Task difficulty, risk, effort and comfort in a simulated driving task—Implications for Risk Allostasis Theory

Risk Allostasis Theory states that drivers seek to maintain a feeling of risk within a preferred range [Fuller, R., 2008. What drives the driver? Surface tensions and hidden consensus. In: Keynote at the 4th International Conference on Traffic and Transport Psychology, Washington, DC, August 31–September 4, 2008]. Risk Allostasis Theory is the latest version of Task-Difficulty Homeostasis theory, and is in part based on the findings of experiments where participants were asked to rate the task difficulty, feeling of risk and chance of collision of scenes shown in digitally altered video clips [Fuller, R., McHugh, C., Pender, S., 2008b. Task difficulty and risk in the determination of driver behaviour. Revue européenne de psychologie appliqée 58, 13–21].The focus of the current research was to expand upon the previous video based experiments using a driving simulator. This allowed participants to be in control of the vehicle rather than acting as passive observers, as well as providing additional speed cues. The results support previous findings that ratings of task difficulty and feeling of risk are related, and that they are also highly related to ratings of effort and moderately related to ratings of comfort and habit. However, the linearly increasing trend for task difficulty and feeling of risk described by the previous research was not observed: instead the findings of this experiment support a threshold effect where ratings of risk (feeling of and chance of loss of control/collision), difficulty, effort, and comfort go through a period of stability and only start to increase once a certain threshold has been crossed. It is within the period of stability where subjective experience of risk and difficulty is low, or absent, that drivers generally prefer to operate.     

Safety impacts due to the incompatibility of SUVs, minivans, and pickup trucks in two-vehicle collisions

This research sets out to estimate the effects of vehicle incompatibility on the risk of death or major injury to drivers involved in two-vehicle collisions.Based on data for 2,999,395 drivers, logistic regression was used to model the risk of driver death or major injury (defined has being hospitalized). Our analyses show that pickup trucks, minivans and sport utility vehicles (SUVs) are more aggressive than cars for the driver of the other vehicle and more protective for their own drivers. The effect of the pickups is more pronounced in terms of aggressivity. The point estimates are comparable to those in the Toy and Hammitt study [Toy, E.L., Hammitt, J.K., 2003. Safety impacts of SUVs, minivans, and pickup trucks in two-vehicle crashes. Risk Analysis 23, 641–650], but, in contrast to that study, we are now able to establish that a greater number of these effects are statistically significant with a larger sample size.Like vehicle mass and type, other characteristics of drivers and the circumstances of the collision influence the driver’s condition after impact. Male drivers, older drivers, drivers who are not wearing safety belts, collisions occurring in a higher speed zone and head-on collisions significantly increase the risk of death. Except for the driver’s sex, all of these categories are also associated with an increased risk of death or of being hospitalized after being involved in a two-vehicle collision. For this risk, a significant increase is associated with female drivers.     

Driving performance during concurrent cell-phone use: are drivers aware of their performance decrements?

Prior research has documented the manner in which a variety of driving performance measures are impacted by concurrent cell-phone use as well as the influence of age and gender of the driver. This current study examined the extent to which different driver groups are aware of their associated performance decrements. Subjects' confidence in dealing with distractors while driving and their ratings of task performance and demand were compared with their actual driving performance in the presence of a cell-phone task. While high confidence ratings appeared to be predictive of better driving performance for male drivers (as confidence increased, the size of the distraction effects decreased), this relationship did not hold for females; in fact, for older females, as confidence increased, performance decreased. Additionally, when drivers were matched in terms of confidence level, brake responses of older females were slowed to a much greater extent (0.38 s) than were brake responses of any other group (0.10s for younger males and females and 0.07 s for older males). Finally, females also rated the driving task as less demanding than males, even though their performance was more greatly affected by distraction. These results suggest that many drivers may not be aware of their decreased performance while using cell-phones and that it may be particularly important to target educational campaigns on driver distraction towards female drivers for whom there tended to be a greater discrepancy between driver perceptions and actual performance.     

