How drivers perceive and react to traffic events is an important issue. In the North American judicial systems it can mean whether you are charged with a criminal offense, pay large amounts in a civil claim, or are denied compensation for a catastrophic injury.
Many research studies have been published over the years and these have been used by expert witnesses at trial to claim one thing or another with respect to someone’s actions. It is not too uncommon for the timing of these actions to be critical in determining to the court whether someone is convicted of a crime or faces some monetary penalty in civil proceedings. The arguments at trial can be broken down to fractions of a second as experts claim that a certain scenario is supported by a certain body of research.
At Gorski Consulting we chose to collect our own data rather than relying solely on the reported research of others. A common source of perception-response data comes from video-taped observations of the reactions of drivers to a changing traffic signal, as shown in the example photo above.
In a typical urban environment there are often two through lanes and a left turn lane that may be occupied by vehicles in each lane. How drivers react to the changing of a red signal to green is an interesting source of research data. The table below shows the results from 24 observations of drivers’ reactions while stopped for a red traffic signal in the passing lane of a roadway. That driver would be the middle one shown in the above photo.
The data was obtained from examining the videotape and noting the timecode when the traffic signal turns green and then noting when the brake lights of the vehicle become extinguished.
Note that, on average, the delay was about 0.40 seconds. That is an unusually fast reaction.
In an oft quoted section of the book entitled “Forensic Aspects of Driver Perception and Response” authored by Professor Paul L. Olson he made the following observation (Page 187):
“Given a reasonably clear stimulus and a fairly straightforward situation, there are good data indicating that most drivers (i.e. about 85 to 95%) will respond by about 1.5 seconds after the first appearance of the object or condition of concern. The evidence also indicates that the minimum perception-response time for this straightforward situation is about 0.75 second. Thus the probable range of perception-response times for reasonably straightforward situations should be 0.75 to about 1.5 seconds. Please note these values are not chiseled in stone on tablets along with other commandments! The fact that an investigator may calculate that the perception-response time for a given individual under what is judged to be a straightforward situation is 1.6 seconds or even 1.7 seconds is not a basis for judging the driver’s behavior to be unreasonable.”
What would the experts, or anyone else say to the data shown in the above table? Does it contradict Dr. Olson? Even without discussing the differences in the data the first thing the observer should consider is the above-noted data is only reporting 24 observations and such a small sample could cause the true mean (statistical “Population Mean”) to reside to the left or right of this sample mean.
One would also need to consider the issue of “primed” reaction. In other words, the drivers stopped at the traffic signal will likely be aware that their reaction (release of the brake pedal) will be required in the very short future. They are also provided with other clues such as the condition of the traffic signal along the opposing roadway. If they are frequent users of the roadway they will become aware of the typical timing of the signal and when it is likely to turn green. Such pre-existing knowledge is what can cause drivers to react much quicker because the required reaction is not totally unexpected. Also the above drivers have only a single reaction to consider (release of brake) and they do not have to may a choice whether other reactions (steering, looking in a mirror, etc.) are required. Thus this decreases the delay in this simple reaction scenario.
Furthermore, note that in observation #12 the driver of a Honda Accord released the brake pedal 4.07 seconds before the traffic signal turned green. This has caused the standard deviation of the mean to be increased to 1.01 seconds and it has obviously affected the lowering of the average, reported mean.
This example demonstrates the importance of the critical evaluation of what has been reported to us. As receptors of information being provided by experts we must be vigilant to the details and be prepared to question conclusions even though they might come from someone who looks like they know something because of their credentials or experience.