Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Feb 1.
Published in final edited form as: Accid Anal Prev. 2017 Dec 14;111:266–270. doi: 10.1016/j.aap.2017.12.002

A new approach for assessing and training drivers’ speed management

Noelle LaVoie a,*, Yi-Ching Lee b, Anna Allison a, James Parker a
PMCID: PMC5780225  NIHMSID: NIHMS928040  PMID: 29248619

Abstract

Motor vehicle crashes are the leading cause of death and injury for teens and speeding is a major contributor, particularly driving too fast for conditions (CDC, 2015, 2013; NHTSA, 2012; Lam, 2003; McKnight & McKnight, 2003). Speed management is a type of tacit knowledge learned through experience that combines speed perception with decisions about safety. Effective training and assessment of speed management requires a safe method for accumulating practice that includes realistic perceptual cues. This study investigated whether speed can be manipulated in an online environment using special effects technology without distorting speed perception. A forced-choice experiment revealed that drivers’ perception of speed was not influenced by the special effects technology, indicating that critical perceptual information was not altered by the speed manipulation of the videos. The experiment also looked at the role of experience in detecting speed differences and found that experienced drivers were able to make judgments about speed more quickly than inexperienced drivers. Implications of these findings for training and assessment are discussed.

Keywords: Speed management, Teen driving, Driving safety

1. Introduction

Motor vehicle crashes are the leading cause of death and injury for teens accounting annually for nearly 3000 deaths, 100 times as many injuries, and over 14 billion dollars in associated costs in the U.S. (Insurance Institute for Highway Safety [IIHS], 2016; CDC, 2015, 2013). Speeding contributes to over 30% of teen driving crashes, as compared to 19% of crashes among adult drivers (NHTSA, 2012). Speeding includes driving too fast for conditions as well as exceeding the speed limit. Driving too fast for conditions includes failure to slow down to compensate for reduced visibility (e.g., rain) or vehicle handling (e.g., ice) as well as driving faster than other vehicles on the road (Aarts and Van Schagen, 2006).

Among teens, driving too fast for conditions is the primary cause of speed-related accidents, rather than intentionally risky driving (Lam, 2003; McKnight and McKnight, 2003). Curry et al. (2011) analyzed crashes involving 15–18 year old drivers from the National Motor Vehicle Crash Causation Survey (NMVCCS), a nationally representative sample of serious crashes. Driving too fast for conditions was the critical error leading to over 20% of crashes among teen drivers.

1.1. Speed perception

Human speed perception is a basic perceptual process, relying primarily on visual cues outside of the vehicle to estimate speed (Recarte and Nunes, 1996). Speed perception is highly sensitive to the amount of visual contrast in a scene, with reduced contrast leading to an underestimation of speed. Indeed, environmental conditions that obscure visual cues, such as darkness and fog, alter drivers’ perception of speed (Recarte and Nunes, 1996; Chatziastoros and Pretto, 2006; Reinhardt-Rutland, 1992). While inexperienced drivers drive slower on average, they do not proportionally decrease their speed to account for poor visibility, leading to a higher rate of collision from this group (Mueller and Trick, 2012).

Speed perception is also influenced by the relative speed of road, vehicle and peripheral objects. When asked to estimate the speed at which a vehicle is traveling participants underestimated high speeds and overestimated low speeds (Hills, 1980; Recarte and Nunes, 1996). More experienced drivers are better able to use speed and distance information to estimate time to collision (Cavallo and Laurent, 1988). Chatziastoros and Pretto (2006) found that the optic flow coming from the road in front of the car was critical in estimating the speed at which a vehicle was traveling. When the speed of the road in front of the traveling car was altered in a driving simulator, drivers adjusted to the speed of the road in front despite the presence of peripheral cues (e.g., traffic signs) traveling at a slower speed. Drivers also experience closing speed adaptation when driving on a straight open road for several minutes before approaching a vehicle, causing drivers to underestimate how quickly they are approaching a vehicle ahead on the roadway (Gray and Regan, 2005).

