This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Formative Research, is properly cited. The complete bibliographic information, a link to the original publication on http://formative.jmir.org, as well as this copyright and license information must be included.
The over-representation of youth in road crash injury and fatality rates is a major public health issue globally. In New Zealand, youth drivers are most vulnerable in the restricted license period when they can drive without the requirement for supervision by an experienced adult. Behavioral change interventions delivered using mobile phone technology to young drivers could serve as a useful mechanism to develop safe driving skills, but this potential remains to be fully explored.
This study aimed to apply behavioral change principles to design and develop a smartphone-based intervention with the aim of helping youth drivers to develop and hone safe driving skills.
An iterative process was used to support development of the smartphone intervention. We reviewed behavioral change literature, identifying fundamental principles and exploring use of behavior change techniques (BCTs) in other areas of public health. We engaged with key stakeholders, including young drivers, government agencies, and relevant organizations. We also took into account technology adoption considerations when designing the app.
We developed BackPocketDriver (BPD), an Android smartphone app that uses in-built sensors to monitor and infer driver behavior. The app implements features that were identified during the design process and are traceable to BCTs and theory. A key feature is messaging, which is used to instruct, motivate, educate, and relay feedback to participants. In addition, messaging addresses attitudes and beliefs. Other features include journey feedback summaries, goal setting, achievements, and leaderboards.
BPD’s design rests on a sound foundation of theory and evidence. With explicit links between theory and features, the app aims to be an effective intervention to change and improve youth driver behavior. The next phase of this study is to run a small pilot study to assess BPD’s effectiveness.
Road safety is a significant public health issue worldwide, with approximately 1.3 million fatalities and 20 to 50 million injuries per year, many of which lead to lifelong disabilities. Internationally, road traffic injuries are the leading cause of death among people aged 15 to 29 years [
According to recently published data [
Over time, young drivers tend to become safer. Drivers aged 16 to 19 years are 6 to 8 times more likely to crash than those aged 55 to 59 years, whereas for 20- to 24-year-old drivers, this drops to 3 to 4 times [
Recent and ongoing initiatives have made progress in tackling the youth driver problem. Such initiatives include legislation and graduated driver licensing (GDL), parental involvement to agree on protective limits on teen driving [
Smartphones offer a low-cost sensing platform that enables many facets of driver behavior to be monitored, including speed, acceleration, braking, and steering. These capabilities form a foundation for monitoring, analyzing, and providing feedback on driver behavior [
Used in the context of driving, the smartphone, nevertheless, is a double-edged sword. Using a mobile phone while driving is a key contributor to distracted driving, which claims the lives of 5000 Americans annually [
Interventions in many areas of public health have been based on behavior change techniques (BCTs) [
This study aimed to apply behavioral principles to design and develop a smartphone-based intervention, BackPocketDriver (BPD), with an aim to help youth drivers to develop and hone safe driving skills. Rather than inventing a feature-set based on intuition, we have reviewed behavioral theory, BCTs, and evidence of their effectiveness to develop an informed smartphone app. BPD represents a step toward developing youth-driving interventions that are more theory-led and grounded in evidence. As a result, we expect BPD to be more effective in changing behavior than other apps that are currently available.
BPD’s development was informed by engaging with key stakeholders, identifying appropriate techniques for behavior change, and relevant design principles for technology-assisted interventions.
This study does not report the outcomes for testing; however, these will be described in the paper on a pilot study of BPD.
The study was given ethics approval by the University of Auckland’s Ethics Committee in February 2016. Informed consent was obtained by participants before participation in the study.
Understanding and incorporating the priorities and preferences of the target audience as well as key stakeholders is important to ensure the success of interventions. Therefore, it is vital to engage with the target population during the design and development process [
In the case of BPD, 3 groups were identified as key stakeholders for engagement: young drivers (aged 16 to 24 years), parents of teen drivers, and relevant organizations. The organizations included New Zealand Transport Authority (NZTA), a government entity with a mission to develop a safe land transport system for New Zealand. New Zealand Police were included because of their prevention-first strategy that aims to reduce fatalities and serious crashes. In addition, the Automobile Association (AA) is an independent organization that is actively involved in initiatives for youth driver safety, including license reforms and young driver education.
Engagement with each of the key stakeholder groups was undertaken either in the form of semistructured phone interviews with representatives from the relevant organizations or less formal discussion-based sessions with young drivers and parents. Topics covered with each group of key stakeholders included understanding the issue—causes and implications of crashes involving young people, the level of interest in a smartphone app for safe driving, barriers to engagement with a driving app, preferred functionality and features, and incentives. When interviews were undertaken, these were recorded and transcribed. During sessions with teens and parents, information was captured on flip charts. Teens and parents also completed questionnaires.
