Analytical Transparency

摘要

Data are ever more available to academic researchers, and their ubiquity offers an evergrowing promise of addressing in a brute-force way many of the business problems firms face today. Despite-or perhaps because of-the ubiquity of these data, there is more than ever a need for theory in the research we as academic researchers do. Whether the theories are behavioral or economic in origin does not matter; theories help us navigate a path through data, determine what additional data need to be collected (lab or field) to test those theories, and interpret our empirical results-assuming, of course, that those theories are understandable. Interpretable and well-written theory provides sharp predictions that can be tested (Shugan 2007).