Why Candidates for Governor and State School Superintendent of Georgia Should Oppose High-Stakes Testing?

Latest Story

"Creative Commons Gravel Hill School 1950" by Erin Nekervis is licensed under CC BY-ND 2.0
“Creative Commons Gravel Hill School 1950” by Erin Nekervis is licensed under CC BY-ND 2.0

In an earlier post, I challenged candidates for state school superintendent to oppose the Common Core State Standards.  Today, I am writing to candidates for Governor and State School Superintendent of Georgia to oppose High-Stakes testing.  If they would, they’d open the door to a new paradigm of assessment that would improve education in Georgia beyond their wildest dreams.

Since 2001, the U.S. Department of Education has mandated the annual testing of children as young as 7 years old in mathematics and reading, and most states have added mandated high-stakes testing in writing, science, and social studies.

The American Education Research Association states that it is a violation of professional standards to make decisions about students’ life chances or educational opportunities on the basis of test scores alone.  Yet schools around the state of Georgia and indeed the rest of the country use end-of-the-year tests to make crucial decisions about whether students move on or not.  Additionally, these high-stakes tests have become an even greater burden on students because they know that the test results will be used to grade their teachers.

There is no easy answer to explain why we have an educational system that puts students in harm’s way by the continuous and unparalleled testing program.  When we read the newspapers soon after the release of international, national, or state tests, the emphasis is on who came in first, or who is at the top of the leader board.  No Child Left Behind and the Race to the Top have perpetuated an educational model based on competition and winning.

In some cases, officials will do what ever it takes to make sure they either win, or make the cut so that they place high on the leader board.  The Atlanta Journal-Constitution exposed the cheating scandal in Atlanta, and revealed that cheating (erasing wrong answers and changing to the correct answers on student test forms) was taking place in most cities and states.

Am I advocating the banning of high-stakes testing because it might lead to cheating.  No.

I am advocating banning high-stakes testing because it does not improve student learning, nor does it help teachers change their instruction to improve student learning.   Most of the those who advocate high-stakes testing believe that American education is failing, and that the fundamental goal of schools is improve achievement scores, and the only way to know if that has occurred is to use high-stakes standardized tests every year, and compare the scores from one year to the next.

But, if we do compare the test results from one year to the next, the results are quite astonishing.  First, we discover that in general, academic performance has gradually increased over time.  Secondly, we do see variation in average scores from one year to the next, but the variation is within expected statistical limits.

To give evidence that you might want to use with your constituents or potential voters, I am going to use a few graphs that were prepared by Mr. Ed Johnson, which were published on this blog earlier this year.  I am also going to use charts from the Anne E. Casey Foundation, and pull together data from various state and federal agencies.

NAEP Trial Urban District Assessments (TUDA)

The National Assessment of Educational Progress (NAEP) provides some of the most reliable data on student learning.  The tests given by NAEP are low-stakes, and an individual student takes only part of the test, so they don’t spend hours sitting for the exam.  NAEP has been studying American education since 1969.

About a decade ago, NAEP launched a study of urban school districts which they refer to as TUDA.  They provide telling results that I think will help you with the case of abolishing high-stakes tests.

The four graphs shown below in Figures 1 and 2 were prepared by Mr. Ed Johnson, an expert on W. Edwards Deming’s system of profound knowledge and how to transform organizations that result in continuous improvement.  He also is an expert in using facts to generate flow charts that help us understand how a system is working.

Figure 1 plots math scores for 21 cities over a ten-year period (for a list of the cities, follow this link).  Note that the scores fall within what are called upper control limits and lower control limits.  In no case do scores fall outside these predicted levels.  Yes, there is variation in the scores.  But they are within expected limits, and the variation is small.  For example, the green and red dots follow the Atlanta Public Schools (APS) and the Austin Independent School District (AISD).  If you show this graph to citizens in Georgia and ask if these graphs support the idea that our schools are failing, what is the answer? The answer is No.

Figure 1. Control Charts for NAEP TUDA Math Grades 4 & 8, 2003 - 2013
Figure 1. Control Charts for NAEP TUDA Math Grades 4 & 8, 2003 – 2013. Source Ed Johnson

Figure 2 plots reading scores for the same 21 cities over 11 years. Again, note that for the most part, the scores for each district fall within the expected limits, except for five points of measurement.  Each is labeled and as you see, only Charlotte and Hillsborough fall outside on the 4th grade reading TUDA results.  Think about this.  On only these five instances can we show significant variation from what we expect on the reading test.  The obvious thing to do, is to ask, what are these two districts doing, and how might what they are doing apply in other places.  It might be worth studying their system of education.

But, the real discovery here is to look at all math and reading TUDA results.  There are roughly 408 points of measurement shown in these four graphs, and in only five instances was the variation outside the range expected.   That is 0.012 percent. The systems of teaching math and reading in these 21 cities is predictable and consistent.

