by Grant Wiggins, Authentic Education
Over the last few months I have worked with a number of high schools and middle schools where the grading and assessment practices simply do not work in a world of standards. The schools are not making local assessment rigorous enough in their concern with demoralizing students through low grades. The solution is straightforward: don’t thoughtlessly translate scores into grades.
The problem. schools have to meet standards, and local assessment should prepare kids to deal with the standards as tested by PARCC and SB. But the new tests are harder and more rigorously scored than most local tests. So, scores will have to be low. (Anyone following NAEP results has known this for years, alas.) This seems to run headlong into a long tradition of grading whereby we do not want to punish kids with low grades (akin to the outrage over sharply-lower school scores on accountability measures this year).
Yet, there seems to be no alternative: to significantly raise local standards of performance seems to mean we have to lower student grades. Or, conversely, we can keep our current average grade of a B for students locally, but then have less rigor than is needed to prepare kids for the tests – and predict results on them (which local assessment should surely do if it is valid and useful).
Note that so-called “standards-based grading” does not inherently solve this problem. Just because we go to standards-based grading doesn’t mean the grading is rigorous. In fact, if you look at schools that use standards-based grading, it is rare for students to get “scores” that are vastly different from the range of “grades” in such schools previously. i.e. we are doing standards-based grading in a norm-referenced framework! The local failure was to assume that assessing against the standards was sufficient to establish rigor. But that is insufficient; it cannot work by itself.
What Is Rigor?
Rigor is not established by the teaching. It’s not established by framing teaching against standards, therefore. Rigor is established by our expectations: how we evaluate and score student work. That means that rigor is established by the three different elements of assessment:
- The difficulty of the task or questions
- The difficulty of the criteria, as established by rubrics
- The level of achievement expected, as set by “anchors” or cut scores.
Many districts and schools don’t even pass the #1 criterion now. Routinely, when my colleagues and I audit local assessment, the tests are much easier than what the external tests test – even in pretty good districts. The usual explanation? The problem of fair grading!
Note, too, from these three elements that even a difficult task and high-quality rubric aren’t enough to establish rigor. The task could be challenging and the criteria demanding – but if the expectations for student products or performance are very low (as established by either specific models or local norms), then the assessment is not rigorous. That’s why having a “cut” score of 40 or 50 on the state tests is a terrible solution – if the goal is to communicate standards-based results vs. finding a way to pass most kids.
Think of the high jump or pole vault in track: you could enter a challenging event and be judged against the true criteria, but if the height you have to clear is absurdly low, then the assessment is not rigorous – even though it is “standards-based” testing and scoring.
Solving The Problem
One solution? Avoid thoughtless calculations based on false equivalencies. Stick with track and field to see the solution: we need not and in fact never do calculate the “grade” for the athlete by mechanically turning the height they jump into a grade by some arbitrary but easy to use formula. To do so, would greatly lower grades and provide powerful disincentives for the less skilled athletes.
On the contrary, we judge progress and performance relative to early jump heights and look for “appropriate” growth, based on effort and gains in height. (I blogged previously about this point at greater length here and here.) However, the expectations for all jumpers are high and constantly increasing.
The same solution is needed locally in academics, if genuine standards are going to be used to alert students as to where they are without discouraging them. (This is the idea behind the SLOs and SGOs in many states.) So, numerous times a year, their work needs to be evaluated against the external standards (as established by high-quality tests and student work samples). “But we have to give grades all year in our online grade book!” I know. But instead of turning their “score” into a “grade” by some unthinking formula, we use our wisdom and judgment to factor in fairness, growth, and effort on some uniform basis.
Suppose, for example, that in a writing assessment done against national standards, we anchor the assessment by national samples culled from released tests. Further suppose that a 6-point rubric is used. Now, assume that in the first administration, say in October, almost all students get a 1 or a 2 (where those are the lowest scores on the scale). Here’s what we might say before the scores are given to students and turned into grades:
“Guys, I’m scoring you against the best writing in the state. So, your first grade this fall will reflect a fair assessment of where you are now. A score of 1 will equal a B-. A score of 2 will equal a B+. Any score above a 2 is an A – for the first semester.
“Next semester, in the winter, to get those same grades, you will have to move up one number on the scale. And by spring, you will have to move up 2 numbers to get those grades.”
This already happens, of course, in AP and IB courses. So, it should be relatively easy to do so in all courses. We have thus solved the problem: grades become fair, standards are made clear, and there are incentives to improve over time.
This article was excerpted from a post that first appeared on Grant’s personal blog; Grant can be found on twitter here; adapted image attribution flickr user vancouverfilmschool; Standards-Based Grading Only Solves Half The Problem