Introduction

Multitasking has become a ubiquitous part of modern life and operating a motor vehicle is no exception. Naturalistic driving research has revealed that drivers spend approximately half of their driving time engaged potentially distracting tasks, unrelated to driving (Dingus et al., 2016; Young et al., 2019b). Driving while distracted impairs a range of cognitive processes and vehicle control skills required for safe driving, including speed and lateral control, situation awareness, hazard perception and reaction time (e.g.,Strayer, 2015; Vieira & Larocca, 2017; Young et al., 2013). In addition, distracted driving has been found to be a major contributing factor in almost 16 percent of serious casualty road crashes resulting in hospital attendance in Australia (Beanland et al., 2013) and in over eight percent of fatal crashes in the United States (Stewart, 2023).

An interesting aspect of distracted driving is that drivers often choose to engage in certain non-driving tasks despite an awareness of their negative consequences (Atchley et al., 2011). The cognitive and neurological processes underlying drivers’ willingness to engage in risky driving, including distracted driving, has been the focus of numerous studies. Individual differences in a complex set of behaviours collectively referred to as ‘executive functions’ have been identified as an important attribute that can explain behavioural differences between drivers. Executive function is defined as ‘‘cognitive abilities for adaptive functioning, allowing for behaviour that is more goal-oriented, flexible, and autonomous” (Spinella, 2005, p. 650). Executive functions include an interrelated set of higher-level cognitive processes and self-regulatory functions carried out by prefrontal areas of the frontal brain system, including planning, organising, inhibiting responses, attentional allocation and control, working memory, reasoning, problem-solving, and monitoring (Burgess & Stuss, 2017; Chan et al., 2008). Impairments in executive functioning can manifest in a range of ways that may impact driving, including poor judgement, attention control, impulse control and self-regulatory abilities (Goldstein et al., 2014).

The findings from several studies confirm the role of executive functioning in risky driving behaviour. In particular, lower levels of executive function have been associated with dangerous or aberrant driving behaviour (Tabibi et al., 2015); young drivers’ risk taking (Hayashi et al., 2018; Pope et al., 2016; Starkey & Isler, 2016; Walshe et al., 2017); and driving performance degradations due to aging (Adrian et al., 2011; Daigneault et al., 2002), neurodivergence (Bednarz et al., 2022; Cox et al., 2016; Patrick et al., 2020) and traumatic brain injury (Narad et al., 2020).

Given its role in the selection and allocation of attention resources and impulse control, executive functioning is an important element in our understanding of the cognitive and neurological mechanisms underlying engagement in distracted driving. Only a handful of studies have examined the relationship between executive function and engagement in distracted driving (Hayashi et al., 2017; Mizobuchi et al., 2011; Pope et al., 2017; Sanbonmatsu et al., 2013; Zhang et al., 2020). Pope et al., for example, examined executive dysfunction and frequency of distracted driving across young, middle-age and older drivers. They found that difficulties with global executive function was related to increased frequency of distracted driving engagement across all age groups and that the relationship between age and distracted driving was partially mediated by global executive difficulty. Zhang and colleagues also found, in a sample of Chinese drivers, that poorer executive functioning was associated with higher engagement in distracted driving and this was true for both young (below 25 years) and mature (25 years and above) drivers. Focusing on texting behaviour among college students, Hayashi et al. found that those students who frequently text while driving demonstrated significantly lower levels of executive function and higher levels of self-reported impulsivity.