We can also see that there are no major swings in the test results.  When we send kids to school, we have a very good idea what to expect.  Another way to say this is that the system is performing as expected.

Or better yet, our teachers are doing it!

But there is always a need for improvement.  In Ed Johnson’s and W. Edwards Deming’s world of human systems, there is always the expectation for improvement.  The methods of improvement do not include the outright firing of department heads, or rank and file workers, any more than would we think that firing principals and teachers and bringing in uncertified and inexperienced teachers would help the situation.   But this is exactly what the Georgia Department of Education mandates when schools “fail” to meet the standards two years in a row.  Schools in this situation are labeled “turnaround schools.”

Here is what you need to know.  The high-stakes testing model is designed to make it very difficult for some schools, especially those schools where most of the students are eligible for free or reduced lunches (a statistic used to identify the poverty level of a school).  We know that students in less affluent schools will not do as well on these tests as students attending affluent schools.  It’s an unsustainable situation because these schools and their neighborhoods are punished by either closing the school or labeling it a turnaround.  But this is a sure ticket for financial rewards for charter management companies and teacher temp agencies including Teach for America and the New Teacher Project.  Follow this link to find out a better way to help these schools.

Labeling schools as failures is not sustainable.  It will not improve instruction. It represents an inaccurate interpretation of testing, and it is perverting a system that should be helping families, rather than punishing them.

Figure 2. NAEP TUDA, Reading, 4th & 8th Grade, 2002 - 2013.  Source: Ed Johnson
Figure 2. NAEP TUDA, Reading, 4th & 8th Grade, 2002 – 2013. Source: Ed Johnson

If We Were to Ban High-Stakes Tests?

Ok.  As a candidate for governor or state school superintendent in Georgia, and you were to go around the state campaigning for the banning of high-stakes tests, the odds are you would be elected.  You will be surprised who will support you, but you will need to tell the rest of the story.

Yes, you will support the idea of banning high-stakes tests.  But you need to clear that you are not suggesting that teachers and administrators all of a sudden stop assessing students.  Nothing could be further from the truth.

If you have been a teacher, you know that the assessment system  you use in your classroom has a major impact on student learning and classroom behavior.  Assessment is an integral aspect of teaching.  As teachers we assess students during every class session, and interaction that we have with them.  Teachers know that assessment, used as part of instruction, does indeed help student learning.  This is not an opinion.  One of the foremost researchers on assessment is Professor Paul Black, King’s College, London and he has found that formative assessment strategies do improve learning for students.  Formative, unlike the high-stakes tests that the government mandates, are embedded in instruction.  In my view, formative assessment is assessment for learning, not of learning.

Formative Assessment

Formative assessment are tools and methods that teachers use to humanize learning, and give students opportunities to apply their learning, and to engage in activities that involve communication, problem solving and team work–the kinds of skills and abilities that are important today, and will be tomorrow.

As a candidate for governor or state school superintendent you should listen to your most important constituency, and this is the professional teachers in public schools.  Last year, there were more than 111,000 teachers in Georgia teaching 1.6 million students.

So, what would happen if you said to nearly 1.6 million students (and their parents) and 111,000 teachers that the state would no longer need high-stakes tests.  Would the education system crumble?  Would students all of sudden not be motivated to learn?

It would thrive.  And it would free up a lot of money that would otherwise go to corporations.

According to a Brookings Institute report, the cost of testing in the U.S. exceeds $1.7 billion.  But according to the report, that is only for payments to testing vendors who also score the tests.  But what is the cost for lost instructional time.  In Georgia, the CRCT exams and high-school end-of-course exams take three-four weeks during the year.  So for about 4 weeks or about 12% of the school year, high-stakes exams dominate the school experience.

What’s the cost of 4 weeks of testing?  Well according to the Brookings Institute, about $600 billion is spent on education in the U.S. per year.  According to the Georgia Budget & Policy Institute, the 2015 fiscal year budget for K-12 education is $7.95 billion.  The cost of high-stakes testing is at least $950 million.

Every school in the state already has experts on assessment, and these educators need to be supported to collaborate with colleagues to develop assessment methods that will improve student learning, and increase student’s love of learning.

We know from many research studies that the best predictor of success in college & career (college & career is the favored purpose of reformers such as Bill Gates) are grades, not test scores.   Teachers are in the best place to assess their students.  Not only are they able to create their own tests, but there are multiple resources available that teachers already use to help their students learn.

Imagine if you were a high school biology teacher, and it was announced that the state would no longer need high-stakes tests.  How would this affect your teaching, and especially your relationship with your students.  One obvious difference is that the curriculum will expand because you no longer would be forced to teach to the test.  No longer would the students in your class be required to take tests that would be used to not only to decide whether they progress to the next science course, but the tests would no longer be used to decide if you keep your job.

In the next post, I’ll go into more detail about what assessment would look like in this alternative paradigm.

In the meantime, are you willing to discuss the possibility of returning the education of students into the hands of professional teachers?