terminants of human behavior, guiding the measurement of these determinants, and shaping the development and measurement of interventions for reducing injuries (National Committee, 1989). It is abundantly clear that injury rates vary according to gender, age, and ethnicity (Chapter 2). Adolescent males, for example, represent one of the groups at greatest risk of injury, especially for motor vehicle and firearm injuries (Fingerhut and Warner, 1997). Behavioral sciences, using a combination of empirical and theoretical approaches, help to explain why gender, developmental age, and ethnicity are important. They examine the influence not only of social and cultural norms, but also of individual and family characteristics that predispose to injury (Irwin et al., 1992; Robertson, 1992). Better understanding of the psychosocial determinants of behavior helps to focus efforts to intervene to change behavior in order to reduce injury risk.
Behavioral research has led to growing awareness of how individuals perceive risk (Kahneman et al., 1982) and how their perception of risk influences their risk-taking behavior (e.g., substance use, recreational vehicle use) that leads to injury (Irwin et al., 1992; Zuckerman, 1993). Risk perception is guided by individuals' understanding of the magnitude of a given risk and also by value judgments about its attributes, such as its familiarity and controllability (Slovic et al., 1982; NRC, 1989). Risk perception and risk decision making are also closely tied to age (Reyna and Ellis, 1992) and gender (Flynn et al., 1994; Barke et al., 1997). This information has helped to explain why facts alone are often insufficient to induce behavioral change for prevention purposes. For example, it is all too clear from the mixed results with HIV (human immunodeficiency virus) prevention efforts that psychosocial determinants, including risk perception and risk-taking behavior, must be taken into account in the design of interventions that strive for enduring behavioral change (IOM, 1994).
Behavioral researchers have also investigated the possibility that people will alter their behavior in response to safety improvements in products or in the environment so as to reduce the safety benefits of these changes. That is, people may increase their level of risk, thereby offsetting some or all of the anticipated safety benefits. It is clear that people behave differently when the risks they perceive change; for example, people walk more carefully when barefoot, drive with greater caution when roads are icy. In similar fashion, consumers are likely to respond in some way to safety-enhanced technology or environmental changes of which they are aware. It may even be that each person has a target level of risk that he or she wishes to maintain, so that, for example, if cars are made safer and the cost of risky driving is lowered, then the driver compensates by taking greater risks. Research seeks to clarify the conditions under which offsetting behavior is likely to occur and to what extent.
Most of the literature on this topic deals with motor vehicle injuries. Evans (1991) outlined some general patterns as to how consumers react to safety-related changes. ''If the safety change affects vehicle performance, it is likely to be used to increase mobility. Thus, improved braking or handling characteristics are likely to lead to increased speeds, closer following, and faster cornering.