Steer Clear From Distracted Driving

Hello, my name Isla Hatch. My science fair project is called Steer Clear from Distracted Driving.

A virtual driving simulator was built, and participants were challenged to race while sending and receiving SMS messages. Driving ability was evaluated with respect to race time, crash frequency, and off-road frequency. Results showed that text messaging negatively affects driving ability. As the messages got more complex, drivers showed an increase in crash events and drove off the road more frequently.

This project was designed to make more people aware of the dangers of texting and driving and hopefully improve the safety of Canadian roads. This project originally targeted children, but it was expanded to include adults making it more realistic and applicable.

Although the experimental phase produced insightful conclusions, the results from the expanded study are in the data collection phase. Complete results will be analyzed and presented during the 2026 Canada Wide Science Fair in Edmonton, AB.

This project investigated the effect of sending and receiving text messages while driving. Forty-seven percent of Canadians admit to using their mobile phones while driving (Blair Nicole, 2026). According to Transport Canada, the percentage of fatal collisions caused by distracted driving have hovered around 20% between the years 2017 and 2023 (Transport Canada, 2025). This project was designed to quantify the effect of texting as a distraction to driving.  It is important for society to realize the dangers of texting and driving.

This project was designed to make more people aware of the dangers of texting and driving and hopefully improve the safety of Canadian roads. This project originally targeted children, but it was expanded to include adults because adults have experience with driving in the real world, making it more realistic and applicable.

The project consisted of two parts; the design phase of building a driving simulator, and the experimental phase, testing how texting impacts driving ability.

The design phase consisted of starting with the goal of converting an existing go-cart into a virtual driving simulator. A Nintendo Switch system was used to provide the driving simulation. To build the go-cart, individual steps were used. Cleaning and modifying were followed by painting and attaching electronic technology. Each step required problem identification, brainstorming of possible solutions, modeling potential solutions, and testing the solutions. The virtual driving simulator consisted of the modified go-cart frame, a 27” MSI computer monitor, a paired set of Mario Kart pedals and steering wheel, a Nintendo Switch with charging port, a set of headphones, and a pair of Apple iPhones.

Many variables were held constant in order to focus on the responding variables. The race course, the cart and character type, and the phones used were held constant across all trials. A standard set of SMS messages were used.

The experimental phase consisted of a series of 5 races. For each race, participants were challenged to race while receiving SMS messages of increasing complexity. Driving ability was evaluated using the factors of race time, off-road frequency, and crash frequency. The data was collected through the cooperation of two people; the first being responsible for sending the SMS message, and the second being responsible for counting the frequency of offroad events, crash incidents, and race time. In the initial study, the results of 6 child participants were studied. In the expanded study, 68 experienced drivers were tested; a more statistically appropriate number.

Background research regarding distracted driving was conducted. Trustworthy websites were identified by selecting recognized authorities on the subject, such as Transport Canada.

The design phase of the project involved building a prototype virtual driving simulator. A number of engineering design challenges and fabrication constraints were identified and overcome. Overall, the prototype worked very well. The controls were very intuitive for those with driving experience. Some of the challenges included accessibility difficulties. It was hard for some participants to get in and out of the kart. Also, the video game platform was initially challenging for those participants with little or no gaming experience. A further challenge was related to moving the simulator between locations. A monitor was actually broken and needed replacement during one of these movements.

The experimental phase of the initial project yielded several results. With respect to analysis, a graphical analysis was used to make conclusions during the initial phase of this investigation. The hypothesis was correct. According to graphs and observations, texting while driving negatively affected the ability of participants to drive. Texting while driving increased race times. Sending and receiving SMS messages also caused more frequent off-road events and more frequent crashes. This is logical because many test subjects took their eyes off of the road and their hands off the wheel to read and answer the text messages. Removing hands from the wheel makes it difficult to steer around corners and obstacles. Taking your eyes off the road makes it difficult to recognize surroundings and terrain. Both negatively affect one’s ability to drive well.

