Not measuring what you think

I know I said that I would try to avoid simply posting interesting articles on this blog, but this is an eloquent rendition of something I have been preaching for years. High school grades and SAT scores better correlate with the quality of the counselors and the race distribution of the school than mental ability.


My grades were in part due to “grade engineering”: the process of maximizing grades with minimal effort and without regard to learning or understanding material. In other words, I received high grades partially by exploiting the weak correlation between grades and mastery...At one time, I suppose, grades might have been an objective and reasonably accurate measure of competence in a given subject. Not anymore. Today, they primarily measure how well a student can game the system. 


Now, there is an argument to be made that there is a correlation between students who are good at gaming the system to get into college and students who are good at gaming the system to graduate or get a job. But I think that's getting away from the real goal of higher education.

It really gets interesting with the proposed solution. 

There is no such thing as “teaching to the AP test,” because fundamental understanding and application of knowledge cannot be mastered by memorizing the answers to past exam questions. 

I disagree with this statement. All of my teachers in AP classes knew roughly what was going to be on the test, and wasted no time in teaching anything else. I didn't study a single vocab word or read a single story that did not appear on an expected or history of AP test questions. And don't even get me started on teaching children to write nothing but 5 paragraph essays so they can pass writing exams. The truth of the matter is that it is really difficult to make tests that measure knowledge. I think these exams are a step in the right direction but come far from solving the problem.