37 Standardized Testing and No Child Left Behind: Solution or Problem – Sara Holman

​

Is standardized testing an unbiased and effective means of assessing students? The No Child Left Behind (NCLB) legislation enacted in 2002 set the goal of all children being proficient in reading and mathematics. Lawmakers chose to measure the goal’s success through standardized testing scores. Many experts see NCLB and the testing practices it put in place as the source of the problem, not the solution. All parties agree that assessment is necessary for both students and schools; however, how to administer or conduct these assessments and the consequences that test scores carry are where they disagree. While NCLB and how it utilizes standardized testing is problematic and not an effective solution, there is no consensus on a better way to handle assessing students and schools.

Lorrie Shepard, currently a Dean of the School of Education and professor at the University of Colorado with a Ph.D. in Research and Evaluation Methods, explored testing history. In her article, Shepard examined testing practices and the theories surrounding them over more than 100 years. Shepard provided her own expert commentary on each of the ages she researched and on testing currently, as well as what it should look like in the future.^[1] Wayne Au earned a Ph.D. in Curriculum and Instruction, and Karen Gourd earned a Ph.D. in Bilingual, Multilingual, and Multicultural Education. Their article, “Asinine Assessment: Why High-Stakes Testing Is Bad for Everyone, Including English Teachers,” presents how the United States testing industry currently creates and grades students standardized tests. Au and Gourd also present a brief history of IQ and achievement testing.^[2] These articles present evidence for the case against the arguments that testing is fair, effective, and unbiased.

William D. Stansfield, an Emeritus Professor of Biological Science, approaches standardized testing and NCLB differently in “Educational Curriculum Standards & Standardized Educational Tests: Comparing Apples and Oranges?” Stansfield argues that the problem is not standardized tests but that the standardized tests and curriculum standards vary from state to state. He argues that “curriculum standards and standardized tests may vary widely from one state to another, making academic-performance comparisons among the states problematic.”^[3] For Stansfield, the requirement differences are the problem since it affects how school effectiveness is compared in each state and nationwide, not the tests themselves. Jennifer Jennings, Ph.D., Professor of Sociology at Princeton, and Jonathan Marc Bearak, Ph.D., take a more neutral tone of the argument. Jennings and Bearak question to what extent standardized testing has affected U.S. education, if teaching to the test was as widely utilized as reported, and how effective a measure it was. However, like the other experts, they point out the need to measure student learning and agree that standardized testing as it is utilized is ineffective.^[4]

In “A National Survey of Middle and High School Science Teachers’ Responses to Standardized Testing: Is Science Being Devalued in Schools?” Mehmet Aydeniz and Sherry A. Southerland report findings from a nationwide survey of middle and high school science teachers. Working off the theory that standardized testing practices affect instructional practices and how teachers assess students, Aydeniz and Southerland questioned teachers about the perceived effects testing has on instruction and standardized testing as a practice.^[5] I conducted a survey to learn where the general public stood on standardized testing and score-based school funding. I based the questions in the survey on the arguments presented by all the authors in their various peer-reviewed articles. Both surveys asked questions relating to accountability and equality.^[6] Aydeniz and Southerland distributed their survey to science teachers in all fifty states and had 161 respondents across fourteen states complete the survey.^[7] My survey had 147 respondents complete the majority of the survey; 108 completed the entire survey, and 25% of all respondents were teachers. A shared limitation of the studies is due to the distribution methods of the surveys. Both surveys were distributed to respondents in all fifty states. Each has its own standards and standardized tests, so respondents are not guaranteed to have the same type of interactions with standardized testing.^[8] While the results of the surveys are significant, they are not comprehensive studies, and more research on the subject of testing needs to be done.