Several models have been developed to explain behavior. Many of the models share common concepts, and awareness of the fundamental ideas is important in developing BPD.
Drawing on the theory of planned behavior, the dual-process approach, and the prototype willingness model [
Emotions can influence performance of target behavior. There are 2 emotion types:
Barriers are obstacles that can prevent a person from acting on his or her intentions to perform a target behavior. Intervention design involves helping people to anticipate and overcome particular barriers. Conversely, facilitators make it easier for people to perform target behaviors.
A person’s willingness or intention to perform a target behavior is governed by the following:
Behavioral models are useful in developing interventions in which they identify a range of psychological elements to address. However, it is often unclear how a particular element can be operated to bring about behavioral change [
A BCT is an observable, nonreducible component of an intervention that is designed to change behavior [
To ensure that the proposed intervention is effective in changing behavior, a review of the literature was conducted to identify relevant BCTs for incorporation into the intervention. We focused on Michie et al’s hierarchical taxonomy comprising 16 categories of 93 distinct techniques [
To illustrate a couple of BCTs,
Behavior change technique categories.
Category | Description |
1. Goals and planning | Setting goals for the target behavior, making plans to achieve goals, and dealing with any barriers |
2. Feedback and monitoring | Monitoring progress toward goals and providing feedback to users |
3. Social support | Providing social support, from friends, family, colleagues, and professionals to help meet goals |
4. Shaping knowledge | Assisting users to better understand their behavior and how to perform target behaviors |
5. Natural consequences | Highlighting consequences of performing particular behaviors, enabling users to see that they would regret not changing behavior |
6. Comparison of behavior | Comparing participants’ behavior with that of others and leading users to consider whether others approve (norms in a psychological model) |
7. Associations | Associating target behavior with positive things and reminding users to perform the behavior |
8. Repetition and substitution | Enabling users to practice and develop skills so that target behavior becomes habitual |
9. Comparison of outcomes | Allowing users to explore the outcomes of exhibiting or not exhibiting the behavior |
10. Reward and threat | Rewarding the target behavior and punishing unwanted behavior |
11. Regulation | Easing the task of performing the behavior, for example, by reducing negative emotions that result from the target behavior |
12. Antecedents | Understanding what triggers unwanted behavior, taking steps to avoid the triggers, and changing the physical environment |
13. Identity | Encouraging users to believe that the target behavior is right for them |
14. Scheduled consequences | Arranging a schedule of punishments and rewards for users performing the target behavior and not the unwanted behavior |
15. Self-belief | Building user confidence that a participant can perform the target behavior |
16. Covert learning | Enabling users to imagine consequences arising from performing a behavior and observing the consequences to others as they perform behaviors |
Behavior change techniques that have generally featured in successful interventions.
Behavior change technique | Description |
1.1 Goal setting (behavior) | Set or agree a goal defined in terms of the target behavior |
2.2 Feedback on behavior | Monitor or observe behavior and provide feedback on performance of the target behavior |
2.3 Self-monitoring of behavior | Establish a method for the person to monitor and record their behavior(s) |
3.1 Social support | Advise on, arrange, or provide social support or encouragement for performing the target behavior |
4.1 Instruction on how to perform the behavior | Agree or advise on how to perform the target behavior |
5.1 Information about health consequences | Provide information about the consequences of performing the target behavior |
BCTs have been widely used in many areas of public health.
Developing youth driving interventions that are informed by behavioral change theory has largely been ignored [
At the time of writing, we were not aware of any smartphone-based interventions for youth driving that have been designed with consideration of behavioral change theory. However, later in this paper, we report on popular apps for youth driving and identify features that can be traced back to particular BCTs.
Developing technology-assisted interventions is not without challenges. As discussed earlier, participants must be willing to engage in the intervention. With a smartphone-based intervention such as BPD, they must also be
Technology acceptance can further be affected by 4 attributes [
A summary of the learnings from the stakeholders is presented in
Young drivers raised a number of risks relating to privacy. Teens did not want their parents to be able to track their movements or to receive real-time alerts of poor driving behavior. Some teens also raised concerns about the data being made available to authorities and used, for example, to issue speeding infringement notices.