While previous research has demonstrated a strong relationship between executive function and distracted driving that spans a range of age groups, these studies have focused solely on the frequency of distracted driving engagement. Deciding whether or not to engage in a non-driving task in a given situation is just one of a number of ways in which drivers self-regulate or adjust their behaviour in response to changing or competing task demands in order to manage their workload and safety risk. Self-regulation of non-driving tasks has been described as occurring across three levels: Planning, Decision and Control (Schömig & Metz, 2013). At the highest planning level, drivers execute global strategies to manage or prevent distracting activities when driving. Strategies at the planning level, do not relate to the decision to engage or not in a task in a specific situation, but rather to more global strategies regarding how drivers deal with non-driving tasks when driving. For example, drivers may elect to perform non-driving tasks before they start driving or to only perform task when they are stationary. Tivesten and Dozza (2015), for example, demonstrated that drivers were more likely to initiate visual-manual phone tasks when the vehicle was stationary, when travelling at lower speeds, or when there were no passengers present in the vehicle. Young and colleagues (2019a) also found that drivers in a naturalistic driving study made a number of strategic decisions regarding when to engage in non-driving tasks, such as only engaging when they were stationary.

At the Decision level, drivers decide, in the moment, whether they will engage in a non-driving task in a given situation and for how long they will engage. Factors such as events occurring in the driving environment, the type of non-driving task, the perceived importance of the task, as well as global strategies implemented at the planning level, are believed to play a role in drivers’ decision making at this level. Once engaged in a non-driving task, the control level commences. At this level, drivers will, to varying degrees, dynamically adjust their driving behaviour and interaction with the non-driving task according to the demands of the driving situation. In terms of driving-based adjustments, research has shown that drivers sometimes reduce their speed and increase following distance when engaged in non-driving tasks in order to maintain adequate safety margins (e.g., Strayer & Drews, 2004). Drivers also dynamically interrupt, delay and resume non-driving tasks in an attempt to share these tasks more safely with driving (Liang et al., 2015). For example, Young and colleagues (Young et al., 2020) found that, under naturalistic driving conditions, drivers interrupted non-driving tasks so they could attend to the driving task and that tasks were more likely to be interrupted if they were longer in duration and had a higher visual load.

This study investigates the relationship between executive function and drivers’ self-regulatory behaviours at the planning, decision and control levels. Specifically, it examines whether difficulties in everyday executive functioning impacts drivers’ ability to self-regulate their behaviour in relation to non-driving tasks and if this differs across the three levels of self-regulation. Previous studies of executive function and distracted driving have relied on the use of questionnaires asking about behaviours occurring across the previous week (Pope et al., 2017; Zhang et al., 2020) or a longer, unspecified time period (Hayashi et al., 2017). This relies on drivers recalling behaviours and intentions, sometimes well after they have occurred. As these behaviours become more automated, this recall may be less accurate. Further, these studies measure only one or two occasions, and present a snap-shot in time, rather than repeated exposures.

This study extends this previous research by using trip diaries completed repeatedly and soon after individual driving trips undertaken over a four-week period. The diary method allows for the collection of longitudinal, naturalistic data close to the behaviour occurring and also permits the opportunity to explore in more depth how and why drivers elect to engage in non-driving tasks, what external factors influence these decisions and how drivers regulate their behaviour to mitigate risk. Diaries have been used successfully in the driving domain (e.g., Wickens et al., 2013), including to examine the factors that contribute to distracted driving (Parnell et al., 2020).

Method

Participants

A sample of 27 participants were initially recruited to take part in the diary study. All participants were recruited from the general public using social media advertising (Facebook, X and Instagram) and flyers placed around the university campus. Inclusion criteria included being aged 18 years or over, holding a valid driver’s licence, being a current driver, defined as actively drive more than twice a week, and currently residing in Australia.

The final sample included twenty-five (7 male, 18 female) drivers who completed the initial demographic survey in full and provided more than one trip diary for analysis. The remaining two participants did not complete either the initial demographic survey or provided an insufficient number of diaries and, thus were removed from the analysis. Participants ranged in age from 20 to 65 years (M = 34.8, SD = 11.5). All participants held a full (80%) or probationary (20%) drivers licence and self-reported in the demographic survey as driving two or more times per week, with the vast majority (92%) reporting that they drove over 10,000 km in the past year. Additional participant characteristics are provided in Table 1.

