Second Chances: How to Upgrade Your Safety Rating

How to get back on track after a disappointing Compliance Review.

What happens now when FMCSA comes to audit your logs?

Check out this excellent article by our friend Deborah Lockridge.

Hours of Service: California vs Federal

Overview of Difference Between Federal and California HOS Rules

Underride Accidents

Dr. Green explains the problems associated with underride crashes.  The photographs of the semi-trailer post-crash in most cases will show the reflective tape more prominent than what the driver saw.  Among many other things, the angle of the trailer across the road may diminish the reflectivity of the tape and alter the size of the side marker lamps which the driver sees.  As discussed previously perception-reaction time is an important factor.  FMCSR require trailers manufactured on or after December 1, 1993 shall be equipped with retroreflective sheeting meeting FMVSS No. 108.  The only way to ensure that this standard has be met is through testing by an expert in this field.  As a motor carrier, you must be aware of the “shortcuts” that experts on the plaintiff’s side will use (overtly or lack of training) when testifying as to lamps and conspicuity systems.

Marc Green

Side underrides occur when the driver collides with the side of a trailer that is blocking the roadway during a turn or while backing up. The driver’s car chassis often goes beneath the trailer and the car’s roof is sheared away. Unlike the trailer rear, the side has no guard to prevent the car underriding the larger trailer.

Here, I explain how to perform a more realistic examination of the factors that contribute to an underride accident. The avoidance task requires the driver to first sense the trailer’s contrast, the difference in brightness between the trailer and its background. Next, the driver must notice and identify the contrast as being a trailer blocking the roadway. The driver then assesses the situation and gains an awareness that a collision is imminent if he takes no action. Next, he selects a response and finally, if time permits, he responds.¹

Driver ability to successfully perform these steps depends on the constraints imposed by the physical, situational factors and by innate human limitations and predispositions. The following discussion describes how these two sets of factors operate when a driver faces a potential underride collision. I also discuss how scene photographs can mislead jurors.

Situational Variables

Tractor Headlamp Illumination

In many cases, the tractor-trailer is stopped with its headlamps pointing in the oncoming driver’s direction. The headlamp beams shine in the driver’s eyes and cause loss of contrast sensitivity. This occurs through two distinct visual mechanisms:

Light Adaptation. Viewers see contrast best when the eye is adapted to the prevailing ambient light conditions. A driver traveling a dark roadway would be adapted to a relatively low light level. His adaptation level, however, undergoes changes as he approaches the bright trailer headlamps. This bright light causes the driver to light adapt his fovea, where contrast vision is best. When the driver looks back down the road he is then mis-adapted for the dark area directly ahead. His ability to see the trailer is reduced.

Glare. A glare source is defined as a light that is significantly brighter than the prevailing light adaptation level. Glare effects fall into two main categories. “Discomfort glare” is the unpleasant visual sensation caused by a bright light. As the driver approaches the glaring headlamps and their brightness becomes very high, he may experience visual discomfort that causes him to look way from the light source. A glare source originating from the tractor in the left oncoming traffic lane may cause a driver to look somewhat rightward rather than directly straight down the roadway. Objects in the road ahead are then seen in peripheral vision, which has lower contrast sensitivity.

The second type of glare is called “veiling,” because light from the glare source enters the eye and scatters around the ocular media. The viewer is effectively looking through a veil to see the world, and loses contrast sensitivity.

The amount of glare depends on several factors. Glare increases as the amount of light entering the eye grows. The closer the driver gets to the tractor, the more light entering the eye. Further, glare increases as the angle between the light source and the sightline decreases. Glare is likely severe on a narrow two-lane road, where the tractor’s left headlamp may lie as little as 6 feet or less to the driver’s left. Lastly, as people become older, the eye clouds, causing more internal scatter and more veiling. I discuss this more fully below.

Several other factors affect glare. One is height of the headlamps. Tractors create large glare effects because they have their headlamps mounted relatively high, putting them close to eye height for a normal car driver. The higher mounted headlamps project farther so they affect the driver at a greater distance down the road.²

Glare also increases when the viewer looks through scratched/dirty surfaces such as windshields and even eyeglasses. These surfaces scatter light more and increase veiling. Further, the scratches become visible and mask the roadway scene ahead. Rain also lowers visibility distances by a half or more.

Trailer Angle

The trailer’s angle across the roadway is also an important variable. Visibility and conspicuity of the retroreflective tape, side marker lamps and the trailer itself decline as the angle departs much from 90^o, when the trailer is perpendicular to the roadway.

Retroreflective Tape. The effectiveness of a given retroreflective tape strip depends on several factors, including “entrance angle,” the angle between the headlamp beam and the retroreflective tape surface, and “observation angle,” the angle between the headlamp beam origin, the retroreflective surface and the driver’s eye (Figure 1).

Figure 1 Entrance and observation angles.

If the trailer angles across the roadway, then the increased entrance angle reduces the percentage of incident light that is reflected back in the beam direction and to the driver’s eyes. Diamond retroreflective material³ is frequently used for tractor trailer markings. It is clear from the table that tape effectiveness falls significant when entrance angle is higher.

Moreover, the effectiveness also falls with increased “observation angle,” the angle formed by the eye, headlamps and retroreflective surface. Drivers sitting higher above their headlamp beams, such as those in a large truck, have higher observation angles.

The lowered reflectance and brightness of angled retroreflective tape also affect distance perception. Even if the driver sees the retroreflective tape, he must interpret its meaning and judge its distance. When objects are very small and approach the resolution limit of the eye, viewers confuse, size, distance and brightness, and they judge the apparent distance based on their brightness. Dirt and tape wear will further lower brightness and increase apparent distance.

Side Markers Lamps. People see objects by sensing the images that they project on the eye. This image size depends on the object’s orientation. Look at your hand in front of your face. Now turn the hand at an angle and notice that the profile it presents to the eye shrinks. The same effect shrinks the effective size of side marker lamps on angled trailers. For example, if the trailer angle is 45^o, then a 3 inch lamp effectively shrinks in size to 2.1 inches wide. At 60^o, it is effectively 1.5 inches wide. Moreover, if the lamp output is directional, then it will aim away from the driver’s eye.

The side marker lamps, like the reflective tape, have a very small size that causes difficulty in judging distance. As they shrink, they act increasingly as “point sources,” whose distance is difficult to judge. They are also easily confused with other point sources such as reflectors on mailboxes, etc.

