Cover Image

Not for Sale

View/Hide Left Panel
Click for next page ( 15

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement

Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 14
14 CHAPTER 3 CONCEPTS OF CRASH RISK Many interacting factors affect commercial driver crash the view that it is a unitary personality trait (Hansen 1988). involvement. Figure 5 is a conceptualization of some major Studies have found a number of different personal traits that interacting factors. The focus of this synthesis is on personal correlate with crash and other accident risk (reviewed in Chap- "constitutional" risk factors, that is, relatively enduring char- ter 4), but many of these traits may be largely independent of acteristics such as health, physical skills, and some personality each other. Moreover, there is an appreciation of the fact that traits. At any given moment, however, a number of other fac- chance (i.e., factors outside the control of drivers) plays a sig- tors and influences are operative. Drivers are influenced by nificant role in crash involvement and that variations in crash fatigue-related situational factors such as amount of prior sleep, involvement within groups of drivers are partly the result of time-of-day, and hours driving (time on task), but also by non- chance (Dewer and Olson 2002). fatigue situational stressors such as pressure to deliver on time While the concept of accident proneness as a unitary trait or recent events causing anger or anxiety (e.g., argument with has been discredited, researchers have discovered that certain boss or spouse). The driver may also be operating a vehicle on personal traits are related to the occurrence of a vehicle crash. a roadway in bad weather. Each of these can become major Rather than using the discredited term "accident proneness," crash factors, although crash investigations have not found a more appropriate term that reflects the empirical evidence vehicle, roadway, or environmental factors to be frequent prin- may be "differential crash risk." To the extent that this dif- cipal causes of crashes for either large trucks or other vehicles ferential risk is enduring, it probably reflects constitutional (Craft et al. 2004, Treat et al. 1979). Other drivers and traffic or other long-term personal traits. To the extent that it varies are significant sources of large truck crash involvement. across time, it may reflect chance variation or changeable Indeed, it appears that the majority of large truck crashes and traits such as age, maturation, or learning by experience. fatal crashes are precipitated by the actions and errors of other Chapter 4 of this synthesis presents numerous examples of involved motorists (Craft et al. 2004, FMCSA 2003, Blower personal traits that correlate with crash involvement. Most of 1998). Commercial driver behavior and responses may be con- these studies were conducted over a short time period, so the sidered a product of all these interacting factors. The principal question of whether or not they document enduring influences interest here is personal "constitutional" risk factors. on crash risk is unanswered. The remainder of this chapter reviews the original concept of accident proneness and describes several models or concepts of crash risk and the driver errors associated with them. 3.2 MODELS OF DRIVER ERROR AND RISK Various models or concepts of driver errors and crash 3.1 THE CONCEPT OF ACCIDENT PRONENESS involvement have been developed (Dewer and Olson 2002, Rimmo 2002). The Indiana Tri-Level Study (Treat et al. 1979, Accident proneness as an industrial safety human factors Treat 1980) posited three major causal categories and clas- concept was first proposed by Greenwood and Woods (1919). sified 420 in-depth light vehicle crash investigation cases The idea spawned much research, and many studies have been accordingly: human (93%), environmental (34%), and vehic- conducted on the subject since Greenwood and Woods's study. ular (13%). Of course, more than one category can be opera- Greenwood and Woods analyzed the accident records of sim- tive in a crash, and so these percentages total more than 100%. ilarly exposed and experienced munitions workers in Britain Within the human category, four subcategories included recog- and found that a small percentage of the workers accounted for nition errors (56% of in-depth cases), decision errors (52%), the majority of accidents. Accident proneness was conceived performance errors (11%), and "critical non-performance" as a unitary, innate trait that resulted in stable differential risk (e.g., blackout, dozing; 2%). Again, these categories are not across time; in other words, the same workers would continue mutually exclusive and thus total more than 93%. Recognition over time to have the greatest risk (Hansen 1988). errors include distraction (which may be from inside or out- While early studies found support for the accident proneness side the vehicle), general inattention (e.g., daydreaming), and concept, more recent studies on the topic have not supported "improper lookout" (looked but did not see). Recognition

OCR for page 14
15 Figure 5. Major interacting factors affecting commercial driver crash involvement. errors are often associated with rear-end crashes and some fied according to the same human factors classification as used intersection crashes (Najm et al. 1995). Decision errors include in the Indiana Tri-Level Study and occurred as follows: conscious decisions to drive unsafely (e.g., speeding, tailgat- ing) and also gap judgment errors resulting in a crossing-path- Recognition error: 14% type crash. Performance errors are motor responses improperly Decision error: 15% executed (e.g., overcorrection following a lane departure). Performance error: 2% Critical non-performance includes both medical causes and Driver non-performance: 3% asleep-at-the-wheel. Risk-taking behavior is most likely to result in decision errors, but any of these four categories could Another common classification for driver errors resulting be sources of chronic driver risk. For example, a medical con- in crashes is as follows (Dewer and Olson 2002): dition could leave one vulnerable to blackout and a non- performance crash. A follow-up analysis of the Indiana data Rule-based (failure to obey rules or regulations) found that young, unmarried males were the highest risk group Knowledge-based (failure to understand required safe and that most of their crashes were caused by poor decision- behavior) making, including overt risk-taking behaviors like speeding. Skill-based (lack of proper skills to perform the task) The FMCSA/NHTSA Large Truck Crash Causation Study (LTCCS) employs a causal classification similar to that of Most fatal crashes involve misbehaviors or rule viola- the Tri-Level Study. In 286 large truck crashes presented as tions such as alcohol use and speeding. Drivers can also preliminary LTCCS data (Craft and Blower 2003), 34% of the make mistakes without obvious misbehaviors, such as fail- crashes had a "critical reason" assigned to the driver of the ure to see another vehicle or misjudgment of a gap in the large truck. For the other 66%, the critical reason was assigned traffic stream. Red-light running may be regarded as a rule- to another involved driver or to a vehicle or environmental fac- based misbehavior if it is intentional, a skill-based mistake tor. The truck-driverassociated "critical reasons" were classi- if it is not.

OCR for page 14
16 Reason (1990) proposed three error categories: violations (deliberate deviations), mistakes (intended action with unin- tended consequences), and lapses/slips (execution of un- intended action). Rimmo (2002) has expanded this by splitting the lapses/slips category into inattention errors (unintended action resulting from recognition failure) and inexperience errors (unintended action resulting from lack of knowledge or skill). Rimmo's classification, with examples, follows: Violations Deciding to drive when known to be very fatigued Deliberately exceeding speed limits Accelerating at green-to-yellow signal change Mistakes Misjudging gap when crossing traffic Misjudging speed of oncoming vehicle Misjudging stopping distance Inattention Errors Failing to notice red light at intersection Failing to see that vehicle has stopped in lane ahead Figure 6. Simplistic model of how different Failing to notice sign types of error contribute to risk. Inexperience Errors Having to check gear with hand Driving in too low a gear ferent driver age levels. Studies of the intercorrelations Switching on wrong appliance in truck among the four error types have found that the correlations between violation behavior (as measured by questionnaire) Figure 6 shows a schematic of how the four error types and the other three factors were less than the intercorrelations contribute to driver risk. Rimmo's (2002) research on these within the other three factors. Different individuals may be variables employed questionnaires asking driver subjects to "violation prone" or "error prone," and both are associated rate the frequency of various driving behaviors and errors in with accident involvement. However, of the two, high vio- these categories. Analysis of the questionnaire data demon- lation scores are more predictive of crash involvement than strated that the four-factor model was applicable across dif- high error scores.