Comment Re:WTF? (Score 1) 493
A lot of math grading has a subjective element. At least when I was in school, you usually had to show your work on math problems, and could get partial credit. For example if you correctly analyzed a word problem, set up the equations correctly, and then made an adding error at the end, you'd get some points despite the wrong final answer. Continues at higher levels, e.g. when doing proofs.
It's possible to reduce some sources of bias by using grading rubrics, specifying precisely what you'll get points for (X points for setting up the equations, etc...). Some people dislike rubrics because they're very mechanical, but in some cases that can be an advantage, since it removes the judgment around "ah well they got it 90% right"... without points assigned in advance, the assessment that someone got it "90%" vs. "70%" right can be influenced in large part by the teacher's prior belief about whether the student understands the material.