Thought-provoking article on educational software thanks to Aaron:
I'm not saying that educational software should encourage mistakes, or have outright flaws, such as buttons that are confusingly labeled and crash the program. But I am saying that programs should be rich enough to allow both right and wrong paths to be followed, and followed in a more than superficial way.
Software can be divided into page-turning vs. simulation-based. Page-turning software, which is very common, allows the student to following only certain pre-determined paths through screens that have been laid out in advance by the authors. Breaking out onto more creative paths is impossible, because there is nothing to break out into: It's a closed box.
Simulation software in contrast has a set of rules and algorithms (the laws of physics; sociological models; geometrical relationships; etc.). It is able to apply these rules to a fairly open-ended set of inputs. Simulation software is inherently more difficult to write; not surprisingly there is little of it out there. In fact, good simulations are so difficult to create that one could almost name all the examples that have ever been done. Page-turning software is incredibly easy to make, and there are countless thousands of titles available, virtually all of very low quality.
Interestingly, while violent video games may be evil, they are largely simulation-based. No page-turning software could hold anyone's attention for the hundreds of hours that video games capture our children. In a good video game, you have a huge world to roam about in, complete with back alleys, multiple levels, and great detail in all the parts. The only problem is that you have to keep killing people to see the next back alley.