Three kinds of data are relevant to diversity, equity, and inclusion in physics and other fields. Quantitative data about demographics (e.g., race and binary gender, and less often sexual orientation, other social identities, and their intersections) are widely available; student degree completion data from Michigan, MIT and other departments will be compared. Quantitative and qualitative data about the learning and work environment are sometime available, as in the 2016 Michigan Campus Climate Survey and in experiences shared by individuals. An institution’s policies and practices represent a third kind of data useful in guiding a department. This presentation will share three case studies on data-driven change from different departments at MIT and will reference recent recommendations from professional societies (APS, AAS, AAAS, NASEM) for improving equity and inclusion.