How do you prioritize tasks when working on multiple datasets in a clinical programming project?