Table 1 Participant demographics and executive functioning scores
Full size table

Participants who completed all study requirements were provided with a $100(AUD) gift card. In addition, to incentivise participants to complete as many diaries as possible, those who completed 20 or more diaries were entered into a prize draw of a further $100(AUD) gift card. The study was approved by the University Human Research Ethics Committee and Informed consent was obtained from all participants.

Study materials

Preliminary survey

Participants completed a preliminary survey to screen for eligibility and collect their demographic and socio-economic characteristics (e.g., age, gender, employment status and highest level of education), driving patterns and experience, and frequency of device engagement when driving during the past six months.

Individual executive function was assessed using the self-report version of the Behaviour Rating Inventory of Executive Function – Adult Version (BRIEF-A; Roth, Isquith, & Gioia, 2005). This scale includes 75 items that address everyday self-regulation behaviours occurring in the past month on a 3-point Likert scale: (1) Never, (2) Sometimes, (3) Often. The BRIEF-A questions capture the extent to which the respondent reports difficulty in nine different domains of executive function, including Inhibit, Self-Monitor, Task Monitor, Plan/Organise, Shift, Initiate, Emotional Control, Working Memory and Organisation of Materials. These scales form two broader indexes: Behavioural Regulation Index (BRI) and Metacognition Index (MI). The BRI captures a person’s ability to freely shift cognitive set across situations and regulate emotions and behaviour via appropriate inhibitory control, while the MI reflects the ability to initiate, plan and organise strategies for problem solving and to self-monitor success. The two composite scores (BRI and MI) are summed together to yield the Global Executive Composite (GEC) score, which provides an overall score of executive functioning. Higher scores are indicative of worse executive functioning.

The preliminary survey also contained a range of psychological tests that were not analysed as part of this study, including the Mindful Attention and Awareness Scale (MAAS; Brown & Ryan, 2003); the NMP-Q Nomophobia Scale (Yildirim & Correia, 2015); the perceived stress scale (Cohen & Williamson, 1988); and the Prospective and Retrospective Memory Questionnaire (PRMQ; Smith et al., 2000). The initial survey was administered online via Qualtrics, with a link to the survey being provided to the recruited participants via email. The survey was completed prior to the diaries being provided to participants and took approximately 30–40 min to complete. The survey was piloted beforehand to ensure it was readable, easy to navigate and was not too long. Participants were free to complete the survey at a time suitable for them and were encouraged to complete it at their own pace and take regular breaks if needed.

Diary surveys

The diary surveys were completed online via the survey software Qualtrics. The diaries comprised predominantly multiple-choice and closed questions to facilitate quick completion. To gather more detailed information about certain behaviours, such as ‘other’ devices or interaction styles, participants were also offered the opportunity to respond to a small number of open-ended questions. Participants were encouraged to complete a separate diary for each device they interacted with during each trip. The diaries collected information on various contextual aspects of each trip as well as technology interaction:

Journey details

Participants were asked to specify details of the trip, including the type of vehicle driven, whether the trip was for work or private purposes, the approximate start time and duration of the trip and whether they had adult or child passengers present.

Device interaction

Participants specified if they interacted with any portable or on-board devices during the trip. If participants selected ‘no’, they were asked to state why they did not interact with a device, including not receiving any incoming alerts from a device, choosing to ignore incoming alerts or calls, not feeling safe to interact, having kids/passengers in vehicle, traffic being too heavy, or performing a driving manoeuvre (e.g. overtaking, parking).

If an interaction occurred, participants selected the type of technology used from the following options: Mobile phone (hand-held or hands-free, including Bluetooth, cradle, voice controlled); iPad/tablet; headphones; smartwatch; onboard satellite navigation; and onboard entertainment system or climate controls. There was an option to specify ‘other’ devices if the device was not listed. The type of actions performed when using the device was also collected, including if drivers physically interacted with the device by pressing buttons or holding. Other activity options included reading, writing, listening, talking, monitoring (e.g., glancing at navigation map) or an ‘other’ option, with participants given the opportunity to select all options that applied. Finally, participants were asked if the interaction was initiated by them (e.g., placing a phone call) or by the technology (e.g., incoming call).

