Performance evaluation are a tricky business. Arbitrary data systems based on student test scores turn excellent teachers into mediocre ones and vice versa. Individual judgments can also be wrong, but systems that have multiple components can be reasonably fair.
Some states are dropping student achievement gain scores and are returning to human judgment systems for teacher evaluations. Which are these forward thinking states?
Florida still uses the unreliable value added model (VAM) scores as part of the teacher evaluation system. Student gains in achievement are averaged and comprise one third of the teacher’s evaluation. Since the new federal education legislation no longer requires VAM scores, some states, however, are dropping its use.
Oklahoma legislators passed a new evaluation system that focuses on teacher professional development and eliminates student achievement growth scores as part of teacher evaluations.
In his decision in a New York lawsuit, Judge McDonough said that student growth scores (VAM were ‘arbitrary and capricious’. He cited the following reasons:
- disproportionate impact for low and high ends of the scale
- small class size made scores more variable
- students at high end cannot gain on test scores but those at low end can
- the swing in growth across years inspite of similar students
- bell curve standards in four categories regardless of the scores generated made gain scores arbitrary
Hawaii dropped the VAM scores from the Smarter Balanced Assessments. Teacher evaluations have included several components such as portfolios, classroom observations, student surveys, and student learning objectives. Student growth scores were 25% of the evaluation of teachers in tested subjects but five percent of the evaluation of other teachers and staff.
States have the authority to design their own systems. Some educators, however, argue that student gain scores are better than nothing. Perhaps not. The American Statistical Society published a statement back in 2014 that legislators should heed:
“Ranking teachers by their VAM scores can have unintended consequences that reduce quality“.