Results from the expanded study are in the data collection phase. Results will be analysed for presentation during the Canada Wide Science Fair (CWSF) in Edmonton, AB during the week of May 23 - 30, 2026.

As part of the extended study, a more in-depth statistical analysis will also be used to make conclusions. Data will be analyzed using separate models customized for the dependent variables. Race time will be analyzed using a linear mixed-effects model (LMM). Course exits and crash frequency will be analyzed using a mixed-effects count model (MEM).

LMM takes into account the repeated measurements collected from each participant. Each driver completed multiple races under different texting conditions, therefore, the observations were not independent. LMM is appropriate because race time changes smoothly (small increases or decreases). The model estimates how average race time changes with texting difficulty and accounts for the fact that some drivers are naturally faster and some are naturally slower.

MEM is appropriate for crash frequency and off-road frequency because these were measured using count data, unlike the continuous metric of race time. This model deals well with whole numbers (0, 1, 2, 3…) and can manage some of the races with zero exit crashes. Count models correctly handle skewed data, zero values and unequal ability between drivers. A negative binomial MEM will be used rather than a Poisson MEM because some drivers exit the course much more often than others (very common in behavior data). This model is designed to estimate how the rate of course exits changes with texting difficulty and controls for repeated measurements.

There are several implications related to the results of this project.

First, the results indicate that SMS distraction has a negative impact on driving ability. The main take-away is that drivers should acknowledge this fact and adjust their driving accordingly. The average person knows this behaviour is dangerous, but having a simulation experience gives drivers a genuine reason to stop using technology during driving. Actual acknowledgment would help them to resist the urge to engage in communication-based distraction while driving. This would allow them to improve the safety of driving, possibly saving lives and/or avoiding injury.

Next, passengers in other cars would benefit because enhanced driver awareness can keep passengers safe. The person in the passenger seat should be the designated texter. By assuming the role of communicator, the driver can be left with the sole responsibility of operating the motor vehicle.

Vulnerable Road Users will benefit by being able to depend on all drivers to pay more attention. Pedestrians and cyclists accounted for more than 621 deaths due to distracted driving in the US in 2022 (NHTSA, 2026). If there was more awareness about the dangers of texting while driving, it would reduce the number of fatalities related to distracted driving.

There are several next steps. Repeating the experiment using a more realistic car simulator could appeal to a greater audience.

Mental distraction (math questions) could be used rather than text-based questions, and drivers would experience a different degree of disruption.

To facilitate data collection, an automated SMS system would be beneficial. By relying on a computer/AI system to send SMS messages, consistency of arrival time would be assured.

This experiment is currently under expansion and a much larger sample size is being collected. Additionally, the number of years of driving experience is also being cataloged.

During this project, many people supported me and my project. Thanks to my dad who helped me build the go kart. Thanks to my mom who helped me with a lot of paperwork and technology. Thanks to my sister who helped me with counting crashes and off roads while testing. Thanks to my grandparents who generously donated a monitor. Thanks to my math and science teacher who is also my scientific supervisor. Thanks to all my participants who gave me their time for testing. Thanks to the Bluewater science fair committee.

Blair, Nicole, 2026, Texting and Driving Statistics, madeinca.ca,

https://madeinca.ca/texting-and-driving-canada-statistics/

Transport Canada. (2025, May 15). Canadian Motor Vehicle Traffic Collision Statistics:

2023. https://tc.canada.ca/en/road-transportation/statistics-data/canadian-motor- vehicle-traffic-collision-statistics/2023/canadian-motor-vehicle-traffic-collision -statistics-2023

US Department of Transportation,2026, Put the Phone Away or Pay, National Highway Traffic Safety Administration, https://www.nhtsa.gov/campaign/distracted-driving#:~:text=Appoint%20your%20passenger%20as%20the,to%20put%20their%20phone%20away.