Self-regulation

Items were included to capture participants’ self-regulation behaviours at the planning, decision and control levels. With regard to the planning level, participants were asked if they did anything before the trip to reduce the number of non-driving tasks they needed to perform when driving. A range of possible response options were provided, with participants able to select as many as applied; including, entering destination into a navigation system; connecting phone to Bluetooth/cradle; turning off notifications or putting phone on silent/do not disturb; putting phone out of reach (e.g., in glovebox/compartment; or participants could specify an ‘other’ option. At the decision level, participants were asked to indicate whether they interacted with one or more devices during each trip. Participants also completed two questions designed to capture information on self-regulation at the control level. These items required drivers to report if, during the device interaction, they changed their driving behaviour or the way they interacted with the device in order to manage the increased risks of device interaction while driving. Multiple response options were provided, along with an ‘other’ option, and participants were instructed to select as many options as applied.

Driving context

Driving conditions at the point where drivers started to interact with (or chose to ignore) the device were collected to assess how a range of contextual factors influenced the decision to engage or not. Factors included road type (freeway, highway, urban road, residential street, intersection, other); traffic density (heavy/bumper-to-bumper, medium, light, no surrounding traffic); and weather conditions (dry, wet/raining, foggy, icy, sun glare, other).

Open-ended question

Participants were finally asked to respond to an open-ended question regarding if there was anything else about their interaction with the device or trip that they would like to disclose.

Procedure

Recruited participants were initially provided with details of the study, their participant ID, instructions about what was required of them during the data collection period and their eligibility for the study was confirmed. Participants could ask any questions about the study at this point and their contact details for sending the links to the survey and diaries, reminders and gift voucher were collected. Participants were emailed a link to the online preliminary demographic survey and completed this before being provided with a link to the diary surveys. Participants were instructed to complete diaries for as many trips as possible over a 4-week period, regardless of whether they interacted with technology during the trip or not. When technology interactions occurred, participants completed separate diaries for each device interaction made during each trip. Data was collected from June to September 2022 and the preliminary survey and diary data were linked via participant IDs. Participants were instructed to complete the diary only at the end of each trip when it was safe to do so. They were encouraged to complete each diary as soon as possible after each trip to ensure accurate recall of the trip and their interaction behaviours. Instructions were given not to complete diaries while driving. Participants were also instructed that the data they provided was confidential, to ensure honest reporting of distracted behaviour. The number of diaries being completed by each participant was monitored during the 4-week period and email reminders sent to all participants each week. Participants were provided with the gift voucher at the end of the data collection period once they had completed all study requirements.

Study design and data analysis

This study used an initial online survey and online trip diaries completed over a 4-week period to examine the influence of executive function on drivers self-reported device use while driving. Data were analysed in SPSS for Windows v.28. The data for items relating to the planning, decision and control levels of self-regulation were converted to binary responses (yes/no) for the purpose of analysis. Descriptive statistics were then used to describe the specific types of self-regulatory strategies drivers engaged in. The associations between executive functioning and the planning, decision and control levels were examined using Generalised Estimated Equations (GEE). The BRI and MI were moderately correlated (r = 0.69), but as they each provide different insights into the potential role of behavioural executive disruptions on drivers’ ability to self-regulate their distracted driving behaviours, both indices were included in the GEE models, rather than the GEC. Four models were conducted for each separate binomial outcome (i.e., engagement in self-regulation (no/yes) for each type of regulation). Models were nested across participant, diary day and diary number with an unstructured correlation matrix. Predictor variables included age (continuous), trip duration (log), gender (man/woman), trip purpose (leisure/work) and presence of passengers (no/yes), MI and BRI. The model for the decision level self-regulation also included planning self-regulation as a predictor. All variables were entered simultaneously into the models and retained in the model.