Trailer Type. The effect of angle depends partially on the type of trailer, tanker, box or flatbed. When light hits a trailer, the direction of scatter depends on the surface smoothness. At one extreme, a perfectly rough matte surface, called a “Lambertian surface,” scatters light equally in all directions. In this case, the angle does not affect the amount of light reflected from the headlamp beam back to the driver. At the other extreme is a perfectly smooth “specular” mirror surface, which reflects light only in one direction according to the rule, “angle of reflectance equals angle of incidence.” Real surfaces generally fall somewhere between Lambertian and specular. A smooth metal tanker, however, will do a good job of approximating a specular surface. Suppose that a trailer is sitting at a 45^o angle in the roadway as shown in Figure 2. The headlamp beams strikes the trailer at a 45^o angle. Most light then reflects at a 45^o angle, which is a 90^o angle to the driver’s eye. The driver will see only a black hole where the trailer is located.

Figure 2 Reflection created by a purely specular tanker trailer.

Since the shiny metal was not perfectly specular, some amount of light might have reflected back in the beam direction toward the driver. However, the reflectance problem was compounded by the tanker’s rounded side. Light hitting the upper curved part would also reflect upward while light hitting the lower part would reflect downward, so both miss the driver’s eye. At most, a driver might see a vague, thin line of light, where the curved surface was perpendicular to the ground. This is likely a very unusual sight that would be difficult to identify.

A white box truck is less specular, so more of the light scatters in directions back toward the driver. The amount depends on smoothness of the finish, and cleanliness of the surface – dirty trailers are more matte, which is good, but reflect back less light overall, which is bad. Lastly, dark colors reflect less light than white. It is irresponsible to paint a trailer any color other than white. Painting trailers with a large colored corporate logo is also a bad idea.

Finally, flatbed trailers are most difficult to see because they reflect no light back to the driver. This expectation has been confirmed in studies on retroreflective tape effectiveness. The ability of retroreflective tape to reduce accidents is greatest for flatbed trucks because the trailer itself presents no information to the oncoming driver⁴. If the flatbed is carrying a load, the cargo may reflect some light.

These are all general principles, but trailer visibility can be determined scientifically with more precision. The technique requires a re-creation with an exemplar tractor-trailer and light measurements using a “luminance photometer,” a specialized device that measures the amount of light reaching a viewer’s eye. Very briefly, the investigator reads the amount of light coming from the background and from the trailer. Next he calculates the contrast, the difference between the trailer and the background. Lastly, he compares the trailer contrast to data showing the amount of contrast that a viewer would require. If the required contrast exceeds the available contrast, then the trailer is not visible. Procedural details are described in Green et al., (2008).

Even if the trailer is visible, however, this does not mean that the viewer is likely to see it. Seeing depends on many factors, including expectation, identification and situational assessment. Drivers do not expect to see objects blocking the road; if seen, the distance and meaning are uncertain. For example, the top of a box trailer might easily be interpreted as the horizon in dim light. I discuss this more fully below.

Driver Variables

Age

Glare Susceptibility. Older drivers are less likely to avoid an underride collision. I have already mentioned their higher glare susceptibility. The International Commission on Illumination method for calculating glare shows that veiling doubles by about age 62 and trebles by about age 74. Older viewers are also slower to recover from glare and to readapt to dim light.

Reaction Time. There is a common myth that age does not affect driver reaction time. It is true that many studies find no slowing of driver perception-reaction time with age⁵. However, these studies are performed in very simple situations where there is little uncertainty or complexity. Many other studies⁶ show that as uncertainty and complexity grow, performance falls for all viewers but it falls faster for the elderly. An older driver confronted with the vague outline of an uncertain shape at an uncertain distance is likely to respond more slowly than a younger driver. The effect will be especially great when confronted with an “avoidance-avoidance” conflict. (See below.)

Other Visual Losses. A full discussion of visual losses with age would require many textbooks, so I’ll just add a few more. First, older viewers have poorer contrast sensitivity, especially in the low illumination conditions of nighttime driving. Second, they have poorer ability to notice objects in the visual periphery. Third, they are poorer at judging motion. All of these can affect their ability to sense, identify and avoid trailers blocking the roadway.

Reaction Time

Many who analyze underride accidents (mis)apply the “standard” 1.5 seconds reaction time. However, reaction times can rise dramatically when a trailer is blocking the entire roadway. First, the 1.5 seconds is not universally applicable. It is derived from studies performed in daylight conditions, not low visibility nighttime viewing. Further, drivers may face an “avoidance-avoidance conflict.”

A person deciding upon response alternatives weighs the possible outcomes of the various responses. Some outcomes are good” and produce a tendency to approach. Other outcomes are “bad” and produce avoidance. If choosing between a good and a bad (approach-avoidance) outcome, the decision is easy and reaction time is fast. If choosing between two good outcomes (approach-approach), then there is some conflict and reaction time is slower. The worst situation, however, is the avoidance-avoidance conflict when both alternatives are bad. The reaction time becomes very long because the viewer vacillates.

A driver who finds himself in imminent collision with a trailer blocking the entire roadway may be in an avoidance-avoidance conflict. The only choices are braking and steering. When very close, there is not enough braking time, so steering is the only alternative. However, the trailer was blocking the entire road. Steering leftward takes him across the road toward the tractor and its glaring headlamps. Steering rightward is possible, but the trailer extends off the roadway blocking the path. There is no escape. The problem is compounded by stress-produced, “perceptional narrowing,” which many authors incorrectly equate with tunnel vision. In fact it is much more. Easterbrook⁷ , who popularized the term, meant that under stress viewers reduce the amount of information, both in the world and in memory, used in decision making and further minimize the set of alternative responses considered.

In sum, reaction time may be very long for the obvious reason that the viewer really doesn’t want to choose an available response because none avoids the problem. All of his escape routes are blocked, so he has no favorable alternative to choose. It is no wonder that drivers in underride collisions frequently make no attempt at avoidance.

Driver Cognition

There is much more to seeing than mere visibility. In order to be consciously perceived, an object must engage attention and then be identified. Highly visible objects are more conspicuous, so the factors that reduce object visibility also reduce likelihood of the object drawing attention. However, even visible objects may not engage attention for many reasons. One factor that determines attention is expectation. Objects blocking the roadway are relatively rare effects, so drivers are less prone to notice them. Moreover, driving is not a series of discrete encounters with various road objects. Rather, it is a continuous task that smoothly proceeds through time. Drivers generally expect the future to unfold in the same way as the recent past. Imagine a driver who has been traveling on a relatively deserted dark road at night. He has encountered few if any conditions that require much attention or change in his steering or braking. He has little reason to expect a sudden road blockage. He would not be looking for it or expect it when it occurs.

