We can start taking statistics seriously when the presentation of the statistics is done meticulously. The bubble chart should be done on a per one hundred thou

We can start taking statistics seriously when the presentation of the statistics is done meticulously. The bubble chart should be done on a per one hundred thousand population, like they do for murders (something JoeAltmaier mentioned). In the never ending war to identify differences in systems, why are we looking out outputs and not inputs? Here we are simply looking at metrics of 14,674 test takers in 2003 and 39,278 in 2014. Who chooses where to offer the AP Computer Science test? AP Computer Science has to be selected by a school to offer the test, the school district administrator has an incentive to do this as they get a bonus based on the number of test takers. The risk is that a school will have all failing students so the administrator and reflect badly on the administrator. The administrator needs to identify the schools that provide the best opportunity for success, where computer classes have already taken root and have a pipeline of qualified student to take a test with an average score of 2.9. Who are the students taking the test? Computer programming requires mathematics, logic, and other functions. It takes hours of personal development, failure, and problem solving in programming. Why do the white / asian / males have a proclivity toward this? A more direct question would be to ask why are the asian females not taking the test, if we could identify this, we could identify inputs that would change the outputs from the system. Please stop looking at the outputs and saying change the outputs, you simply find metric abuse. [1] http://en.wikipedia.org/wiki/AP_Computer_Science http://en.wikipedia.org/wiki/AP_Computer_Science