Contemporary teacher evaluation systems are built upon multiple measures including, primarily, teacher-level value-added and observational estimates. While researchers have conducted examinations of these systems and indicators, researchers have not investigated how using these systems might distort the validity of the inferences being drawn, especially when one indicator is trusted over the other. In this conceptual piece, we introduce three ideas predicated by Campbell's Law (1976), and defined by how those with power within such systems (e.g., principals, administrators, others who observe teachers) might (1) artificially inflate or (2) deflate teachers' observational estimates when used alongside teachers' value-added estimates. We also discuss how said professionals might (3) artificially conflate teachers' observational and value-added estimates to purposefully (albeit perhaps naïvely) exaggerate perceptions of validity, as per the engineering of inflated correlation coefficients between these indicators over time. Correspondingly, we call scholars to engage in further research surrounding these validity-relevant issues.