In the testing debate, all parties also agree on the need for assessment for students and schools, even if they do not agree on how to administer the assessments or how it affects students. However, they do agree that NCLB and the testing practices it put in place as a problem, not the solution. William Stansfield starts his argument with, “In 2002, the “No Child Left Behind” (NCLB) legislation was enacted, requiring every K–12 pupil to attain the seemingly impossible goal of becoming “proficient” in reading and mathematics by 2014.”^[9] Shepard, Au, and Gourd are on opposite sides of Stansfield on the effectiveness of standardized tests. However, they all agree that the NCLB legislation is not effective or practical. Stansfield is not the only expert who argues that the goal of NCLB of every student becoming proficient is unattainable. Au and Gourd argue this in their article, pointing out the absurdity of using standardized tests to obtain this goal. They argue that for standardized tests to be considered accurate, some students have to fail them; if any test has the 100% pass rate the NCLB requires, the tests’ validity would be questioned.^[10] The use of the term proficient for Stansfield is also an area of concern in the law since each state is allowed “to define its own concept of proficiency.”^[11] These arguments raise several questions. One, what does proficiency mean, and what does it look like? How can a school let alone states reach 100% proficiency and prove it has done so? Au and Gourd are correct; if any test given to students has a 100% pass rate, the results would question the accuracy of the results, yet that is the goal states were required to meet by 2014. Experts on policy creation contend that goals should be measurable and objective; this goal does not meet this requirement.

Another agreed-upon failure of the NCLB is using test scores as a source of accountability and school funding. NCLB requires states to test students to proficiency in state standards.^[12] These scores are used for accountability: “Test scores are used as both incentive for improvement and measurement of student progress.”^[13] In my study, the general population disagrees with standardized testing being tied to school funding. When asked whether state governments should base schools’ funding on standardized test scores, over 88% of respondents disagreed with the idea, and over 80% disagreed with schools losing funding due to poor scores. Lawmakers and policymakers decided to tie test scores and funding together to ensure educational equality. Lawmakers believed that if they required school systems to report how minority students perform in testing, it “…will encourage school systems to pay increased attention to the academic progress of these historically disadvantaged aforementioned populations and pressure school systems to allocate sufficient resources.” ^[14]Unfortunately, NCLB policies have had the opposite effect. Underperforming schools lose their funding, making it hard to hire quality teachers or get restructured in a way that teaching focuses on increasing scores. Aydeniz and Southerland argue that most of the students in this group do not respond to the type of teaching that would increase test scores. The NCLB mandates harmed the majority of lower-income and minority background schools. The result was that equality is only viewed through increased test scores.^[15] Au and Gourd put it this way: “… Closing achievement gaps really just means that equal numbers of rich and poor kids pass and fail…”. ^[16]

These practices reported by Aydeniz and Southerland are found in the teachers’ responses to their survey. The majority of teachers surveyed stated that their school administration encourages them to teach to the test and change how they assess students in the classrooms because of testing.^[17] Like Stansfield, Jennings and Bearak call for tests to be aligned to set standards and inform everyone what skills are important.^[18] Jennings and Bearak inform audiences that tests are not aligned with the standards that NCLB has required states to put in place and test. Education departments are informing testing contractors what standards need to be tested but not informing them of the weight or importance each standard should carry.^[19]

Each author argues that assessment in some form is necessary to ensure that children are learning and that schools and teachers are educating students effectively. According to Shepard, this led to the creation of the first published standardized test in 1908.^[20] However, this is also where the disagreements begin. While every author agrees that assessments are necessary in their articles and studies and call for the testing process to be changed, they all have different opinions on those changes. Stansfield is a proponent of teaching to the test, and he argues that standards should be uniform. If lawmakers insist on comparing state test scores, use the same test and tell educators what standards students are expected to know.^[21] He argues that testing should help teachers see how effective they are and where they need to improve, assess students’ performance, and allow students to see what they got wrong by showing them the correct answers to missed questions. These three elements for Stansfield are what standardized testing should aim to do, which is why he also argues that test questions should also be made available to teachers, students, and parents.^[22] For Stansfield, the problem is not the tests but how NCLB allows each state to define its own standards and tests.Stansfield, 389 and 392-393.

Shepard, Au, Gourd Aydeniz, Southerland, Jennings, and Bearak agree that testing policies need to change. However, for these experts, standardized tests are also the problem. Shepard argues that standardized testing is harmful, not helpful, when used for assessment.^[23]  Shepard, Au, and Gourd argue that standardized tests themselves are flawed. Shepard argues they were created at a time and by experts who believed in complete objectivity.^[24] Au and Gourd argue that the tests were developed and based on IQ tests that were expanded past their original scope and used for racist purposes.^[25] Jennings and Bearak report that tests are flawed because they are not aligned with the standards that NCLB has required states to put in place and test. Instead, the education departments of each state are informing testing contractors what standards need to be tested but not informing them of the weight or importance each standard should carry.^[26]

Like Stansfield, Jennings and Bearak call for tests to be aligned to set standards and inform everyone what skills are important.^[27] However, this is where the similarity with Stansfield disappears. In a more neutral vein of the argument, Jennifer Jennings and Jonathan Marc Bearak, like other experts, point out the need to measure student learning and agree that standardized testing is ineffective. The general public in my survey also agreed that standardized testing is ineffective. Overwhelmingly the majority of my survey respondents disagreed with the idea that standardized tests were unbiased or an effective means of assessing students or teachers.

