Percentages are objective. Sure, if you mean how to calculate them, but everything else requires some level of subjectivity. Who picks the questions? Who decides how and if to award partial credit for a problem? Elliot and Starch demonstrated this in 1912 study. Matt Townsley from Iowa did a recreation of the study in his district and posed the following question (left) to 40 certified math teachers and gave them four student answers (right) with the following directions: The question is worth 10 points. How many points would you give each student? (original post)