Opportunities and Issues in Multiple Data Type Meta-Analyses

When tackling complex problems to help with decision-making, we may often have access to multiple sources of data, each of which provide partial information to answer a primary question of interest. By considering the totality of data simultaneously, instead of performing analyses on each data type separately, we can leverage across all types of data to deepen our understanding, appropriately calibrate the uncertainty in our estimates and predictions, as well as potentially reveal weaknesses in our underlying theory. We explore some of the objectives and complications associated with data combination, analysis, and design of experiments in meta-analyses by considering three examples from diverse applications.