Cross-cultural differences in driver risk-perception

This study investigated differences in risk-perception among U.S., Spanish, West German, and Brazilian drivers. Subjects estimated the risk involved in slide-projected traffic scenes. The scenes, photographed in the United States and Spain, were rated for the amount of risk by using a seven-point scale. The subject groups in each country included younger, middle-aged, and older nonprofessional drivers, as well as middle-aged professional (bus, taxi, or truck) drivers. In the data analyses, the independent variables were subjects' country, age, professional driving experience, and sex, and 23 dichotomously coded characteristics of the traffic scenes. The following are the main findings: (1) Spanish drivers reported the highest risk, while U.S. drivers reported the lowest risk; (2) younger drivers tended to report lower risk than middle-aged and older drivers; (3) nineteen of the 23 analysed characteristics of traffic scenes contributed significantly to risk ratings, even after simultaneously controlling for the effects of all other scene characteristics; (4) ten scene characteristics had a differential effect on the risk ratings in the four tested countries; (5) two scene characteristics had a differential effect on the risk ratings in the four tested subject groups; (6) none of the variables affected differentially the risk ratings of professional vs. nonprofessional drivers and males vs. females. These findings provide information concerning the desirable country- and age-specific emphasis in driver education, driver licensing, and public information campaigns.     

Young drivers' optimism bias for accident risk and driving skill: Accountability and insight experience manipulations

This study aimed to determine whether two brief, low cost interventions would reduce young drivers’ optimism bias for their driving skills and accident risk perceptions. This tendency for such drivers to perceive themselves as more skilful and less prone to driving accidents than their peers may lead to less engagement in precautionary driving behaviours and a greater engagement in more dangerous driving behaviour. 243 young drivers (aged 17–25 years) were randomly allocated to one of three groups: accountability, insight or control. All participants provided both overall and specific situation ratings of their driving skills and accident risk relative to a typical young driver. Prior to completing the questionnaire, those in the accountability condition were first advised that their driving skills and accident risk would be later assessed via a driving simulator. Those in the insight condition first underwent a difficult computer-based hazard perception task designed to provide participants with insight into their potential limitations when responding to hazards in difficult and unpredictable driving situations. Participants in the control condition completed only the questionnaire. Results showed that the accountability manipulation was effective in reducing optimism bias in terms of participants’ comparative ratings of their accident risk in specific situations, though only for less experienced drivers. In contrast, among more experienced males, participants in the insight condition showed greater optimism bias for overall accident risk than their counterparts in the accountability or control groups. There were no effects of the manipulations on drivers’ skills ratings. The differential effects of the two types of manipulations on optimism bias relating to one's accident risk in different subgroups of the young driver sample highlight the importance of targeting interventions for different levels of experience. Accountability interventions may be beneficial for less experienced young drivers but the results suggest exercising caution with the use of insight type interventions, particularly hazard perception style tasks, for more experienced young drivers typically still in the provisional stage of graduated licensing systems.     

Late detection of hazards in traffic: A matter of response bias?

In this study, results from two different hazard perception tests are presented: the first one is a classic hazard-perception test in which participants must respond – while watching real traffic video scenes – by pressing the space bar in a keyboard when they think there is a collision risk between the camera car and the vehicle ahead. In the second task we use fragments of the same scenes but in this case they are adapted to a signal detection task – a ‘yes’/‘no’ task. Here, participants – most of them, University students – must respond, when the fragment of the video scene ends, whether they think the collision risk had started yet or not. While in the first task we have a latency measure (the time necessary for the driver to respond to a hazard), in the second task we obtain two separate measures of sensitivity and criterion. Sensitivity is the driver’s ability to discriminate in a proper way the presence vs. absence of the signal (hazard) while the criterion is the response bias a driver sets to consider that there is a hazard or not. His/her criterion could be more conservative – the participant demands many cues to respond that the signal is present, neutral or even liberal – the participant will respond that the signal is present with very few cues. The aim of the study is to find out if our latency measure is associated with a different sensitivity and/or criterion. The results of the present study show that drivers who had greater latencies and drivers who had very low latencies yield a very similar sensitivity mean value. Nevertheless, there was a significant difference between these two groups of drivers in criterion: those drivers who had greater latencies in the first task were also more conservative in the second task. That is, the latter responded less frequently that there was danger in the sequences. We interpret that greater latencies in our first hazard perception test could be due to a stricter or more conservative criterion, rather than a low sensitivity to perceptual information for collision risk. Drivers with a more conservative criterion need more evidences of danger, thus taking longer to respond.     

