(Update: Adding comment from Pearson)

The standardized testing onslaught that has dominated public schools in recent years — and a growing resistance movement by parents, students and educators — has persuaded some states to review the load of  K-12 assessments with an eye toward eliminating exams deemed unnecessary. In fact, some states and school districts have already announced that they have dropped some tests, and some tests are being shortened.  None of that means, however, that standardized testing is going away.

As Bobson Wong writes in the following post, standardized testing is here to stay — even though the quality of many of the exams are poor or at best average in determining the depth of a student’s knowledge and skills. New federally funded Common Core tests were meant to be more sophisticated than older exams used around the country — and are — but a lack of time and sufficient funds meant that they don’t come close to being at the level many had hoped.  Wong (@bobsonwong), who has taught high school math in New York City public schools for nine years, explains how he thinks standardized tests could be made better — and fairer to students. He is a recipient of the Math for America Master Teacher Fellowship and the New York Educator Voice Fellowship. In this post, he writes about standardized tests and how to make them fairer to students.

By Bobson Wong

Amid the recent debate over new teacher evaluation systems, one thing remains clear: standardized tests won’t be going away anytime soon. Whether you support or oppose the use of student test scores to evaluate teachers, we should all agree that our students deserve high-quality assessments that fairly and accurately measure what they know. Unfortunately, much of the debate around the new teacher evaluation system hasn’t focused on the quality of the standardized tests that we give to our students. My experience as a high school math teacher has enabled me to see firsthand what elements of our current testing system work and what don’t.

New York has a generally successful model for developing and sharing questions for the Regents Exams given to high school students. Educators play a key role in the process – all Regents test questions are created and reviewed multiple times by qualified New York State educators, who have at least a master’s degree in the relevant subject area and who are chosen through an open application process. Furthermore, old Regents exams are available online for people to see. As a teacher, I often use previous Regents questions in my instruction.

In contrast, New York used a different model when it contracted Pearson to write its controversial Grade 3-8 assessments. Although teachers have been involved in reviewing these tests, Pearson has much less stringent requirements for its test item writers – the company’s own web site states that a bachelor’s degree is “preferred, not required.” Instead of encouraging public scrutiny of its work, Pearson not only keeps its tests secret but also bans any discussion of exam questions after they are administered. Reviewing and analyzing exam questions – even flawed ones – can be a valuable teaching tool.

The biggest flaws with the Regents exams lie in the ways in which student scores are reported and used. Raw scores are converted into a scaled score ranging from 0 to 100. Most people are unaware that the state sets the scale so that two-thirds of students receive a passing grade (65 or above). The scale misleads many students into thinking that their Regents grade reflects the percentage of questions that they answered correctly.

For example, I have found that students who receive a 65 on the Algebra Regents believe that they got 65 percent of the exam correct even though their true percentage is actually only 35 percent. Manipulating score grades so that a certain percentage of students will pass undermines the higher expectations that we are trying to set with the Common Core Standards. Furthermore, students and teachers are only told the final score. Without finding out what questions students got wrong, we won’t know what topics to focus on for future instruction.

In addition, standardized tests fail to accurately measure many of the incremental gains made by students. For example, I have found that while I can improve the skills of struggling children, whose skills are often years behind, these gains don’t usually result in significant gains on standardized tests. A ninth-grade student who improves from a sixth-grade to a seventh-grade level usually shows no significant improvement on a ninth-grade test. This flaw in our current state tests raises serious questions about their ability to measure achievement gains for all students.

To ensure that our children take tests that correctly convey what they know, we need to improve both the quality of test questions and the way in which results are interpreted. Test item writers should be experienced educators with advanced degrees in the subject area. Test questions should be made freely available to the public after being administered so that students and teachers can learn from them.

In addition, Regents test scores should be reported as performance levels (as Grade 3-8 tests are), not scaled out of 100. These performance levels should be clearly defined to reflect the amount of content mastered, not pre-determined passing percentages. Students and teachers should also be told what questions students got wrong to guide future instruction.

Finally, we need to recognize that our current standardized tests don’t accurately measure growth for all students and find better ways to assess them.

Here’s a response from Pearson:

We couldn’t agree more with Bobson Wong’s 5/28 blog post where he states that “our students deserve high-quality assessments that fairly and accurately measure what they know.” As my colleague Alfred Binford wrote in a recent Answer Sheet blog, Pearson is committed to assessment systems that provide better, quicker feedback on how schools are preparing students for success in the real world.
With respect to two assertions that Mr. Wong makes about the New York Grades 3-8 State Assessments, we wanted to clarify what we see as misunderstandings or misconceptions about our work there. First, Mr. Wong wrongly asserts that Pearson has less stringent hiring requirements for test developers than our competitors. We hire well-qualified, talented individuals to join our test development team in line with requirements set forth by our state partners. Our process to develop tests for New York students – also done in collaboration with and under the direction of our partner, the New York State Education Department – is quite rigorous, multi-faceted, and must adhere to standards that ensure the tests are appropriate for our students. In fact, you may be surprised to know that it can take up to three years to develop a standardized test (see our infographic[pearsoned.com] on testing). We rely on not only department staff but also educators from across the state to contribute to each stage of the test development process to help us design the best tests possible.
Second, Mr. Wong incorrectly alleges that Pearson “keeps its tests secret” and “bans any discussion of exam questions after they are administered.” We believe wholeheartedly in transparency as a fundamental component to test fairness. But each state must make its own decision about how and when it publicly releases its test questions – as states own the test questions, not Pearson. We believe that tests must provide fair and reliable measures of what students know and how they can apply what they learned to solve real-world problems.
Finally, Mr. Wong is a dedicated and passionate educator, and his students have been lucky to have him at the head of their classroom. We share his goal, and the goal of all educators to ease the burden on our students and do our best to make certain that every student is prepared to succeed in life on his or her own terms.