Results

Overview of technology engagement

A total of 710 trip diaries were completed by participants across 604 trips. Each participant completed an average of 28 (SD = 33.2) diaries across the 4-week study period. On average, participants completed the trip diaries within 6.4 h (SD = 3.5, range: 25 min to 13 h) of undertaking each trip. Of the 604 trips, 408 (67.5%) involved one or more device interactions, with an average of 1.3 (SD = 0.6) interactions per trip (514 interactions in total). The vast majority of these interactions (93%) were initiated by the driver, with only 7% device initiated (e.g., incoming call or notification). Given the small number of device-initiated interactions, data for the driver- and device-initiated interactions were analysed together. The vehicle’s entertainment system was the device most frequently engaged with by drivers (27.6%), followed by climate controls (12.3%), on-board satellite navigation (11.5%), hands-free phone (11.0), and hand-held phone (9.2%).

When a device interaction did not occur during a trip, this was because drivers ‘did not receive any incoming calls, messages, notifications’ (55.9%); ‘ignored any incoming calls/messages/notifications’ (6.3%); ‘had kids/passengers in car’ (6.3%); were ‘performing a driving task (e.g., turning, parking)’ (4.2%); ‘traffic was too heavy’ (3.8%); or they ‘did not feel safe’ (3.4%).

Planning level self-regulation

A GEE model examined whether the Behavioural Regulation and Metacognition indexes of executive function predicted driver engagement in the planning level of self-regulation. Planning self-regulation was measured using the item “Before starting your trip, did you do anything to help reduce the number of non-driving tasks you needed to undertake when driving?”, with the outcome being a yes/no response. Drivers reported that they engaged in self-regulation at the planning level in 273 (37.5%) of the 710 trip diaries.

Table 2 displays the factors associated with the odds of drivers planning their interaction with devices while driving. Results showed that the Behavioural Regulation and Metacognition indices of executive function had different relationships with planning regulation. Drivers who scored highly on the Behavioural Regulation index were more likely to report that they undertook actions prior to driving to reduce the number of non-driving tasks that they engage in. Conversely, drivers who scored highly on the Metacognition index were less likely to report engaging in self-regulation at the planning level.

Table 2 Factors associated with drivers’ self-regulation of device interaction at the planning level
Full size table

In terms of other factors, the odds of performing planning self-regulation behaviours were significantly higher when driving for leisure, compared to when driving for work. Men were also more than two times less likely than women to adapt behaviour at the planning level. Driver age, trip duration or the presence of passengers were not significantly related to the odds of self-regulating behaviour at the planning level.

In cases where participants indicated that they engaged in self-regulation at the planning level for a trip (n = 273), details regarding the type of planning activity or behaviour were collected. Participants could specify more than one planning level behaviour per trip. The most common behaviour performed before starting a trip was to enter destination details into the satellite navigation system (n = 174, 57.6%), followed by connecting their mobile phone to the vehicle via Bluetooth or a cradle (n = 80, 26.5%). Putting their phone in the glove box (n = 5, 1.7%), on silent or do not disturb (n = 5, 1.7%), or turning off notifications (n = 2, 0.7%) were also reported. Participants also reported engaging in ‘other’ planning level behaviours before driving (n = 36, 11.9%), including adjusting heating, ventilation, air conditioning (HVAC) controls, setting up their music on the radio or phone and turning their phone off.

Decision level self-regulation

Self-regulation at the decision level was measured using the item “During the trip did you interact with any type of portable device or onboard technology when driving (including when stopped at traffic lights)?” Responses were recorded as yes/no. As noted, there was a total of 514 device interactions across the 604 trips made.

The factors associated with the odds of drivers deciding to interact with devices while driving are displayed in Table 3. Higher scores on the Behavioural Regulation and Metacognition indices of executive function were associated with significantly higher odds of engaging with a device while driving. Younger drivers also had greater odds of engaging with devices when driving. Self-regulating at the planning level, having passengers present in the vehicle and longer trips were associated with lower odds of engaging with a device while driving. Drivers’ gender and trip purpose were unrelated to engaging with a device while driving.