Explicit and implicit self-enhancement biases in drivers and their relationship to driving violations and crash-risk optimism

Previous research has found that drivers tend to consider themselves superior to their peers on both driving ability and driving caution, as well as judging themselves as at less risk of a crash (crash-risk optimism). These studies have relied on explicit measures by getting drivers to respond to written items. The current study measured 158 New Zealand drivers’ explicit and implicit attitudes towards their own driving attributes in comparison with others. Implicit attitudes were measured using a computer-based reaction time task, the Implicit Association Test (IAT). Both explicit and implicit self-enhancement biases were found in driver ability and driver caution. Implicit biases were considerably stronger than explicit biases and men demonstrated stronger self-enhancement biases in driving ability than women. Explicit and implicit ratings of driving ability and explicit ratings of driver caution predicted crash-risk optimism. Explicit and implicit ratings of driving caution predicted a measure of driving violations. The implications for safety interventions and research on drivers’ mental processes are discussed particularly in regard to the ability of implicit measures to bypass social desirability effects.     

A rear-end collision risk assessment model based on drivers’ collision avoidance process under influences of cell phone use and gender—A driving simulator based study

Driver’s collision avoidance performance has a direct link to the collision risk and crash severity. Previous studies demonstrated that the distracted driving, such as using a cell phone while driving, disrupted the driver’s performance on road. This study aimed to investigate the manner and extent to which cell phone use and driver’s gender affected driving performance and collision risk in a rear-end collision avoidance process. Forty-two licensed drivers completed the driving simulation experiment in three phone use conditions: no phone use, hands-free, and hand-held, in which the drivers drove in a car-following situation with potential rear-end collision risks caused by the leading vehicle’s sudden deceleration. Based on the experiment data, a rear-end collision risk assessment model was developed to assess the influence of cell phone use and driver’s gender. The cell phone use and driver’s gender were found to be significant factors that affected the braking performances in the rear-end collision avoidance process, including the brake reaction time, the deceleration adjusting time and the maximum deceleration rate. The minimum headway distance between the leading vehicle and the simulator during the rear-end collision avoidance process was the final output variable, which could be used to measure the rear-end collision risk and judge whether a collision occurred. The results showed that although cell phone use drivers took some compensatory behaviors in the collision avoidance process to reduce the mental workload, the collision risk in cell phone use conditions was still higher than that without the phone use. More importantly, the results proved that the hands-free condition did not eliminate the safety problem associated with distracted driving because it impaired the driving performance in the same way as much as the use of hand-held phones. In addition, the gender effect indicated that although female drivers had longer reaction time than male drivers in critical situation, they were more quickly in braking with larger maximum deceleration rate, and they tended to keep a larger safety margin with the leading vehicle compared to male drivers. The findings shed some light on the further development of advanced collision avoidance technologies and the targeted intervention strategies about cell phone use while driving.     

Speed maintenance under cognitive load - Implications for theories of driver behaviour

No theory of driver behaviour has yet managed to achieve widespread acceptance and use in the field of Traffic Psychology, partly due to the difficulty in testing many of the theories. However, one class of theories, the motivational theories, can be usefully split into two groups and the differences between them can then be examined. One group posits the constant monitoring and targeting of a certain subjective variable, often risk, as the controlling factor in driving. The other group however states that subjective variables such as risk are only relevant once a certain threshold has been passed.In this study we aimed to examine this difference by manipulating both speed of travel and the amount of cognitive load participants were under. Participants were asked to initially drive at their preferred speed for 1 min in a driving simulator. Participant's speed was then automatically increased or decreased by 10, 20 30 km/h or left unchanged. Participants were then required to maintain the new speed for 1 min. After this 1 min the speed was again automatically changed and had to be maintained for one more minute, but this time participants also carried out a secondary mental arithmetic task. Finally participants were asked to again drive for another 1 min at their preferred speed. This procedure was repeated seven times, once for each speed manipulation; −30, −20, −10, +0, +10, +20 and +30 km/h. After each 1 min interval verbal ratings of task difficulty, effort, feeling of risk and the typicality of the speed were collected.The results show a threshold effect in ratings of task difficulty, effort and feeling of risk, with no significant difference given between the ratings during the baseline period and the experimentally decreased speed periods until after participant's preferred speed of travel had been exceeded. Furthermore, even when under cognitive load the threshold relationship was still apparent, if diminished. Finally it appears that when under cognitive load drivers have difficulty maintaining a travelling speed which is lower than the speed at which they would prefer to drive. However, driving at a speed in excess of their preferred speed appears to be easier to maintain, at least in the short term.     

