Motor vehicle crashes remain the leading cause of mortality among teenagers worldwide [1] and unintentional death in persons aged 15-24 years in the United States [2]. Young novice drivers are especially vulnerable to crashes or near crashes related to distracted driving behaviors (DDBs) such as texting while driving (TWD) because of their inexperience, poor risk assessment skills, and ubiquitous use of cell phones [3-5]. DDBs, and TWD in particular, have negative effects on driving performance through the detrimental effects of shared attention, task switching, inattentional blindness, and increased cognitive load while driving distracted [6]. To mitigate these harms, nearly all US states have implemented laws prohibiting TWD in some way, often imposing more restrictions for younger or more inexperienced drivers (ie, graduated licensing restrictions) [7].

Despite teens acknowledging the dangers of distracted driving and generally supporting laws limiting TWD [8], they also continue to engage in DDB [9,10] and are more likely to be in crashes involving distracted driving [11,12]. Phone blocking apps have been found to be more effective than a control condition in reducing TWD among teen drivers [13], but they do not eliminate it, in part because of a lack of motivation to use the apps [14-16]. Unfortunately, even if motivated, novice teen drivers often lack the ability to comply with texting bans due to poor impulse control [17] and, relatedly, problematic cell phone use [18]. These issues, combined with the perceived advantages of TWD and the perceived utility of compensatory strategies [9], may outweigh the acknowledged risks among young novice drivers who engage in DDB [19].

The educational paradigm for teenage driver’s education has remained static, with most evaluation research demonstrating limited effectiveness in reducing crashes among young novice drivers [20], although there are recent exceptions [21]. Despite the pervasiveness of technology related to drivers’ training [22], driver’s education still relies on classroom-style textbooks, lectures, videos, and low-technology simulation (eg, a video screen connected to a simple wheel and pedal). Our goal was to develop an intervention to augment traditional in-person driver’s education.

Head-mounted display–based virtual reality (VR) is an exciting technology used for gamification and is increasingly used for learning and safety training in multiple fields [23-27]. Providing immersive, experiential evidence of how distractions adversely affect performance through VR could help address problematic teenage and novice driving behaviors [28-30]. Research involving VR and drivers training to date has featured primarily postlicensing skill assessment targeting generalized improvement [29,31], such as through hazard perception and mitigation training [28,32] and error identification [31] of adult drivers. Only Jakab [29] specifically recruited those with “having close to no experience with driving in real life.”

To the authors’ knowledge, just one early VR study by Morley et al [33] focused on distracted driving specifically. Morley et al [33] engaged a small number of participants, aged 20-40 years, in driving around a virtual track. Once familiar with the track, they were asked to interact with a virtual smartphone, which triggered a large oncoming truck accompanied by loud horn sounds. This resulted in an unavoidable crash with crash sound effects and “violent movement from the force feedback” technology. From in-session feedback, 20 (86%) out of 23 participants said that the event reinforced or changed their perceptions of the dangers of phone use while driving. No longer-term feedback on the effects of the VR experience was collected. As Morely et al [33] and other VR driving studies demonstrate, there is definite potential for VR to address TWD as part of driver’s education.

VR is specifically recognized as a promising approach to novice driver education and training that can meet the new Novice Teen Driver Education and Training Administrative Standards [34]. VR-based games have several potential advantages over traditional video or low-technology simulators as a teaching tool: (1) presence, or the idea that the user is physically present in the virtual environment; (2) embodiment, when the user feels they are genuinely inhabiting the virtual character and having their actions mirrored within the virtual environment; and (3) physicality, where the user’s actual degree of physical activity is substantially increased during gameplay, creating an increased immersive, “learning-by-doing” experience [34,35]. By experiencing the consequences of in-game distractions, young novice drivers may reevaluate their beliefs about TWD and perceived ability to text while driving and reassess their perceived evaluation of the risks they take when engaging in TWD, which affects their attitudes and intention to engage in TWD.

The primary aim of this study was to evaluate the utility of Distracted Navigator and facilitated discussion using a mixed approach in a pre- and postintervention randomized controlled trial. We sought to test the following hypotheses: (1) intervention group participants who experience the VR game will have significantly reduced intentions to engage in TWD and (2) the intervention group will have statistically significant changes in other theory of planned behavior (TPB) constructs regarding TWD. Additionally, we used open-ended and focus group interview data (1) to gain participant insight into the utility of their VR experiences and themes supporting the quantitative data and (2) to identify qualitative themes for future development and enhancements to the VR game.

By having teens experience a novel VR game featuring DDB (Distracted Navigator) and engage in a facilitated educational discussion relating gameplay to real-world consequences, we sought to make more effective education on how DDB and TWD impact driving performance.

