When trying to learn something new, it's very useful to have goals not just in terms of reading and absorbing content (i.e. read this physics textbook) but also to have new things that you can do (i.e. be able to solve any momentum transfer problem). "Be able to pass this test" is a more action-y and exciting way to view knowledge acquisition than just "read this book".

Unfortunately, as everyone knows, standardized tests usually suck. But given how many of them there are, there must be some good ones out there- even if the score doesn't tell you as much as you'd like, then at least you can gauge how you're doing by how confusing or difficult it is for you to work through the questions. (I feel like it would be a good idea to work any standardized test you were studying for your own purposes without looking at the multiple choice answers to really test yourself.) For my purposes, I would say that a good standardized test is one such that, if you actually know the material that the test is testing, you can confidently say that you know the field.

I would say that the Physics GRE is probably an example of a good standardized test. I never took it, but I was a physics major as an undergrad, and looking over the questions I can see that it cuts a pretty good swath across the field. If you know enough to answer all those questions and know why your answers are right, then you have a very solid grounding in the field. I, for one, would need to review some stuff I've forgotten if I were to take it for real (namely optics).

For starting to learn new programming languages, a less conventional "standardized" test is solving Project Euler problems in that language.

I'm particularly interested in examples of this for cognitive science and/or neuroscience, since I'm trying to read a lot about those fields and I'd like to know what's expected of someone educated in them. Also there tend to be many fewer resources for workable problems in these fields than in the harder sciences, which I'm more used to.