Communicating quantitative information: tables vs graphs

In applied statistics and computational econometrics a key task for researchers is to bring the sizable but unstructured body of numeric evidence, for example from Monte Carlo simulation, in a form ready for introducing to scientific dialog. At their disposal they find established means of arrangement: narrative text, tables, graphs. Employing classical principles of communication to evaluate their suitability graphical devices seem optimal. They absorb large quantities of data, and organize content into a productive tool. Graphs confirm the advantage when put to work in a standard simulation exercise. However, theory and application contrast with the norm observed in peer-reviewed journals – by a wide margin and with considerable persistency researchers prefer tables.