Distracted Navigator was developed iteratively with Preview Labs [36], informed by VR PLAY [37] and other game development guidelines. Using a VR head-mounted display, a player is placed virtually inside a spaceship cockpit and has control of the speed and the direction of the ship, which they navigate through an asteroid field (Multimedia Appendix 1). Distracted Navigator provides vibrational haptic feedback via the player controller when asteroids hit the spaceship with noticeable damage to the spaceship’s windshield.

After a 5-minute onboarding tutorial and practice session, the user plays their first round of the game (without any distractions) for ~5 minutes, navigating the spaceship to avoid oncoming asteroids. They receive a score depending on how successfully they avoided damage to the ship. In round 2 of the game, distractions are introduced, as the player is exposed to several “emergency” tasks. While still navigating the ship around incoming asteroids for ~5 minutes, players must manage emergency tasks including (1) typing in a specific number sequence into a keyboard, (2) looking at orbs of lights on the sides of the cockpit to reactivate the ship’s lighting system, (3) pulling a specific sequence of overhead levers, and (4) plugging devices into the console. The game ends with comparing the round 1 (without distractions) and round 2 scores (with distractions) with an illustration of a spaceship demonstrating the level of damage received during round 2 of play.

Distracted Navigator was designed specifically to (1) be immersive and fun while experiencing the performance-reducing effects of DDB, (2) not incentivize players to improve their score while engaging in DDB, and (3) leverage the familiar-to-teen experience of watching live streaming gameplay by broadcasting individual play to other group members. Collectively, Distracted Navigator tasks simulate the effects of (1) texting, (2) taking eyes off the road, (3) multitasking, and (4) plugging in devices to demonstrate the negative effects of shared attention, task switching, inattentional blindness, and increased cognitive load while driving distracted [38]. The game also features a cartoon rooster that flies across the screen during gameplay to demonstrate inattention blindness like the classic “invisible gorilla” experiment [39].

Both the intervention and the control group participated in the discussion, facilitated by a content expert familiar with working with teens (author TV), which educated participants on the dangers of DDB. Importantly, the facilitated discussion was implemented using an interactive approach akin to motivational interviewing to elicit the teens’ own perspectives and to better engage them in collaborative, thoughtful discussion [40].

The content was adapted from didactic educational material on distracted driving offered by the longstanding nonprofit service organization American Automobile Association (“triple A,” colloquially) [41] and the local pediatric hospital’s DriveSmart campaign (C.S. Mott Children’s Hospital, University of Michigan Health System Pediatric Trauma), including web-based advice to parents to support their teen’s safe driving (eg, conversation tips). The material was further enhanced by the inclusion of TPB concepts related to TWD (Figure 1).

Since the evaluation of risks, peer norms, and perceived control beliefs all play roles in the complex decision to use or not use a phone while driving [42,43], especially for young drivers [9,44,45], the TPB captures the complexity of distracted driving better than other models (eg, health belief model) [46]. The facilitator acknowledged the pressure to reply to messages while driving, and teens were asked to share whether they thought habitual phone use could make avoiding distracted driving difficult. The facilitator then elicited strategies to combat this pressure and offered other strategies, if not suggested by the teens themselves, to enhance control beliefs and perceived power. In addition, teen perceptions of multitasking ability and the utility of risk-compensation strategies were addressed. The intervention group had additional material that tied Distracted Navigator gameplay experiences to everyday driving situations and reinforced these TPB concepts.

Game onboarding and facilitated discussion were beta-tested with 2 teens recruited from a local children’s hospital volunteer patient advisory committee. Teenagers were recruited via email, and consent procedures were conducted with parents in a virtual meeting room. We then recruited future and novice teen drivers from a Michigan high school to pilot-test the effect of a virtual gaming intervention and accompanying discussion on TWD perceptions and attitudes from July to August 2022 and in July 2023. Authors ANH and VC liaised with a local high school to recruit teen drivers. Teens were eligible if they were aged 14-17 years. Internet and email literacy were implicit eligibility criteria. Email contact was collected for both the students and a parent or guardian. Author VC met with the teens and their parent or guardian via a virtual meeting room where eligibility was assessed, study procedures were reviewed, and written participant assent and parent or guardian consent were obtained.

Teens were recruited in blocks of up to 7 and randomly assigned at the block level to intervention and control groups approximating a 2:1 ratio. All teens were asked to complete the baseline survey before the session. All other study activities took place face-to-face. In the intervention group, participants played the Distracted Navigator VR game and engaged in the facilitated driver’s education curriculum on the effects of distracted driving with references to their gameplay experiences, completed the postsession survey, and then participated in the focus group to provide additional feedback. As described earlier, the control group engaged in a similar moderator-facilitated driver’s education curriculum on the effects of distracted driving with no reference to the gameplay. They then completed the postsession feedback survey and were then allowed a free period to play the VR game for fun. Following this gameplay, the control group engaged in a second focus group session to provide feedback on the VR game. The participant study flow is shown in Figure 2. Author AG assisted in fitting study participants with VR headsets and gameplay training.

