Examining driving behavior of autistic individuals to develop and evaluate an innovative driving instructor training program in Qatar

Abstract

Autism is a neurodevelopmental condition characterized by challenges in communication and interaction, repetitive behavior, and restricted interests, with manifestations appearing early in development. Autistic individuals often receive various forms of support such as transportation services provided by family members or caregivers, particularly during childhood. As they transition into adulthood, they often lack access to the adequate services that were previously available to them. This transition period from childhood to adulthood can present challenges for autistic individuals. Transportation challenges can hinder autistic individuals from independent mobility, limiting their access to employment, education, and social opportunities. Learning to drive, driving a car, and managing personal mobility independently all contribute to quality of life. Obtaining a drivers' license is an essential step toward accessing education, job opportunities, maintaining social networks, and improving the quality of life for autistic individuals. However, learning-to-drive can be stressful for those individuals due to the interference of autism-related characteristics during the learning process. For example, difficulties with planning, attention, monitoring, motor coordination, concentration, communication, and anxiety can impact driving abilities. These challenges may make it harder for autistic individuals to learn to drive safely. In Qatar, attention has been given to improving the quality of life among autistic individuals, as indicated in the National Health Strategy 2022 and National Vision 2030. However, autism-tailored driving training programs were not available in Qatar. Due to the absence such programs, instructors may lack the expertise and skills to quickly notice autism characteristics in their trainees and fail to apply a personalized approach to those trainees' demands. In the doctoral project described in this dissertation, the driving behaviors of autistic individuals were examined to inform the development and evaluation of a training booklet for driving instructors in Qatar. The booklet was developed by integrating information from the existing resources and empirical findings from the studies presented within this dissertation. More specifically, the objectives were threefold: (1) To understand the driving behavior of autistic individuals in Qatar when they drive under the influence of cognitive load, frustrating environment, and driving hazards. (2) Based on the outcome of the local studies and existing resources, develop a one-day training program for driver instructors, together with a training booklet. (3) Evaluate the training booklet for driving instructors by comparing the practices of trained instructors when they train autistic learners against the non-trained instructors. This dissertation contains six chapters. Chapter 1 (general introduction) provides the theoretical and background understanding of the research, outlining the doctoral study's rationale. It includes detailed information about autism, covering its etiology, prevalence, and functional implications, as well as an exploration of autism and driving globally and within the context of Qatar. Moreover, this chapter outlines the dissertation objectives and the methodology used to address them. Chapter 2 presents a study investigating the impact of cognitive load on a Lane Change Task (LCT) among male autistic individuals. The study combined a driving simulator-based LCT with a secondary task that induced cognitive load. Specifically, an N-back task generated verbal working memory (WM) load. Seventeen autistic and thirty-four non-autistic individuals participated in the study. Participants completed the LCT in a baseline condition and while performing a three-level auditory-verbal response N-back task (0-back, 1-back, and 2-back). The included driving parameters were mean deviation in the lane change path (MDEV), percentage of correct lane changes (PCL) in response to a lane change sign, and lane change initiation (LCI). The percentage of error rate (PER) was included to measure participant's performance on the secondary task. The results showed that dual-task performance of both groups deteriorated with increasing cognitive load, but this effect was more pronounced in the autistic group. Specifically, the performance of both groups on MDEV, PCL, and PER suffered from the increasing cognitive load. Nevertheless, neither PCL nor LCI differ between autistic and non-autistic participants. Notably, LCI also deteriorated with increasing cognitive load for non-autistic participants, but not for autistic participants. Similar to previous research, it is suggested that distracted driving should be eliminated as much as possible before occurring in the first place. Chapter 3 addresses a study that examined the behavioral and physiological responses of autistic male individuals during driving frustration. The responses of autistic and non-autistic participants were compared when subjected to a series of frustrating simulated driving events. Behavioral (i.e., driving behaviors) and physiological measures were complemented by self-report to allow insights into underlying mechanisms of driver responses. A total of sixty participants, twentythree autistic and thirty-seven non-autistic individuals, mostly pre-drivers, took part in this study. Results revealed that autistic partipants' driving bahavior was impacted by the frustrating events, as indicated in an increased maximum deceleration. However, they also displayed safer driving behaviors, as indicated by a higher mean following distance. Taking all driving measures and known safety cut-off values into account, the impact on traffic safety was comparable between both groups. However, autistic participants experienced higher electrodermal activity (EDA) than their non-autistic counterparts during the