The wheel reinvention factory

How long does it take to develop a course syllabus? For a course you taught for many times, it still takes a couple of days: to change the calendar, to tweak assignments and grading rubrics, to update reading lists, to order books, etc. If it is a new course, or a significantly revised one, it can take weeks. That is exactly what I did this week: revised a graduate course I taught once before. It is a labor-intensive process, and results of it are often imperfect. It took me at least five years to get my undergraduate Social Foundations class where I wanted it to be, and I had more time to play with curriculum back then. For this course, I do not have all the time in the world, and simply cannot invest two weeks in it. Considering I may or may not teach it in the future, this might not be a vise investment of time.

Of course when I see something takes too much time or effort, my brain starts churning. After all, every time we do it, we reinvent the wheel. The mental process of an instructor follows more or less the same patterns: What do I want them to learn? (Learning objectives). What kinds of activities and assignments can work to help them learn? What sources have the main concepts and fact? How can I make sure they learned what they need to learn? How to make grading system clear and fair? How to space all of this throughout the course in a logical manner? And finally, how to explain it all in a syllabus? Those are fairly common, repeatable questions, which from a semblance of an algorithm. Moreover, there are only a limited number of answers to each question, and each choice narrows the choices in the next step. For example, most courses' learning objectives include mastering certain concepts, and certain facts, as well as development of certain skills. While the number of concepts is unlimited, the skills are limited: those are skills of reasoning/thinking, and performance kinds of skills (how to present information, to speak, to show, etc.). The facts can also be broken down into, say, dates, names, statistics, cases. But then, if your course is heavy on concepts and their applications, there are only limited choices on how to help people learn concepts. For example, you can give a case, and show the concept at work. Or, you can define it. Then you always want people to use concepts in context different from the original example, so you can be sure students learned to use the concept. Then you can test how sophisticated is their use of a concept by asking to use it in a more difficult case, etc., etc. Class schedule narrows down the choice of activities. A 3-hour class cannot contain only lecture or only discussion; it must have some combination of various activities. Also, there are really a limited number of activities, and they are all good for a specific task. Let's see, there are lectures, demonstrations, large class discussions, small group discussions, debates, simulations and simulation games, skits, teaching segments… OK, probably about a dozen more. Still, not unlimited. Those can be classified by engagement level, and focus on content, by instructor-led vs. student-led. I am just giving some examples, so do not expect complete lists.

This really does look like an algorithm. Can a computer program take guesswork out of this, so I don't forget any important steps? For example, imagine smart software like this: You want to design a course?

• Step One: Please enter the key concepts you want to teach (it looks for definitions, so you only need to pick one out; it also looks through Goggle Scholar to suggest key readings). OK, which skills from this list are the most important? Great. Which facts do you want students to know and understand: Enter names (it churns out an internet search on them), historical events (does the same). Will they need to know any demographic, economic or other facts? Or copy and paste appropriate profession standards, so we can analyze them for concepts, skills, and attitudes.
  
• Step two: it tells you your course is overloaded with concepts and facts; please reduce to make it appropriate for sophomore level.
  
• Step three: Identify your time budget: contact classroom hours, homework hours, grading hours. Here is a list of appropriate activities. Each takes so much time from classroom contact time, so much home work, and so much grading time. If you select one, it reduces the budget until you exhaust it all.
  
• Step three: here is the recommended mix of activities and assignments, based on your schedule. And here is a list of suggested assessments; select from the list.
  
• Step four: check your course, click here to print the syllabus. Click here to generate and edit classroom activities handouts.
  
• After each class, rate activities and assignments, so the system learns which ones actually work, and which do not. Share the learning curve with other professors teaching the same course.
  

The point of automating tasks is that more time can be spent on actually creative, deep thinking about teaching. It is also to minimize omissions, miscalculations of time and effort, poor grading practices, repetitive activities, vagueness of objectives, mismatch between objectives and assignments, between what we teach and what we assess. This is no more complicated than the TurboTax; it uses the same level of algorithmic complexity.

Anyone wants to go in business with me? We can pitch this to a venture capital firm, raise 2-3 mln, and pay ourselves nice salaries for 5 years, waiting for this to succeed. Or fail.