Fields from political science to critical education policy studies have long explored power relations in policy processes, showing who influences policy agendas, policy creation, and policy implementation. Yet showing particular actors' influence on specific points in a policy text remains a methodological challenge. This article presents a five-stage, embedded mixed methods design for establishing influence on educational policy moving from a policy text outward. I use an example analysis of Australia's policy making on boys' education--the report "Boys: Getting it Right" (2002)--to show how data transformation measures, both quantitizing and qualitizing, within a larger qualitative study helped identify influence. This mixed design, I argue, can be useful in other research contexts, with variations for data availability and researcher resources.