Shepard argues for less testing in lower grades and changing the style of testing that is done. Testing, according to Shepard, should be done in the National Assessment of Educational Progress (NAEP) style or use a merit badge style system presented by Al Shanker in 1988.^[28] Shepard argues that progress had been made in the 1990s to improve testing and accurately assess students, but NCLB “undercut the progress that had been made… in improving the substantive quality of state tests.”^[29] The cost associated with the more effective testing is too high for states to implement it, primarily due to the required testing volume.^[30]

As for policy changes, Shepard puts forward an accountability system proposed by Robert L. Linn in 2003. Linn argued at the time that lawmakers must share responsibility and accountability with school administrators and teachers. He argued that if lawmakers do not give schools and teachers adequate resources to reach the set standards, then the lawmakers should also be held accountable for the failure. Interestingly the majority of my survey respondents agree with Linn. While over 80% of respondents do not feel funding should be based on scores, the majority do believe more funding should be directed to schools to meet standards.

Aydeniz and Southerland were the only authors who questioned teachers on standardized testing. Only 15% of their respondents believed that standardized tests showed teachers’ effectiveness, and 141 teachers did not believe the tests measured what students learned.^[31] Compared to my respondents, educators, and the general population, the numbers are very similar. Aydeniz and Southerland argue that “…policymakers should focus on investing resources into curriculum and professional development…[with the] potential to improve the learning of historically underachieving students and of the students who need further challenge…” and call for more studies like theirs to be done so that educators can present hard data to policymakers. They also challenge educators to endorse and be more involved in creating better policies.^[32]

The common vein seen in every article and study is that experts and the general public agree that NCLB has not been effective. When looking at the future of both testing and education reform, there need to be more studies done on what methods of testing/assessment give the most effective and accurate representation of knowledge and skills students have gained. Finding a reasonable solution to the assessment problem will be necessary before real policy change can take place or be seen as effective. Removing the connection between test scores and school funding and changing the testing methods entirely are the most called for and possibly the most effective solutions. However, the changes look different in every group, with no one solution found on each side. Much like NCLB, this is both the solution and the problem.

Bibliography

Au, Wayne and Gourd, Karen. “Asinine Assessment: Why High-Stakes Testing Is Bad for Everyone, Including English Teachers.” The English Journal. Vol. 103. No. 1. (September 2013) pp. 14-19 https://www.jstor.org/stable/24484054

Aydeniz, Mehmet and Southerland, Sherry A. “A National Survey of Middle and High School Science Teachers’ Responses to Standardized Testing: Is Science Being Devalued in Schools?” Journal of Science Teacher Education. Vol. 23. No.3. (2012) pg. 233-257. https://www.jstor.org/stable/43156646

Jennings, Jennifer L., and Bearak, Jonathan Marc. ““Teaching to the Test” in the NCLB Era:  How Test Predictability Affects Our Understanding of Student Performance.” Educational Researcher. Vol. 43. No. 8. (November 2014) pp. 381-389. https://www.jstor.org/stable/24571340

Shepard, Lorrie A. “Testing and Assessment for the Good of Education: Contributions of AERA Presidents, 1915-2015.” Educational Researcher, 45, no. 2 (2016): 112–21. http://www.jstor.org/stable/43996906.

Stansfield, William D. “Educational Curriculum Standards & Standardized Educational Tests: Comparing Apples and Oranges?” The American Biology Teacher. Vol. 73 (September 2011) pp.389-393. http://www.jstor.org/stable/10.1525/abt.2011.73.7.4.

Appendix A: Distributed Surveys

• Survey 1: https://iu.co1.qualtrics.com/jfe/form/SV_cC76vE7pT75zqWq
• Survey 2: https://iu.co1.qualtrics.com/jfe/form/SV_25E5J51R9Ehu7d4