Abstract
The criterion problem of what is considered evidence for the effectiveness of driver education in the U.S., the comparing of different types of education programs and the expectations from research in this field are discussed. The author ends by suggesting some possible criteria.