Methodological approaches and problems in research into alcohol-related accidents and injuries.

A classification system for methodologies and problems in research on alcohol-related accidents and injuries is introduced. On the basis of this system, possible designs and statistical methods are presented, and their possible contributions to research on alcohol-related accidents/injuries are evaluated. Special consideration is given to potential biases in secondary data sources, and to methods of overcoming these biases. Finally, the need for more theory-driven research is established.