Visual characteristics of the roadway have an impact on driver’ decisions about what speed to drive, with drivers choosing slower speeds on narrower roads (Charlton and Starkey, 2016), hilly and curvy roads, when lane boundaries are difficult to see (Edquist et al., 2009) and at night (Kockelman and Ma, 2007). Experienced drivers are also better able to select appropriate speeds for conditions that require reduced speeds, such as slowing for objects on the side of the road and on congested freeways (Kockelman and Ma, 2007), when there is a high likelihood of vehicles entering the roadway (Edquist et al., 2009), and when there are complex visual scenes to process. Inexperienced drivers are less able to compensate for complex visual stimuli (Borowsky et al., 2010) and are more likely to fixate on actual hazards (such as a pedestrian) and underestimate the danger of potential hazards (such as low road visibility or closely following the car in front of them) which leads young drivers to underestimate the danger of decreased visibility, such as fog or darkness.

1.2. Social influences and driver characteristics

Social influences impact drivers’ choice of driving speed. Drivers who report that both family and friends approve of speeding are more likely to exceed posted speed limits than drivers who believe their friends and families do not approve (Fleiter et al., 2006). Social influence interacts with driver characteristics such that young male drivers who believe their friends speed are more likely to speed (Moller and Haustein, 2014).

Drivers who experienced an incident that made them angry were more likely to speed in a driving simulator, even miles after the incident (Roidl et al., 2014). Drivers are also more likely to drive too fast for conditions under time pressure and the impact of time pressure may be explained by both the driver’s emotional state and by a cognitive bias that leads drivers to underestimate their speed when under time pressure (Coeugnet et al., 2013). Experienced drivers are more likely to adopt strategies to mitigate the tendency to speed under time pressure than are novice drivers (LaVoie et al., 2008).

Driver characteristics are consistently associated with speed choice, with males tending to drive somewhat faster than females (Hassan et al., 2017; Anastasopolous and Mannering, 2016), drivers under 40 years of age driving faster than older drivers (Anastasopolous and Mannering, 2016), and driver’s with a high income driving faster than those with low or middle incomes (Kweon and Kockelman, 2006; Hassan et al., 2017; Anastasopolous and Mannering, 2016).

1.3. Speed and tacit knowledge

Foss et al. (2011) determined that driving too fast for conditions declines during the first 24 months of driving, closely following a power curve and declining more quickly than overall crash rates. This indicates that learning, rather than intentional risk-taking, aggressive driving or overconfidence, is the cause of improvement.

The Safe Speed Knowledge Test measures drivers’ tacit knowledge (knowledge gained through experience) about appropriate speeds in a variety of contexts (Legree et al., 2003). These include environmental conditions such as bad weather or road conditions known to increase crash risk as well as personal conditions such as emotional states (e.g., anger, anxiety) or fatigue, factors that are prevalent and associated with increased risks of traffic accidents (Alonso et al., 2017; Useche et al., 2017). Novice drivers experience a greater detriment from these factors than more experienced drivers (Paxion et al., 2014). The Safe Speed Knowledge Test successfully distinguished between drivers with a safe driving history and those with a history of car crashes, establishing that knowledge of appropriate speed is associated with reduced crashes. Thus, training which improves speed management has significant potential to reduce crashes among teen drivers.

Few options exist for providing novice drivers with opportunities to acquire tacit knowledge of safe speeds. Digital video editing and special effects techniques, such as those used in television and movies, may be used to manipulate videos of vehicles driving so that vehicles appear to moving at different speeds. Advances in web browsers have increased the amount of interaction users have with videos embedded in web sites. Recent changes include the ability to speed up or slow down a video. Changing the speed of a video of a vehicle taken from a driver’s point of view makes the vehicle appear to moving at faster or slower speeds.

This paper describes an experiment conducted to determine whether digital video editing and special effects techniques can be used to manipulate vehicle speed in digital videos designed to be shown in a web-based environment while maintaining perceptual realism. This was tested using a forced-choice perception experiment and a series of driver’s point of view videos, either edited with VFX to increase and decrease speed, or left unedited. A secondary question was whether experienced drivers would be more sensitive to differences in speed than inexperienced drivers. A third question was whether there would be an interaction between VFX editing and driving experience, such that experienced drivers would be more likely to detect differences in vehicle speeds in the edited videos.

2. Method

2.1. Participants

Participants were recruited from the community and through email blasts sent to parents of adolescents who received care at a large pediatric care network. Sixteen teens (10 females) and 16 adults (9 females) participated in the experiment. Participants met the following requirements: teens were between 14 and 17 years old and had less than one year of driving experience, and adults were between 28 and 55 years old and had at least 10 years of driving experience with no moving violations or at-fault accidents in the past 5 years.

