## 10 March 2010

Woodberry Forest is on a trimester system, which means that I just yesterday finished my grades and comments for the fourth marking period.  (Comments?  Yes, I write a short missive about every student, 5-6 times a year.)  The prize for finishing the week-long process of writing exams, giving exams, grading exams, calculating grades, writing comments, and entering grades and comments into the database is... two weeks of spring break.  And I am thankful.

Now is as good a time as any to discuss the meaning of the grades that I assign.  Yeah, sure, many schools  have some sort of official policy stating what percentage represents what grade.  Such policies are as fungible -- and as useful -- as a steaming pile of dirt in a cow pasture.  It is my job to funge students' grades so that the letter I assign communicates meaningfully to students, parents, and other interested parties (read: colleges and scholarship committees).

The overriding philosophy of my grades is TRANSLUCENCY.  No, no, not "transparency"... I don't want students to know their grades every day.  Regular posting of grades leads to slackage from the top students, hopeless near the bottom of the class.  And I don't want the grading process to be so much of a mystery that rumors begin to circulate of Jacobs throwing lab reports down stairs -- if assignment of grades is opaque, if standards seem vague, why should anyone in the class work hard?

No, I truly mean "translucent."  I want a grading system that allows a student to calculate his grade with reasonable precision any time during the marking period.  I also want the system to be intricate enough to discourage students from actually carrying out that calculation.

I personally use a weighted average of labs (10%), quizzes (15%), homework (20%), and tests (50%) to determine an overall course grade.  Each component is converted to the official school scale via the "square root curve" -- take the square root of the percentage score and multiply by 10.  This makes an 81% into an A, 64% into a B, 49% a C, and so on.  I do all these calculations in a rather intensly complicated spreadsheet that I have developed over the years.  Students can figure out their course grade, too -- if they keep all their assignments and then use the correct weighting without messing up data entry into the TI-84.  You see?  Translucent.

I have a sense of what the final grades should mean.  In AP physics, my grade reflects a probability of earning a score on the official May exam, with A translating into a 5 and B- translating into a 3 or 4.  In the general physics course, an A student has learned physics to the AP standard, but has been asked to master three times fewer topics than an AP student.  In general physics, a C is the old "gentleman's C" -- the student worked hard, did what I asked, but couldn't quite put together more than a cursory understanding of basic problem solving.  (Or, a C student is someone capable of As if he could be bothered to study.)

Now, I don't include in my calculations any sort of official "fudge factor" like a class participation grade or the like.  There's nothing inherently wrong with evaluating class participation, I just don't feel like subjecting my students to such a subjective evaluation, nor do I feel like dealing with the fallout from students who aren't given perfect class participation scores.  So, what do I do when my calculated grades don't seem to exactly match student performance?

I do my fudging by dropping some number of homework or quiz scores each marking period.  I look at each student's calculated grade.  I never merely change one person's grade -- bumping individuals up from, say, a B+ to an A- actually got me in considerable trouble my first years teaching.  So, I look at the overall grades.  If they seem generally low, I'll drop one more quiz.  If they seem high, I'll drop one fewer homework.  This way, I can't be accused of playing favorites, since the entire class gets bumped slightly up or down.  But, I can determine what happens to students right on the border between grade cutoffs.

Grades are necessary because they motivate students and because they communicate about performance.  Grades must be assigned fairly, and must be based as much as possible on objective criteria.  Nevertheless, I don't ever want a student focused on his grade to the detriment of the development of his physics skills.  That's why I make my grading translucent.