Driver Behavior Questionnaire

From this assumed relationship between violations and crashes, it is only a small step to hypothesize that drivers who by their own admission commit many violations and unsafe driving actions should be over-involved in crashes. To test this. Reason and his associates (Reason, 1990 Reason et aL, 1990) developed die Driver Behavior Questionnaire (DBQ), as a tool for validating the theory of planned behavior (see Chapters 3 and 8). A version of the questionnaire, modified for American drivers by Reimer et al. (2005), is reproduced in Table 9-2. It consists of statements that the driver responds to by giving a score ranging fixim 0 to 5, where 0 indicates that the person very rarely engages in this behavior and 5 indicates that the person engages in this behavior nearly all the time. There are three types of questions in the questionnaire, indicative of three types of driving-related inappropriate behaviors ... 

Table 9-2. Reason s Driver Behavior Questionnaire (adapted for U.S. drivers) with statements identifying errors (E), lapses (L), and violations (V). The items are rated from 0 (rarely) to 5 (nearly always) (reprinted from Reimer et a ., 2005, with permission from Elsevier).

Table 9-5. Correlations between parents and children on the different factors in the Driver Behavior Questionnaire (reprinted from Bianchi and Summala, 2004, with permission from Elsevier).

Questionnaire-based measures of aggressive driving. The rationale behind traffic laws and regulations is that the transportation system is just that a system. Its tolerance - from the perspective of expected driver behavior - is defined by the list of traffic violations or infringements. It is also implicitly assumed that drivers v/ho commit many traffic violations are more likely to be involved in crashes than those o do not. This assumption has partial validity from studies that examined die relationship between violations and crashes, though die strength of the relationship varies as a fimction of the violations studied (Cooper, 1997). 

Like all questionnaires that purport to reflect stable characteristics, it is important to determine just how stable or how consistent is a person s score over time. Ozkan et aL (2006) administered the DBQ to 622 drivers and found that over a three-year period the test-retest reliability of the scale scores was only 0.61. This is a fairly low level of reliability and it means that the DBQ can vary significantly over time, as one would expect given the findings that violations tend to diminish with age while lapses tend to increase with age. With a shorter -three-month - test-retest interval the reliability appears to be slightly higher (Parker et aL, 1995), and with a one week interval the test-retest reliability was higher still (Li et al., 2004 obtained r>.80 in China ). The utility of the DBQ, and its close relationship to scales of driver violence and vengeance (Li et aL, 2004) has also prompted the recent development of a Motorcycle Rider Behavior Questionnaire (Elliott et aL, 2007), in order to better understand the causes of motorcycle crashes (see Chapter 16). 

Figure 9-10. The factor loadings (correlations between the communal factor and the individual variables) of general aggression and driving aggression (labeled as road rage ) based on three different questionnaires Driving Anger Seale (das). Driver Vengeance Questionnaire (dvq), and Driver Behavior Inventory (dbi), and among subscales of these questioimaires. (from Van Rooy et aL, 2006).

Reimer, B. L., A. D Ambrosio, J. Gilbert, J. F. Coughlin, J. Biederman, C. Surman, R. Fried and M. Aleardi (2005). Behavior differences in drivers with attention deficit hyperactivity disorder The driving behavior questionnaire. Accid. Anal. Prev., 37,996-1004. 

In addition to rewarding drivers for safe behavior, the questionnaire sought management reaction to the use of disciplinary actions in view of unsafe driving behavior. 

But simply replicating the results does not validate them. The issue of validity is most often involved when we assume intervening variables and rely on surrogate measures of safety (rather than crash involvement). Thus, we should always question the validity of findings that are based on research in driving simulators and in studies relying on drivers self-reports or responses to questionnaires. In neither instance do we measure actual driving behavior, and in neither case do we know how to consider the accidents relative to real ones. 

Motivational factors. The theory of plaimed behavior (Ajzen, 1991. See also Chapter 3) would suggest that speeding - just like any other behavior - is the end product of intentions and the amount of behavioral control over their execution. But what is the relative impact of the different components on the speed we finally choose De Pelsmacker and Janssens (2007) attempted to answer that question by administering a questionnaire addressing the various components of the model to 333 Belgian drivers. A key measure in the questionnaire was the... 

---