Further, the tractor headlamps may mislead the driver. The driver must interpret the meaning of the oncoming tractor headlamps as he approaches. How far is it? Is it just the horizon? What does it mean? The visual ability to perceive “looming,” the movement of objects directly toward the eye is limited until the object is very close. In the absence of looming cues, the oncoming driver will be unlikely to interpret the headlamps as belonging to a vehicle that is moving toward him at a normal rate of speed rather than to a stopped vehicle that is backing across or turning out of a side road or driveway.

Arousal

Many accidents occur when drivers are in a low state of arousal, which lowers performance. The low arousal stems from two main sources. One is circadian rhythms, the normal 24 hour arousal cycle that all people experience. For people on a normal sleep-wake cycle, they will have a major arousal starting around midnight until 4 or 6 in the morning. Fatigue may also lower arousal during this period, but even a well-rested person should be expected to suffer declines in attention, object identification, situational awareness and perception-reaction time. The second source of lowered arousal is “vigilance decrement.” People who have been performing a monitoring task, such as drivers on the roadway, typically exhibit a loss in vigilance in as little as 30 minutes, especially if the task is monotonous. A person driving a dark road with little to see is a prime candidate for such a decrement.

Documentation By Nighttime Photographs

Scene investigators often make extensive photographic documentation of the collision. These should not be used in any attempt to portray visibility conditions at the time of the accident. They are not only useless, but they are highly misleading. The trailer and especially the retroreflective tape will generally appear far more visible and conspicuous in the photograph than they would have been to the driver. As a result, they should not be shown to jurors.

The reasons that nighttime photographs are misleading lie in both the photographs themselves and in the responses that they elicit from viewers. I have discussed this issue in detail elsewhere⁸, so the following is just a brief overview.

First, some or all of the pictures may be taken with added light sources, such as a flash or the headlamps or flashers of other vehicles which arrived after the collision.

Second, cameras and photographs create distorted images, especially of night scenes. Camera light meters set exposure on the assumption scenes are evenly illuminated and that average scene reflectance is 18%. The camera sees a mostly dark field and sets exposure to accommodate this low light level. Bright areas, such as side marker lamps and reflective tape, are then drastically overexposed and have unrealistically high contrast. As a result, lights appear far more visible than they would have been to an actual viewer.

Lastly, perception is a function of the viewer as much as the image. I have already described how the expert had many advantages in seeing the trailer. The jurors have the same advantages and perhaps more. The jurors’ sensory systems are different than those of the driver. I have already explained how factors such as mis-adaptation and veiling glare would have lowered driver contrast sensitivity. Jurors also get to inspect the photographs at their leisure with no time constraints. Moreover, the jurors knew what was there, what was going to happen, and what the outcome would be.

Conclusion

The task faced by a driver approaching a trailer blocking a dark road is formidable. He must sense, and notice the trailer. He must correctly identify it and assess its distance. He may have to perform these tasks while facing the glare of tractor headlamps and be impaired by the effects of aging, bad weather or low arousal. He must then select a response by attempting to resolve an avoidance-avoidance conflict while under stress and perceptual narrowing.

It is relatively easy to point out the physical variables that cause these difficulties. Headlamps change the eye’s adaptation state and create scattered glare in the driver’s eye. Trailer angle reduces the effectiveness of the retroreflective tape, shrinks the size of the side maker lamps and reduces the amount of headlamp luminous intensity reflected back to the driver. Some viewer variables are also obvious. Aging increases glare susceptibility and lowers contrast sensitivity. Contrary to common lore, aging can also dramatically increase perception-reaction time in uncertain situations, such as identifying the trailer and resolving the avoidance-avoidance conflict.

Endnotes

¹ See Green, M. et al. 2008. Forensic Vision: With Applications To Highway Safety. Lawyers & Judges Publishing: Tucson.

² Headlamp aim is a key factor in creating glare. Headlamps misaimed upward or leftward cause more glare. However, it is usually impossible to go back and to determine the actual headlamp aim at the time of the collision.

³ 3M Product Bulletin 4000, 2005.

⁴ Morgan, C. (2001) NHTSA Report Number DOT HS 809 222.

⁵ “How Long Does it Take to Stop? Methodological Analysis of Driver Perception Brake Times,” Transportation Human Factors, Vol. 2, pp. 195-216, 2000.

⁶ E. g., Simon, J., and A. Pouraghabagher 1978. The Effect of Aging on the Stages of Processing in a Choice Reaction Time Task. Journal of Gerontology, 33. 553-561.

⁷ Easterbrook, J. A. 1959. The effect of emotion on cue utilization and the organization of behavior. Psychological Review 66, 183-201.

⁸ Supra note 1.

Why Perception-Response Time (PRT) Is Not Like Gravity

Why Perception-Response Time (PRT) Is Not Like Gravity

Marc Green

It is sometimes easier to explain a concept by saying what it is not, e.g., seeing is not a homunculus in the head looking at a screen. Similarly, it is useful to show that PRT is not by nature like gravity, especially since the previous chapter unintentionally reinforced the errant impression that PRT is exactly like gravity.

For present purposes, gravity has three important properties.

• First, it is a variable, but it can reasonably be treated as a constant. Its assumed value is always 32.2 ft/sec², even though, strictly speaking, its exact value depends on several factors. In most applications (on earth), the error is too small to matter. Since it is a virtual constant, it may be applied in a cookbook manner with no need to understand the origin, history, effects of various circumstances or underlying scientific basis.
• Second, a value for gravity is always necessary when performing braking and similar accident reconstruction calculations. It is not possible to simply say that the value is too uncertain to determine, to use qualitative terms such as more or less acceleration or to deem it irrelevant.
• A third difference is that gravity has no variability. It is a simple, single number. Humans have an innate urge to reduce cognitive complexity, i.e., find a simple answer to a complex problem. It would be nice if collisions could be analyzed with a single number. There is a certain amount of time to avoid collision and either the driver’s PRT is longer or shorter than this time. If it is shorter, he avoids collision. If it is longer, he does not. What could be simpler?

It is perhaps natural that accident reconstructionists and engineers, people trained in the physical sciences, should treat PRT like a cookbook physical constant. They need gravity to compute braking as well as to perform many other physical analyses. They also want a simple number that they can use without any deep understanding. Psychology and behavior is not their business and they don’t want to be diverted into the extensive time and effort required to obtain a real understanding of the phenomenon.

