In this study, we examined how the relationships between one observation protocol, the Protocol for Language Arts Teaching Observation (PLATO), and value-added measures shift when different tests are used to assess student achievement. Using data from the Measures of Effective Teaching Project, we found that PLATO was more strongly related to the Stanford Achievement Test (SAT-9), the alternative assessment used by MET to assess more ambitious outcomes. We also found that the SAT-9 is more instructionally sensitive to the PLATO factor of Cognitive and Disciplinary Demand than the state tests used in MET study. This difference suggests that PLATO factors designed specifically to identify ambitious instructional practices are especially sensitive to which test is used to construct value-added scores.