Consent to participate was obtained in-person: teen participants provided written assent and adult participants provided written consent. Upon consenting, eligibility of participation was verified. The recruitment, enrollment, and screening processes were approved by the large pediatric care network’s Institutional Review Board.

2.2. Stimuli. Baseline and manipulated videos

Digital video editing and special effects techniques were used to manipulate vehicle speed in a series of digital videos. Baseline videos were filmed from a driver’s point of view through the front windshield of a vehicle on a closed course track at a steady speed that captured the forward roadway and the sides of the road. Videos were created using a camera mounted inside a vehicle. Videos were filmed at vehicle speeds of 15 mph, 20 mph, 25 mph, 30 mph, 35 mph, 40 mph and 45 mph. These original videos were then manipulated to create a series of videos where the vehicle appears to be moving at different speeds. Videos were cleaned of minor artifacts (e.g., windshield dust) and trimmed of initial acceleration and final deceleration, prior to applying any speed manipulation effect. No adjustments were made for exposure, color, or contrast to any of the footage in order to avoid confounding the speed manipulation. Manipulated videos were created from each baseline video by adjusting the speed in 5mph increments. For example, the baseline 20 mph video was manipulated to create several new videos in which the vehicle appears to travel at 15 mph, 25 mph, 30 mph, 35 mph, 40 mph and 45 mph. Manipulated videos were created by applying a time adjustment effect to the baseline video that best described how each pixel changed from frame to frame in order to create the most realistic speed adjustment (as opposed to just removing or blending whole frames together). In order to accomplish this, a commercial off-the-shelf video effects system was used to adjust the pixel motion (i.e., parts of the image that are in motion). The manipulated videos appear to be identical to the baseline videos except that the vehicle is moving either faster or slower than in the original video.

2.3. Visual comparison validation

This process was validated using the baseline videos at each speed to make a visual comparison of actual versus manipulated speed changes. The videos were aligned using visual landmarks at the start and end of each clip. Then the manipulated video’s speed percentage was adjusted until the two videos’ start and end points were aligned. The resulting percentage change was recorded and compared to the expected table of conversions based on the mathematical function needed to change one speed into another (e.g., 200% to increase a 15 mph video to 30 mph). Differences between the actual amount that each video needed to be adjusted and the expected amount it should be adjusted showed very minor discrepancies for most video clips. The range of differences across all speeds was −4% to +4%.

Videos, both baseline and manipulated, were programmed into a psychological testing software, E-Prime, using a forced-choice paradigm. Each trial (see Fig. 1) consisted of two videos, presented sequentially, and one question. Each video was 6 s long.

Fig. 1.

Fig. 1

Open in a new tab

Sequence of presentations in each trial.

Table 1 lists the available pairs using one original video and one manipulated video. The speed difference in each pair could range from 5 mph to 15 mph, with the assumption that having a speed difference beyond 15 mph would make the comparison too easy and thus meaningless for our purpose.

Table 1.

List of paired comparison in the Original vs. Manipulated comparison.

Original videos Manipulated videos (baseline speed→adjusted speed) (mph)
15 mph – 15→20 15→25 15→30 – – –
20 mph 20→15 – 20→25 20→30 20→35 – –
25 mph 25→15 25→20 – 25→30 25→35 25→40 –
30 mph 30→15 30→20 30→25 – 30→35 30→40 30→45
35 mph 35→20 35→25 35→30 – 35→40 35→45 –
40 mph 40→25 40→30 40→35 – 40→45 – –
45 mph 45→30 45→35 45→40 – – – –
Open in a new tab

The trials consisted of 6 speed differences embedded within the three types of comparisons (manipulated – manipulated, manipulated – original, original – original). Most of the speed differences and types of comparison combinations had at least five repetitions, except where fewer repetitions were available (e.g., a speed difference of 15 mph within the manipulated vs. manipulated comparison had only 1 available repetition, namely original speed of 45 adjusted to 30 vs. original speed of 30 adjusted to 45, given that 15 mph was the maximum of video manipulation).

2.4. Design

A within-subjects design was used. Each participant completed 75 trials, presented in 4 sessions with the order of the trials randomized within the session. There were 30 pairs in the original vs manipulated condition, 30 pairs in the manipulated vs. manipulated condition, and 15 in the original vs. original condition. Trials were balanced so that it was equally likely that the first video or the second video would have a faster moving vehicle.

