Are situation awareness and decision-making in driving totally conscious processes? Results of a hazard prediction task

Abstract

Detecting danger in the driving environment is an indispensable task to guarantee safety which depends on the driver's ability to predict upcoming hazards. But does correct prediction lead to an appropriate response? This study advances hazard perception research by investigating the link between successful prediction and response selection. Three groups of drivers (learners, novices and experienced drivers) were recruited, with novice and experienced drivers further split into offender and non-offender groups. Specifically, this works aims to develop an improved Spanish Hazard Prediction Test and to explore the differences in Situation Awareness, (SA: perception, comprehension and prediction) and Decision-Making ("DM") among learners, younger inexperienced and experienced drivers and between driving offenders and non-offenders. The contribution of the current work is not only theoretical; the Hazard Prediction Test is also a valid way to test Hazard Perception. The test, as well as being useful as part of the test for a driving license, could also serve a purpose in the renewal of licenses after a ban or as a way of training drivers. A sample of 121 participants watched a series of driving video clips that ended with a sudden occlusion prior to a hazard. They then answered questions to assess their SA ("What is the hazard?" "Where is it located?" "What happens next?") and "DM" ("What would you do in this situation?"). This alternative to the Hazard Perception Test demonstrates a satisfactory internal consistency (Alpha=0.750), with eleven videos achieving discrimination indices above 0.30. Learners performed significantly worse than experienced drivers when required to identify and locate the hazard. Interestingly, drivers were more accurate in answering the "DM" question than questions regarding SA, suggesting that drivers can choose an appropriate response manoeuvre without a totally conscious knowledge of the exact hazard. AcknowledgmentsWe are grateful to the Spanish participants who volunteered for the tests as well as to the two anonymous reviewers whose comments enabled us to improve the manuscript and to our English editor Barbara Lamplugh for revising and improving the English. We are also indebted to the Junta de Andalucía (Proyecto Motriz P11-SEJ-7404), the BS-14-164 I+D+I Research project from the CEI-Biotic, University of Granada and the Spanish Government, MICINN (PSI2013-42729-P), all of whom gave us financial support. This research was supported by the Spanish Dirección General de Tráfico -DGT (0100DGT21263 and SPIP2015-01782) and the three driving schools in Granada (Autoescuelas la Victoria, Luna and Genil) from whom we obtained our sample of participants. We much appreciate their contribution to the study. We gratefully acknowledge financial support from the computational resources supplied by the Centro de Servicios de Informática y Redes de Comunicaciones (CSIRCUniversidad de Granada). Its design, data collection, analysis and interpretation were carried out independently of the f...

Modelling human cognition of abnormal machine behaviour

“…Operators' knowledge of the components and their interrelationships may be organised on different levels of detail, called Blattices^, and Moray (1990) suggests that at certain levels, these lattices may differ considerably from the actual system and may even constitute wrong knowledge. In a recent study by Gugliotta et al (2017), drivers are shown video clips ending with a hazardous situation. Drivers are more accurate in answering a question asking what they would do in a similar situation, than they are in answering a question on their awareness of what the hazard is, where it is located and what happens next.…”

Hazard Perception and Prediction test for walking, riding a bike and driving a car: “Understanding of the global traffic situation”

“…On the other hand, risk estimation can be measured with the question: To what extent does the situation you have seen appear dangerous? Respond 1 if it seems slightly dangerous, 6 if it seems highly dangerous [20]. Previous studies have concluded that young drivers are particularly susceptible to "optimism bias", defined by Weinstein (1980) as the tendency to believe that one is more skilled and less likely to experience a negative event than one's peers, and higher perceptions of driving skill are associated with lower perceptions of accident risk.…”

“…Previous studies have concluded that young drivers are particularly susceptible to "optimism bias", defined by Weinstein (1980) as the tendency to believe that one is more skilled and less likely to experience a negative event than one's peers, and higher perceptions of driving skill are associated with lower perceptions of accident risk. [20][21][22][23][24]. found that novice drivers scored significantly lower than experienced drivers when they estimated their driving skills, such as driving ability, awareness of others and self-confidence in driving, although at the same time, they estimated traffic situations as more dangerous.…”

Development and validity of a hazard prediction test for Chinese drivers

“…Within the SA model, there are three levels, namely, perception (level 1), comprehension (level 2), and projection (level 3) [ 21 ], which respectively approximate the source, location, and projection of the road hazard. A number of studies have applied the SA model to hazard prediction assessments and found that it successfully discriminates among driver groups with varying experience [ 19 , 22 – 24 ].…”

“…The three questions are “What is the hazard?” (question 1), “Where is the hazard?” (question 2) and “What happens next?” (question 3). During the experiment, participants are required to select a response from the given choices, or orally respond to questions when a video clip is cut to black or the final still image of the clip remains on the screen [ 10 , 18 , 22 ]. Compared to the reaction time paradigm, the WHN method can prevent the adverse effects of risk threshold and response criterion on participants’ response [ 16 , 25 ].…”