Ask any group of teachers if they grade on the curve and you will receive an almost universal, resounding no! Now, I believe teachers when they say they don’t grade on the curve; however, what has become apparent in recent years is that shedding some of our traditional habits—our normative grading tendencies—is easier said than done. Even those who have moved to a more standards-based approach to grading can, if not mindful, fall back into habits misaligned with a modern assessment system.
The crux of a standards-based instructional environment is a criterion-referenced approach to assessment where students are assessed on the quality of their work as it relates to the established performance criteria. The standards movement of the 1990s brought about a shift (at least in theory) away from norm-referenced assessment, where students are compared to one another to determine performance levels, to a criterion-referenced approach. Within a norm-referenced system, producing the top performance in a class meant you earned the highest grade, regardless (to a point) of its quality. Of course, there were minimal requirements for the work students were producing, however, teachers often used other students’ responses as a reference point for scoring. This occurred mostly in performances where teacher judgment (e.g. writing) was involved. In the case of assessments, where each point was awarded for each expected fact or step, this might not occur as often.
3 Normative Grading Practices to Avoid
Few (if any) teachers consciously take a normative approach to grading and reporting, but remnants from the past still linger, which is why it is critical that we be hyper-aware of the normative tendencies that influence how we judge student performance. Whether it’s through leaving room, restricting access, or curving backwards we can inadvertently fall short of fulfilling the promise that students be judged solely on the quality of their work.
Leaving room
If we’re not careful, the order in which we assess student performance can influence how we determine the applicable levels. Think of Olympic figure skating or gymnastics. In either of those sports, performing early in a competition can be a disadvantage, as the judges often leave room in their scoring in case another athlete performs a little better than the earlier athletes. Giving away too much too soon can make the judges job more challenging since the role of the judge is to rank the athletes to award medals.
That is not the goal of classroom assessment. If we take that same approach—intentionally keeping score tamped down to leave room for later demonstrations—we’ve lost the plot. The order in which teachers assess student performances is often random, so it is imperative that we not let this randomness play any role in influencing our decisions. Olympic judging is inherently normative; classroom assessment shouldn’t be. Students should be judged on the quality of their work as it compares to the established performance criteria, period. The fact that a performance was one of the first or last assessed should be irrelevant to the determination of quality.
Restricting access
Restricting access to certain levels of performance is another normative habit that can influence teachers’ decisions. Again, I don’t believe teachers consciously choose to restrict access, but this habit has been so embedded in our collective assessment mindsets that it is difficult to shake. The bell curve is a random distribution model that balloons in the middle. When applied to grades, we would see the fewest number of As and Fs, a few more Bs and Ds, and mostly Cs. Ignoring the fact that teaching is neither random or haphazard, this bell curve mentality can be summed up through the notion that it is possible to have too many As. Imagine a scenario where two-thirds of your students consistently reached the highest levels of performance and, therefore, all earned As for the semester. Would you share this news publicly with your colleagues? Sure, some of you would, but many more of you wouldn’t for fear of the predictable criticism that would likely be sent your way: that your class is too easy, not rigorous enough, or dumbed down. We want all students to succeed at the highest level, or so we say, but if they were to do that, we would be more likely to think there is a problem. Thinking there are too many or too few of any level does not square with a criterion-referenced, standards-based learning environment. Once clear performance criteria have been established, how student performances compare to that criteria is all that matters.
Curving backwards
Imagine a high school ELA teacher’s ninth-grade students have just submitted their first argumentative essays. The teacher was effective in co-constructing criteria with the class, thorough in using exemplars to contextualize that criteria, and methodical about providing exceptional feedback on first draft papers. Now that the final papers have been collected, it’s time to score them. Using the four-point holistic rubric, the teacher begins to examine the quality of the writing. Jeremy submitted an exceptional writing assessment that his teacher determined was unquestionably a 4; the top-level of the rubric. Maria, however, is an incredibly gifted writer whose writing is akin to someone in their senior year. The teacher, after reading Maria assignment, pauses to reflect on the differences between Maria’s and Jeremy’s writing. She thinks to herself, “If Maria’s assignment is a 4, how can Jeremy’s be a 4?” She then decides that since Maria is a 4, Jeremy must be a 3. See the issue?
Maria’s exceptional writing should not diminish Jeremy’s accomplishment. He met the criteria for the top-level. As soon as the teacher examined Jeremy’s work in comparison to Maria’s she was norming and, by reducing Jeremy’s score, she curved his score backwards. Curving is about advancing scores up (i.e. the highest score in the class would be advanced to 100, then all other scores would distribute from there). For example, if the highest score on an assignment was 23/30, then the 23 would have 7 added to it—as would all other scores—and grades would be determined from there. What Jeremy and Maria’s teacher did is sort of the opposite; she changed the criteria after it had already been established. Student scores should not be dependent upon who else is in the room. Again, the only comparison that matters is how the student compares to the established criteria.
A standards-based instructional environment demands that student grades neither be dependent on who the teacher is nor be dependent on who else is in the room. Establishing clear performance criteria is essential to valid and reliable assessment interpretation; however, this clarity can be for not if we fall back into normative habits that undermine the gist of what a criterion-referenced system is all about. Teachers who collaborate and calibrate to secure valid and reliable assessments will prevent one another from slipping back into old habits. Just being aware of the potential influence these (and other) normative habits can have is already half of the battle.