Instrumented vehicle studies provide contemporaneous measurement of participants’ objective behaviour and situation awareness during a drive. This provides insight into their actions, but additional methods are needed to fully explain why they chose that course of action. To delve deeper into road users’ decision-making, researchers can conduct retrospective cognitive task analysis interviews that probe the key aspects of the decision-making process. Cognitive task analysis interviews can be used in conjunction with other methods, such as in the current project where on-road study participants completed cognitive task analysis interviews after their drive, or as a stand-alone research tool. In some circumstances, stand-alone interviews are the only feasible approach for data collection. However, respondents may provide idealised or non-specific answers that cannot be verified without an objective record of the events in question (Klein 1993). For this reason, the optimal approach is to use a combination of data collection methods.

There are several approaches to cognitive task analysis, but for studies concerned with naturalistic decision-making processes, a leading method is CDM (Klein et al. 1989, Klein and Armstrong 2005), which uses a series of structured prompts to aid recall of past events and explore factors that shaped decisionmaking. The CDM prompts explore how people assess the situation, determine that a decision must be made and formulate appropriate actions (Klein 1993). This process highlights the most critical information that users employ to make judgements and decisions, which should ideally be readily accessible through the system’s design. Thus, a major advantage of studying naturalistic decision-making is that it can help designers to create systems that better reflect users’ information processing needs (O’Hare et al. 2000).

CDM has been used to describe and assess naturalistic decision-making across varied domains, including aviation (Plant and Stanton 2013), military operations (Rafferty et al. 2012) and health care (Galanter and Patel 2005). Notably, these domains all involve highly trained personnel in safety-critical situations. Road transport differs from the traditional CDM context in that road users are not necessarily ‘experts’ at using a given transport mode. However, road users often face situations that resemble classic naturalistic decision-making paradigms, especially at rail level crossings: they must make the decision to stop or proceed under time pressure and dynamic conditions, with incomplete information, ill-defined goals and often poorly defined procedures. For this reason, CDM-based approaches have been gaining popularity as a method for understanding road user decision-making (Young et al. 2015). Further, CDM was chosen given the interest in novice-experienced driver differences and the role of expectancy in decision-making at rail level crossings, issues that we explored through applying Klein’s (1993) recognition-primed decision (RPD) model.

< Prev   CONTENTS   Source   Next >