Purpose of Grading
Presumably, the purpose of grading is to assess how well a student knows something or can do something (or both).
Historical background
When I was a student here, the average grade was around 75, and the high honor roll (average over 90) was about 10% of the class (ish). A few years ago, the median senior average was 92. Part of the discrepancy is probably because seniors can take mostly electives if they choose, and part is probably because the quality of the student body is academically higher than when I was here (standardized test results support the latter claim). Nonetheless, I think part of the discrepancy is from what is commonly called grade inflation.
Relative or absolute standards?
It’s fairly easy for a seasoned reader, for instance, to read a set of English papers (say) and put them in rank order (or rank “piles” if several are similar). Assigning grades, then, for relative merit is usually easy—though see the caveat below. The question becomes, then, where to peg the relative ranking on an absolute scale. What grade should the median paper be given? Determine that, as well as the desired spread, and the grades usually fall into line fairly easily.
Ideally, at least, one could use “standards-based grading,” where some external authority (like AP) sets the criteria for grading rather than the course teacher. Except that after setting these standards, College Board carefully “norms” the scores so that the grade distribution stays relatively constant over time. But in theory, one could set standards and judge each student’s work accordingly. The question of whether and if so how those standards should change is one we can defer.
One could also use relative grading, where one grades a student compared to the work produced by other students. Traditionally, this scaling is done for the current year only, but there’s an argument to be made for also taking into account how well students have done on similar assessments in the same course in years past. Maybe this class is weak compared to others and so its median grade should be lower. Maybe it’s better than most of the ones in the past, with the result that a higher median is appropriate.
Comparative arbitrariness of assessment
So, how are grades arrived at? In spite of what many people think, grades are just as arbitrary in math/science courses as in humanities courses. An example I commonly use is that if you give me any topic in math, I can write a test where the average will be 90. I can also write one where the average will be 40. The difference is that students don’t see what goes into the construction of the test, so they are not as aware of the arbitrariness of what’s tested.
What goes into an individual grade?
Now, let’s look at the components of a graded assignment. Rarely is only one skill or knowledge of one specific fact measured.
Take an English paper. In general, the desideratum is a persuasive paper that’s well written and, if an analytical one, supported by examples from the literature. Often, the student who thinks well writes well, and so there’s not a large discrepancy in the two aspects of the grade (which are often called “content” and “mechanics”). Still, if you’re assigning one grade for the paper, how should one balance the style of writing with the merit of what is said? 25% mechanics and 75% content? Should the proportions be different at different grade levels? Or in different courses?
Take a math test. Somewhat analogous to the English paper situation, there are the conceptual aspects of the material and the “working-out-the-details” aspects. You don’t, for instance, want to drive over a bridge designed by an engineer who typically makes computational errors. But without someone who understands physics and engineering, you’ll never make any progress in designing new bridges. So, how should the trade-off (if there is one) between conceptual understanding and computational accuracy be handled? Some partial credit? No partial credit, where only the answer counts?
Risk-taking and entrepreneurship
We (SJS) often say we want students who are intellectual risk-takers, who are creative, who are entrepreneurial. Such people have historically not been treated well by traditional schools.
Part of the reason for that situation is that for teachers to encourage students to take risks, they themselves have to be encouraged to take risks. Taking risks means ipso facto that failure will occur. As long as teachers are “called into the office” when something they try turns out to be politically unpopular, few will be willing to take such risks except in a very minor way.
Another part of the reason for the situation is that teachers tend to like to control things and to be perceived as experts in a subject or skill area. Thus, some won’t take “off-topic” questions because of the risk of being wrong or looking foolish in front of students (or peers or superiors). Creativity involves risk-taking, which we’ve already said requires tolerance of failure. And most entrepreneurs fail one or more times before doing something amazing.
Caveat from above: assigning grades for relative merit is difficult with outlier students: those who have brilliant ideas but poor execution, for instance.