It implements deep learning, outcome-based learning, formative feedback, metacognitive awareness, and other things.

The project is going well, but to truly succeed, it needs a group of geeks, teachers, and authors. Right now, it's just me, with some fans cheering in the background.

If you like what you see on this site, maybe you can help get a community started. Just a little seed of one, to begin with. Any ideas on how to make this happen? Want to be a partner? Please let me know.

Frustrated studentUniversity students learn skills, like programming, writing, and problem solving. Skills are important not just for employment. Problem solving, for example, matters in all aspects of life.

Doing is what matters. After a skills course, students should be able to do tasks without help. Ask them to write an article, and they should be able to research the topic, draft an article, and write it. Independently.

It's broken

At many schools, courses like programming and math don't work very well. The courses have high failure rates. Even students who pass don't learn what they should. Give them a task slightly different from the exercises in the textbook, and they're flummoxed.

It's not just tech courses. Students should learn how to work with other people, communicate, and think critically. These "soft" skills aren't easy to learn. As the saying goes, "the hard things are easy, and the soft things are hard."

We can do better.

We can help students learn skills effectively and efficiently. Researchers have figured out how, but few universities use their findings. We should change that.


  • Student time and money. We need to give students as much as we can for every hour and dollar they spend. Fortunately, we can improve skill learning without breaking students' backs or bank accounts.
  • Faculty time. It's hard enough for professors to keep up with research in physics, computer science, finance, etc. Ask professors to be experts in learning research as well? It's not realistic. Fortunately, we can improve skills courses without driving professors crazy.
  • Administration. Change is hard for any organization. For universities, with tenure, academic silos, traditions of independence, etc., change is almost impossible. Innovations relying on significant change may be doomed from the start. Fortunately, we can improve without changing universities' administrative habits. Semesters, courses, sections, credits, all stay the same.

The bottom line

The CyberCourse initiative can help universities help students learn skills. It starts with existing learning research, which shows us how to help students learn. However, learning principles have to be implemented with care.

  • We need an instructional model that doesn't burden students, professors, or institutions.
  • We need a business model to create and maintain the instructional model, without increasing costs for anyone.
  • We need tech to pull this off. However, we can't add a new burden to university IT departments. Many are stretched too thin as it is.

You're saying to yourself, "Self, this can't be done!" Yes, it can. It takes some creative thinking. We need to unbundle professors' jobs, rethink course design, replace the textbook with something better, and support a new kind of educational entrepreneur. Read on to learn more.

Perhaps the most difficult change is to be realistic about what it takes for average students to learn skills. Real skills, not memorizing enough to pass a multiple choice exam. If anything stymies CyberCourse, it will be the belief that what passes for skill education today is sufficient.


Imagine Web sites called CyberCourses that replace textbooks. CyberCourses implement ideas from education research, like:

  • Deep learning. Don't just memorize facts. Learn how to do things with those facts.
  • Outcome-driven learning. Focus on what students need to learn. Drop everything else from the course.
  • Formative feedback. No multiple-choice tests. Students do tasks. They get specific feedback on how to improve.
  • Personal interaction with experts. Students have one-on-one help when they get stuck.
  • Nudging. If a student falls behind, she gets an email. "Are you OK? Do you need help with the course?"

What's in a CyberCourse? It varies, according to topic, student preparation, etc. I've made CyberCourses for learning geek skills. They have text, images, videos, and lots of exercises. Courses work like this:

  • Students read content online.
  • Students do many exercises. They submit solutions through the CyberCourse.
  • Students get feedback on each exercise. Not just a score, but a list of things to improve.
  • Grading is done by instructors, teaching assistants, or outsourced to cheaper workers. An optimized work flow means grading is fast, and standardized.
  • Students get one-on-one help when they get stuck. There can be regular class sessions, or help-on-demand. CyberCourses are good for flipped classes.

The result? Students get value for every hour they spend. Less busy work, more learning.

Instructors also use their time well. They only do things that only people can do, like showing students how to improve. Computers do the rest.

Dolfins I've made are at CoreDogs.Com.

I made CyberCourses, and have used them in my university classes for several years. Other people have used them, too. You can try my CyberCourses.

The CyberCourses work well, but have one Big Problem. I created the authoring and feedback tools just for my own use. The tools are OK for geeks, but not for normal humans.

I'm rewriting the software, and will release it as open source. Authors could write CyberCourses on their own topics, for their own audiences. They could give away their work, or charge for it.

The software is not enough by itself. To make a good CyberCourse, authors need to know about learning research. That's the second part of the project, a CyberCourse on CyberCourse design.

This would all work better if authors, instructors and students shared their knowledge. We need a community of CyCos. (That's better than CyberPeople. Too much Doctor Who there.)

That's the CyberCourse project:

  • Rewrite the software. Make it open source.
  • Write the CyberCourse CyberCourse. Make it open as well.
  • Start a community. Neighbors helping neighbors write, teach with, and learn from CyberCourse.

It's going well. The software is about half written. You can try a version at http://demo.cybercour.se.

The project won't take off without a community of geeks, teachers, and authors. Building community is not my strong suit. Is it yours? Want to help make it happen? Have other ideas? Please let me know.

AuthorWho writes CyberCourses? You. You choose the topics. Make your CyberCourses free, or charge for access. If you charge, you get all of the revenue.

You could supplement your income. You could even make your entire living this way. You would have time to make the CyberCourses excellent.

Write for mass markets, or niche markets. Imagine:

  • A math CyberCourse for Detroit high school students, with Detroit examples, Detroit assignments, and Detroit personalities.
  • A business writing CyberCourse for aerospace engineers, with exactly the writing tasks they need.
  • A biology CyberCourses for one university in Florida, featuring species that are around campus.
  • A human resources (HR) CyberCourse for one company, helping employees learn the policies and workflows they need to follow.

Publishing companies can't afford to serve niches this small. You can, because you don't have printing costs, distribution costs, a sales force, middle managers, a board of directors, or a big building downtown. You can charge a low price, and still make a good income.

This isn't a project for one person. It needs a community of geeks, teachers, and authors.

The trouble is, I'm not good at organizing things. Thinking, building, writing - I can do those. Organizing - not so much.

Any ideas on how to make this happen? Please let me know.

What now?