Quality of feedback was an important concern raised by young drivers. Youth drivers wanted reassurance that any feedback would be useful and effective. In addition, they felt that they would be stressed by negative or
Young drivers also expressed
The Police identified that youth who are interested in using the app are unlikely to be those who engage in criminal behavior. There exists a correlation between criminality, antisocial behavior, and car crashes, with risk taking and poor decision making being contributory factors. Appealing to this demographic subset, given its
Constructive feedback was important to young drivers. Although negative feedback poses a risk, feedback that is encouraging and addresses both good and bad driving, allowing users to discover what they are doing wrong, was viewed as something that young drivers could
Parents were interested in monitoring both routes driven and driving behavior of their teens. Similar to young drivers, parents saw value in the app providing their teens with driver education and instruction. Parents favored automatic deactivation of the phone while driving; however, teens wanted this tempered, for example, to be aware of when a short message service (SMS) text message had arrived but having to stop the vehicle before reading the message.
Key findings from stakeholder engagement to inform app development
Finding category | Young drivers | Parents of teen drivers | Relevant organizations |
Risks to adoption | Threats to privacy; Negative or inaccurate feedback on driving; Battery and mobile data consumption; Excess use of push notification or audio alerts; Cost | App being used as a source of distraction; Appeal of app to most at-risk drivers; Potential to subvert intervention; Cost | |
Gain enablers | Constructive feedback; Safe driving education; Peer competition | Ability to monitor teens’ driving and behavior; Automated deactivation of phone while driving; Suggestions to improve driving | Sticky intervention; Data analytics based on crowd-sourced data |
Incentives | Recognition of achievements; Use of app data as proof of safe driving; Endorsement by relevant organizations, for example, NZTAa; Esthetics and ease of use | Integration with licensing process | Material rewards schemes, for example, fuel discounts; Automated starting or stopping of journey monitoring |
aNZTA: New Zealand Transport Agency.
Mapping objectives to behavioral elements and behavior change technique categories.
Objective | Behavioral elements | Behavior change categories |
1. Improve driving skills | Emotions; Control; Barriers | 2. Feedback and monitoring; 4. Shaping knowledge; 7. Associations; 8. Repetition and substitution |
2. Strengthen intentions to perform target behaviors | Facilitators | 1. Goals and planning; 10. Reward and threat; 13. Identity |
3. Increase positive attitudes toward performing target behaviors | Attitudes; Norms; Barriers; Facilitators | 3. Social support; 5. Natural consequences; 6. Comparison of behavior |
4. Manage self-identity | Self-identity | 4. Natural consequences |
5. Address the mismatch between perceived and actual driving skills | Facilitators | 13. Identity |
The AA reported that conventional driver training decays over time. Conversely, a driving app has the potential to remain supportive and of value to youth over time. To do so, it needs to be
BPD has 3 target behaviors for young drivers:
To drive within speed limits.
To perform maneuvers safely and in a controlled manner.
To not use a mobile phone while driving.
These behaviors lend themselves to BPD’s smartphone-based delivery platform because they can be automatically tracked by the smartphone. On the basis of the gathered data, the app can generate tailored responses to help participants develop the wanted behaviors.
We have identified 4 objectives from the target behaviors. For each objective, we have identified the relevant behavioral model elements to operate on. In selecting particular BCTs for each objective, we have considered which BCTs have been proven to work in other interventions. In addition, we have considered which BCT categories are best placed to meet particular objectives. To increase skills, for example, category 8 (repetition and substitution) is appropriate, whereas category 5 (natural consequences) is well suited to changing attitudes [
For objective 1, to improve driving skills, BCT category 2 (feedback and monitoring) plays a key role by offering BCTs that can be used to monitor driving behavior and relay feedback to participants. On the basis of feedback, areas to focus on can be identified, enabling participants to practice and improve on these aspects. Category 4 (shaping knowledge) can be employed to assist with improving skills through BCTs that educate participants, for example, by providing instruction on how to perform maneuvers and antecedents to performing the target behaviors poorly. Category 7 (associations) includes BCTs for prompting wanted behavior at particular times. BCTs from category 8 (repetition and substitution) help with honing target behaviors through practice. They also facilitate formation of good habits.
A person’s emotional state can affect his or her driving behavior. As discussed earlier, experienced feelings, such as being upset, can negatively impact performance of target behaviors even when a person has strong intentions and a positive attitude toward the target behaviors. Category 4 is of further value for objective 1 in which it has BCTs that can be used to educate participants in recognizing and managing emotions. Moreover, category 4 can be used to raise participants’ awareness of barriers to performing wanted behaviors, for example, poor time management, drugs and alcohol, tiredness, and phone use. In shaping knowledge, the intervention can suggest how to deal with barriers.
