Counting to several

My local newspaper recently headlined that students’ scores had marginally declined on Vermont’s statewide science assessment.

As with most standardized tests, no matter where you live, scorers use subjective rubrics to decide, for example, whether a student’s answer reflects a “thorough,” “general,” “limited,” or “minimal” understanding. According to officials, scorers can make those precise determinations because a “general” answer, for instance, includes an unspecified number of “errors and omissions,” as opposed to a “limited” response, which includes “several errors and omissions.”

Can you count to “several”? I can’t.

On a four-point scale, an answer with “errors” earns three points, while “several errors” earns two points. That’s a twenty-five percent score variation based on the difference between “I don’t exactly know howmany” and “several.”

Further compromising the “data,” students work on parts of the test in randomly assigned groups. Do you think a student’s individual score might be somewhat influenced by how smart the other students in his group happen to be?

The insignificance of marginal variations aside, if a statewide average markedly declines, or improves, from one year to the next, there exist two alternative explanations. Either the statewide student body got remarkably dumber, or smarter, from one year to the next, which isn’t likely, or despite test promoters’ claims of statistical consistency and precision, the test accidentally got harder or easier. Publishers and education officials have had to swallow plenty of that assessment crow over the past two decades of our national testing frenzy.

Welcome to the world of modern assessment.

Don’t misunderstand. When I’m grading an essay, I’m not perfectly scientific either. The difference is I don’t claim to be, and I don’t base a whole year’s grade on one test. Also my grades cost a lot less.

Somehow officials aren’t overly troubled by these repeated assessment fiascoes. Critics have instead again focused their wrath on letter grades. Reformers charge that A-F report cards provide “hodgepodge” grades that are “impossible to interpret” and “rarely present a true picture of a student’s proficiency.” Some complain that letter grades inappropriately mingle academic competency, effort, and progress, and that teachers should award a separate grade in each category.

You don’t need a separate grade in “progress” to determine if a student hasmade any. You just need to track his grades in succeeding quarters and years. As for effort, we gave effort grades inmy school for decades until our new computer grading system made awarding them a clerical nightmare. We also awarded checkmarks in areas that contribute to academic success, like homework completion and class participation, until our new computer program eliminated our ability to do that.

I won’t pretend that all the complexities of academic performance can be fully captured in a single alphabet character. Letter grades are shorthand. The question is, are they effective.Any grading system short of a comprehensive narrative, whether it consists of letters or numbers, is shorthand. Teachers’ comments have long provided a brief narrative when letter grades need further explanation. For example, I commonly add a note when incomplete work and missing assignments have lowered a student’s overall grade. I also add comments about everything from strong effort to exemplary class participation.

I’ve found that for most parents this is enough. Those who want more information can and do call, write, or arrange for a conference. Over the years I’ve talked with many parents, and a face-to-face conversation almost always answers any lingering questions they have.

Reformers’ latest recycled marvel, standards-based grading, rests on the assumption that parents aren’t satisfied with a summary A through F in English, for example, but instead want multiple 4 through 1 grades in specific English categories. I’m skeptical as to the validity of those category grades, especially since so many assignments simultaneously involve and assess a combination of skills. I’m also not sure howmany parents actually want separate grades assessing their child’s performance in “narrative” as opposed to “explanatory” writing, just two of the ten new standards-based, English grades each elementary student inmy school will begin receiving this year.

Promoters insist these elaborate standards-based changes are a response to parental demands for better communication and reporting. Ironically, as part of the standards-based move to better communicate with parents, my students’ parents will now be receiving three report cards a year instead of four. According to advocates, you can’t really compute meaningful standard-based grades over a nine-week marking period, prompting the shift to twelve week trimesters.

A grading systemthat can’t produce meaningful results in nine weeks isn’t much of a grading system. Changing how you report what students know doesn’t change how much they know. It’s also hard to reconcile proponents’ claims that standards-based grading is a response to parental demands when in districts where it’s been implemented, so many parents don’t like it.

Yeah, but what’s my kid’s grade?

Like the Common Core, standards-based grading isn’t the grassroots idea its boosters claimit is. It didn’t originate with parents. It was born in the fevered imaginations of experts and theorists who are strangers to the real world of kitchen tables and classrooms.

For howlong will so few be permitted to visit so much folly and harm on somany?

DAVID GOLDMAN — THE ASSOCIATED PRESS In this April 18, 2013, file photo, Burgess-Peterson Elementary School principal Robin Robbins, center, meets with students during an after-school study program in Atlanta, in preparation for state standardized testing, soon to begin.