2.5. Procedure

An individual cubicle with a desktop computer was set up for this study. Participants were asked to follow the instructions presented in E-Prime on the desktop computer: “Welcome to the speed perception study. You will see pairs of videos, one at a time, and then be asked to judge which vehicle in the videos appears to be driving faster. Please press 1 or 2 to indicate the vehicle in the first or second video that appears to be driving faster. Let's practice this with two trials. Please press any key when you are ready to begin.” For each trial they were asked to view and compare the pair of videos and decide which video had the vehicle that appeared to be faster. Once a trial ended, participants were prompted, “Which vehicle in the videos appears to be driving faster?” Participants pressed 1 or 2 on the number keypad on the keyboard to indicate their answer. The E-Prime program was set up to collect accuracy and reaction time data. Accuracy was determined by participants correctly indicating the number of the video that had a faster vehicle. Reaction time was determined by the time it took participants to press the 1 or 2 key, that is, either from the time the video was stopped by the participants or from the time the video ended to when participants indicated their answer.

Participants began with a 10-trial practice, and then completed four sessions of trials for a total of 75 trials. The order of the trials in each session was randomized. The entire study lasted approximately 45 min.

3. Results

Accuracy and reaction times were recorded and trials with the same properties were averaged within each participant. The average accuracy and reaction times were compared between age groups, as well as between speed difference categories and types of comparison (e.g., original vs. manipulated video) using a repeated measures ANOVA.

Results showed there were no significant main effects for type of comparison on accuracy or reaction time, indicating that the manipulated videos were not perceived differently from the original videos by participants in either age category. This provides a strong indication that the manipulated videos introduced no perceptual artifacts that impact how drivers judge speed, regardless of experience (see Table 2).

Table 2.

Accuracy and reaction times for three types of comparison and two age groups.

Type of comparison Experienced drivers
(Adults)		 Novice drivers (Teens)
Accuracy Reaction time
(ms)		 Accuracy Reaction time
(ms)
Original vs manipulated 0.89 1032.72 0.89 1545.83
Manipulated vs manipulated 0.88 924.25 0.87 1816.74
Original vs original 0.90 1062.42 0.87 2007.83
Open in a new tab

There was a main effect of age group on reaction time, F (1,35)=16.22, p < 0.0001, with the experienced drivers responding faster (M=973 msec, SD=881) than the novice drivers (M=1694 msec, SD=2876). There was also a main effect of speed difference on accuracy, F(5,35)=11.83, p < 0.0001. Participants were more accurate when the difference between the videos was large and were the least accurate when asked to compare two videos with only a 5 mph speed difference (see Table 3).

Table 3.

Accuracy and reaction times for six speed differences and two age groups.

Speed difference (speed
of video 2 – speed of
video 1)		 Experienced drivers
(Adults)		 Novice drivers (Teens)
Accuracy Reaction time
(ms)		 Accuracy Reaction time
(ms)
5 0.90 1071.99 0.83 2201.45
−5 0.70 1174.57 0.79 1573.06
10 0.94 883.33 0.91 1772.38
−10 0.88 1027.62 0.92 1553.27
15 0.99 880.69 0.94 1483.08
−15 0.93 842.45 0.92 1655.37
Open in a new tab

4. Discussion

Digital video special effects (VFX) available in commercial software packages offer a potential technology for presenting vehicles at realistic speeds in an online environment (Okun and Zwerman, 2010). A forced-choice experiment determined that drivers perceived the speed of the vehicles in the digital videos in the same way, whether they were manipulated (e.g., a video of a vehicle traveling at 30 mph was speeded up to 40 mph) or at the originally filmed speed. The experiment also demonstrated that experienced and inexperienced drivers perceived the manipulated and original videos in the same manner, ruling out the possibility that there were artifacts in the videos that only experienced drivers could detect. Using digital video preserves the original perceptual cues such as weather, lighting and road curvature (Chatziastoros and Pretto, 2006) while permitting the perceived speed of the traveling vehicle to be manipulated with VFX. This suggests that embedding digital driving videos in a website would allow users to interactively adjust the perceived speed of a vehicle in the video while maintaining accurate visual cues.