A necessary element to improving driving skills is self-belief—participants must believe that they are capable of performing the target behaviors. BPD can strengthen participants’ self-belief through applying BCTs in category 8. This category includes a BCT for graded tasks, where tasks become more difficult over time. As participants work through a grade or level, they become more proficient and prepared for the next. In terms of driving to speed limits, for example, successive levels might lower the speeding tolerance for achieving a speed-focused goal.
Regarding the second objective, to strengthen intentions to perform target behaviors, we recognize that while many participants have a positive attitude toward the target behaviors, without goals they might lack the impetus to engage and develop the wanted behaviors. BCT category 1, goals and planning, is appropriate to draw on as it provides BCTs for participants to set and track progress with goals associated with target behaviors.
Category 10, rewards and threats, can also be used to incentivize participants. Social rewards recognize that participants have performed a target behavior well and provide a sense of achievement. Similarly, category 13, identity, includes a role-modeling BCT where a participant can be elevated to a role model after performing well in a target behavior. This can bring a sense of kudos to the participant, fostering their motivation and engagement.
Although attitudes of many young people align closely with safe driving, there are others who hold less positive attitudes toward BPD’s target behaviors. Hence, for some participants, the intervention needs to change their thinking (instrumental attitude). This is the motivation for objective 3 to increase positive attitudes toward performing target behaviors. BCT category 5, natural consequences, includes BCTs that can be applied to help participants see the consequences of performing wanted or unwanted behaviors. Related to consequences is the notion of anticipated regret, which involves having a participant think about how they would feel if they did not change their behavior and continued to perform an unwanted behavior, for example, speeding. Thinking through an undesirable outcome may contribute to change in instrumental attitude.
Descriptive norms influence attitudes. To show that it is normal for other young people to perform the target behaviors, BCTs from category 6, comparison of behavior, can be used. Social comparison involves bringing to the attention of participants other participants who they consider to be part of the same social group and who are performing the target behaviors well. Category 6 also includes BCTs for addressing the approval of others (injunctive norms). In addition, where celebrity figures who are respected by young people endorse the target behaviors, this might also contribute to changing attitudes and meeting objective 3.
BCT category 3, social support, is also appropriate to consider for the third objective. The BPD app could include social networking functionality allowing participants to support one another in developing the target behaviors. Another supportive role for BPD would be to address barriers to performing the target behaviors. Barriers include peer pressure and triggers, for example, racing or using mobile phones while driving. The app could deliver advice on how to deal with such barriers.
Objective 4, to manage self-identity, recognizes that a person’s attitude might be opposite to the target behaviors. For BPD,
The final objective concerns the mismatch between perceived and actual driving skills. Young drivers tend to overestimate their safety margin, resulting in more risk taking [
Having identified the subset of BCT categories that are applicable to BPD,
Relevant behavior change techniques (BCTs). There is strong evidence that the BCTs shown in italics (outlined in
Behavior change technique | Description |
Mutually agree on short-term goals to be achieved, such as “This week I will brake more gently.” | |
1.2 Problem solving |
Prompt participants to analyze behaviorally influencing factors and develop strategies for overcoming barriers. For example, “So it seems you’ve been having trouble with your speed. How do you think you could try to change that next time you go out?” |
1.3 Goal setting (outcome) | Facilitate longer-term goals, such as “Be a safe driver,” “Get my full license,” and “Avoid accidents.” |
1.4 Action planning | Prompt participants to plan their driving, including factors such as context, frequency, and duration. |
1.5 Review behavior (goals) | Review behavioral goals together with the participant and consider modifying them based on progress. For BPDa, goals can be reviewed and modified by the app. |
2.1 Monitoring without feedback | Record behavior with the participant’s knowledge. Driving behavior data captured by the app could be made available to a participant’s parents. The knowledge that their driving behavior is being observed can influence their behavior. |
Monitor and provide informative feedback on performance. BPD could provide feedback in terms of poor driving behavior, suggestions on how to improve, and recognition of good behavior. | |
Establish a method for participants to monitor their own behavior. BPD could provide the ability to review earlier feedback and to identify behavioral trends. | |
2.7 Feedback on outcomes | After periods of prolonged safe driving, BPD might inform participants that they are now statistically less likely to be involved in an accident than when they started the intervention. |
Arrange for participants to receive support from others. In BPD, this could take the form of a social network connecting participants and friends. | |
Provide advice on how to perform a behavior. BPD could present how-to messages, describing techniques, and practices that help participants to perform the target driving behaviors. | |
