Grading can be a difficult proposition. In a math course, the mechanics of a solution, meaning the correct addition and subtraction, may be correct. However, a student may have misapplied a solution or not correctly understood the application of the mechanics to the problem in question. As a result, it can be better to grade these different aspects differently. In a writing course, this would be akin to grading spelling and grammar separately from organization of the argument.
Grading rubrics, which outline the different facets of a solution and provide tiered assessments appropriate to the problem provide a solution to this for instructors. Historically, I have given my undergraduate math students a 5-question quiz each week of class. I grade these on a five-point scale (from 0 to 4) where 4 is a perfect solution, 3 is some small math (mechanical) error was seen, 2 is a good start, 1 is something somewhat relevant was written down, and 0 is nothing meaningful was stated. This is a type of informal grading rubric. It is also one-dimensional, which causes some issues.
As the University of Maryland University College converts from its custom WebTycho platform to Desire2Learn, I have noticed that formalized grading rubrics are baked into the platform and can be associated with each assessment. Then, when grading, the rubric guides the process. So I created my first formal rubric and attached it to this my quizzes. This grading rubric essentially formalizes the informal grading rubric I had already used:
While formalizing the prior process is useful, it is self-evident this should be revised to reflect the multidimensional grading process used in college-level mathematics.