While experienced and inexperienced drivers were equally accurate in their judgments about vehicle speed, experienced drivers were able to make these judgments significantly faster. Considered in the context of a speed-accuracy trade-off, this finding is consistent with findings that novice drivers are less accurate than experienced drivers at perceiving vehicle speed, which in turn is thought to contribute to novice drivers’ problems with speed management (Mueller and Trick, 2012; Recarte and Nunes, 1996). This certainly supports the idea that one aspect of speed management that is gained through practice is quick, accurate speed perception. This is consistent with research that shows the importance of driving practice for developing speed management (Foss et al., 2011).

Interactive videos that incorporate VFX have potential for driver’s education, specifically teaching novice drivers about speed management. A web-based driver’s education system created with this technology could provide novice drivers with the type of repetitive practice required to develop tacit knowledge implicated in good speed management (Legree et al., 2003; Foss et al., 2011). By creating videos which depict vehicles in a range of environments, teens could experience a wide range of situations that require active speed management, such as limited visibility due to rain or nighttime, or visual complexity in scenes with crowded roadsides and traffic congestion. It may also be possible to address non-perceptual causes of speeding, such as peer pressure or social influence, by encouraging teens to use a training system together so they can see others learning to make safer speed choices.

It is generally difficult to provide perceptually realistic driving practice in a classroom setting, or without undue exposure to risk, since it requires a hands on approach. While driving simulators offer this opportunity, they are too expensive for many driving academies and high schools and so are not available to most teen drivers. Driving simulator technology also has limitations on perceptual accuracy, for example distortions in simulated weather conditions, such as fog and more complex road configurations (Hurwitz et al., 2005; Chatziastros and Pretto, 2006; Culham, 2012), which lead drivers in a simulator to drive an average of five miles an hour faster than in the field (Hurwitz et al., 2005). A web-based system with actual on-road footage that drivers can interact with may be able to overcome these issues with distorted perceptual cues.

It typically takes years of driving practice to develop the speed management skill of an experienced driver, but training may help to shorten novice driver’s learning curve. Novice drivers are particularly likely to be involved in speed-related crashes during the first six months after receiving their license (Foss et al., 2011). Targeted speed management training offered as part of driver’s education maybe able to shorten this learning curve, reducing the number of crashes particularly during the first six months. If this approach transfers to on-road driving improvements, the impact on crash rates could be substantial, as speed management is implicated in 30% of teen driving crashes (NHTSA, 2012).

Videos embedded in an online environment could also be used to assess a driver’s knowledge of speed management. Legree et al. (2003) found that the Safe Speed Knowledge Test, which evaluates self-reported speed adjustments in different conditions, was strongly related to drivers’ crash history, suggesting that this type of assessment is associated with overall driving skill. An interactive version of this test, with realistic perceptual cues, may improve the predictive validity of the assessment. Indeed, experienced drivers in the perception experiment were able to make accurate judgments of speed faster than inexperienced drivers, lending some support to this possibility.

4.1. Limitations

The current experiment tested the perceptual equivalency of manipulated and original digital videos filmed on a closed track rather than a variety of normal driving situations. While this eliminated possible visual confounds due to varied environments, it means the results of the current study may not generalize to all driving scenes. The size of the experiment and specific participant populations included also limit the generalizability of the study. Next steps include replicating this experiment with videos depicting a range of realistic driving scenes.

Digital manipulation of vehicle speed in videos has potential for applications in training and assessment. Future work should examine the use of this technology for training novice drivers to improve their speed perception and speed management, including measuring transfer to on-road driving. Evaluating drivers’ speed management skill using an interactive video assessment may predict drivers’ ability to drive at appropriate speeds for conditions. Future work should investigate the predictive validity of an online video evaluation of speed management.

Acknowledgments

This research was supported by the Eunice Kennedy Shriver National Institute of Child Health & Human Development of the National Institutes of Health under Award Number R41HD082894. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