All study staff and investigators completed the University of Michigan Minors as research participants and followed the participant contact protocols. The CONSORT (Consolidated Standards of Reporting Trials)-EHEALTH checklist [47] is available in Multimedia Appendix 2.

This research project was reviewed and approved by the University of Michigan Institutional Review Board (#HUM00213233). The University of Michigan Institutional Review Board has full accreditation by the Association for the Accreditation of Human Research Protection Programs, Inc. All research staff and investigators have completed the University of Michigan–mandated training on conducting research with minors in addition to human subject protection certification. No exemptions or alterations were made to the informed consent process. Written informed consent was collected virtually using a secure instance of the SIGNNow document service. As all participants were aged 14-17 years, a separate abbreviated assent document was not used in this study. Both informed consent and assent were collected with the use of the same form. A complete informed consent and assent process were conducted via a secure Zoom (Zoom Video Communications) video call with both the adolescent and parent facilitated by a Michigan Medicine research coordinator or a study investigator. In addition to consent to conduct primary research data collection and analysis, the consent document also contained permissions to use deidentified study data for future research studies and distribution without additional consent. Participants were assigned a study number, and collected data were stripped of identifiers. Audio recordings were transcribed by the researchers, and any potentially identifying information was removed. Participants who completed all study procedures were provided 1-time compensation of a US $100 gift card and provided with food and beverages during the session as incentives. These gift cards were obtained through the University of Michigan Human Subjects Incentive Program and distributed to the participants by the investigators at study completion. Identification of the participants would not be possible through any information provided in this paper or any supporting documents. All identifying or potentially identifying information is housed and maintained by the University of Michigan and is only accessible to the research team.

Although there was no statistically significant impact on novice drivers’ DDB attitudes and intentions, our findings show success in developing an intervention to augment traditional driver’s education. Participant feedback suggests that the immersive experience of Distracted Navigator reinforced the negative effects of TWD and other DDB in ways that standard drivers’ education lectures cannot.

According to the protection motivation theory [53,54], novice drivers who have limited driving experience may have an inaccurate perception of the risks or consequences and therefore may engage in behaviors like TWD more often. Our feedback suggests that Distracted Navigator was successful in providing a safe and immersive setting for teens to experience such consequences and thus could influence the perceived risk of TWD behavior. Indeed, despite most participants reporting never experiencing the consequences of DDB (eg, ticketed or a crash), participants readily connected the “emergency tasks” to real-world behaviors and the negative effects of distracted driving.

While our focus group feedback underscored the efficacy of the combined programming—facilitated discussion and Distracted Navigator—in demonstrating the repercussions of DDB and changing participants’ intentions, the quantitative data did not show analogous statistically significant outcomes. In addition to the small sample size, there are 2 primary other explanations for this discrepancy.

The first is the possible confounding effect of the facilitated discussion in isolation. All groups, control and intervention, agreed that the facilitated discussion was superior to traditional classroom-based drivers’ education. While the synergistic effect when paired with the game was clear, the benefit of the discussion alone, thoughtfully engaging teen drivers in discussion of these behaviors and related impairments, can still heighten awareness of their dangers with or without the VR gameplay.

Second, in both the preintervention survey and the focus groups, most participants expressed their intention and self-reported habits to not text and drive. This “floor effect” allowed for little movement toward a statistically significant impact. Even so, nearly three-quarters of participants had said the program changed how likely they were to text and drive when asked outside of the TPB items.

As a pilot study, our findings show promise for Distracted Navigator as a new technology-based educational tool for novice drivers. Though our study was limited by size and the possible confounding effects of a specifically designed discussion, the feedback was positive with concrete recommendations for future improvements, highlighting potential areas for future research. Another potential benefit of the VR approach is the capability to live stream VR gameplay, a popular and familiar format for teens to engage in the play by proxy, enhancing the scalability of the intervention [55,56]. Larger, longitudinal studies with greater statistical power are warranted to evaluate the potential for VR to have a lasting impact on driver’s education.

Our study was constrained by the small pilot sample size, which limited statistical power and contributed to the imbalance in the baseline differences in important TWD behaviors. The study relied on self-reported data for texting behavior, which could have been susceptible to demand characteristics, and did not include a longer-term follow-up to evaluate the lasting effects of the intervention. Most participants had prior experience with VR, and all had their own cell phones, so results may not be generalizable to the teenagers living in underserved areas without exposure to technology or access to cell phones. However, we captured candid feedback that offered valuable insights into participant experiences with the VR game and facilitated discussion to inform the intervention and VR game development.

The growing availability of VR systems enables the easy integration of educational VR games like Distracted Navigator to enhance driver training education programs. Insights gained from participant feedback will be used to improve the VR game and the facilitated discussion material in preparation for a larger, prospective, randomized controlled trial to evaluate their impact on DDB beliefs and objective DDBs using smartphone technologies.