GothamSchools — daily independent reporting on NYC public schools

More Thoughtful

Debunking Standards Issue #5: Tests Matter; Standards Do Not

This and next week I am raising objections to the idea that new standards — particularly new national standards — are worth the attention they get. It is ridiculous to think that they can be a meaningful lever of broad educational improvement. In fact, I do not think that they can have any significant impact at all.

Problem #5: Tests Matter; Standards Do Not

As much as they may hate them, teachers do respond to tests. Not always well or in good faith, but teachers and schools feel the pressure of high stakes and public reported tests.

Tests, of course, are usually supposed to be based on standards. State tests are specifically supposed to be based on state standards. State tests might, in the future, be based on the Common Core standards.

But that’s not really true, not exactly. You see, tests only include the standards that we know how to test and are capable of testing relatively cheaply.

The recent draft of the Common Core ELA standards actually begins, “A crucial factor in readiness for college and careers is students’ ability to comprehend complex texts independently.” However, psychometric requirements for reliability, combined with reasonable limits on the time and money it costs to take an exam, prevent the inclusion of complex texts if only because text length is an issue. Later, the draft says, “Students must be able to revisit and make improvements to a piece of their writing over multiple drafts when circumstances encourage or require it.” There is also a whole section for “Speaking and Listening.” Does anyone expect that that standardized tests based upon the Common Core standards will include a speaking section or will include students’ ability to revise their own work?

And so, even if tests do a good job of evaluating students on the standards that they attempt to include, they do not actually represent the larger set of standards fairly. Moreover, in the absence of real advances in testing, I do not see how changes in the standards will lead to changes in the tests. Test developers will continue to test what they know how to test, regardless of what the standards say or how they have been changed.

Previous: Problem #4 — Classrooms
Next: Problem #6 — Local Control

  • Jason Becker

    I love this post. Poor testing, poor data collection, and poor data use are all major problems that standards do not address at all. I just happen to think you need standards first with which you can align these tests. A major problem I have with a lot of social programs is that they can have poorly defined goals which turn them into Rorschach tests for policy makers, implementers, and stakeholders alike. Education without a well-defined purpose ends up with disparate goals (even in big-picture mode) for practically everyone in the system. It seems to me that testing is a poor way to judge the performance of a system if its goal is not to improve the faculties necessary to perform well on these exams.

    Testing-reform is an important component that I think cannot be decoupled from any standards movement because without well-designed tests which are aligned with the accepted standards, there is little incentive to implement these standards and no reliable data to assess how well students are doing and in what areas.

  • Gideon

    You’re dead on about how state tests are limited to certain standards, and often measure only low-level skills. However, these are not the only tests out there. Many other formative and summative assessments are used to evaluate student learning by teachers and schools, which does allow for the more nuanced types of assessments you call for. Good teachers regularly assess their students’ speaking and listening, as well as their reading and writing. So it is important to have clear standards so teachers know what they should be measuring.

    Also, I think computer-adaptive testing is coming along, and will allow for assessments that more accurately measure high and low level students and can cover more standards more efficiently. Maybe some day they can even assess speaking skills.

  • Michael M.

    I would suggest that both “standards” and “tests” may be missing the point of a comprehensive and well-rounded education. And selling good teachers short.

    Case in point: I just went to my 4th grader’s family morning. After an intro by the teacher who described a process by which family lore stories could be easily remembered, shared, fleshed out, and transitioned from conversation to essay, the teacher had the kids in pairs describe their stories to each other, outline them, then write them up. Pairs of parents too.

    I thought this was terrific teaching — and learning. And teaching parents how to echo the in-school teaching.

    So… what standard was this? How would it be tested? More importantly, how does this get VALUED?

Tips, questions, feedback?

Contact us at .

Word from Our Sponsor

Follow GothamSchools

RSS
Subscribe to the daily email digest:

Chalk It Up

Recent Comments

2 comments so far today

Archives

May 2013
M T W T F S S
« Apr  
 12345
6789101112
13141516171819
20212223242526
2728293031