In a paper entitled "Simulator and Field Measures of Driver Age Differences in Left Turn Gap Judgments" presented at the 74th Annual Meeting of the Transportation Research Board (January 1995), Loren Staplin of The Scientex Corporation described results of a study with a twofold purpose:

1. To test prior research findings suggesting that older drivers undergo a sharp decline in their ability to judge distances, closing speeds, and safe turning and crossing intervals and
2. To determine the best way to approximate real-world conditions in a driving simulation test.

The Study

Researchers recruited a sampling of 79 paid test subjects from Pennsylvania driver license centers: 25 "young/middle-aged" drivers (mean age 33.3.); 29 "young-old" drivers (mean age 65.1); and 25 "old-old" drivers (mean age 79.4).

Simulator Tests

After administering a battery of tests to ensure that the subjects' cognitive performance and visual acuity fell within "age norms," researchers exposed them to a set of three simulator tests, spaced a month apart to eliminate learned responses.

To create the test simulations, researchers took 35mm film of a white Mercury Marquis approaching an intersection at 30 miles per hour. The vehicle approached from around a bend that was 1.6 miles from the intersection. During simulation tests, researchers asked subjects to respond to the image of this approaching vehicle as if they were motorists waiting at the intersection to make a left-hand turn. The approaching Marquis was presented in three formats:

1. On videodisc through a 20-inch television monitor that lacked not only the resolution but the accurate size and space cues of the original film,
2. On videodisc through a large-screen television projector that preserved the size and space cues but not the resolution of film, and
3. Projected onto a screen from the original 35mm film with a full resolution of more than 3,000 horizontal lines--as compared with less than 400 for videodisc-generated images.

In all three tests, subjects tapped the horn when they first became aware of the oncoming vehicle and depressed the brake at what they judged to be the last safe moment to make a left turn in front of the Mercury. The tests were then rerun with the Mercury approaching at a simulated 60 mph.

Field Tests

The remaining subjects (there was some attrition) were then taken to the site where the Mercury was originally filmed. This time subjects sat in a drivable car equipped with simulator controls, and researchers asked them to repeat the test responses as the Mercury was actually driven by in the oncoming lane.

Results

As shown in Figure 1, older drivers felt the need for a larger gap than did younger drivers before venturing a turn in front of an oncoming vehicle. One alarming finding was that older drivers' estimates of the minimum safe gap decreased when the speed of the oncoming vehicle doubled. Younger drivers, on the other hand, seemed to maintain a constant time interval, which automatically lengthened the gap as closing speed increased.

Responses evoked by the 35mm film projection simulator were most consistent with those produced by controlled field tests. The television monitors ran a poor third, with target recognition distance being especially degraded. For all subject groups except the oldest, however, the study concluded that the three formats produced sufficient information, with adequate lead time, for the subject to make a valid judgment. Apparently the oldest test group had suffered a significant decline in their ability to process visual cues of distance and closing speed and to make valid judgments based on those clues.