Drupal distro for designing courses

So what you are proposing is a system to create and serve content like that provided by Carnegie Mellon's Open Learning Initiative: https://oli.cmu.edu/

Not really. Two differences:

• OLI (last time I looked) was about tools and infrastructure. It wasn't about good learning. That is, authors aren't given a course building method that's based on learning science principles. For Dolfinity, the Drupal distro is only 1/3 of the project. There's also (1) a guide to course building that improves as learning research advances, and (2) a community of dolfineers.

• Dolfins are not open. Authors can charge for them. IMHO, that is essential for creating sustainable, high quality dolfins. Note that because authors get all the revenue, they can afford to charge low prices, and still make a decent living.

(Aside: when all content is free, the only people who can afford to create content are those whose efforts are subsidized somehow. That's not a stable situation.)

Suppose I, as an author and instructional designer, want to create interactive learning objects in Dolfinity and make them available to others via LTI. Is Dolfinity going to handle brokering these objects? Or do I need to set up a payment and fulfillment system?

My vision is that Dolfinity would support LTI out of the box. That solves many problems.

Not sure exactly how the charging system would work, but Dolfinity would have a basic mechanism in place. The trick is to link an institutional OAuth account to a license. That should be in Dolfinity, but I'm not sure how.

In any case, I'm imagining Commerce would be part of the distro, with instructions on how to set up a couple of payment gateways.

With the model you are proposing, I think it's going to be a struggle to get higher ed institutions involved unless faculty can just designate a collection of "dolfins" as the textbook for their courses, and require students to purchase access or whatever. Among other things, that means there will need to be a facility for faculty to review content and see student feedback, learning effectiveness data, etc.

Agree with all of that. Independent profs are the key, those who can adopt a "textbook" without committee approval. Fortunately, there are many such humans, at least in universities.

Community colleges often have centralized adoption decisions. However, with the flipped nature of dolfins, and their ability to offer quality, branded courses in remote locations, there will be good reasons for admins to consider dolfins.

Content review features should - maybe not in the first version, but eventually - allow faculty to:

• Annotate dolfins and share their annotations with students. Include links to their own material.
• Mark sections and exercises as required.
• Add their own favorite exercises, that would be merged into the standard exercise system.

If by "feedback" you mean "students' opinions about Dolfinity," I can tell you that some will be negative. Dolfins, ideally, have lots of exercises. My experience in teaching with dolfins over the past few years (http://coredogs.com was the first dolfin - you can try it) is that students are surprised by the amount of work. IMHO, it's a Good Thing that dolfins keep students on task, but some students aren't used to working.

Of course, this is a problem for schools that only use student opinion surveys to evaluate faculty teaching. Nothing we can do about that.

(I don't know as I have any particular interest in developing content for this model, honestly, as I'm focusing more on developing rich simulations, rather than static content objects.)

Here's something to think about: are students getting the most out of your simulations? Some instructors will use them as busy work. "Here, do this set of steps, and take a quiz." Students don't learn about the simulated processes.

Suppose you wrote a short dolfin, that helped students learn about the simulated processes. Guided them through the Big Ideas. Helped them explore pedagogically important scenarios. Would that be worthwhile?

May I ask what simulations you write?

I've always enjoyed writing simulations. I'm one of the dudes behind Visual Statistics, a stats learning package in the '90s. We did a lot with simulation. For example, students set up a group comparison scenario, then ran ANOVA simulations, and got the distribution of the test statistic. Fun stuff. Also simulators for basic probability, regression, quality control, other stuff.

Actually, we wrote a workbook for Visual Statistics. Exercises to help them learn about stats. Otherwise, they would press the buttons, and "Ooooo" at the pretty pictures. :-)

Thnx for the comment. It helps.

Regards,
Kieran

Aye, I do know about peer feedback research. Not an expert, though. Mazur's research on that is quite interesting.

Do you know about Francom et al. (2009) in TechTrends? Part of that paper was peer interaction in Bio 100. Just got the paper, haven't read it yet. Merrill is a coauthor, so it should be interesting. If you want the paper, email me at kieran@dolfinity.com - I'd upload it here, but am not sure that's kosher.

Sorry, rambling. Anyway, the first dolfin was CoreDogs at http://coredogs.com. It's been in use for a few years. It had peer evaluation built-in from the beginning, to some extent at least. There's a little about that at http://dolfinity.com/coredogs-internals.

However, it's only a start. I would LOVE to make peer interaction a focus in Dolfinity. The research is too strong to ignore it. Not sure the best way to implement it. We'd have to work that out.

Kieran

Good questions.

What kinds of artifacts can students submit as evidence of learning?

Students complete exercises. Exercises are embedded in the content. There are screen shots at:

http://dolfinity.com/coredogs-internals

http://coredogs.com/article/feedback-system

Exercises are active. For my Web courses at CoreDogs, students make pages or parts of pages, and upload them to a server. Then they submit URLs. My most recent class submitted 97 solutions per student.

For CoreDogs, text was sufficient. But...