That's close enough—A threshold effect of time headway on the experience of risk, task difficulty, effort, and comfort

Subjective impressions of task difficulty, risk, effort, and comfort are key variables of several theories of driver behaviour. A point of difference between many of these theories is not only the importance of these variables, but also whether they are continuously present and monitored or only experienced by individuals at certain critical points in the driving task. Both a threshold relationship and evidence of constant monitoring of risk and task difficulty have been found for speed choice. In light of these conflicting findings this study seeks to examine a different part of the driving task, the choice of time headway.Participants (N = 40, aged 19 to 30) drove in a simulator behind a vehicle travelling at 50 km/h at set time headways ranging from 0.5 seconds to 4.0 seconds. After each drive ratings of task difficulty, risk, comfort, and effort were collected. In addition participants were asked to drive at the time headway they preferred. In order to assess familiarity participants also drove on both the left and right hand side of the road and the role of driving experience was also examined.The results show support for a threshold awareness of task difficulty, risk, effort, and comfort in relation to time headway. Participant's ratings of these variables tended to be low or nil at large time headways, but then around the 2.0 second mark began to noticeably increase. Feelings of task difficulty, risk, and effort were also found to be highly correlated with each other. No effect of driving experience or side of the road was found.     

Effects of measurement assumptions on estimations of risk of collision

This paper provides a replication and extension of previous research describing the relationship between risk of fatal collision and driver blood alcohol content. Risk of collision indices are derived through a comparison of the blood alcohol levels of nighttime Canadian drivers and nighttime driver fatalities. In the context of these data sets an assessment is made of the effects of different treatments of missing data upon estimations of risk of collision. Assumptions regarding the distribution of alcohol among non-respondents in roadside surveys, as well as among driver fatalities who are not tested for alcohol, are shown to have substantial implications for the magnitude of the risk factors derived. Nevertheless, although the magnitude of the relationship between driver blood alcohol and risk of fatal collision varies as a function of the assumptions utilized, the directionality of the relationship corresponds consistently to the results reported in previous investigations.     

The effect of distractions on the crash types of teenage drivers

Teenage drivers are overrepresented in crashes when compared to middle-aged drivers. Driver distraction is becoming a greater concern among this group as in-vehicle devices, opportunities for distractions, and teenage drivers' willingness to engage in these activities increase. The objective of this study was to determine how different distraction factors impact the crash types that are common among teenage drivers. A multinomial logit model was developed to predict the likelihood that a driver will be involved in one of three common crash types: an angular collision with a moving vehicle, a rear-end collision with a moving lead vehicle, and a collision with a fixed object. These crashes were evaluated in terms of four driver distraction categories: cognitive, cell phone related, in-vehicle, and passenger-related distractions. Different driver distractions have varying effects on teenage drivers' crash involvement. Teenage drivers that were distracted at an intersection by passengers or cognitively were more likely to be involved in rear-end and angular collisions when compared to fixed-object collisions. In-vehicle distractions resulted in a greater likelihood of a collision with a fixed object when compared to angular collisions. Cell phone distractions resulted in a higher likelihood of rear-end collision. The results from this study need to be evaluated with caution due to the limited number of distraction related cases available in the U.S. GES crash database. Implications for identifying and improving the reporting of driver distraction related factors are therefore discussed.     

Comparative perceptions of driver ability--a confirmation and expansion

This study aimed to investigate earlier Swedish and United States research which has shown that drivers tend to rate themselves as more skillful and less risky than the average driver, and to extend the range of driver characteristics on which such ratings are made. A total of 178 male and female drivers rated "me as a driver," "an average driver" and "a very good driver" on the following eight relevant bipolar semantic differential scales: foolish-wise, unpredictable-predictable, unreliable-reliable, inconsiderate-considerate, dangerous-safe, tense-relaxed, worthless-valuable, irresponsible-responsible. The results confirmed expectations that a substantial majority of drivers, up to 80%, would rate themselves above average on a number of important characteristics, but also showed that they rated themselves below "a very good driver." The ratings did not vary significantly across demographic categories. Implications for road safety are briefly discussed.     

