Working Backwards to Assessment
Feb 9th, 2007 | By Eric Hoefler | Category: Education/LiteracyI’m saying: we should teach to the test, as long as it’s an appropriate test, and to discover what an appropriate test is for each discipline and each course, we have to work carefully backwards.
I’ve thought this for a while, but the last two posts over at Friends of Dave gave me a way to put some of these thoughts together.
In a quick overview of an article reviewing One Nation, One Standard by Herman Badillo, Dave says:
Mr. Crouch [the reviewer] rightly points out that schools often reduce requirements to make life “easier” for poor or minority students when in fact these students are actually being harmed. Friends of Dave - Someone Else Who Gets It
He then goes on to quote from the review:
He [Herman Badillo] rightly believes that Latinos and the nation at large will benefit from the imposition of high standards and the removal of supposedly sympathetic but imprisoning lower standards. Friends of Dave - Someone Else Who Gets It
This is absolutely right. Of course, we must be cognizant of each student’s “zone of proximal development” or we risk losing him or her. However, the end goal is the same for all: excellence. Why would it be anything less? Are we trying to lead anyone to mediocrity? Sometimes, it appears that way.
We must also ensure that our perception of a given student’s “zone of proximal development” is accurate and not the product of our prejudices or the student’s smokescreen. (As most teachers know, students have many reasons for appearing less capable than they are: the simple impulse to avoid work, the fear of failure, past encounters with poor teachers, social stigma. We must see through these and find ways to help the student to move past them.)
In a related post, Dave quotes from Jonathan Alter’s article “Stop Pandering on Education.” I’m moving out-of-order here, but I want to give a quick “amen.” Alter is discussing the problem with tenure and says:
It’s time to move from identifying failing schools to identifying failing teachers. Friends of Dave - Alter Gets It
I think tenure is ridiculous. Poor teachers who cannot or will not improve after repeated efforts to give assistance should be removed from education. That this is not already the case is astonishing to me.
The problems starts here:
basic assessment is both right and popular: “I don’t think parents see reliable data as punitive.” (emphasis added) Friends of Dave - Alter Gets It
And continues here:
Each political party is about half right. Republicans are right about the need for strict performance standards and wrong in believing that enduring change is possible without lots more money from Washington. Democrats are right about the need to pay teachers more but wrong to kiss up to teachers unions bent on preventing accountability. (emphasis added) Friends of Dave - Alter Gets It
Let me state clearly before I shut anyone down: I am not anti-assessment or anti-standards.
The problem I have with the whole argument around NCLB and the standards/assessment debate is one of balance. I don’t see it. And I don’t see enough understanding about how problematic and complex those two words are: “standards” and “assessment.”
And the reactionaries on both sides won’t shut up, sit still, listen to one another, and think long enough to address those terms. If you start asking for more thoughtful definitions, you’re perceived as being anti-standards/assessment … just another lazy teacher who doesn’t want to be held accountable. Meanwhile, the pro-standards/assessment party charges righteously ahead without any clear sense of their direction.
People are arguing the wrong argument. It’s not about whether or not we need standards and assessment, but rather what those standards should look like and, to a greater degree at this point, what forms of assessment are most valid.
As Alter points out, both sides are partially correct. A balance between the two extremes is necessary (as is the case in most everything in life).
Where he misses it, along with many others, is his easy acceptance that the forms of assessment we’re currently using are the most appropriate.
I discussed some of my specific objections to the “one-size-fits-all” standardized test approach in two earlier posts. I also dealt with similar ideas in my post about context and relevance. I won’t repeat any of that here, but I will briefly outline what I think is the necessary approach for creating appropriate assessments.
We must work backwards:
- We start with: “What is the purpose of this discipline?” and “What is it important that students know (understanding) and can do (application) as a result of this course?” These are difficult, complex, and evolving questions with implications far beyond the classroom. Moving too quickly here weakens everything else.
- Then to: “How can I assess whether or not students are gaining this understanding and acquiring the necessary knowledge and skills in the most authentic manner?”
- Then, we should ask the question: “How can I prepare students for that assessment?” In that sense, we are teaching to the test–but a test that has been carefully designed to fit what is being taught. And we certainly are not teaching to the same kind of test for every discipline and every situation.
Multiple-choice tests are certainly appropriate tools for testing some kinds of knowledge and skills. They are dreadfully inadequate for other kinds. The same can be said of essays. The same can be said of any assessment. The assessment must fit the standard, and the standard must connect the content and skills to the understanding and application.
I believe that we are selling students short by relying only on one or two forms of assessment. The current methods of standardized assessment let everyone off easy. If we truly want higher standards, then we need to develop assessments that can test them.
Related posts (auto-generated):
- Testing Teachers From the CNN article “Commission urges tracking of teacher progress,” notes about a “special commission” involved in the revision of NCLB: Teachers should be...
- A Positive Stance While I disagree with the way NCLB has been conceived and enforced, I agree with this approach from teachers to students: We decided it...
- Response to “In Defense of NCLB” I found Dan Meyer’s blog post “In Defense of NCLB” through Chris Lehmann’s post “A Smarter Mind than Mine Takes on NCLB.” I wasn’t...
- Fourteen Percent Mike Petrilli, in a recent post on the Flypaper blog, comments on a study in Philadelphia that measured the impact of a “healthy-eating” initiative...
- More on Fairfax’s Anti-NCLB Stance Marc Fisher’s article is appropriately biting, and as far as I’m concerned, he nails it. He’s reporting Jack Dale’s refusal to administer standardized reading...
RSS

































I’ve mentioned elsewhere recently that I had the opportunity to spend a day writing test items for our state’s new science assessment. A group of teachers from various parts of the state, and from both secondary and elementary levels, worked in grade-level teams with representatives of a large test development company. It’s an industry - test item writing. They hire retired teachers and psychometricians to do this work.
They wanted us to help them understand the regional culture so that they could write questions that would minimize bias. Not easy to do (maybe impossible for a state the size of Alaska - with hundreds of geographically isolated and culturally distinct communities.) I asked, Did you know that many communities don’t have trees? Or that they hunt entirely different animals? They hadn’t thought of that.
Many of the high school science teachers were dissatisfied with the multiple choice options for the more complex concepts. A couple of things that stand out for me: The test company folks agreed that multiple choice questions were not ideal, but those were the best we could do with a test that would be administered on a large scale. One of the “experts” also said, This isn’t a reading test. Yet it was a paper and pencil test that we were developing.
You are right that we invest a lot of faith in technologies that are far from perfect, and which may in fact be misleading.