Two recent reports have found serious problems with the two most prevalent methods used to evaluate teachers — classroom observation and the use of test scores — to establish an instructor’s effect on student achievement, reports Matt Barnum of The SeventyFour.
In a statement by the American Educational Research Association (AERA) this month, the association questions the use of “value-added” models that are statistical measures that try to isolate a teacher’s influence on student growth. Using value-added models to assess teachers, principals, and the programs that train them has “considerable risks of misclassification and misinterpretation.” The report says teacher observation may be “a promising alternative.”
However, a paper published a few days after this study finds that classroom observations have many of the same problems that value-added measure have.
Teacher assessments can contain bias due to the students in a class or the manner in which students are placed into a particular group. Matthew Steinberg of the University of Pennsylvania and Rachel Garrett of the American Institutes for Research found that teachers with the best math students, for example, got the best evaluations. And teachers with strong classroom management skills may be assigned students with the most behavioral problems, making it seem to observers that he or she has allowed her students to be unruly, thus earning him or her a lower score.
Criticism of the value-added approach has received greater attention from the media and researchers than the teacher observation method, once again giving observers an incorrect view of the value-added procedure.
“I think [value-added] opponents tend to set unreasonably unattainable targets for what [it] has to achieve in order to be used at all,” says Morgan Polikoff, a University of Southern California professor.
Researchers have gathered evidence that shows a strong value-added relationship to long-run student outcomes, such as income and college attendance. No such findings exist for teacher observations.
The fact remains that teachers continue to be routinely identified as effective or highly effective. This month the National Council on Teacher Quality released a report that identified 42 states and the District of Columbia that have procedures requiring unbiased systems of measuring student outcomes as one portion of teacher evaluations. Five states do not have this policy; three have evaluation policies, but in name only because of waiver requests to the federal government, writes Dian Schaffhauser of THE Journal.
Gov. Rick Snyder of Michigan passed legislation to update teacher evaluations last week. The new law will be adopted during the next academic year. It will not, however, modify requirements in the teacher tenure reform law approved in 2011, which included terminating teachers based on their evaluation rating.
The new bill allows districts to use evaluation techniques of their individual choosing, but school officials must use the method they choose at all schools in the district. The bill does not feature student growth and test results as much as in earlier measures.
“The law before this bill was passed had student growth (accounting) for 50 percent of a teacher evaluation,” said Wayne Roedel, Fowlerville Community Schools superintendent. “Now, 25 percent will be based on student growth for the next three years.”
Also, half of student growth will be measured by standardized test scores while the other half will be identified by local measures or assessments, says Livingston Daily’s Abby Welsh.
Carol Burris, a former principal of South Side High School in New York, wrote in September of 2011 that the new New York State teacher and principal evaluation system would fail. And is has. Burris explains that the evaluation system, rushed into approval by Governor Andrew Cuomo, will be on hold for the 2015-16 school year because the State Education Department, and 90% of the state’s school districts, could not agree on a new evaluation system by the November 15th deadline.
Last year Cuomo said that too few teachers received bad ratings. His solution was to increase the weight of test scores in teacher evaluations and made the legislature comply. That meant that the Common Core standardized test scores rose to 50%, and 50% would be based on observation. Burris says this resulted in teachers “teaching to the test.”
“These tedious and ineffective teacher evaluation systems were funded in some school districts by the Bill and Melinda Gates Foundation, and Duncan’s Education Department coerced states to adopt them with Race to the Top dollars and No Child Left Behind waivers. But there was never any evidence that they would work, or serve the best interest of kids,” said Burris.