4.2 Information about antecedents | Provide information about situations, events, or emotions likely to cause poor performance of the target driving behaviors. |
Provide information about the positive or negative health consequences of wanted or unwanted behavior. BPD could deliver messages concerning the benefits associated with target behaviors. | |
5.2 Salience of consequences | Use methods to emphasize consequences for |
5.5 Anticipated regret | Have participants imagine how regretful they would feel if they perform unwanted behavior, for example, speeding and something negative happens. |
6.2 Social comparison | Draw attention to performers of good behavior to allow comparison with a participant’s own performance. For example, BPD could maintain a |
6.3 Information about others’ approval | Provide information about what other people think about good and bad behavior. BPD could provide informational messages about the negative social perception of unsafe drivers (or vice versa). |
7.1 Prompts or cues | Introduce stimuli to encourage good behavior. BPD might provide NFCb sticker tags that participants can place in their vehicles to remind them to use the app and put their phone away. |
8.3 Habit formation | Prompt rehearsal and repetition of good behavior in the same context repeatedly, so the context elicits the behavior. Having finished using BPD, participants should continue to perform the target behaviors they have developed habitually. |
8.7 Graded tasks | Set easy tasks and then gradually make them harder as participants improve. BPD could offer goals at varying difficulty levels and ensure that participants make progress through the more challenging goals. |
10.1 Material incentive (behavior) | Inform participants that a material reward ( eg, money or vouchers) will be given in exchange for demonstration of the target behavior. BPD might seek partnership with businesses and organizations to provide such rewards. |
10.4 Social reward | Similar to 10.1, but rather than a material incentive, the incentive would enhance a participant’s standing in some way. Performing target behaviors in BPD could earn participants achievements. |
10.11 Future punishment | Inform participants that punishment or loss of reward occurs if poor behavior continues. BPD might simply raise awareness of legal or social punishments in response to detecting prolonged poor driving behavior. |
13.1 Identification of self as role model | Inform participants that their good behavior is an example to others. BPD could promote demonstrably safe drivers to others, offering a level to aspire to. |
13.3 Incompatible beliefs | Draw attention to discrepancies between current or past behavior and self-image to create discomfort. BPD could use messaging to highlight differences in actual versus perceived driving skills and incompatible beliefs over driving practices. |
aBPD: BackPocketDriver.
bNFC: near field communication.
In addition to the BCTs for which there is strong evidence that they have led to behavioral change in other interventions, BCTs 4.2 information about antecedents, 5.5 anticipated regret, 7.1 prompts or cues, 8.3 habit formation, and 13.3 incompatible beliefs are seriously worth considering because they are founded in behavioral change theory [
On the basis of the design considerations, as discussed above, we have identified several features for the BPD smartphone app. The features are informed based on the selection of BCTs that are appropriate for the intervention. Each feature is described below.
BPD uses social rewards
Goals are fundamental to BPD
On the basis of a participant’s prior driving performance, BPD suggests particular goals that users can modify in terms of difficulty. Users are encouraged to choose more difficult goals as their driving performance improves
At the conclusion of each driving episode, users can review their performance
In addition to postjourney feedback, users have the opportunity to view feedback on previous journeys
BPD makes liberal use of messaging as a means to meet many of the intervention’s objectives introduced earlier. Messages serve many purposes, including providing information relating to instruction, consequences, antecedents, anticipated regret, feedback, other’s approval, and incompatible beliefs.
Messages are generally framed in terms of gain as opposed to loss, which has been shown to be more effective in leading to behavioral change [
BPD displays messages to users at different times in response to different stimuli.
Users can connect with elected friends who are also using the app, facilitating social support
The app operates a leaderboard with which users can compare their own progress with that of other participants. This facilitates social comparison
Smartphones are capable of detecting many aspects of the driving environment, for example, time of day, type of road, and prevailing weather. Detecting driving conditions is a feature that enables automated generation of a driving log, including hours spent driving on different road types. Such logs are a requirement for learner drivers in some jurisdictions. Automated logging protects against the possibility of fraudulently entered manual log entries. As part of the stakeholder engagement, the NZTA viewed logging positively.
BPD implements near field communication (NFC)–initiated journey monitoring. Participants stick an NFC tag on their dashboard and swipe their phone over the tag to commence monitoring. The tag additionally serves as a cue
An alternative approach would be to automatically detect, without any participant action, the start of a journey. This would promote usability and would also ensure that all journeys are monitored. However, automated detection requires the device’s accelerometer to be activated at all times, which causes significant battery drain. As this sort of interference was seen as a risk to adoption by the target demographic, BPD does not implement the automated detection.
