ABSTRACT
In traffic psychological examinations psychologists want to predict respondents’ traffic safety based on their results in a psychological assessment. The assumption that traffic safety can be predicted premises that one can measure an individual’s degree of traffic safety (Groeger, 1997). Thus, the measurement of respondents’ traffic safety is of significance to traffic psychology (Klebelsberg, 1982). In particular, earlier studies (eg Spoerer, 1965; Undeutsch, 1962) deal with accidents as measures of traffic safety. According to Spoerer (1979) and Beierle (1995), in about ninety percent of all traffic accidents human failure can be assumed as a cause. This is possibly the reason for the persistent interest in validating traffic-psychological test batteries on this crucial criterion. (Groeger, 1997; Risser, 1997). According to Risser, Zuzan, Tamme, Steinbauer and Kaba (1991) accidents can be defined as a system status, which emerges if two or more drivers recognize a hazard too late to prevent a collision. However, recent studies in traffic psychological assessment often resort to alternate measures of traffic safety, such as standardized driving tests (eg Bukasa, Wenninger and Brandstätter, 1990; Bukasa, Christ, Ponocny-Seliger, Smuc and Wenninger, 2003; Karner and Neuwirth, 2000; Sommer, Arendasy, Olbrich and Schuhfried, 2004). According to Risser (1997) this tendency is due to the fact that several objections have been raised against the use of this criterion measure in the past. These objections rank from the disadvantageous statistical properties of accidents, which are due to the poisson-distribution of this measure (Risser, 1997), to the low validity coefficients obtained in empirical studies using accidents as criterion measure (Bukasa et al, 1990).