• For a database design course, students would submit diagrams as well as text. The SVG module is way cool, or they attach files from, e.g., Visio.
• For a math course, where students must show their work, could be (1) using a drawing tablet (Bamboo, for example), and submitting the resulting graphic, or (2) taking a photo of handwritten work with a cell phone, and texting it in.
• For stats, it would depend on the tools instructors have students use. Say they use Excel. WYSIWYG editor, plus file attachment. Maybe they use R. Editor and file attachments. Using Google spreadsheets? Text, with URLs.

Will Dolfinity support self-assessment, peer assessment, assessment in a discourse context?

Peer assessment: that's part of CoreDogs already, to an extent. That would need to be improved, with input from people like yourself.

Self-assessment: depends what you mean. I'm thinking that there would be a quiz feature that students could use for self-diagnosis of understanding. Is that what you mean?

Discourse: already been speaking to people about teaching writing with Dolfinity. It's a good fit. I don't see why conversations couldn't be assessed, if you can make a rubric.

And if Dolfinity is presenting some kind of prompt for a user response, which the instructor will then manually review and give feedback, will Dolfinity track submissions and status (e.g. what still needs to be reviewed, who needs to be reminded to submit their work)

Aye to all. One proviso: the instructor may not be the one giving feedback. Could be a TA, or even a grader in another country.

pass that data to an LMS, or leave it up to the instructor to decide?

Pass the data to an LMS, is what I am thinking.

discussions that I initiate, monitor, and encourage, and real-world projects largely designed by my students, in which they submit an annotated spreadsheet that answers a question they are prompted to ask for themselves.

You seem more pedagogically advanced than most statistics instructors. Kudos!

I've been imagining dolfins replacing the textbooks and lectures of more traditional, textbook-bound statistics courses. However, your approach would be more effective in helping students learn proactive problem solving (I'm guessing). It'd be great to add to Dolfinity features that would let people create dolfins for courses like yours, to spread that approach.

I'm hoping that people like yourself will help determine how Dolfinity is constructed.

Currently I do all this in Moodle.

There is nothing you couldn't do in Moodle that you could do in Dolfinity. It's a question of effectiveness, efficiency, and revenue.

• Moodle does not help you use learning science research findings to design content and course. Dolfinity will. The Drupal distro is only 1/3 of the Dolfinity project. There's also (1) a guide to designing dolfins, and (2) a community to help. Authors' creations would implement Deep learning, Oucome-based Learning, formative Feedback, Interaction with experts, and Nudging students - DOLFIN.
• The Dolfinity feedback system is specifically designed to let graders give formative feedback quickly, and in a standardized way. You can do the same with Moodle, but it isn't optimized for that.
• There's a business model that is an important part of the Dolfinity project. Dolfinity will have Commerce built-in, so authors can charge for their work. You can do that with Moodle, but it's not a main thrust of the product. The business model is discussed at http://dolfinity.com/new-business-model.

What would I use from Dolfinity in this scenario?

You could be an author. You could package your content and methods in a dolfin, and let other instructors use it. You could give it away, or charge for it. You would get all of the revenue. That means you could afford to charge low prices, and still have a good income.

Instructors use your dolfin instead of a textbook. They get good pedagogy without having to invent it themselves. They could annotate your dolfin, and share the annotations with their students. They could stop lecturing, and become problem solvers. More fun for them, and better for students.

Thanks for your comments. Useful. I hope you'd consider being part of the project.

Regards,
Kieran

I can rant, that's so. But it's considered ranting, not random.

There is a ton of research into learning and providing tools for educators and students.

This is true. I've read a ton of it myself. Dolfinity is based on it. Well, a subset, anyway.

These methods are based on the end goal of obtaining skills set through standards and research such as Common Core State Standards (CCSS).

IMHO, this is not true.

Common Core claims to be research based. If that's true, they've been very selective in choosing research. I won't rant about it, though I did a while back (can't find that article just now). My conclusion: CC hurts skill learning.

To your main point: publishers have processes for authors. There are standards, reviews, and other stuff.

I know, I've been through those processes, as an author, and as a reviewer. The problem is, they're broken processes.

None of the people involved in making a typical college textbook know much about learning. The authors - usually profs - are experts in their field, but have not studied learning. The research you mentioned? They don't know it.

The reviewers are other profs, with the same limits. The technical editors don't know about learning. The copy editors don't. The media people don't. The project editor doesn't.

This doesn't affect sales, at least not much. Why? The publishers are selling to profs who don't know much about learning either.

There is only one publisher I've worked with that uses learning research: O'Reilly, for their Head First series. You can read about their approach at http://shop.oreilly.com/category/series/head-first/formula.do .

I want to replace the normal textbook building process with one that works. "Works" means that it creates textbook-like-things that help students learn skills effectively and efficiently. A process based on learning research.

Keep in mind, Dolfinity is not just about a Drupal distro. That's only 1/3 of the project. There's also (1) a guide to content and course design, updated as learning research advances, and (2) a community to help authors create effective content.

I've not written a K12 textbook. But, as a parent, I've been on the receiving end. My daughter's 11th grade math text was, er, flawed, to say the least. Flawed compared to the way math is taught in Japan or Singapore. BTW, Google "Singapore math" for products that are used in a few US schools.

Regards,
Kieran