To them, it must seem a small step to cross over into human factors and to analyze avoidability where they will need a PRT value just as they needed a gravity value to perform the purely physical analysis.

It is not surprising then that they view PRT as a human factors analog to gravity – a fixed number where variability, context and scientific basis can be ignored. If AASHTO says that PRT is 2.5 seconds, then there is no need to bother reading the Green Book in order to learn the method that AASHTO used to determine the value, the assumptions they made or what they intended it to represent.

There is no need to read the original studies upon which AASHTO relied in reaching their conclusions. In fact, learning about such issues is critical to anyone attempting to apply this value to the real-world. I explain this in depth later when I discuss how the AASHTO derived their values and the problems with their choice of base data.

Unfortunately, PRT is not like gravity at all, and crossing from physics to human factors, i.e., experimental psychology, is a huge step. Psychology is far more complex¹ than physics, and PRT is a far more complicated concept than gravity.

It is not reasonable to treat PRT as a constant, or even a small set of constants, that apply universally across all situations. Mean/median PRT varies wildly due to circumstances and PRT distributions often have very large variability and large skew. The number of factors affecting PRT is far greater and more difficult to quantify than the factors that determine gravity. In sum, to assign a value for gravity, it is not necessary to be a physicist or to have read the underlying physics research. To assign a PRT value to a real situation, or even know whether it is feasible to assign a number, it is absolutely necessary to know a great amount of the underlying research in both PRT and in behavioral psychology in general.

Sometimes, the key question is not the speed of PRT but whether PRT is even a relevant issue. Such an assertion may seem unintuitive since a driver response is required for avoidance. This is again thinking in terms of the gravity analogy. In reality, there are several reasons for ignoring PRT in many cases.

1. The relationship between PRT and accident causation is weak.

One study (Muttart, 2005) found no relationship between PRT and accident involvement while another (Mihal & Barret, 1976) found that individual differences in PRT had no relationship to accident rate. A third (Ayres, & Kubose, 2012) concluded that for a given TTC, PRT often failed to distinguish between those who crashed and those who did not. Further, Malaterre, Ferrandez, Fleury & Lechner (1988) concluded that under extreme emergency, instinctive reflexes take over so that all drivers “become equal”. In fact, most accidents involve drivers who have good driving records and who have not had a previous major crash (Campbell, 1959). The most likely explanation for such results is that situational factors, speed, time, distance, light and contrast, and not PRT, primarily determine whether an accident will occur. Whether a driver can avoid a collision is often due to plain old luck, i.e., it “is “rather like taking a bet” (Prynne & Martin, 1995) and “therefore a matter of chance combination of circumstances” (Baker, 1960). In sum, there is little evidence that accidents occur because there are “bad apples” who respond abnormally slowly.

Many will find this conclusion difficult to accept because it runs contrary to a strong human cognitive bias, “fundamental attribution error” (Ross, 1997). When a person judges the cause of an event, he can assign it to either dispositional factors inside the person or to situational factors outside the person’s control. People are heavily biased toward blaming individual disposition even though most people act the same way in the same situation. Fundamental attribution error is very powerful and highly resistant even to strong evidence that environmental constraints were the primary cause. It is also one of the prime promoters of hindsight bias.

2. Long PRT is often a side effect, not the cause, of avoidance failure.

That is, the cause is the factors that resulted in a PRT that was too long, and not the PRT itself. Think about it this way. Assume the fastest possible driver PRT is 1.5 seconds. The driver responds in 2.5 seconds on a dark road at night to a pedestrian. If there were research studies that would contain PRT data for a pedestrian in his specific clothing, on his specific location on the roadway, with the specific street lighting etc. then that might be used directly. Unfortunately there are no such data and likely never will be. At this point, there are only a few ways to proceed. One is to make up a number, which in my experience is all too common. (“Well, the standard PRT is 1.5 seconds in daylight, so I added another second for nighttime”). Another is to use an illuminance criterion, such as the 3.2 lux twilight value. This is another cognitive complexity reducer since it is simple and requires no understanding. The chapter on contrast detection has already explained why this method is unreliable.

So what is the alternative? Perhaps the best strategy is to shift away from PRT to the factors that determine PRT. After all, response is just the final event in a chain of human information processing, i.e., sensing, identification, situational awareness and response selection. It is an effect of these factors and only a manifestation of them.

To see this, consider how PRT might be determined. Under absolutely ideal conditions, humans can reliably respond to an unexpected road event in roughly 1.0-1.5 second, depending on whom you ask. This sets the absolute limit on what a driver can achieve. If the available time is less than this, then avoidance is simply impossible. If it is longer, then inevitable uncertainties arise. The real world is never ideal so the question is always, how much more is the PRT going to be. There are many factors that may inflate the PRT: visibility, conspicuity, expectation violation, complexity, novelty, weather, etc.

The amount by which each of these factors individually, let alone in groups, raise PRT is generally difficult to say with much certainty. The important point is that the issue is not the driver’s PRT but rather the conditions that caused the increase to whatever PRT that driver actually produced.

The real performance determining factors should be examined, and not the PRT that they produce. If lighting and contrast are low, then PRT will be long. If the driver fails to see the object in time to avoid because of lighting conditions, the issue is visibility, not PRT (at least directly). If the driver fails to see the visible object in time because other road objects attracted his attention, or because he is texting on his smart phone, the focus should be the events that controlled attention, not PRT. Certainly, a specific number like 2.0 seconds results in a nice fuzzy feeling and the pretence of scientific precision, but in most cases this is merely illusory confidence.

In sum, the process should work backward from the way people normally think of it. The question accident investigators typically ask is “What is the expected PRT”? The better question is “What factors caused the driver’s PRT, what ever it was, to be insufficient?” The task is to explain what actually happened and not to hypothesize some specific number in the absence of real scientific evidence. The fundamental cause of most collisions is likely “late detection” (Rumar, 1990). Anyone who has investigated collisions knows that drivers commonly say that they never saw the pedestrian, etc. or saw him just a fraction of a second before the collision. What is the point in considering PRT in such cases? The real issue is the reason that detection was late.

3. Variability is critical

Variability is irrelevant to applying a value for gravity but absolutely central to interpreting PRT. Gravity is always 32.2 ft/sec², end of story. In PRT, variability is as critical as the measure of central tendency (mean/median), since it defines the range of normal behaviors. Research studies are little help for determining variability, beyond setting a lower limit. As I have explained, research studies are designed to minimize variability, so they typically underestimate the uncertainty of real-world behavior. This is most important in situations that require thinking, i.e., ones that are not highly reflexive.

