Symbolic Programming Vs. the A.P. Curriculum | Hacker News Reader

Symbolic Programming Vs. the A.P. Curriculum

el.media.mit.edu

18 pointsb-man5 comments

For anyone interested in teaching programming, this one is a must read

> Now they need a kind of apprenticeship: real tasks to work on, the freedom to experiment and to risk failure, but in an environment that models respect for commitment.

As someone who just got out of an AP Computer Science class, I could not agree with this statement more despite the fact that it was written over 15 years ago. While I had an excellent teacher and really enjoyed the class, I found that in the textbook and on the test that we were simply required to demonstrate that we knew the basic components of the Java API, not necessarily how to implement them. Even on the free response questions, the test writers held our hands through unrealistic problems that tested whether or not we knew what a HashMap was and if we knew how to iterate through it.

The GridWorld case study doesn't help either. In their failed attempt the create a more "real world" problem, they created more questions that could be answered on the ability to memorize what classes in the program do what and not on how well you understand programming concepts.

Sure, it makes sense for Google to "test" their job applicants with a programming problem, but I honestly doubt they ask "The following program will throw a (blank) exception".

Perhaps The College Board should structure their curriculum similar to their Studio Art curriculum, where students send in pieces of their portfolio to be judged and "graded". Students should be given a real world application to create or a difficult puzzle to solve that must be turned in at a certain date to be given a grade. Obviously there would be a bit more involved thinking in that process, but I really think that would provide a much more meaningful experience.

Any person can memorize a bunch of rules from an API, but a real "5" student can actually apply those rules and manipulate them to solve any problem thrown at him.

Although they obviously have room for improvement, rigid schools do not lend themselves to a more flexible approach like this one.

To use a more extreme example, let's say a general wants to introduce more flexibility in the way his military is run. How far can that really go?

That's exactly the point of constructionism (the epistemological school that the article's author subscribe to). The objective of the educational system is to form citizens capable of critical thinking and independence, something that a soldier is very much discouraged to do, so your analogy is exemplary of an incorrect approach for the educational system (which is our current rigid model, of course).

I don't oppose constructionism at all, but I believe state-run schools are fundamentally incompatible with real flexibility. The author's website suggests he is a Marxist -- I don't know if that means state socialist or anarchist in this case, but if it's the former, he would probably disagree with me.

I take a libertarian stance on that too man. But I also follow a pluralist approach on how people organize their own systems, so if people want to trust the government (specially in very miserable and risk intense areas I do believe the government can play a very important role to set up a framework that could be built by incremented the individuals).

Reply on news.ycombinator.com