Behavioral epidemiology of food additives.

Behavioral toxicology in the natural environment can be considered a special branch of epidemiology. Behavioral epidemiology, because it typically relies on complex functional criteria, faces all of the problems of behavior measurement posed by uncontrollable variation, and amplified even further by chemical exposure. Many such issues arose in a study of behavioral responses to artificial food colors in children. Difficulties in employing Applied Behavioral Analysis in such a context run the gamut from selection of retrospective criteria to appropriate statistical models.