Every student deserves multiple chances to demonstrate his or her learning. This may be through two or more different assessments that assess the same outcome(s). It may also take place using redos/retakes of the same assessment (the teacher may require acceptable evidence that the student has undertaken subsequent learning before the redo). It's not always easy, but it might also involve providing the student with an alternative form of assessment - better yet, that alternative is suggested by the student.
In either case, how can a teacher justify averaging the student's earlier "on the way" evidence of learning with later evidence that the student has progressed? No matter how you slice it, that's punishing the student for not learning fast enough. Why would we care if it takes a student 2 weeks, 5 weeks or 15 weeks to learn something? What matters is that they get there.
Let's address some practical issues that might get in the way:
- Most importantly perhaps, common "grade books" encourage or practically mandate averaging grades throughout a course. So? If the practice is wrong, it's wrong, and while these grade books may be a practical barrier, they're not insurmountable. In an earlier post, I published a spreadsheet that we created and use instead of a traditional grade book. We'll share with anyone who'd like to contact me. In addition, though we didn't find any stand-alone software that didn't automatically average throughout a term, our division is moving to PowerSchool as our student management software next year, and we understand that the capability to sidestep this averaging is built in. If it's not, we'll sidestep that and find another way.
- This leads to accepting the role of teachers' professional judgment in determining a student's grade. If it's not by averaging, how do we determine a student's final standing for a particular outcome? The answer is to trust to teachers' judgment. This determination should be based on the most recent and most consistent demonstrations of student learning. It should "triangulate" that data with the teacher's conversations with the student, and with observations of the student's learning outside of formal summative assessments (though these latter should be primary determinants). There should be no "math" involved in determining a student's learning for an outcome. In determining an overall grade for a course, these outcome by outcome judgments will be appropriately weighted and combined to give that overall grade, but that's the only occasion for using math in grading.
- Administrators, are you uncomfortable with relying on teachers' professional judgment? We put teachers in charge of students and their learning every day. If these teachers can't be trusted, then we need to take responsibility for that, work with them to improve, and/or find better teachers. We also need to acknowledge that teachers influence grades, profoundly, every day in their teaching, and every time they create an assessment toward that grade. Teachers can move grades 10, 20, 30% or more up or down simply by changing the assessments themselves. That's silly, but it's true, so let's not put teachers in a straightjacket, even if it's sometimes of their own making.
- Percentages - how does a teacher exercise that judgment to pick a number between 0 and 100 to characterize learning? The answer is that they don't. We use a simple 4 point scale (again, described in an earlier post, and based on the province's report cards). 3 if the outcome has been met as prescribed, and 4 if the student exceeds the outcome or is extraordinarily proficient. 2 if the student has a "basic" but not yet all the way there level of learning, and 1 if it's just somewhat acceptable. 0, or better, "not yet" is used as a placeholder only, not a grade to be averaged. We only convert to a percentage twice in a semester, for the sake of the report cards, and then only by provincial mandate. No reasonable person can say they can discern the difference between 76% and 78%, and let's stop pretending that we can.
- What is the role of a final exam or other form of final assessment? Well, it's a great opportunity to give a student that one last chance to demonstrate learning, if we do as we should, and assess each outcome separately from every other. Our senior math/sciences teacher gives the exam to students in pieces, one outcome per piece. Students choose which outcomes will comprise their final exam in advance, as they know where they stand at all times in their learning, and choose these outcomes based on those which can stand to be improved upon. This way, students are not required to reassure the teacher that they have learned that which they have already amply demonstrated. They get to focus on the learning that is most important at that point instead. Also, the final exam/other has no set weight. Why should it? Suppose a student has struggled with trigonometry throughout the course, has a breakthrough and "gets" it, and proves that on the final exam. It doesn't make sense that this should only count for 30 or 40% of the final grade. The student has learned it, as surely as the student who got it right from the beginning, and should be credited fully with that learning.
There's more, of course, and we're all getting at it in our #sblchat's and other forums. There are a ton of good resources for learning more, online and in print.
The bottom line is that I don't believe this is really open to debate. If we are averaging students' grades and giving significant weight to early-on demonstrations that they haven't yet learned something, then we are at fault. That's not how it works in the "real world", so let's just fix this practice, once and for all.