Archive Hosted by the AFL-CIO

These Tests Stink

June 02, 2011 By Kaye Thompson Peters, Fellow

“We have to test,” says U.S. Senator Al Franken, as the esteemed panel of policymakers and educators gathered in St. Paul, including U.S. Secretary of Education Arne Duncan, nodded. Testing has become the magic pill, the accepted tool, by which we will improve education, it seems. And yet, those of us who work and learn in our public schools doubt it.

When I asked my ninth grade honors English students what they would want me to ask Secretary Duncan, they overwhelmingly said to ask why we have to test so much. They have been the first ninth graders to take longitudinal tests in St. Paul and just finished their third reading test of the year. (Longitudinal tests evaluate a student over a period of time rather than once.)

They also have taken three math tests. That’s six tests—six hours of missed class—in addition to the two hours they spent on the MCA writing test they must pass for graduation.

In asking the class what concerns them, I unleashed a flood of frustration and resentment. These are good students. Some are stronger than others, but my strongest students reported that they in fact had seen their scores go down over the course of the year. The very tests that are supposed to measure growth had failed them.

That could be seen as a reflection on my teaching, but the students know they have learned how to better analyze literature this year and I have seen them grow. I asked them why they think their scores have dropped. They had a number of reasons, showing they have been paying attention in class. The students recognize that the tests are not aligned with the curriculum. The tests are also varied—lack a control—one student suggested. They are tired of testing and lose concentration.

Content is also a challenge. In the math test, they find the material so below what they are learning now that they don’t remember how to do it. In the reading test, if they do well on the first round, the computer program boots them up to ever-higher levels. If the text were just more sophisticated, it would be one thing, but the tests actually begin asking about philosophical and literary movements that are typically not covered until college.

Even the best students do not have the vocabulary to be successful. They are asked to identify literary devices they will not learn until 10th grade and to interpret text by multiple choice. In other words, the test asks them to do what any literature teacher worth his or her salt tells them is wrong. There is not just one interpretation of literature.

Essentially, my students are able to diagnose all of the variables that affect their test performance—and yet the policymakers of this country continue to insist that testing is a stable, credible means of assessing student learning.

Why is testing necessary? Because policymakers do not trust teachers to make accurate evaluations of their students’ work. Now, there is no doubt that some teachers have “written off” students, have dumbed down curriculum in response to their students’ inability to do the work.

The variations between Texas, Minnesota, Mississippi and Massachusetts have been widely reported and documented. Yet, my students and I want to know if testing is really the tool for remedying the disparities.

The buzz words in education these days are “data-driven” and “research-based.” If I can’t link it to published evidence or statistics, my lesson plans are questionable. My judgment as an educator is not trusted unless I can tie a number to it. I beg to differ.

In fact, when I got the first test results back on my students last fall, I could quickly spot the errors in the testing.

A student who was reading Catcher in the Rye with ease and recognizing the construction of the tone of Holden’s voice scored as reading below grade level. I knew what had happened. Upon asking the student, he said he was tired of the tests. It didn’t count toward anything for him and so he had just marked random choices.

Another student scored well, above what I would have suspected, because of some good guessing. One of my students told me she had conscientiously gone through her test trying to do her best and the girl next to her who was marking in patterns scored just as well.

Standardized testing is not the answer. I would argue it is not even a valuable tool. What is valuable is assessment embedded into the curriculum, which brings us back to a teacher and a student.

So, how do we meet the need to ensure every student has an equal opportunity for a rigorous education? The Common Core standards that have been adopted by 43 states may be the place to start. Set a clear, reasonable bar for what students should know in a reasonable sequence and then trust the teacher and schools to design curriculum and assessments that will get them there. That would set both the rigor and flexibility that Sec. Duncan said is desirable—and it would give my students the education they deserve.

Thanks for participating! Commenting on this conversation is now closed.

3 Comments:

  • @ed_Advocate says:

    June 2, 2011 at 12:54 pm

    Common Core does provide strong common ground as a starting point. I also agree that standardized testing must not be the only evaluator of effective schools and effective teaching. That said, despite the limitations of standardized tests, they do provide an overall snapshot that can be a useful tool in addition to the embedded assessments you describe. Perhaps a redesign of the standardized testing…one that requires more than filling in blanks would help to make the tests more reliable and useful. No, standardized tests are not the answer to all of educationdom’s ills, but they could provide part of the solution.

  • Alec says:

    June 2, 2011 at 3:16 pm

    Kaye,
     

      Please read Standing on the Shoulders of Giants and enlighten people. Our competitors, the successful ones, are doing less and less testing while we do more and more.

    http://www.ncee.org/wp-content/uploads/2011/05/Standing-on-the-Shoulders-of-Giants-An-American-Agenda-for-Education-Reform.pdf

    Also, read studies of our own testing from the very reputable national academies of science, medicine, education, and research.

    Testing has decreased graduation rates without raising achievement.

    http://www8.nationalacademies.org/onpinews/newsitem.aspx?RecordID=12521

    Finally, this article from Harvard Research:

    http://www.edweek.org/ew/articles/2011/05/25/32henig_ep.h30.html?tkn=UOMF71cRoaH+MqvanmWckpe7vTJtO2dfV9hi&intc=es

  • Alec says:

    June 2, 2011 at 3:22 pm

    Problem is that we grade what is easiest and cheapest to grade. The MCA’s used to have open response and gridded response, but you had to pay people to come in and grade them. What is easy to grade does not always assess what is being taught.

    Also, there is often little correlation between test scores and effectiveness even from class to class. I’ve had two classes, same course, same grade, same everything. One set of test scores would have made me teacher of the year and one set were disappointing.

    Test scores, and value added scores are fabulous and I love them when used to guide my instruction and pin point where I might need to make changes, do things differently, or change focus. Even in Shanghai, which is one of the few places that uses test scores to evaluate teachers, the scores are 1/16th of the formula. A holistic approach like that is fine with me.

    We don’t want a holistic approach. We want tests to be at least 50% of our evaluation? Anyway, from year to year a teacher can go from top to bottom and back to top, based on a variety of statistics.

    Using tests for formative purposes and professional growth is good, but they shouldn’t count for 50% of an evaluation.