Acceptance of authority and teaching to the test are two entirely different issues, and these guys are teaching to modern American standardized tests, not the r

Acceptance of authority and teaching to the test are two entirely different issues, and these guys are teaching to modern American standardized tests, not the recitation-style tests described in the Feynman excerpt. On the standarized physics tests I took, you typically got a somewhat simplified physical situation and were asked a question about it. You had to decide what physical principles were involved, what kind of reasoning to apply, and how to set up the equations (if math was involved.) Sometimes the problems were quite concrete. Sometimes they were abstract. Sometimes they only required simple application of an equation (but you had to figure out which equation to apply!) and sometimes they required deep reasoning. Very often the situations described resembled laboratory experiments, and solving the problem correctly required recognizing factors that the problem description intentionally failed to call your attention to -- like the inertial component of the ball-rolling exercise described by Feynman. The school in the article is being compared to other schools on the basis of standardized tests that reflect decades of agonizing over how to design standardized tests so that they can't be gamed through regurgitation. Besides cultural bias (which I think is a real, valid problem, but which doesn't materially affect science tests) the primary criticism you hear of standardized tests these days is that tests are too holistic and can't separate out all the variables that affect a kid's ability to choose the right answer. For instance, a kid might have a wonderful abstract understanding of chemistry but not be able to balance a chemical reaction, read an electron dot diagram, find "Cl" on the periodic table, recognize the word "titration," add two numbers, or figure out which end of the pencil to use on the Scantron. Modern testing techniques have not yet developed to the point where they can tease out that magic variable of "conceptual understanding" from the background noise of "functional incompetence." Sarcasm aside, and acknowledging that lab skills are much too expensive to test and can safely be neglected by test-obsessed teachers, teaching to modern multiple-choice tests isn't such a bad way to teach science.