Voluntary risk taking and skill deficits in young driver accidents in the UK

In absolute terms, young drivers have three to four times as many accidents per year as older drivers; and even allowing for their relative numbers in the population, their accident involvement is about 2.5 times higher than older drivers. A sample of 3437 accident reports was considered, including 1296 in detail, from midland police forces in the UK, involving drivers aged 17-25, and covering the years 1994-1996 inclusive. Four types of accident were identified as being of particular concern due to their high frequency: 'cross-flow'-turns; rear-end shunts; loss of control on bends; and accidents in darkness. (The term 'cross-flow' is used in relation to turns to denote an intersection accident where a driver is turning across the path of oncoming traffic, i.e., left turns in the US and continental Europe, but right turns in the UK and other countries where driving on the left side of the road is the norm.) An examination of driver risk taking behaviours as revealed in police interviews gave an insight into some of the motivational factors underlying young driver behaviour. Young driver accidents of all types are found to be frequently the result of 'risk taking' factors as opposed to 'skill deficit' factors. It had previously been thought that one of the main problems that young drivers have is in the area of specific skills needed in the driving task. However, it appears that a large percentage of their accidents are purely the result of two or three failures resulting from voluntary risk taking behaviour, rather than skill deficits per se. It is shown that specific groups of young drivers can even be considered as above average in driving skills, but simultaneously have a higher accident involvement due to their voluntary decisions to take risks.     

Aging and the detection of imminent collisions under simulated fog conditions

The present study examined age-related differences in collision detection performance when contrast of the driving scene was reduced by simulated fog. Older and younger drivers were presented with a collision detection scenario in a simulator in which an object moved at a constant speed on a linear trajectory towards the driver. Drivers were shown part of the motion path of an approaching object that would eventually either collide with or pass by the driver and were required to determine whether or not the object would collide with the driver. Driver motion was either stationary or moving along a linear path down the roadway. A no fog condition and three different levels of fog were examined. Detection performance decreased when dense fog was simulated for older but not for younger observers. An age-related decrement was also found with shorter display durations (longer time to contact). When the vehicle was moving decrements in performance were observed for both younger and older drivers. These results suggest that under inclement weather conditions with reduced visibility, such as fog, older drivers may have an increased crash risk due to a decreased ability to detect impending collision events.     

The effects of driver training on simulated driving performance

Given that the beneficial effects of driver training on accident risk may not be an appropriate criterion measure, this study investigates whether professionally trained and experienced drivers exhibit safer driving behaviour in a simulated driving task compared with drivers without professional driver training. A sample of 54 police trained drivers and a sample of 56 non-police trained drivers were required to complete two tasks. Firstly to overtake a slow-moving bus on a hazardous stretch of single-lane road with bends and hills and secondly to follow a lead vehicle travelling at 55mph in a built-up section with a speed limit of 30mph. Results showed that in comparison with non-police trained drivers, police drivers were significantly less likely to cross the central division of the road at unsafe locations during the overtaking task and reduced their speed on approach to pedestrians at the roadside in the following task to a greater extent. Police drivers also adopted a more central lane position compared with non-police trained drivers on urban roads and at traffic lights during the following task. Driver group differences in simulated driving performance are discussed with reference to the implications for driver training assessment and skill development.     

Changes in self-reported driving intentions and attitudes while learning to drive in Great Britain

Novice drivers are overrepresented in traffic collisions, especially in their first year of solo driving. It is widely accepted that some driving behaviours (such as speeding and thrill-seeking) increase risk in this group. Increasingly research is suggesting that attitudes and behavioural intentions held in the pre-driver and learning stage are important in determining later driver behaviour in solo driving. In this study we examine changes in several self-reported attitudes and behavioural intentions across the learning stage in a sample of learner drivers in Great Britain. A sample of 204 learner drivers completed a self-report questionnaire near the beginning of their learning, and then again shortly after they passed their practical driving test. Results showed that self-reported intentions regarding speed choice, perceptions regarding skill level, and intentions regarding thrill-seeking (through driving) became less safe over this time period, while self-reported intentions regarding following distance and overtaking tendency became safer. The results are discussed with reference to models of driver behaviour that focus on task difficulty; it is suggested that the manner in which behind-the-wheel experience relates to the risk measures of interest may be the key determining factor in how these change over the course of learning to drive.     

