I opened my email the other day and saw this subject line: “3 important critiques of standardized assessment” on my daily Edweek Update email.
If you missed this piece, definitely take the time to read it. Matthew Lynch’s article and the readers’ comments are all worth reading (as is the article linked in a subheading from the 6/28 email What’s a High-Quality Assessment Item?).
I especially appreciated the criticism of several commenters which I would paraphrase as “if not tests, then what?”
While I tend to think that the “test-every-kid-every-year” solution is an overcorrection, we do know that our most vulnerable students are the ones who suffer most in the absence of standards which are clearly communicated and assessed. My favorite perspective on this is Lisa Delpit’s, but we can also look at data that show some significant improvement among low achievers and students of color after No Child Left Behind’s mandate for yearly standardized testing.
I work at a school in the New York Performance Standard Consortium, schools that are actively involved in de-coupling those two terms. At Harvest Collegiate High School we collect data, measure student growth, and break that data up by subgroup (students with IEPs, racial groups, gender). In other words (to quote the commenter “NJB”) we assess “the knowledge, competency, and achievement of large groups of students,” but we do so with authentic performance assessments instead of tests.
We need to leverage both tests and performance assessment in thoughtful ways to create a set of healthy
assessment habits which don’t have the giant footprint of the current system. This would include the use of representative sampling, grade span testing, some benchmark years where every student was tested (think of this the analogue to a pediatrician “check up”), and lots of support for teachers to develop standardized performance assessments that give them insight into their students throughout the year.
Below I’ve copied Lynch’s “3 important critiques of standardized assessment” and included how performance assessments can help address the issues they raise.
1. It encourages teaching to the test.
The common answer to this critique is that it’s not necessarily bad to teach to a test as long as it’s a good test. What the “pro-test” responders don’t tell you is that we now have clear data that suggests that even the SAT (a much more well-researched educational assessment than any state standardized test) is less predictive of success in college than teacher grades.
The current regime asks “standardized tests” to do something that they are not set up to be able to do. If the College Board cannot create a single tests that is predictive of student success then we certainly can’t expect strapped state education departments to do so.
The people who we can expect to do this are teachers. Teacher-created performance assessments can be targeted and precise in a way that standardized tests taken by millions of students cannot be. Teachers (when they are well-supported and compensated for the work) can create assessments worth “teaching to,” then they can teach to them. This shrinks the feedback loop, giving teachers access to the kinds of information they need to help their students move from where they are to where we want them to be.
2. It sacrifices means for the end.
I think this is my favorite of Lynch’s critiques. I don’t know about you, but I don’t know many professionals who go into work to find a supervisor saying to them “here’s a number 2 pencil, answer this set of 50 multiple choice questions in the next 45 min, you may begin.”
But I also hear the commenter’s concerns. We owe parents and taxpayers well evidenced information about what students have learned.
Performance-based assessments give students the opportunity to do meaningful tasks like develop a science lab or write a historical argument. In my classes, students are assessed on their ability to use math and physics to analyze situations and develop arguments. For example, in one performance assessment students designed containers to deliver supplies to disaster victims then wrote letters to the Red Cross arguing why their container was the best.
This gives students the chance to engage in more authentic critical thinking (“means”) AND gives teachers the opportunity to collect data on how well they are doing it (“ends”).
3. Its power in education is too broad.
This critique is more about the stakes associated with tests than the test themselves. We could imagine a world where end of grade tests were treated more like NAEP tests — students take them and the data is analyzed without direct impact on the students who generated the data or their teachers.
A place where both Lynch and his critical commenters ought to agree is that we need more data, not less, in order to make good decisions about how well students are learning and schools are educating. You don’t have to be a policy wonk to realize that the current evaluation system in many states that requires art teachers to be assessed by student performance on math tests is unreliable (and insane).
Performance-based assessments like the ones we use at Harvest allow our evaluators to use data from our classes to assess our teaching. We use data to start conversations not end reviews. This is the kind culture that every school could create if we stopped emphasizing a single assessment at the end of the year and started valuing the ways in which teachers collect data on student learning every day.
Assessment of something as complex as learning presents huge logistical challenges. Lynch and his commenters express interesting critiques and counter-critiques on the state of assessment in schools today which are all valuable and worth considering.
One of the ways we can bridge the gap between them is to start to question some of the assumptions they bring to the conversation about what makes an assessment. As always, I’m happy to hear your assessment of my perspective on Twitter or in the comments.
Photo by Oregon Department of Transportation. We wouldn’t expect a paper-and-pencil test alone to show whether or not someone can drive safely. Performance assessment can be standardized and can give insight into what students can do in a way that tests alone cannot.