The study examines reliability and validity evidence of observational systems for evaluating teacher effectiveness and fostering professional development conversations in schools. Specifically, this study compared the Framework for Teaching's (FFT) validity evidence using the traditional scoring approach with a new composite scoring approach that averages the components nested within each domain. The study was conducted with a sample of 85 teachers and 10 school administrators from five high-poverty charter schools. Overall, the findings build on previous research, offering additional evidence for using alternative scoring methods for the classroom observational measures. The FFT composite scores are internally consistent at the domain and total levels, as well as more stable across time compared to the traditional scoring approach. Implications for research and practice are discussed.