According to a new working paper released by Matthew Kraft of Brown University and Allison Gilmour of Vanderbilt, new teacher evaluations are no better at identifying those with poor teaching skills than the ones previously used in classrooms across the country.
Seven years ago, The New Teacher Project released a study that found less than 1% of teachers in the United States had received a rating of "unsatisfactory" on their annual evaluations. As a result of the study, a nationwide movement to overhaul teacher evaluations grew in an effort to better reflect teacher performance in the classroom.
However, the new evaluations, many of which include test scores and other measures of student learning, were found to have increased the number of teachers identified as being ineffective to just 3%. While the authors considered this to be an improvement, they added that it was "not a landmark change in ratings."
A wide variation was found across states, with less than 1% of teachers in Hawaii being rated ineffective to 26% of teachers receiving that rating in New Mexico.
A large difference was also seen at the other end of the scale, as only 3% of teachers were given a proficient rating in Georgia in comparison to 73% in Tennessee.
Kraft said the results create an issue for those who believe that evaluations should be used to help teachers improve, as well as for those who would like to see the evaluations used to fire poor performers, writes Emma Brown for The Washington Post.
"Both avenues require evaluations that are both accurate and meaningfully differentiate among teachers," he said. "We have to know who the struggling teachers are if we want to help them improve, and we have to know who they are if they continue to struggle, and they should be in another line of work."
The authors also surveyed 100 principals in an urban school district that had adopted a new evaluation system in the 2012-13 school year, finding that, on average, principals believed 28% of teachers to be performing below proficiency standards in the first year. However, they also predicted that they would be handing out low ratings to only 24% of teachers, adding that some scores would be purposely inflated. By the end of the year, it was found that just 7% of teachers actually received poor ratings.
Some principals said they did not feel comfortable telling teachers bad news, while others said they did not have enough time to properly handle the documentation and support that is required when a teacher receives such a rating.
Others said teachers' scores were inflated in an effort to offer encouragement to those who they felt had the potential to perform at a higher level or who were working toward improvement. Others said they did not want to put the effort into hiring a replacement and did not believe a higher performing teacher would come to work at the school, or that they would spend less time pushing for a teacher to find a new job somewhere else than they would if they were to give that teacher a low rating.
Data available to the public on teacher performance ratings was used for the study by Kraft and Gilmour for a total of 19 states that use new evaluations.