Homework, exam problems and solutions in a global social network

Here is a simple idea that will be transformative to how we teach physics (and probably many other fields).


In physics, working problems is a key part of learning. When a student is having trouble understanding a particular thing, often the best advice is to work more problems on that topic. As a teacher, it is often difficult to come up with as many good homework problems as the students could use. Yes, there are books of problems out there in some areas, but we are likely to think that only some of them are good ones, and it might take some doing to identify which ones will help a particular student who is struggling on a particular topic.


I propose a global social network that will operate in the following way:

  • Each physics instructor will upload all the problems they create (along with solutions if they have them) to the network
  • Some kind of social network technology is used for the system (perhaps just users checking off keywords and rating how similar problems are) to learn which problems are like other problems, so students can work as many problems as they need on a particular topic.The system can also learn which problems different instructors are most likely to assign.
  • Instructors can personally annotate problems and solutions, and students can access the annotations of any (or all) instructors.
Further Thoughts
  1. Clearly the job of assigning problems for credit (exams, homework or whatever) will have to change (this will be necessary anyway, since the above arrangement is likely to happen one way or another, and in fact element of it area already in place). My first thought is to have special exam rooms that are Faraday cages and are offline.Then exam problems can be assigned straight from the network, but students will not have access to the solutions.
  2. Students will have to confront the temptation of sloppy study habits (turning too quickly to the solutions) but if the examination process is rigorous, they will have to learn what really works, perhaps at first the hard way.
  3. The fact that a student might have already done a problem that appears on the exam will be mitigated by the large number of problems that are present in the pool (the large phase space). This approach is already used in lower tech implementation by some instructors.
  4. I would much rather put my time into contributing several really great problems per quarter to this system, vs scrambling to create or find as many problems as I think my students should have.
  5. I expect that in such a system, the roles of instructor and student could potentially blur somewhat, but some system would still need to be in place to separate good stuff from bad.
  6. I am convinced that should such a system be built, essentially all of my physics colleagues would gladly join in. We all seem to see this aspect of teaching as very much a team effort, not something proprietary to be carefully protected.
  7. Whereas there are lots of open questions about the many ways technology may or may not change teaching in the future, I suspect this idea is sure to work, because problem solving will necessarily be part teaching physics (and many other fields) no matter how other aspects of the teaching process may change.
  8. I believe that what I propose would be possible with existing technology, and I’m hoping it would not be too hard and that some expert might just go and do it (as happened with the ArXiv).I do not have the tech skills to do that, but I would be happy to contribute in some appropriate manner. I would be happy to help apply for NSF funding, for example (although I suspect this project could just emerge as someone’s side project, without an official grant or status.)Please get in touch if you think I might be able to help, or just go ahead and make this thing happen!

First posted July 2015, and updated periodically