Classroom observation is a significant component to many of the new teacher evaluation systems being designed nationally, yet it is has not received the same level of attention and scrutiny as value-added measurement, even though it is weighted more heavily in many evaluation systems. Researchers at the Brookings Institute examined the use of classroom observation in four urban districts and found a troubling bias in observations that favored teachers of high achieving students and punished teachers of struggling learners, including ELLs. The researchers suggest approaches for taking demographics and classroom conditions into account in observations, as well as the inclusion of trained outside observers, in order to assure a more equitable system that does not discourage teachers from working with diverse student populations.
Whitehurst, G., Chingos, M., and Lindquist, K. (2014). Evaluating Teachers with Classroom Observations: Lessons Learned in Four Districts. Brookings Institution.