Data synthesis: No picnic, but no need to panic

Data synthesis integrates heterogeneous datasets to produce innovative science. First, however, researchers must explicitly ask: “Is data synthesis the best approach for our question?”. If so, we encourage affording data synthesis the same attention devoted to other research methods. Broadly, successful data synthesis depends on planning, persistence and patience. Workflow pilots are valuable for organizing the plan and facilitating collaborations across career stages and institutions. Such workflows must often evolve to accommodate the unexpected, such as new variables or data structures. Finally, data synthesis requires a patient, flexible mindset, recognizing that the process of data synthesis is no picnic.