Moreover, there is still the additional problem in defining normality even if variability could be convincingly determined. Suppose the mean PRT is 1.5 seconds and a driver responds in 2 seconds. Is this within the realm of normal driver behavior? The question is unanswerable without knowing variability. If the standard deviation is .25 second, then he is two standard deviations above the mean and is in the 5% of slowest drivers. If the standard deviation is .5 second, then he is only one standard deviation above the mean and is in the slowest 16% of drivers. Where are the limits of “normality?” The one or the two standard deviations above the mean driver? In contrast sensitivity, the Adrian contrast model sets the value at the 99.63% detection level (and then adds multipliers). That is more than three standard deviations above the mean and is a far looser definition of normality. Which definition is correct? Of course, PRT distribution exhibit strong kurtosis toward longer values, so standard deviations may greatly underestimate the number of the slow times.

There is one final problem in defining variability as well as mean/median. That is the problem of defining PRT itself. In counterfactual thinking, it is always easy to say that the driver could have avoided the collision if he had responded faster.

This may be trivially true, but then the important questions are

• 1) what is meant by “responded”? and more importantly
• 2) responded faster to what?

Let’s Get Real About Perception-Reaction Time (PRT)

Marc Green

Imagine a trial about a botched surgical procedure. A surgery “expert” takes the stand to give his opinion. Upon examination, he says that no, he has never done any surgery himself. Nor has he ever studied the underlying scientific disciplines of anatomy or physiology. He says that he qualified, however, because he has read a book chapter on surgery by the noted physician Paulski Olsonovich and that he once took a 2-day chiropractor course that included a 2 hour discussion of surgery.

Would this surgery “expert” be allowed to testify? Not likely. But substitute the phrases “perception-reaction time” for “surgery,” “vision” for “anatomy” and “cognition” for “physiology” and apparently, voila, the “expert” is qualified.

This probably explains why there is no area of “expert” opinion on road accidents that has more misinformation, more inappropriate use of canned numbers, more misunderstanding and, to use scientific terminology, more good old fashioned BS. Like the surgery expert above, most PRT “experts” have never actually done the task that they feel free to opine about.

They have likely never measured a reaction time nor done any other behavioral scientific research and do not understand the complexities of scientific research and how much the methodological details determine what a scientific research study can actually tell you. In short, science, like surgery, is something you do and not just something you know.

Moreover, most “experts” have never read the original source data and don’t have the background to evaluate and interpret the studies if they did. Like the surgery expert, they have no training or experience in foundational scientific areas, human learning, memory, perception, decision-making, etc., to put the results into a broader behavioral context. They rely on secondary sources that omit many of the critical methodological details necessary to interpret the data.

Accident reconstruction is about physics – speeds, time, distances, etc. Accident reconstructionists, however, sometimes feel compelled to go beyond physics and to give an opinion on causality and accident avoidability.

Here is where the trouble starts. The accident reconstructionist cannot give an avoidability opinion without providing a PRT value. This clearly goes beyond physics into the realm of human behavior, i.e., the field of psychology. With no scientific experience in psychology, however, the “expert” simply parrots a value that he has heard in a course, read in a secondary source book, pulled blindly from a computer program, etc. although he has no real understanding of where it comes from or what it means.

The article below, published in Collision 2009 and elaborated in (Green, 2017), demonstrates why such an approach is inadequate. It also shows why it is necessary to actually read the original source research and why a background in basic perception, cognition, etc. is necessary to understand what the research is really saying. Lastly, it shows why a background in having actually performed behavioral research is essential to opine on topics such as perception-reaction time.

Perception-Reaction Time: Is Olson & Sivak All You Need To Know?
Collision, (2009), 4, 88-95.

Accident reconstruction often requires a driver “perception-reaction time” (PRT), the interval between obstacle appearance and driver response initiation, i.e., the foot just touches the brake pedal and/or the hands just start turning the wheel. The PRT number is often a critical factor in establishing causation and subsequently in assigning blame.

There are two popular opinions and rationales for PRT.

• The first opinion is that the PRT is 1.5 seconds. The usual basis for 1.5 seconds is that the reconstructionist read it in an accident reconstruction book, learned it in a class or simply believes that it is the “accepted value.”
• The second opinion is that the PRT is 1.1 (or 1.6 for the 95th percentile driver) based on “Olson.” This opinion almost invariably means that the accident reconstructionist has read one of Olson’s secondary sources, such as a chapter in Forensic Aspects of Driver Perception And Response (2003), or has simply seen it cited somewhere. In fact, the data stem from Olson & Sivak (1986), but Sivak remains anonymous because few have read the original research. For simplicity, I will simply refer to the data as “Olson,” which is how they are usually referenced.

In either case, the rationale is inadequate. PRT is a very complex, situationally-dependent phenomenon that cannot be captured in the canned numbers that are so typically employed. Few who reconstruct accidents know much about the underlying science, where the numbers they quote originate, how they were obtained or what they really mean.

This article addresses the misuse of canned numbers (including the AASHTO 2.5 seconds and computer programs) in general but focuses primarily on the “Olson values.” There are three main problems.

• First, every research study has limited generality because it is conduced under a specific set of conditions. An accident reconstructionist wishing to apply Olson, or any other research study, to a specific accident should understand the differences between the driving situation in the study and the specific accident. These differences will be smaller in some cases and larger in others, but there will be always differences. In some cases, Olson does not apply at all. In fact, the concept of PRT itself may not even apply. The discussions below of visibility and the tollbooth problem are examples.
• Second, Olson is a perfectly fine study, but it is only one of perhaps 100+ driver PRT studies. These other studies provide data for other sets of conditions. Knowledge of these studies allows the reconstuctionist to better interpolate and extrapolate PRT for a broader set of conditions. However, there are no existing data for many common accident scenarios, so it is often impossible to determine a PRT with much precision. For example, there are almost no data for PRT at night, when accidents are frequent.
• Third, assigning a reasonable PRT requires knowledge that goes even beyond the PRT literature. The issue of driver PRT cannot be snipped off from the larger topics such as perception, memory and learning and examined independently. This is especially true in accident scenarios where no PRT data exist. The assessment of older driver PRT, as discussed later, is a good example.

What Does Olson Actually Say?

In order to properly use Olson (or any other study) as a basis for estimating real world PRT’s, the first step is to carefully analyze the experimental procedure. The second step is to determine the differences between research conditions and the accident conditions. The last step is to compensate for the differences. This is the most difficult problem.