Detecting driver behaviors other than speed and smoothness, for example, following (stopping) distances, is possible by using additional mobile phone sensors. This is discussed further in the Discussion section.
By offering material incentives
Parents showed interest in being able to monitor their children’s driving behavior, although the target demographic viewed parental involvement as a risk to adoption and use of BPD. A parental interface has not been implemented.
Sample messages derived from objectives and behavior change techniques (BCTs).
Message | Objective | BCTa |
1 | 4.1 | |
1 | 4.1 | |
1 | 4.2 | |
1 | 4.2 | |
1 | 4.2 | |
1 | 4.2 | |
1 | 7.1 | |
3, 4 | 5.1 | |
3, 4 | 5.5 | |
3, 4 | 6.3 | |
4 | 5.5 | |
4 | 4.2 | |
5 | 13.3 |
aBCT: behavior change technique.
Feature wish list for BackPocketDriver (BPD).
Feature | Behavior change techniques applied | |
Location and speed detectiona | —b | |
Acceleration, braking, and turning detectiona | — | |
Phone usage detectiona | — | |
Achievements | BCT 10.4 |
|
Goal setting | BCT 1.1 |
|
Journey summaries | BCT 1.5 |
|
Messaging | BCT 2.7 |
|
Journey detection | BCT 7.1 |
|
Friends | BCT 3.1 |
|
Leaderboards | BCT 6.2 |
|
Detection of driving conditions | — | |
Additional driving behavior detection | — | |
Rewards scheme | BCT 10.1 |
|
Parental interface | BCT 2.1 |
aNecessary for target behaviors: 1
bNot applicable.
Before starting a journey, users choose goals to work toward, for example,
Having processed the data, the Web service sends feedback to the BPD app. Upon receipt, a notification appears in the device’s notification tray. Users click the notification or navigate back into the app to view the journey summary (
All messages generated by BPD are viewable at any time on the screen shown in
Screenshots of the BackPocketDriver app.
The next step is to run a small study to assess BPD’s potential for effectiveness in developing safe driving skills among youth. A before-after study is currently underway involving 20 participants, aged 16 to 24 years, and on their restricted or full license. Participants are monitored using a minimal BPD app for 1 month to classify their driving behavior. They then switch to the full BPD app that includes the behavioral-change feature set for a second month. Following the study, any change in driving behavior will be identified based on app-generated data, and participants will have an opportunity to provide feedback on the intervention.
This study outlines the design of a smartphone-based intervention for developing safe driving skills among youth drivers. Although other researchers have investigated the use of smartphone technology in monitoring driver style and behavior, work that has sought to
There is a plethora of driving-related apps available from app stores. Many apps target a particular aspect such as preparing for licensing theory tests (eg, Theory Test Kit and New Zealand Driving Theory Test), logging journeys (eg, DrivePad), and blocking messages, calls, and notifications during driving (eg, DriveMode, Shut Up and Drive, and Safe Ride). There are also apps that, similar to BPD, aim to assist young people to develop safe driving practices. We have selected 6 popular apps that appear to have overlapping objectives with BPD and which offer more than simple blocking or logging functionality.
For each app, we have examined its features and identified any BCTs to which features are attributable.
LifeSaver is a blocking app that automatically silences a user’s smartphone on detecting driving. Journey feedback (BCT 2.2) is limited to reporting on the unwanted behavior of using the phone, for example, to text while driving. At the end of a journey, the app displays a percentage score where 100 indicates that the user did not use their phone while driving. LifeSaver supports a family view, which is essentially a leaderboard that ranks the family members according to their scores. As the leaderboard shows each family member’s score, teens are aware that their parents are monitoring them (BCT 2.1). Through location tracking of family members, the app facilitates social support (BCT 3.1) as users can see when their family members are driving and defer calling them until they have finished their journey.
TrueMotion Family is similar to LifeSaver in that it is also a family-oriented app with a leaderboard that publishes each family member’s driving score. In addition to phone use, for example, texting and calling, TrueMotion Family also factors aggressive driving and speeding to generate a user’s score. The app pinpoints unwanted behavior events on a map allowing the user to see where the events occurred; it also allows users to review their driving behavior over time (BCT 2.3).
Behavior change technique feature matrix for popular youth driving apps.
