This weekend, education experts and advocates from around the state and the country will be gathering in Raleigh for the 2016 Network for Public Education National Conference. The following essay by a New York attorney named Bruce Lederman highlights a subject of one of the panels at the event — the controversy surrounding computerized evaluations of teachers based on student test scores. Lederman is representing his wife in litigation against the state of New York over its use of this evaluation method.
Misguided obsession with computerized teacher evaluations based upon students test scores
By Bruce H. Lederman
On April 16, the Network for Public Education will be holding its annual meeting in Raleigh. One of the panels will feature a discussion of the New York case Dr. Sheri Lederman, v. John King, Jr, NY Commissioner of Education, in which a single teacher is standing up and saying that her computerized evaluation based upon student scores is simply irrational. Just like in New York, North Carolina teachers are evaluated (called Standard 6) based on how well their students perform on standardized tests which are used to demonstrate academic growth from one year to the next year.
With the nation divided over Common Core, and with parents opting children out of testing to protest its improper use, it’s worth pausing and considering why the political obsession with computerized teacher ratings based upon student scores, particularly with complex computer evaluation models, is bad policy and bad for students.
The teacher evaluation process in question, which is used in one form or another by numerous states, is a complex statistical program known as a value-added model (VAM) or “growth scores.” VAM or “growth scores” were originally developed to study how to improve growth of agricultural seeds by analyzing the effects of factors like fertilizer, sunlight, water and temperature. The problem is that children are much more complex than agricultural seeds and rating teachers by a complex computer program designed to measure seed growth simply does not work. In North Carolina, the value-added model used is called EVASS (Education Value-Added Assessment System) and performs in the same manner by generating an expectation of how students should perform during the year and holding the teacher accountable for lower than expected student performance
A problem discovered as a result of the Lederman litigation demonstrates why parents in North Carolina should think twice when politicians say that they can use complex computer programs to identify and get rid of failing teachers. One of Dr. Lederman’s 4th grade students got a score of 98 on his standardized math test, which meant that he got two questions wrong on three day test. Instead of praising Dr. Lederman for teaching a student who got a score of 98, the computer program labeled her as “ineffective,” with a “growth score” of 22 out of 100, because the student got a score of 100 on his third grade math test. Would you want to punish or fire teachers whose children score 98 on standardized tests? This shows the danger of reliance on complex computer programs which claim to be able to predict performance and rate people’s job performance based upon complex statistical formulas.
Dr. Lederman decided to stand up for herself and fight her “ineffective” growth score in court. Dr. Lederman received perfect ratings from her principal based upon actual observations and went to court with a stack of affidavits from parents and former students saying that she was one of the finest educators that they had ever met. Dr. Lederman’s research into how children learn has won awards. She will be speaking at the NPE conference on how to improve education, particularly when children suffer trauma. People need to understand that complex computer programs which rely upon “big data” to evaluate teachers can unfairly and irrationally identify top teachers as failures, which will demoralize them and force them out of education. That is bad policy and bad for children.
The American Statistical Association (ASA) – the nation’s leading professional association for statisticians and related professionals – issued a position paper that explained some of the problems inherent in using VAM for evaluations of teachers. Their study states: “Ranking teachers by their VAM scores can have unintended consequences that reduce quality.”
Statistical consulting companies are making millions of dollars selling VAM programs to state education departments. These mega corporations seem to have convinced politicians that a computer program can actually work and accurately rate individual teachers. As common sense tells us, however, the issues influencing how children score on a single test on a single day are so complicated that undue and inflexible reliance upon Growth Scores does a real disservice to our children.
Politicians will claim that they want to attract and keep good teachers, but insistence on use of computerized teacher evaluations takes the humanity out of education, demoralizes teachers and is ultimately counterproductive. Dr. Helen Ladd of Duke University agrees that “VAM type evaluations are unreliable for individual teacher evaluations and should not be used for high stakes purposes.”
The success of the “opt out” movement in New York and elsewhere, demonstrates that parents are realizing that increased reliance on testing for the purpose of rating and firing teachers is bad for children. Parents realize that an evaluation system which prizes teaching to the test over good educational practices is bad policy. Anxiety over testing is replacing love of learning. The opt out movement is growing because the goal of good educational practices is being abandoned by politicians in favor of a witch hunt to find and fire allegedly bad teachers.
Politicians should reconsider the obsession with firing teachers based upon computerized ratings which are unreliable and irrational. It is bad for teacher and worse for our students, in New York and in North Carolina.
Bruce H. Lederman is a member of the law firm D’Agostino, Levine, Landesman & Lederman, LLP, which is counsel for Dr. Sheri Lederman in an Article 78 proceeding to declare her Growth Score arbitrary and capricious.