simulated frustrating driving events, and a higher percentage of autistic participants reported negative emotions (e.g., anger and anxiety) as compared to their non-autistic counterparts. The results together suggest that while autistic participants can cope with frustrating driving situations, they likely experience higher levels of stress. The current study provides important insights for inclusive driver education programs, which could focus on the best ways to deal with frustrating driving events. Chapter 4 describes a driving simulator study in which the visual attention and driving behavior of autistic individuals in response to hazards was examined. The study included different hazard types, including attention-dividing and attentionfocusing (DF), environmental prediction (EP), and behavioral prediction (BP) hazards. Both objective (i.e., eye-tracking and driving simulator) and subjective (i.e., self-report) measures were included. A total of fifty-three (nineteen autistic and thirty-four non-autistic) individuals participated in the study. The measures included time to first fixation (TTFF), frequency count (FC), first fixation duration (FFD), average fixation duration (AFD), brake reaction time (BRT), minimum timeto-collision (minTTC), speed change immediately before encountering the hazard, and crashes. Self-reported assessments of difficulty in managing hazards were also collected. The results indicated that both groups detected hazards at similar rates, and exhibited comparable TTFF, BRT, speed changes, and self-reported difficulty in handling hazards. However, autistic participants demonstrated shorter gaze durations but higher FC for hazards. They also experienced more crashes and shorter minTTC. The findings varied depending on the type of hazard, with autistic participants showing higher, lower, or similar attention to some hazards compared to non-autistic participants. There were no significant group-by-hazardtype interactions in driving behavior, except that autistic individuals were more likely to experience crashes for BP hazards than non-autistic individuals. However, a difference in crash rates was observed among non-licensed participants. Autistic non-licensed participants experienced a higher crash rate compared to their nonautistic counterparts. The study may reflect that pre-driver autistic participants could benefit from hazard perception training, particularly in dealing with BP hazards. Chapter 5 presents a study on improving the learning-to-drive process for autistic learners in Qatar, conducted in three progressive phases. First, the driving instructors' prior knowledge and practices regarding autism and driving were assessed? Second, the driving instructors' knowledge of autism and driving before and after an evidence-based training was explored. Lastly, the learning-to-drive process for autistic learners following twenty-eight days of driving lessons from trained driving instructors to non-trained driving instructors was compared. Eighty-four male driving instructors and six male autistic learners were included in the study. Results from the assessment phase indicated that most instructors lacked theoretical and practical knowledge about autism and driving. During the training phase, the training improved the driving instructors' knowledge of autism and driving. After the practice phase, trained driving instructors reported that they could better match their lessons to their autistic learners than non-trained driving instructors. Moreover, autistic learners who received driving training from the trained instructors showed a more positive attitude toward driving, had fewer driving concerns, and experienced less perceived stress compared to those who received training from non-trained instructors. These findings support the development of an autism-specific training program for driving instructors to better meet the needs of autistic learners. Chapter 6 provides the general conclusions derived from the studies presented in this dissertation. It summarizes and synthesis the key findings related to the driving behaviors of autistic individuals, considering factors such as cognitive load, frustrating driving conditions, and hazard perception. Additionally, the chapter includes the main findings of a study that evaluates a training booklet designed to support the learning-to-drive process for autistic learners. This chapter also presents information regarding the limitations and directions for future research. Despite the specific limitations and future directions identified in each study, a common suggestion across all studies is to incorporate a larger sample size. All participants were male, as access to female participants was unavailable during the diagnosis and recruitment process. This suggestion should include an adequate number of individuals with and without prior driving experience, as well as a balanced gender proportion, to strengthen statistical power and enhance the potential for the findings' generalizability. To this end, technological and training interventions for reducing cognitive load impact on lane change tasks, improving performance while driving under frustration driving environment, and improving hazard perception skills were addressed in the current chapter. The chapter also highlighted practical implications for tailoring driving training programs to meet the needs of autistic learners in contexts beyond Qatar. Additionally, the chapter provides in-depth insights into autistic drivers and their adaptation to the transition toward automated driving, taking into account the growing presence of autonomous vehicles in the transportation system. The studies included in this dissertation primarily focused on pre-driver participants, offering valuable insights into the learning-to-drive process. However, these findings cannot be generalized to more experienced autistic drivers, especially since research indicates that experienced autistic drivers often show driving performance comparable to that of non-autistic drivers.