Table 3 Factors associated with drivers’ self-regulation of device interaction at the decision level
Full size table

Control level self-regulation

Self-regulation at the control level was measured using two items: one measuring regulation with respect to changes in driving behaviour (“When interacting with the device, did you change your driving to keep safe?”) and one measuring changes in device interaction mode or style (“Did you change the way you interacted with the device to keep you safe?”). Responses to both items were coded as yes/no.

In terms of drivers reporting changing their driving behaviour when engaged with devices, neither the Behavioural Regulation nor Metacognition indices of executive function were related to changes in driving when interacting with a device (see Table 4). Indeed, trip duration and passenger presence were the only factors examined that were related to this control behaviour, with longer trips and not having any passengers present in the vehicle related to an increased odds of changing driving when interacting with a device.

Table 4 Factors associated with drivers’ control level self-regulation
Full size table

Drivers reported that they changed their driving behaviour once engaged with a device on 181 (35.2%) of the 514 recorded device interactions. The most common behaviour changes were slowing down (n = 82, 30.4%), followed by only interacting with devices when stopped at traffic lights (n = 78, 28.9%) and avoiding changing lanes or other risky driving manoeuvres (n = 40, 14.8%). Avoiding interacting with other devices (n = 33, 12.2%), increasing distance to vehicle ahead (n = 25, 9.3%) and pulling over to the roadside or parking space (n = 8, 3.0%) were other ways in which drivers modified their driving in order to reduce their risk when engaging with a device.

Drivers changed their style of interacting with devices once engaged on 386 (75.1%) of the 514 recorded interactions. The factors associated with the odds of drivers self-regulating their interaction with a device itself once engaged are displayed in Table 4. The Behavioural Regulation and Metacognition indices of executive function both had different relationships with self-regulation relating to changing the way drivers interact with devices. Drivers who scored highly on the Behavioural Regulation index were more likely to report that they change the way that they interact with devices in order to manage their risk. Conversely, drivers who scored highly on the Metacognition index were less likely to report changing the way that they interact with devices once engaged. Other factors that were related to changing device interaction included trip purpose and passenger presence, with work-related trips and having any passengers present in the vehicle related to an increased odds of drivers changing the way they interact with a device to minimise risk.

Drivers could report multiple ways in which they adjusted their interaction style, with keeping interaction with the device as short as possible being the most common (n = 285, 39.6%). This was closely followed by pausing interaction with the device to glance back at the road (n = 239, 33.2%) and pressing buttons without looking at the device (n = 143, 19.9%). Other control level adaptive behaviours included making a call instead of texting (n = 19, 2.6%), using alternative interaction modes (e.g., hands-free, voice operation) (n = 11, 1.5%), only taking ‘easy’ calls (e.g. those that do not involve decisions or emotion) (n = 10, 1.4%) and asking a passenger to interact with the device instead (n = 5, 0.7%).

Discussion

This study explored the relationship between drivers’ level of executive functioning and their self-regulatory behaviour in relation to interacting with technology while driving. Three levels of self-regulation were explored, ranging from the planning stage, through to deciding to engage and then behavioural adjustment once engaged. In line with much other research, our findings confirm that, across the age span, device interaction when driving is frequent, with over two-thirds of the trips undertaken including one or more device interactions (Dingus et al., 2016; Klauer et al., 2014; Young, Osborne, et al., 2019a). Also of interest, was the finding that the vast majority of device interactions were initiated by drivers, rather than being in response to the device providing notifications, messages or incoming calls. This finding is in line with Parnell et al. (2020), who found that almost 85 percent of device interactions were driver initiated and suggests that drivers are proactively seeking to engage with devices.