The distraction effects of phone use during a crucial driving maneuver

Forty-two licensed drivers were tested in an experiment that required them to respond to an in-vehicle phone at the same time that they were faced with making a crucial stopping decision. Using test track facilities, we also examined the influence of driver gender and driver age on these dual-task response capacities. Each driver was given task practice and then performed a first block of 24 trials, where one trial represented one circuit of the test track. Half of the trials were control conditions in which neither the stop-light was activated nor was the in-vehicle phone triggered. Four trials required only stop-light response and a further four, phone response only. The remaining four trials required the driver to complete each task simultaneously. The order of presentation of specific trials was randomized and the whole sequence was repeated in a second block giving 48 trials per driver. In-vehicle phone response also contained an embedded memory task that was evaluated at the end of each trial circuit. Results confirmed our previous observation that in the dual-task condition there was a slower response to the light change. To compensate for this slowed response, drivers subsequently brake more intensely. Most importantly, we recorded a critical 15% increase in non-response to the stop-light in the presence of the phone distraction task which equates with increased stop-light violations on the open road. These response patterns varied by driver age and driver gender. In particular, age had a large effect on task components that required speed of response to multiple, simultaneous demands. Since driving represents a highly complex and interactive environment, it is not possible to specify a simplistic relationship between these distraction effects and outcome crash patterns. However, we can conclude that such in-vehicle technologies erode performance safety margin and distract drivers from their critical primary task of vehicle control. As such it can be anticipated that a causal relation exists to collision events. This is a crucial concern for all in-vehicle device designers and for the many safety researchers and professionals seeking to reduce the adverse impacts of vehicle collisions.     

Statistical implications of using moving violations to determine crash responsibility in young driver crashes

Traditional methods for determining crash responsibility – most commonly moving violation citations – may not accurately characterize at-fault status among crash-involved drivers given that: (1) issuance may vary by factors that are independent of fault (e.g., driver age, gender), and (2) these methods do not capture driver behaviors that are not illegal but still indicative of fault. We examined the statistical implications of using moving violations to determine crash responsibility in young driver crashes by comparing it with a method based on crash-contributing driver actions. We selected all drivers in police-reported passenger-vehicle crashes (2010–2011) that involved a New Jersey driver <21 years old (79,485 drivers < age 21, 61,355 drivers ≥ age 21). For each driver, crash responsibility was determined from the crash report using two alternative methods: (1) issuance of a moving violation citation; and (2) presence of a driver action (e.g., failure to yield, inattention). Overall, 18% of crash-involved drivers were issued a moving violation while 50% had a driver action. Only 32.2% of drivers with a driver action were cited for a moving violation. Further, the likelihood of being cited given the presence of a driver action was higher among certain driver subgroups—younger drivers, male drivers, and drivers in single-vehicle and more severe crashes. Specifically among young drivers, those driving at night, carrying peer passengers, and having a suspended or no license were more often cited. Conversely, fatally-injured drivers were almost never cited. We also demonstrated that using citation data may lead to statistical bias in the characterization of at-fault drivers and of quasi-induced exposure measures. Studies seeking to accurately determine crash responsibility should thoughtfully consider the potential sources of bias that may result from using legal culpability methods. For many studies, determining driver responsibility via the identification of driver actions may yield more accurate characterizations of at-fault drivers.     

Predicting injury risk with “New Car Assessment Program” crashworthiness ratings

The relationship between crashworthiness ratings produced by the National Highway Traffic Safety Administration's (NHTSA's) New Car Assessment Program (NCAP) and the risk of incapacitating injury or death for drivers who are involved in single-car, fixed-object, frontal collisions was examined. The results are based on 6,405 such crashes from the Motor Vehicle Traffic Accident file of the Texas Department of Highways and Public Transportation. The risk of injury was modeled using logistic regression taking into account the NCAP test results for each individual model of car and the intervening effects of car mass, age of the driver, restraint use, and crash severity. Three measures of anthropometric dummy response, Head Injury Criterion (HIC), Chest Deceleration (CD), and femur load were used to indicate vehicle crash test performance. The results show that there is a significant relationship between the results of the NCAP tests and the risk of serious injury or death in actual single-car frontal accidents. In terms of overall injury, chest deceleration was a better predictor than the Head Injury Criterion. For restrained drivers, crash severity, driver age, and chest deceleration were significant parameters for predicting risk of serious injury or death; the risk of injury decreased as chest deceleration decreased. The results were similar for unrestrained drivers although vehicle mass and femur load were also significant factors in the model. The risk of overall injury decreased as chest deceleration decreased but appeared to decrease as femur load increased.