Yet another document related to national standards has been released: the SMARTER balanced assessment consortium report. There are some nice points in it about the different kinds of evidence needed, given the Standards; and there are (finally) sample assessments in mathematics, albeit only for Grade 8. Alas, one again looks in vain in these national policy papers for a discussion of what has always been the linchpin of effective local reform: how teachers should give grades.
To me, it is inarguable that getting local grading practice and policy right is key to actually achieving standards in a standards-based system. Surely, if students are going to reach standards, their work has to be graded against standards throughout the year. Put the other way around, if every external test result is a surprise in which test scores bear little relationship to local grades – as is now often the case – then reform can never systematically occur. I think it is morally unconscionable that students can get honor grades all year only to flunk the state test – and I think this regardless of my distaste for over-reliance on one-shot external tests.
I know, no one really wants to deal with this touchy subject in schools. But it is high time we faced an ugly truth: most teacher, school, and district grading schemes are indefensible. Grading schemes and norms vary wildly across teachers and schools in terms of how one counts homework, effort, labs, projects, re-do’s, and quizzes. Extra points are sometimes given for fancy covers; points are sometimes taken away for using the wrong type of paper. More unfortunate is when educational apples and oranges are averaged together (e.g. effort and achievement) in a single grade – all of this utterly unmoored from feedback against external standards.
Think of it: few teachers try to and know how to validly grade against state standards, even though state standards are their statutory obligation. If a parent or student asks during the year: Where does the student stand against state and national standards at this point in time? Few teachers can validly answer the question. What is common knowledge in every sport or artistic endeavor is mostly unknown in academic performance. Even worse, of course, is that the only standards-based feedback that now does occur – on external tests – comes when the school year is over.
An athletic analogy makes more obvious how dysfunctional are the habits we take for granted. Suppose in soccer, there is only one game, at the end of the season. During the weeks leading up to the game, teachers run players through countless soccer drills, and grade them on each drill. Suppose, in addition, that different teachers use their favorite drills and their own differing grading systems to assess “progress.” Throughout the season they give whatever grades seem proper and “average” them all to assign a single mid-term and final soccer grade.
Now imagine game day, during the last week of the season. Suppose the rules for the one big game vary slightly from year to year (but coaches and players don’t know those changes until game day), as determined by state soccer officials in reference to soccer standards. Suppose, during the game, that no one can see if the ball goes into the net on a shot on goal; only the official scorer can see this but the result is not reported during the game. Finally, imagine if the only feedback players and coaches get from the game is delivered four weeks later as a final score. Who would improve under such a system? Who would meet soccer Standards other than gifted players under such a system? Yet this is how assessment works in school: there is little correlation between local grading and state standards and tests, feedback from standards-based tests comes too late to use for improvement, and there is too little detail in the feedback.
As the metaphor suggests, two things must be designed to happen locally that now do not, if the situation is to improve: the students must be asked to perform and get feedback against the final Standards throughout the local “season” and the feedback must de derived from the “game” of complex performance, not just based on local low-level “drills.” There is no other way to prepare students to meet performance standards with any realistic hope of success.
I have long written about the need for more performance assessment and greater transparency in external assessment through the release of tests with extensive item analysis. (And, in fact the two national assessment consortia are developing such tasks as part of the system). Here, I wish to focus attention on the grading side of things – on what WE can control locally if we choose to. For, unless teachers grade student work locally against real standards-based performance, week in and week out, no one will know where one stands – teacher or students – until it is too late.
What, then, are practical solutions to the problem? In this blog entry I only summarize some feasible options. The chief mechanisms involve calibration and moderation, valid anchors of tasks and scored performance, and more defensible grading calculations. In my next blog entry I will provide more detailed explanations and rationales for some of these ideas.
- The two assessment consortia develop assessments designed to be used locally to establish local standards that are aligned with national standards. Their only purpose is calibration of local and national standards, not individual student grades or scores.
- Local policy is changed to ensure that students receive at least 20% of their grades from educators beyond just their own teacher grading in isolation. Pairing teachers to exchange papers, team grading and departmental grading would be the primary mechanisms; occasionally outside experts and auditors might grade papers, too (e.g. employers, college professors, consultants).
- Staff develop a set of anchor papers that they believe reflect both the work and level of performance that each of the Standards demand. Those anchors would then be validated by district and state officials as reflective of the Standards. The papers would then be published and used to anchor all standards-based grading locally. (This should also be a recommended state policy and practice as part of national standards).
- Once per month, teachers report out individually how students are doing against 2-3 targeted standards in their subjects using these validated anchor papers and rubrics.
- Once per semester, groups of teachers at each school division level (elementary, middle, high) would score student work together against key standards, using performance assessments. For example, all students K-12 might write a brief paper in which they identify the main idea of a piece of non-fiction and argue for their claim; the papers would then be scored against rubrics for main idea, and clear and supported writing. The teacher-judges would not know either the name or the grade of the student. They would be scored against “exit” standards for that school level.
- Teams of teachers across 2 grades score assessments given to students that are borrowed from valid state and national sources. For example, teachers might download released items in math and science from the web sites of states like Massachusetts or Florida (where full tests are available along with item analysis) or use NAEP released items.
- Teacher grades are not considered final until all the grades are reported, discussed in district or regional teams, and perhaps adjusted in light of common standards (a process called “moderation” long used in other countries, and in AP and IB scoring).
- Report cards are developed locally for “scores” separate from “grades” where “scores” are meant to be valid reflections of only where the student stands against national standards and where “grades” reflect traditional normed judgments about progress, effort, and achievement.
- HS and Middle school teachers give distinct grades for effort, progress, and achievement so that the feedback is more helpful and the data more valid.
- Grade-giving is audited on a regular basis at the district level to assure consistency and validity of grade schemes. The state department of education might demand such audits as part of school accountability to ensure that there is a constant concern about alignment of local grading and state standards. Principals and Department Heads would be evaluated on the quality of the grades under their purview.