Our findings show that executive function plays an important role in distracted driving and its regulation at multiple levels. With respect to self-regulation at the planning level, drivers who scored highly on the Behavioural Regulation index (e.g., had poorer behavioural regulation) were more likely to report that, before starting a trip, they undertake actions to reduce the number of non-driving tasks they engage in. While this finding appears counterintuitive at first, it suggests that these individuals might have a level of awareness regarding their difficulty with inhibiting device engagement when driving in the moment and, thus, they seek to minimise the possibility that they will be exposed to devices or notifications once driving by undertaking strategic planning tasks such as muting notifications or storing the device out of reach. This is supported by the finding that people with worse (higher) behavioural regulation scores were more likely to report engaging with devices when driving, as discussed below. Given that the majority of reported device interactions were proactively sought out (driver initiated), this indicates that, once the opportunity to engage with devices while driving is present, individuals with poorer behavioural regulation struggle with inhibitory control or the ability to resist the temptation to engage.

We also found that drivers with poorer metacognitive functioning were less likely to engage in self-regulatory behaviour at the planning level. Individuals who have difficulties with metacognition typically experience issues with planning and managing future events and implementing appropriate steps ahead of time to carry out a task (Goldstein et al., 2014). These individuals may, therefore, have problems anticipating future device engagement when driving and, thus, are less likely to undertake preparations ahead of time to minimise future interactions. Further, deficits in executive functioning is also characterised by a failure to recognise or be aware of errors made during activities, suggesting that the association between poorer metacognitive functioning and lack of pre-planning in this study may reflect an inability to recognise the driving errors made when previously engaging with devices, resulting in these drivers being less likely to plan ahead to avoid similar errors in the future.

At the decision level, drivers who reported higher levels of executive difficulty engaged more often with devices when driving. This aligns with previous research, which found a significant correlation between executive function difficulty and frequency of engagement in a range of distracted driving behaviours (Hayashi et al., 2017; Mizobuchi et al., 2011; Pope et al., 2017; Sanbonmatsu et al., 2013; Zhang et al., 2020). These findings suggest that individuals with difficulties in executive function are more prone to engaging with devices when driving due to their inability to effectively manage attention and inhibit or regulate their behaviour. In particular, problems with inhibitory or impulse control, or the ability to inhibit or resist competing actions and stimuli and stop certain behaviours at an appropriate time, appear to play a significant role in drivers’ choice to engage with devices when tempted (e.g., Hayashi et al., 2017; Zhang et al., 2020). Further, the ability to sustain attention over time and stay ‘on task’ is an important component of executive function that underlies the ability of drivers to maintain focus on the driving task and refrain from engaging with devices.

Interestingly, we also found that self-regulation at the planning level was associated with a lower likelihood of engaging with a device once driving. This suggests that training in how to effectively pre-plan trips so that devices are already set up (e.g., navigation programmed before driving) or removed from the driver’s sight and reach may be a valuable tool to minimise device engagement once driving, particularly for drivers with poor impulse control, who may struggle to resist the temptation to interact.

With respect to self-regulation at the control level, drivers were more likely to report changing the way they interacted with the device itself to mitigate safety risk, rather than adjusting their driving behaviour. This may be because drivers have a greater level of control over their device interaction style, whereas changes in driving behaviour may be constrained by the prevailing driving conditions. Executive function had a less clear relationship with drivers’ self-regulation at the control level. Executive function was not related to drivers changing their driving behaviour when interacting with devices, while opposite relationships existed for the Behavioural Regulation and Metacognition indexes with changing device interaction. Drivers with poorer behavioural regulation were more likely to report changing the way that they interact with devices in order to manage risk. Conversely, drivers with higher levels of metacognitive difficulty were less likely to change their interaction style. The reason for these results is unclear, but it may be that once engaged, factors other than executive function take precedence in determining whether drivers try to minimise their risk, such as whether the interaction is work-related or if passengers are present in the vehicle.

