The Framework for Teaching (FFT) is one of the most widely used observational systems for evaluating teacher effectiveness and driving professional development conversations in schools. This study contrasts reliability and validity evidence relevant to the FFT as traditionally scored with evidence relevant to a composite scoring approach that connects to specific practice feedback. The FFT is typically interpreted at the domain level and at an overall total level using four categories of teacher effectiveness--unsatisfactory, basic, proficient, and distinguished--scored without computational rules governing relationships between the 22 components and four domains. For this study, the composite scoring approach was computed by averaging the components nested within each domain. A sample of 156 teachers and 34 trained school administrators from 12 high-poverty charter schools used the FFT as part of regular evaluation practices, yielding an extant set of de-identified data. Results indicate the composite scores were internally consistent at the domain and total levels. In comparison with traditional scores, the composite scores were more stable across time, as well as more predictive of student growth in reading and mathematics achievement. Implications for professional development and educator evaluation are discussed.