In order to shed more light on safe driving as a visually connected rational process, an experimental study with 100 driver subjects was undertaken. The purpose of the study was to determine if the driver's ability to perceive and rationally deal with pending collisions could be taught in a classroom situation using projected threedimensional traffic interaction scenes. The experiment was undertaken and tested the following null hypotheses:​

• Drivers who have had visual training recognize pending collisions no more accurately than those who have not had such training.
• Drivers who have had visual training recognize no greater quantity of collision producing situations than those not trained.
• Drivers recognize pending collisions no more accurately when trained tachistoscopically utilizing threedimensional slide images than when two-dimensional images are used.

Results of the study indicated that the null hypotheses could be rejected and thus, conclusions which stress the importance of the visual processes to driving may be drawn. The effects of this experiment tend to alter the present emphasis on manipulative skills in driving to a more visually connected decision-making process.