Earlier this year, Michigan Department of Education announced to raise the cut scores for the Michigan Educational Assessment Program (MEAP) tests. While there are mixed responses from teachers and superintendents, it is generally believed that this is a step towards education reform in Michigan. Arne Duncan, US secretary of Education commented on this action as “Michigan being honest with students on where they stand” and this action will put “kids first by taking critical steps to help them compete in a global economy”.
. What Duncan refers to as not being honest with students is the low standards in standardized tests in Michigan. Before this change, students can score as low as 40 percent correct to be “proficient” on the MEAP tests. Other studies also confirm this belief. A report by Education Trust compared MEAP test with the National Assessment of Educational Progress (NAEP) test and found a wide gap between the results.
“While 84 percent of Michigan fourth graders are proficient by state reading standards, only 30 percent are proficient on the NAEP standards.
In math, our children’s performance is also inﬂated: 70 percent of our eighth graders score proﬁcient on the state math test, but only 31 percent are proﬁcient on the national test.”
Upon receiving these remarks, one may ask” What do the results mean?” To answer that question, it is important to know about the tests: how they are designed, how they are implemented and how the test results are interpreted. I will attempt to explain the above questions through the following paragraphs.
Standardized tests have been used continually to evaluate the performances of students, either in the United States or elsewhere in the world. In United States, the situation gets more complicated due to the politics of education. First of all, there is no national curriculum in the United States. Each state can set its own curriculum in math, reading, etc. It is interesting to note that no two states have the same curriculum. Secondly, most education policies are implemented in the state or district level. Even though there have been recent federal involvements in education, such as the No Child Left Behind Act or the Race to the Top, the right to adjust education programs and agendas remain to those of the states or districts.
Efforts to evaluate students statewide or nationwide were realized as MEAP or NAEP tests. While both being standardized tests, they are two different tests and serve different functions.
NAEP is approved by the US congress. It was tested on a sample of students in the country. NAEP was designed by the National Assessment Governing Board which consists of members from all groups including educators, scientists, businessman, and parents. NAEP does not relate to any teaching standards. NAEP is administered by independent test companies. NAEP is tested on a wide number of subjects including reading, mathematics, science, art and social science. NAEP tests what students “should know”.
On the other hand, MEAP is a project by the Michigan Department of Education. It was tested on all students in Michigan. It designed by the Michigan Department of Education. MEAP is related to the curriculum set by the Michigan Department of Education. MEAP is usually administered by teachers or superintendents. MEAP only focuses on reading and mathematics. In addition, MEAP tests “basic” skills of students. In other words, MEAP tests what students “have learned”.
Another important difference between NAEP and MEAP is the achievement level. Achievement levels were a recent addition to the interpretation of test results. While earlier NAEP test only report raw scores from tests, parents and educators wanted to know what the raw score means. The response is the achievement levels. In addition to the raw scores, NAEP now reports the test scores in terms of achievement levels. The achievement levels for NAEP are basic, proficient and advanced. On the other hand, MEAP wasn’t originally designed to report any scores. Earlier MEAP tests only report if a student reaches basic level of learning or not. The current standards to 2009 begin in 2005. MEAP results were now in four categories: level 1, level 2, level 3, and level 4. Scholars often compare level 1 to advanced, level 2 to proficient and level 3 & level 4 to basic in NAEP standards. Proficient levels, often mentioned in the news, is referring to the students reaching proficient or above in NAEP or level 2 or above in MEAP.
The proficiency levels in state tests are often used by state officials to show their progress. On the other hand, NAEP results have often been used to compare state tests results. NAEP was often viewed as the “golden standard” for standardized tests. As NAEP is calibrated close to international standards set by developed nation, we can see if states set world-class standards by comparing state tests results with NAEP results.
A series of studies headed by Paul E. Peterson explores how large the gap is between different states. They used the data in 2003, 2005, 2007 and 2009 for NAEP and state assessments. Peterson used the difference in proficiency levels and the standard deviation from the average of the difference. While this report does not address the inherent difference between NAEP and state assessments, the reports showed the relative difficulty of standards between states.
Peterson pointed to the No Child Left Behind (NCLB) act to be the main incentive for states to not set world-class standards. Under the NCLB act, states were required to test their students from 3rd to 8th grade annually. Each state would report the test scores to the public. If schools did not achieve levels of growth under the NCLB act, schools would lose part of their funding and ultimately be forced to restructure.
While NCLB act is well intended, it is impossible for the states to follow. First of all, NCLB set unrealistic goals. The ultimate goal for NCLB is for all American students to be proficient in reading in math and reading by 2014. Scholars have simply noted that the ideal of “everyone is proficient” under any “challenging” standards is simply an oxymoron. (Rothstein et al, 2006) Secondly, the growth in test scores NAEP requires is unrealistic. Thirdly, NCLB relies only on test scores for evaluation. Lastly, NAEP allows state to use test results from the state assessments for the NCLB requirements.
Under the above circumstances, it is not surprising to find states to make changes to the test rather than improve on the quality of education. Michigan has been pointed out as one of examples for lowering test standards to meet NLCB requirements. We can see the effect reflected on the gap between MEAP and NAEP scores.
In Peterson’s report, Michigan has a large gap comparing to other states. In 2009, Michigan was marked D minus, with gap only smaller than three other states. We can also see the trend of the lowering standards in Michigan. Michigan was ranked C, C minus, D and D minus from 2003 to 2009.
If we look only at the gap between NAEP and MEAP, the results are staggering. The gap between reading tests is quite large but stable from 2005~2009. On the other hand, the gap between math tests is growing. The difference between the percentage of students proficient between NAEP and MEAP enlarges from 34.5, 47.51 to 52.77 in 4th grade math and from 32.71, 39.18 to 43.97 in 8th grade math.
The above results combined are alarming. Test standard for Michigan has lowered substantially comparing with previous years and other states. While the exact outcome of the Michigan Department of Education raising the cut score is unclear, this action is moving us towards the right direction.
Rothstein, Richard, Rebecca Jacobsen, and Tamara Wilde. 2006. “'Proficiency for All' – An Oxymoron” Education Week, November 29
Education Trust Special Report 2011, Becoming a Leader in Education: An Agenda for Michigan
. 2005. Johnny Can Read…in Some States
, Education Next
. 2006. Keeping an Eye on State Standards
, Education Next
. 2008. Few States Set World-Class Standards
, Education Next
. 2010. State Standards Rise in Reading, Fall in Math, Education Next