A close reading of the Olson & Sivak study reveals the important methodological details. Olson tested two groups, a younger group of 49 drivers with an age range of 18-40 and an older group of 15 drivers with an age range of 50-84. The drivers were told only that they would be a study of driver behavior. They drove the test vehicle during daylight at about 27-31 mph with the experimenter sitting in the rear seat. The route took them over a rural road chosen so that there would be no distractions or possible hazards. After 10-15 minutes, the vehicle came to a hill. The experimenters had placed an obstacle, a 6″ x 36″ block of foam, in the left side of the lane directly in front of the driver. As the driver ascended the hill, the obstacle came into view. The sight distance to the obstacle was about 150 ft (46 meters), which translated to about 3.3-3.8 seconds time-to-collision (TTC). Instruments measured the time/location at which the driver released the accelerator and pressed the brake.

In order to determine the PRT, the driver had to re-travel the route and tell the experimenter where he had first seen the obstacle. Olson then calculated the putative PRT time by measuring the distance from the location where the driver claimed to have first seen the obstacle to the location where he released the accelerator. PRT is this distance divided by speed.

Their results show a median PRT of about 1.1 second to press the brakes, with no difference between younger and older drivers. The 5th percentile drivers responded in .8 second while the 95th percentile driver responded in about 1.6 seconds. Olson has published these results in several later book chapters but without the methodological details.

First, the research was not exactly a study of PRT to an unexpected obstacle. The PRT determination required the driver to return to the scene and to say where he first saw the obstacle. At this point, the obstacle was expected and not a surprise.

This is a very unusual way to determine PRT. Usually, the PRT clock starts counting at the moment the signal is presented. It is unclear how accurately drivers could say where they were when they first saw the obstacle, so there are questions about what this study actually measured. However, one thing is certain: if the PRT clock had started counting at the moment when the driver first had a clear sightline to the obstacle, then the PRT would have been longer.

Reading the actual study reveals that the methodology was biased to produce short PRT’s. There are many other procedural factors that further promoted very short PRT and that limit the study’s generality for assigning PRT to real accidents.

1. Drivers were alerted. The term “alerted” unfortunately has two senses, which often creates confusion. Some authors use the term “alerted” to mean that the driver knew that there was an obstacle or even a particular obstacle ahead. In this sense, “alerted” means “expecting.”

The other sense of “alert” refers to general arousal level. Drivers in the Olson study may not have been expecting a particular obstacle, but they certainly were alert and had a very high arousal level: they were participating in an experiment where their behavior was being monitored. There was even someone sitting in the back seat watching them. Moreover, they had been driving only 10-15 minutes before encountering the obstacle. Arousal level is related to driving time.

The well-known phenomenon of vigilance decrement (Mackworth, 1948), a rapid decline in detection, typically starts within a half hour after task initiation. Further, research (Philip, Taillard, Klein, Sagaspe, Davies, Guilleminault, & Bioulac, 2003) has shown that time spent driving is a better predictor of decrease in driver performance than hours without sleep. The short driving time in the Olson study gives the test drivers a significant arousal advantage over a real driver who may have been on the road for an extended period.

In sum, the Olson drivers’ high arousal level likely produced shorter PRT’s than would occur under many normal driving scenarios. Olson was fully aware of this likelihood when he noted that “The subjects in this study were possibly alert relative to the general population of drivers” and that “the results are probably conservative (i.e., lower) to what would be found in the real world.”

2 The testing occurred during the day. Olson does not specify the times of his testing, but it is likely that much of it was performed when drivers are at a moderate or high point on their “circadian rhythms,” the normal 24-hour cycle of arousal that all people experience.

For most people, the arousal cycle has lows in the late afternoon and especially in the early morning hours. During these periods, many performance measures, such as accident rates and PRT are at their worst. One study (Wylie, Shultz, Miller, Mitler, & Mackie, 1996) of long haul truck drivers, for example, found that accidents correlated with time-of-day, early morning hours, but not with time-without-sleep. As a rule of thumb, in fact, it takes about 24 hours before people exhibit major sleep-deprivation losses.

Olson’s drivers then likely had this additional arousal advantage over normal drivers in the early morning hours who are at a low point on the circadian rhythm. However, drivers who habitually work nights may have their rhythm “phase shifted,” so the peaks and lows are at different times than normal drivers.

3. The obstacle appeared at the point of fixation. Olson placed the obstacle on the roadway at the crest of a hill and directly in front of the driver. It likely the exact location where the driver was fixating at the moment he reached the 46 meter sight distance. Location of an obstacle in the visual field can affect PRT. The optimal location is along the sightline at the point of fixation. Objects located here cast their images on the fovea, the retinal area of sharpest vision and the focus of attention. Olson placed the obstacle in the ideal visual field location.

In contrast, many collision scenarios involve a lane incursion where a vehicle or pedestrian approaches from the side. The obstacle then first appears in peripheral vision, where visual sensitivity is lower and attention is weaker. Moreover, when a viewer detects an object in peripheral vision, he most likely makes a saccadic eye movement toward it.

The saccade requires time to move the eye plus a “dwell time” for the viewer to perceive the scene. The total saccade time about is 1/3 second in good day visibility. At night, the time is likely to be longer. The first saccade may miss the object, so viewers may have to make more than one saccade to “home in” on the target. When the new fixation requires a significant change in distance, such as shifting gaze from a mirror to an obstacle a few hundred feet down the road, the eye’s change of accommodation and vergence and reacquisition can drive the time up to as long as a second (Travis, 1948.)

Lastly, if the target is more than 15^o from the sightline, the driver will likely also have to make a head turn. Imagine a driver approaching an intersection or railroad track. He must turn his head to look one direction and then the other. It takes the driver 85th percentile driver .7 seconds to turn his head one way and then another 1 second to turn back the other (Long & Nitsch, 2008). This 1.7 seconds search time is on top of the PRT.

Visual field effects likely explain why Olson & Sivak found a 1.1 median PRT second while studies (Green, 2008a) using lane incursions typically find slower mean PRT’s of about 1.5 seconds. (About .1 second of this difference is likely due to the difference between using median and mean as measures of central tendency.) Olson and Sivak’s 95th percentile level was 1.8 seconds while the 95th percentile lane incursion PRT would be about 2.4 seconds, which is also near value used by AASHTO in geometric road design.

In sum, the Olson study optimized the PRT by placing the obstacle at the fixation point directly ahead of the driver. PRT will be longer when objects approach from the side as well as for other reasons discussed in subsequent sections.

