I still fail to see why there is always this choice proposed that you can either have a educational system driven by test scores or have teachers who are able to teach students correctly, but not both.

Time and time again I hear the argument that teachers now have to "teach the test". If it's possible for a teacher to "teach the test", and students can pass it just by route memorization, well then surely these are some pretty horrible tests!

It's rare to hear people want to reform the actual format and content of the tests to ensure that "teaching the test" is less possible. You could argue, for example, allowing more free-form answers, or creative test formats that allow students to display critical thinking skills to the grader would do just as much to reduce the "teaching to the test problem" as throwing out tests as the primary means of evaluating performance, but you don't really hear this argument all that often.

Admittedly I know little about it, but I find it pretty hard to believe that the best teachers, who do engage their students and provide them a solid education, end up with classrooms of worldly students that end up failing standardized tests.

I agree, the issue as it's argued tends to become a false dichotomy. I've always found testing to be one of the most effective techniques for learning. That's not to say standardized tests are most effective types of tests for learning, but they certainly have their place in a public educational system.

But hell, what do I know? I also think rote memorization gets a bad rap.

Standardization is orthogonal to the issue of a well designed test. A simple illustrative example, based on a test of height:

Poor test design: no rule against standing on your toes.

Standardized test: all results must be reported in meters, as opposed to subjective units ("a wee bit taller than the person giving the test").

In one sense, of course, you are right. But I was using the term as a shorthand for the state standardized tests used to measure academic progress grade by grade (e.g., CAT, TAKS). These state standardized tests tend to measure the minimum requirements for a grade level.

But for a test to be a learning technique, it must challenging enough to make the students study new material. Otherwise, it's just an excercise in "monkey see, monkey do."

Personally, I think the future of standardized tests will be like ACT's Compass college placement test, which progressively adapts to the level of the student as he/she is taking it, yet still results in a comparative standardized score.

What? I know Ivy League graduates that can't pass state issue driving tests. I've read articles about computer scientists that can't configure a thermostat. Badly designed marginal tasks are not an accurate measure of person's performance.

The opinion card trumps the opinion card ad nauseum. However, I prefer well written (and civil) opinionated film reviews if for no other reason a strong opinion creates discussion and interest.

Further, specific and concrete examples of "good teachers" presented by either the film maker or reviewer would have been absurd (but the review's point was the the film didn't even attempt to address the issue of a good teacher). The notion of a "good teacher" is absurd, I think, because it doesn't exist. Consider your kindergarten teacher and advanced math prof. One was probably nurturing and kind, the other an expert in a complex subject. What do these two people have in common that can be communicated in the 90 minutes alloted to a film or the 5 minutes alloted to a blog post?

How does one master the material yet fail the test?

That never happens. But I think you can pass the test without ever mastering the material. Sheer memorization will allow students to pass tests and only have a minimal understanding of what they have covered.