Using Statistical Evidence to Prove Causality (i.e., Causation) to Non-Statisticians
16 Pages Posted: 6 Jul 2007 Last revised: 2 Feb 2017
Date Written: July 5, 2007
Abstract
Many writers claim that statistics have become increasingly important in litigation. However, no comprehensive contemporary guide exists for attorneys who want to use statistical data to create effective demonstrative evidence - an issue that is especially important when non-statisticians use statistics to make inferences about causality. In this paper we outline a new theory explaining the perception, comprehension, and recall of quantitative graphs. As an outgrowth of that theory we propose that four cornerstones are essential for inferences of causality in most disciplines: Sufficiency, Necessity, Proximity, and Plausibility. Consistent with this theory, we contend that four hallmarks of causality are critical. These hallmarks show positive evidence of Association, Prediction, and Dose-dependence, as well as negative evidence Ruling Out Alternative Explanations. We close by using recent advances from research in experimental psychology to formulate best-practice examples of how these four hallmarks can be shown in quantitative graphs during litigation.
Keywords: empirical methodology, law and psychology
JEL Classification: C00
Suggested Citation: Suggested Citation