4. The visibility conditions were good. Olson tested drivers in daylight and good visibility, so obstacle visibility was not a limiting factor in driver behavior. PRT is likely to increase at night and under other low visibility conditions.

In fact, when visibility is sufficiently low, the concept of PRT becomes irrelevant. After all, if the driver can’t see the obstacle, then he can’t respond to it. For example, assume that PRT for a pedestrian cutting left-to-right across the driver’s path in good visibility conditions 1.5 seconds. In this case, driver first sees the target in peripheral vision. At night, the same pedestrian wearing dark clothing emerges from outside the driver’s headlamp beams. When the pedestrian is far to the left, he receives little headlamp illumination and is invisible.

As pedestrian and vehicle approach, more headlamp illumination falls on the pedestrian. At some point, driver sees the pedestrian. In theory, the 1.5 seconds reaction time clock starts when the pedestrian first becomes visible in the periphery. But when is that? [Note: Olson didn’t start timing PRT until the point at which the driver actually saw the obstacle.] In order to state a well-defined PRT, it would be necessary to know the exact point at which the pedestrian became visible. Even if this could be calculated, then it would still be necessary to specify the point where the pedestrian became conspicuous enough to draw attention and eye movement. This point is likely unknowable with great precision.

It is impossible to precisely estimate of the amount of slowing that will occur at night. However, some qualitative statements are possible. For the same pedestrian walking the same path, driver will have less time to avoid the collision at night because the pedestrian will likely have to be much closer in order to achieve the required visibility. The difference between day and night PRT will depend on factor such as street lighting, pedestrian clothing, background clutter, etc. A pedestrian wearing white clothing, for example, will often have better visibility and more approximate daylight visibility conditions than a pedestrian wearing dark clothing. However, there are exceptions (Green, 2008b).

Lastly, low visibility conditions also slow cognitive processing by creating uncertainty and by impairing recognition. I explain this further in the next section.

5. The obstacle appeared suddenly and unambiguously. Olson’s drivers responded reflexively and did not have to think much because the situation was very clear. There was minimal cognitive processing, little uncertainty and no complexity, so PRT was very short. Moreover, the variability is very small because people are relatively uniform in their speed of making reflexive responses.

Situations that are more ambiguous or which develop more gradually require conscious thinking that slows response and drastically increases variability. For example, a driver traveling at night who approaches red and white dots (e.g., the rear reflective tape on a truck) at some ill-defined distance must gain “situational awareness.” He must identify the lights, determine the distance, search memory for previous similar experiences, decide what is going to happen if he responds and if he doesn’t respond, choose a response, chose how hard to make the response, etc. (Green, 2008). Moreover he must consider his ability to control vehicle speed and direction.

The “tollbooth problem” (Fajen, & Devaney, 2006) provides a good example. Imagine a driver on a high-speed limited-access road traveling 65 mph. Suddenly, he sees a tollbooth up ahead about a mile away and realizes that he will have to stop. Does he start braking immediately? The answer, of course, is no. Immediate braking wastes time arriving at the tollbooth. Rather, the driver has an internal model of his vehicle’s braking capabilities and has learned the mount of time/distance needed to stop at a comfortable deceleration (or even at an uncomfortable deceleration.) Eventually he reaches the critical distance and begins to brake.

Theoretically, PRT would be the time between first sighting of the tollbooth and the pressure on the brake pedal. However, this is not a “reaction” in any conventional sense, so the concept of PRT doesn’t really apply. The driver does not brake because there is no need to act. While the tollbooth problem might seem trivial, drivers face similar problems frequently. Up ahead, they see brake lights or unidentifiable objects, some dim dots of red and light. Should the driver brake immediately or wait until he is sure of the situation?

The point of the tollbooth example is that there is much more to PRT than perception. Drivers have a mental model of their ability to control their vehicle. The decision to act is always based partly on this mental model. The model’s constituents are the “safe field of travel” and “stopping distance” (Gibson and Crooks, 1938). As a driver travels down the road, he is surrounded by obstacles, cars ahead, curbs and other barriers, pedestrians crossing the road, etc. which define a safe field of travel. This field changes constantly as new vehicles, pedestrians, etc. appear and change position.

The driver also has a mental stopping distance and steering model of his ability to brake/swerve his vehicle. This area is like a cocoon that surrounds the driver, providing a buffer zone with obstacles. Drivers believe that they can avoid collision with obstacles outside the cocoon. Ideally, the driver steers his vehicle through the cocoon’s center, adjust speed and direction as the safe field of travel dynamically changes.

For this scheme to work, the driver must accurately assess object distance, speed and stopping distance (or time). However, distance perception is highly fallible, especially for small points of light, unfamiliar objects, foggy atmosphere and some other situations. Drivers are also poor at judging their own speed (Denton, 1980) and there are many situational factors that can cause them to underestimate how fast they are going, I e., fog and, low edge rates (Denton, 1980.) Drivers may also err in their belief of stopping ability when driving an unfamiliar vehicle or on wet or icy roads, sharp downgrades, dark conditions, etc.

Moreover, most drivers have likely had little or no experience making sudden stops, especially at high speeds. They base their cocoon size on their experiences stopping at lower speeds. Since stopping distance increases with the square of speed rather than linearly with speed, they are likely to underestimate the needed distance.

Even if the driver decides to respond, the choice of response is sometimes unclear. A driver heading toward a tractor-trailer blocking the road may find that there is no time to brake and that steering to the left will take him into oncoming traffic while steering to the right will put him in a ditch. This is termed an “avoidance-avoidance” conflict where the driver must choose among a set of bad alternatives.

In such cases, PRT typically is very, very long. Often, the driver can’t decide and fails to respond at all before collision. The common example is the underride accident where there is an unfortunate tendency to assume the driver’s failure to respond because he had fallen asleep. In fact, the driver may have been caught in an avoidance-avoidance crisis.

6. The drivers were traveling slowly. Olson’s drivers traveled at speeds ranging between 27-31 mph. At such slow speeds, sudden, abrupt braking or steering is less likely to cause an unrecoverable loss of control and to have dangerous consequences. In contrast, drivers traveling at 65 mph on a freeway may to hesitate to make sharp swerves or go to full-out braking because of potential control loss. They have to weigh the hazard of a collision with the hazard created by a loss of control that sends the vehicle over a median or guardrail, into other traffic or that initiates a side-skid and rollover. It is a type of avoidance-avoidance conflict that will likely lengthen PRT.

