This document discusses approaches to developing comprehensive teacher evaluation systems using student achievement data. It outlines five propositions for validating such systems and the design claims and evidence needed to support each proposition. It then describes different approaches used across the US, including student learning objectives, subject- and grade-alike measures, universal pre-/post-tests, and value-added composites. Both advantages and disadvantages are provided for each approach. Finally, it discusses examples from Minneapolis and Edina that use value-added measures and standardized tests to evaluate teacher effectiveness.