Causal inference in environmental impact studies

Justification of a cause-and-effect relationship in environmental impact studies is complicated by inherent qualities of impact data. Lack of randomization and replication invalidate the use of inferential statistics for inferring a causal link, and place special demands on descriptive arguments for causation. Assembly rules for causal arguments have been developed in epidemiology and provide a rigorous structure for descriptive analysis. Explicit use of assembly rules for making causal arguments allows investigators to efficiently organize, study, and present available evidence. Within this framework, statistical tests can be used to determine if populations at study sites were different after the impact, but the establishment of a causal link between the impact and the observed difference is based on an argument. Causal inference by means of argument is consistent with the scientific method of strong inference and increases the likelihood of correct conclusions.