The fear of losing control is likely why drivers frequently resort to two-stage braking (Prynne & Martin, 1995). They initially push the brake pedal down part way and then monitor the situation, hoping that they can avoid the collision without and extreme response that risks loss of control. If collision is still likely, then the driver might go to the extreme maneuver.

7. The “older” drivers were not all old. Olson somewhat surprisingly failed to find any slowing in their “older drivers.” This has caused many to claim that aging has no effect on PRT. However Olson’s “old” group included drivers as young as age 50. While visual abilities start their decline in the early 40’s, the significant effects do not begin until viewers enter the 60’s. Olson does not give the ages of the individual drivers, so it is impossible to know the number who were in their 50’s and early 60’s where aging effects are small. However, it is very possible that Olson found no aging effect, in part, because their “older” drivers were too young.

Olson’s task further likely minimized aging effects. As discussed elsewhere (Odom & Green, 2008), studies in the basic research literature have repeatedly found that impairments of aging (and other conditions such as distraction and alcohol use) are more pronounced when perceptual and cognitive abilities are taxed under conditions such as low visibility, uncertainty and complexity. The simple, virtually automatic avoidance task in the Olson study required little cognition. It was performed in good visibility, so perceptual abilities were not a limiting factor.

Moreover, research with older subjects always raises the issue of representativeness. Olson does not state how he recruited the subject drivers. However, most researchers would routinely screen their subjects, especially older ones, for any visual or other health problems. The older subject drivers are then likely to be healthier, more active, in better visual and cognitive condition than the population as a whole. Moreover, the drivers very likely agreed voluntarily to be in the study.

Only the relatively healthy and “spry” senior is likely to volunteer for a research study. In sum, research on screened, self-selected older drivers likely overestimates abilities of the older population as a whole. In any event, the “older” group consisted of only 15 drivers.

This discussion of older driver PRT highlights the point that PRT assignment often requires knowledge of the general psychological literature and of scientific methodology.

• First, it is necessary to actually read Olson’s study in order to learn that he placed drivers as young as age 50 in the old category. This is not a detail that is ever mentioned in secondary sources.
• Second, the effects of complexity and uncertainty on the size of age-related deficits do not appear anywhere in the driver PRT literature or any computerized PRT program. It is basic science published in basic science sources.
• Third, the important issue of representativeness would not be apparent to anyone who was not intimately familiar with the way scientific research is conducted.

Conclusion

Accident reconstructionists should take the Olson results for what they are – the fastest that a driver can avoid an “unexpected” obstacle in highly optimized conditions.

Any deviation, such as low visibility, peripheral visual field location, complexity or uncertainty is almost certain to increase PRT. The finding that there is no loss of PRT with age is not generalizable and depends on specific conditions. Lastly, real drivers are unlikely to be as alert as the drivers in the study. Lower arousal level may produce longer PRT’s, especially at low points in the circadian rhythm and after driving for extended periods.

Each of the 7 factors described above would doubtless add time to Olson’s optimized 1.1/1.8 seconds PRT but assigning a precise number is difficult. I have sometimes seen opinions where someone arbitrarily adds 0.5 or 1 to compensate for nighttime or complex conditions. While essentially guesswork, these estimates are doubtless closer to reality than the simple, foveal, daytime, high-arousal values taken at face value. However, there are few if any PRT data for many of these conditions. This is why it is so important to have general knowledge about perception, attention and memory to fall back upon. They are often the only available guides.

Despite the lack of data for many situations, however, I can draw two practical conclusions about assigning a driver PRT.

• First, estimates will often cover a very broad range because precision is impossible.
• Second, estimates will often be very high – much higher than are normally seen. Low visibility and violated expectation make the obstacle disappear and even a moment’s hesitation to search or to think or to decide upon response can eat up seconds. The AASHO Redbook’s view of PRT written in 1973 remains true today:

“Whenever the driver is confronted with a complex traffic or highway situation and is required to make choices, judgments, and decisions, his response time may increase to 2, 3 or even 5 seconds” (p. 278).

References

American Association of State Highway Officials (1973). A Policy on Design of Urban Highways and Arterial Streets. Washington, DC: AASHO.

Denton G. (1980). The influence of visual pattern on perceived speed. Perception, 9, 393-402.

Fajen, B. R. & Devaney, M. C. (2006). Learning to control collisions: The role of perceptual attunement and action boundaries. Journal of Experimental Psychology: Human Perception and Performance, 32(3), 300-313.

Gibson, J.J. & Crooks, L.E. (1938). A theoretical field-analysis of automobile driving. American Journal of Psychology, 51, 453-471.

Green, M. (2008a) “How long does it take to stop?” Methodological Analysis of Driver Perception-Brake Times. In M. Green, B. Abrams, M. Allen, Forensic Vision With Application To Highway Safety, 379-406. Tucson: Lawyers & Judges Publishing.

Green, M. (2008b) “Pedestrian accident analysis” Methodological Analysis of Driver Perception-Brake Times. In M. Green, B. Abrams, M. Allen, Forensic Vision With Application To Highway Safety, 329-348. Tucson: Lawyers & Judges Publishing.

Long, G. & Nitsch, A. (2008). Effect of dead turning on driver perception-reaction time at passive railroad crossings. Transportation Research board 2007 Meeting CD.

Mackworth NH (1948). The breakdown of vigilance during prolonged visual search. Quarterly Journal of Experimental Psychology, 1, 5-61.

Odom, J. & Green, M. (2008). Aging Vision. In M. Green, B. Abrams, M. Allen, Forensic Vision With Application To Highway Safety. Tucson: Lawyers & Judges Publishing.

Olson, P. & Farber, E. (2003). Forensic Aspects of Driver Perception and Response. Tucson: Lawyers & Judges Publishing.

Olson, P.L. & Sivak, M. (1986) Perception-response time to unexpected roadway hazards. Human Factors, 28, 96-99.

Philip P., Taillard J., Klein E., Sagaspe P., Charles A., Davies W., Guilleminault C., & Bioulac, B. (2003). Effect of fatigue on performance measured by a driving simulator in automobile drivers. Journal of Psychosomatic Research, 55, 197-200.

Prynne, K. and P. Martin 1995. Braking behavior in emergencies. SAE Technical Paper 950969.

Travis, R. (1948). Measurement of accommodation and convergence time as part of a complex visual adjustment. Journal of Experimental Psychology, 38, 395-403.

Wylie, C. D. Shultz, T. Miller, J. C. Mitler, M. M. and Mackie, R. R. (1996). Commercial motor vehicle driver fatigue and alertness study: Project report. Technical Report FHWAMC- 97-002, Federal Highway Administration, Washington DC.