Our finding that executive functioning plays an important role in distracted driving and its regulation enhances our understanding of the cognitive processes underlying this risky behaviour. Indeed, the findings highlight that there is much more to the cognitive processes underlying distracted driving than the sharing of attentional resources across tasks. Specifically, the results provide insight into the cognitive mechanisms underlying why some drivers may be less adept at resisting the urge to engage with technology while driving or be less able to adjust their engagement to mitigate risk once engaged. Importantly, problems with inhibitory control, planning, sustaining attention and monitoring one’s behaviour and failures appear to play a significant role in device engagement and its regulation at various levels. Practically, our findings could help to refine and target intervention strategies so that they better address the underlying causes of device engagement by drivers. Given that issues with impulsivity appears to play a role in technology use, interventions that target impulsivity-related behaviours, such as Cognitive Behaviour Therapy (CBT), may have some benefit in assisting drivers to control the urge to engage with certain forms of technology. Mindfulness-based approaches and training may also assist drivers who have problems regulating their attention, maintaining awareness and regulating their emotions. Indeed, research has shown that drivers who have higher levels of mindfulness are less likely to engage in a range of distracted driving behaviours (Feldman et al., 2011; Young et al., 2019a, 2019b, 2019c). Further research is required to determine the exact form of CBT or mindfulness training that will be most effective at reducing distracted driving and how to incorporate this into existing driver training and education programs.

Several limitations of the current study should be noted. First, the sample size was relatively small, which may limit the generalisability of the findings to different driver groups that have varying levels of driving experience and motivations for engaging with devices. For example, inexperienced drivers in their first year of driving who lack the self-regulatory skills to safely manage device use, or commercial or fleet drivers who may have additional pressures to engage with technology when driving. While each participant completed the required number of diaries, there was large variability in the number of diaries completed across participants. Based on reported driving patterns in the initial survey, it was clear that some participants only recorded a trip diary for a small subset of the trips driven across the four-week period. This means that individual differences in the propensity to engage with devices may have had more of an influence on the data for those drivers with a greater number of trips recorded. This could also cause an issue for data accuracy if those drivers who failed to complete diaries regularly were also those who engaged more or less frequently with devices. The reasons behind the limited diary completions for some drivers are unclear, but may be due to motivational issues (e.g., feeling like they could not spare the time to complete the diary for certain trips) or some participants forgetting to complete diaries despite receiving reminders.

Due to the inherent drawbacks with self-reported measures, the accuracy of measures of executive functions and distracted driving is reliant on the drivers’ self-evaluation and recall of their own behaviour. To reduce failures in recall, participants were requested to complete the diaries as close to the relevant trip as was practicably and safely possible, with the vast majority of the trip diaries completed within 6 h of the end of the trip. Participants were also assigned a unique code and assured of the confidentiality of the data to facilitate honest reporting of distracted driving behaviours.

Finally, the study was focussed on examining driver interaction with technology-based devices. However, there is a plethora of non-technology sources of distraction that drivers are exposed to (e.g., eating, drinking, talking with passengers) and there may have been trips where drivers were also engaging in non-technology related tasks that were not captured in the current diary data. In order to gain a more comprehensive picture of the impact of executive function on distracted driving behaviours, future work using trip diaries should consider requesting that drivers also record their engagement with non-technology based sources. This would allow for the exploration of possible interactions of everyday, non-technology tasks with device use, including whether engaging in non-technology based tasks means that drivers are more or less likely to engage with devices. In particular, understanding how the presence of passengers influences device use would be valuable.

Conclusions

The current study is the first known to the authors to examine the role of executive function in drivers’ self-regulation of distracted driving behaviours across a range of levels. Our findings showed that executive function, and, in particular, the processes of inhibitory control, sustained attention and monitoring one’s actions and behaviour, have important links to distracted driving and its regulation at the planning and decision levels. These findings provide a unique opportunity to target drivers who are more likely to engage in distracted driving behaviours and develop countermeasures that not only reduce engagement with devices, but to also facilitate and enhance the positive self-regulatory behaviours they already engage in.