Behavior change technique | LifeSaver | TrueMotion Family | Mojo | DriveSmart | EverDrive | Steer Clear | |
1.1 Goal setting (behavior) | |||||||
1.3 Goal setting (outcome) | ✓ | ||||||
2.2 Feedback on behavior | ✓ | ✓ | ✓ | ✓ | ✓ | ||
2.3 Self-monitoring of behavior | ✓ | ✓ | ✓ | ||||
3.1 Social support | ✓ | ✓ | ✓ | ||||
4.1 Instruction on how to perform the behavior | ✓ | ✓ | |||||
5.1 Information about health consequences | |||||||
4.2 Information about antecedents | |||||||
5.5 Anticipated regret | |||||||
7.1 Prompts or cues | |||||||
8.3 Habit formation | ✓ | ✓ | ✓ | ✓ | ✓ | ||
13.3 Incompatible beliefs | |||||||
6.2 Social comparison | ✓ | ✓ | ✓ | ✓ | |||
6.3 Information about others’ approval | |||||||
10.4 Social reward | ✓ | ✓ | ✓ | ✓ | |||
13.1 Identification of self as role model | ✓ | ✓ | ✓ | ✓ | |||
1.8 Behavioral contract | ✓ | ||||||
2.1 Monitoring of behavior by others without feedback | ✓ | ✓ | ✓ | ||||
10.2 Material reward | ✓ | ✓ | ✓ | ||||
10.11 Future punishment | ✓ |
Mojo, similar to LifeSaver and True Motion Family, employs a leaderboard that ranks teens among their friends based on points earned while driving. As with LifeSaver, scoring is based on unwanted phone use alone. LifeSaver breaks down feedback, for example, providing a count of swipes and taps that a user makes on their phone while driving. Mojo differs by employing material rewards (BCT 10.2). Users who have amassed high scores are invited to spin a wheel for the chance to win a voucher. Mojo’s feedback, rather than being limited to a score, also offers tips for improving behavior. For example, if the user has been making calls while driving, Mojo displays a message to tell the user that they will improve their safety and score by not making calls during future journeys.
DriveSmart monitors driver’s behavior and generates a percentage score based on their braking, cornering, and speeding. Similar to TrueMotion Family, DriveSmart plots driving events on a map and allows users to review their behavior over time. Similar to Mojo, DriveSmart has rewards partners and can offer material rewards in exchange for good driving behavior. Unlike the above apps, DriveSmart does not offer any collaboration features such as a leaderboard; instead, it is intended to be used by an individual and not in a group context. As part of feedback, DriveSmart uses loss-framed messages, alerting users to future punishment (BCT 10.11), for example, for speeding.
EverDrive has a feature set similar to TrueMotion Family. It monitors a driver’s acceleration, braking, cornering, speed, and phone use. Instead of providing feedback through percentage scores, it uses a 5-star scheme.
Unlike the above apps, Steer Clear does not monitor or provide feedback on driver behavior. It includes logging functionality that allows individuals to record their driving hours in different conditions. In addition, it has unique features: a behavioral contract (BCT 1.8), goal setting (BCT 1.3), and videos to share experiences of other users, a form of social support (BCT 3.1). When a user starts using the app, they make a pledge to drive safely; the pledge forms the basis of a behavioral contract. In using Steer Clear, users work toward the outcome goal of completing the set of Steer Clear modules. Once complete, users are eligible for insurance discount (BCT 10.2).
Many of the apps are group oriented involving family members and/or peers. They include a leaderboard feature, and the way that leaderboards are used is effective in exercising several BCTs. The leaderboards allow social rewards (BCT 10.4) in the form of stars and percentage scores to be publicized to the group, facilitating social comparison (BCT 6.2). They also enable higher scorers to identify themselves as role models (BCT 13.1). In addition, the leaderboards make users aware that they are being monitored by other group members in a way that does not involve feedback (BCT 2.1). Of these, BCTs 10.4 and 13.1 are particularly appropriate for strengthening intentions to perform wanted driving behaviors.
Of the apps that offer feedback on behavior, they do so in the form of a numeric score. Only Mojo and DriveSmart include textual feedback to supplement scores, and even here, the messages do not address health consequences, antecedents, anticipated regret, or incompatible beliefs—that either are proven or theoretically informed BCTs. Furthermore, none of the apps employ goal setting for behavior (BCT 1.1), which is a proven BCT. Similarly, instruction on performing wanted behaviors (BCT 1.4), another BCT for which there is strong evidence that it is effective, is employed very sparsely.
