Several years ago, we established a goal to share data generated through a very large (200+ schools, 1,500 teachers, tens of thousands of students) randomized controlled trial of an elementary mathematics teacher professional development program in Florida. Our purpose for sharing data were (and are) driven by the following ideals: 1. Enhance the transparency and reproducibility of our findings, 2. Create opportunity for secondary analysis using these data (to leverage the investment in the data for the greater good), and 3. Enable researchers to use our measures and leverage the large data sets that were used to create item parameters for the benefit of their own studies and to increase comparability across studies To reach these goals, we decided to share source data and data dictionaries, replication code, output files, summaries describing our procedures, and a record of decisions made. We intended to share quantitative data created through several waves of data collection and using many different instruments. We wanted to find a template or model that would be replicable and robust to different types of data and also increase the likelihood that subsequent users of the data would be able to make sense of what we shared. We looked for a model or template for sharing educational data in OSF (and other repositories) and struggled to find a model that seemed to align with our vision of how data could be shared. Consequently, we decided to create our own working model/template. We have now used this model to share data from four waves of data collection over a three-year period. Each wave includes data from several different measurement instruments administered to more than 1,200 elementary-level educators. Our model for preparing and organizing the data for sharing is still a work-in-progress, but we think it is useful, and we look forward to presenting it to attendees and discussing ways to improve it. In this workshop, we will start by describing the work we have done so far and our guiding principles. We will share links to data we have shared through osf.io and introduce our working model for preparing and sharing those data. We will invite attendees to take a close look at the material we have shared. After providing the opportunity for attendees to analyze and discuss what we have shared (and how we are sharing it), we will invite attendees to share examples of other models to review and discuss. We will then lead a discussion that focuses on ways to improve the model/template and modify it in the spirit of continuous improvement and robust design.