In her latest MUST READ commentary piece, fellow public education advocate, Wendy Lecker, lays out the facts about Governor Malloy’s unfair, inappropriate and fatally flawed teacher evaluation system. Like the junk bonds that helped take down Wall Street, Connecticut’s teacher evaluation system is based on junk science and false assumptions.
The question is not whether the state should have a comprehensive teacher evaluation system, but whether the corporate education reform industry will continue to stand in the way of developing one.
While recent attempts by the teacher unions to “correct” the flaws in the Malloy administration’s teacher evaluation system are laudable, the fact is that Malloy’s system cannot be fixed because it is inherently dependent on standardized test scores that fail to evaluate teachers on factors that are within their control.
As Wendy Lecker explains, the present system is not only unfair but is actually a useless waste of a massive amount of time and money. It needs to be repealed and replaced by a teacher evaluation program that actually enhances the quality of education in our schools.
Wendy Lecker’s piece from this weekend’s Stamford Advocate and Hearst Media Group papers is entitled, Solution to failed tests is not more tests
Fact: Connecticut’s teacher evaluation plan, because it relies on student standardized test scores, is fundamentally flawed. Student test scores cannot measure a teacher’s contribution to student learning. In fact, the president of the Educational Testing Service recently called evaluation systems based on student test scores “bad science.”
Rather than admit failure, the Malloy administration is trying futilely to “fix” the fatal flaw. Last week, PEAC, the panel charged with developing Connecticut’s teacher evaluation system, working under the direction of Commissioner Stefan Pryor, approved a change which calls for more standardized tests to be included in a teacher’s evaluation.
The commissioner’s “solution” is to add interim tests to a teacher’s rating. Determining what tests will be used, how they will be aligned to the standardized tests, and how all the test scores will be rolled into one “score” for teachers, will likely render this change completely unworkable.
However, there is an even larger issue at play. Will the addition of more tests in a teacher’s evaluation help us measure whether a teacher is effective?
According to the Connecticut Supreme Court, Connecticut’s public schools must prepare children “to participate in democratic institutions, and to prepare them to attain productive employment and otherwise to contribute to the state’s economy, or to progress on to higher education.”
Thus, we want our children to acquire the skills and knowledge that will enable them to succeed in college and in life. We want teachers who will help our children develop these skills.
Standardized tests have no bearing on college success. Moreover, although standardized tests are supposed to measure cognitive skills, research from MIT has shown that increasing test scores does not increase cognitive skills.
Even more striking is that cognitive skills, while important, are not the most important skills in determining success either in college or in life after college. Research has shown again and again that non-cognitive skills such as self-discipline, taking responsibility, and listening skills are more critical.
A recent comprehensive study by Northwestern Professor Kirabo Jackson found that children with teachers who help them develop non-cognitive skills have much better outcomes than those who have teachers who may help them raise test scores. Jackson found that every standard deviation increase in non-cognitive skills corresponds to a significant decrease in the drop-out risk and increased rates of high school graduation. By contrast, one standard deviation increase in standardized test scores has a very weak, often non-existent, relationship to these outcomes. Test scores also predict less than 2 percent of the variability in absences and suspensions, and under 10 percent of the variability in on-time grade progression, for example.
Increases in non-cognitive abilities are also strongly correlated with other adult outcomes, such as a lower likelihood of arrest, a higher rate of employment and higher earnings. Increased test scores are not.
In short, focusing on non-cognitive abilities, those not measured by test scores, are more important in predicting success in high school and beyond.
Jackson also found that a teacher’s supposed effect on test scores is not related to how well that teacher can improve non-cognitive skills.
Moreover, a new statement by the American Statistical Association reminds us that ranking teachers based on test scores does not even work for measuring their effect on cognitive skills.
ASA notes that teachers account for 1-14 percent of the variability in student standardized test scores. The majority of variability in test scores results from “system-level conditions”; meaning everything affecting a student outside the teacher’s control: the child’s socio-economic status, parental background, language barriers, medical issues, student mobility, etc. Rating systems cannot eliminate the “noise” caused by these other factors.
ASA further states that test scores at best “predict only performance on the test.” This conclusion confirms Jackson’s results, i.e that tests cannot predict how well a student will succeed in school or life.
In the context of this evidence, what does the PEAC change mean?
By adding more tests of the same skills in the same subjects, PEAC merely added more meaningless “noise.” This addition will not give us any better picture of how well a teacher teaches.
Worse still, adding more tests increases the focus on tests, increases the frequency of testing, and distracts us from considering the skills teachers should be helping children develop. And since Connecticut’s evaluation system completely ignores these non-cognitive skills, they will be de-emphasized in school.
Meaningful evaluations systems can be developed, but relying on faulty measures is simply rearranging deck chairs on the Titanic. Connecticut’s students, parents, teachers and taxpayers deserve better.
You can read the full commentary piece here: http://www.stamfordadvocate.com/news/article/Lecker-Solution-to-failed-tests-is-not-more-tests-5449394.php