The apps have limited support for increasing positive attitudes toward wanted driving behaviors. The leaderboard feature, linked to BCT 6.2 (social comparison), can help address norms and demonstrates to a teen that others in their social group do exhibit the wanted behaviors. However, many of the other BCTs discussed earlier for addressing attitudes, managing self, and dealing with the actual or perceived skills mismatch are not associated with the surveyed apps’ features. Hence, it seems unlikely that the surveyed apps can lead to long-term behavioral change.
In recent years, much work has been conducted to validate use of smartphones in providing a low-cost sensing platform and to supersede the older in-vehicle data recorder (IVDR) units that necessitate a fixed installation [
Today’s smartphones include inertial sensors that enable smartphone driver support systems (SDSS) to detect driving events such as acceleration, braking, turning, and lane changing [
Beyond a smartphone’s inertial sensors, other in-built sensors include cameras and microphones that are being used to detect whether drivers are drowsy or distracted. CarSafe uses both the forward- and rear-facing cameras to monitor the driver’s face and eyes along with the road ahead [
In an early SDSS study [
Parental involvement is a contentious issue for SDSS. Key findings for IVDR systems that involve parents, for example, [
Gamification, the use of game elements in nongame contexts [
With mobile phones known to be a source of distraction when driving, the role of SDSS in blocking incoming calls and text messages has been investigated. A study involving teen drivers [
BPD’s feature set has been derived through the application of behavioral theory and BCTs and consideration of the evidence relating to BCT effectiveness elsewhere. Beyond monitoring and classifying driver behavior, which is where much of the existing work in SDSS stops, BPD’s design includes a suite of features that have a clear mapping to distinct BCTs. A key feature is postjourney feedback. Messaging is also employed for many purposes: to instruct, motivate, educate, and relay feedback to participants and to address participants’ attitudes and beliefs. BCTs and gamification are complementary; BPD’s design combines elements of gamification with BCTs in offering goal setting and review and achievements and leaderboards. Leveraging the smartphone sensor platform, the design also allows for monitoring of additional facets of driving behavior and automated detection of driving conditions.
BPD’s design has also taken into account technology acceptance considerations. The app enables youth to perceive gains through constructive feedback, education, and social comparison or competition. These features further contribute to addressing delay discounting and social influence in that they help to retain user engagement and interest over time, not only for individuals but also for peer or social groups.
Key risks have been mitigated. Unlike other SDSS and IVDR systems and apps, BPD does not present percentage scores or include raw or absolute figures for acceleration and speed when providing feedback. This ensures that BPD cannot be used to subvert the intervention, for example, by teens using the app to record and share race times and dangerous driving events. In addition, BPD allows users to control how they share their data in the interests of privacy. Moreover, the app conserves device resources—using no mobile data and minimizing power consumption—so as to have minimal perceived effect on users’ smartphones for daily operation. Furthermore, BPD does not offer real-time feedback, contrary to many SDSS and IVDR systems. This is both to avoid distraction, which has been linked to real-time feedback in other studies [
Regarding usability, the app is easy to use. It requires no calibration before use, and during a journey the smartphone can be kept in the driver’s pocket (the app functions accurately without requiring a fixed dashboard mount). The app does not provide a completely seamless experience in that it does not detect the start of a journey and begin monitoring automatically. Implementing this feature would increase the risk that youth would not use the app because the necessary power-draw would interfere with the normal operation of the smartphone. This represents a conflict between requirements; a key conflict among stakeholders is the issue of parent involvement versus youth privacy. In developing BPD, we have sought to minimize risks to intervention adoption.
BPD is complementary to fundamental research aimed at understanding the underlying issues that contribute to youths’ over-representation in crashes. In Shope et al’s study [
BPD is a smartphone-based intervention that aims to improve driving skills in youth. Critically, BPD’s design has been informed by behavioral theory and behavioral change expertise. Stakeholder feedback and technology acceptance considerations have also been factored into the design. Having implemented the app on a sound theoretical foundation, the next step is to evaluate its potential to be effective in changing youth driving behavior. A small study involving 20 youth participants is currently underway, and we expect to report on the results in the near future.
Automobile Association
behavior change technique
BackPocketDriver
graduated driver licensing
in-vehicle data recorder
near field communication
New Zealand transport agency
smartphone driver support system
short message service
user interface
This study was funded by the University of Auckland, Faculty Research Development Fund Post-Doctoral Fellowship project 3706763, “A Smartphone-Based Intervention for Promoting Behavioral Change in Young Drivers.” The authors would like to thank TomTom for supporting our research. Use of the TomTom Web service for accessing speed-limit data without restriction has been necessary to run the ongoing pilot study.
None declared.