References

  1. Aarts L, Van Schagen I. Driving speed and the risk of road crashes: a review. Accid. Anal. Prev. 2006;38:215–224. doi: 10.1016/j.aap.2005.07.004. [DOI] [PubMed] [Google Scholar]
  2. Alonso F, Esteban C, Sanmartin J, Useche S. Reported prevalence of health conditions that affect drivers. Cogent Med. 2017;4:1–15. [Google Scholar]
  3. Anastasopolous PC, Mannering F. The effect of speed limits on drivers’ choice of speed: a random parameters seemingly unrelated equations approach. Anal. Methods Accid. Res. 2016;10:1–11. [Google Scholar]
  4. Borowsky A, Shinar D, Oron-Gilad T. Age, skill and hazard perception in driving. Accid. Anal. Prev. 2010;42:1240–1249. doi: 10.1016/j.aap.2010.02.001. [DOI] [PubMed] [Google Scholar]
  5. Centers for Disease Control and Prevention. Teen Drivers: Get the Facts. 2013 (Retrieved from http://www.cdc.gov/motorvehiclesafety/teen_drivers/teendrivers_factsheet.html)
  6. Centers for Disease Control and Prevention. CDC Study Finds Annual Cost of Motor Vehicle Crashes Exceeds $99 Billion [Press Release] 2015 (Retrieved from http://www.cdc.gov/media/pressrel/2010/r100825.htm)
  7. Cavallo V, Laurent M. Visual information and skill level in time-to-collision estimation. Perception. 1988;17(5):623–632. doi: 10.1068/p170623. [DOI] [PubMed] [Google Scholar]
  8. Charlton SG, Starkey NJ. Risk in our midst: centrelines, perceived risk, and speed choice. Accid. Anal. Prev. 2016;95(A:192–201. doi: 10.1016/j.aap.2016.07.019. [DOI] [PubMed] [Google Scholar]
  9. Chatziastros A, Pretto P. Changes in optic flow and scene contrast affect the driving speed. Paper Presented at Driving Simulation Conference Europe; Paris, France. 2006. [Google Scholar]
  10. Coeugnet S, Miller H, Anceaux F, Naveteur J. How do time pressured drivers estimate speed and time? Accid. Anal. Prev. 2013;55:211–218. doi: 10.1016/j.aap.2013.02.040. [DOI] [PubMed] [Google Scholar]
  11. Culham JC. Motion Perception: New ideas on how drivers perceive speed emerge from the fog. eLife. 2012 Oct 30; doi: 10.7554/eLife.00281. Retrieved from https://elifesciences.org/content/1/e00281. [DOI] [PMC free article] [PubMed]
  12. Curry AE, Hafetz J, Kallan MJ, Winston FK, Durbin DR. Prevalence of teen driver errors leading to serious motor vehicle crashes. Accid. Anal. Prev. 2011;43(4):1285–1290. doi: 10.1016/j.aap.2010.10.019. [DOI] [PubMed] [Google Scholar]
  13. Edquist J, Rudin-Brown CM, Lenne MG. Road Design Factors and Their Interactions with Speed and Speed Limits. MUARC Report Number 298, 15 May. 2009 (Retrieved from: www.monash.edu.au/muarc/reports/muarc298.pdf)
  14. Fleiter J, Watson B, Lennon A, Lewis I. Significant others, who are they? Examining normative influences on speeding. Proceedings 2006 Australasian Road Safety Research Policing Education Conference; Gold Coast. 2006. [Google Scholar]
  15. Foss RD, Martell CA, Goodwin AH, O’Brien NP. Measuring Changes in Teenage Driver Crash Characteristics During the Early Months of Driving. AAA Foundation for Traffic Safety; Washington, DC: 2011. [Google Scholar]
  16. Gray R, Regan D. Perceptual processes used by drivers during overtaking in a driving simulator. Hum. Factors. 2005;47(2):394–417. doi: 10.1518/0018720054679443. [DOI] [PubMed] [Google Scholar]
  17. Hassan HN, Shawky M, Kishta M, Garib AM, Al-Harthei HA. Investigation of drivers’ behavior towards speeds using crash data and self-report questionnaire. Accid. Anal. Prev. 2017;98:348–358. doi: 10.1016/j.aap.2016.10.027. [DOI] [PubMed] [Google Scholar]
  18. Hills BL. Vision, visibility, and perception in driving. Prevention. 1980;9(2):183–216. doi: 10.1068/p090183. [DOI] [PubMed] [Google Scholar]
  19. Hurwitz D, Knodler M, Jr, Dulaski D. Speed perception fidelity in a driving simulator environment. Paper presented at Driving Simulation Conference; North America, Orlando, FL. 2005. November 30–December 2. [Google Scholar]
  20. Insurance Institute for Highway Safety. Fatality Facts. 2016 (Retrieved from http://www.iihs.org/iihs/topics/t/teenagers/fatalityfacts/teenagers)
  21. Kweon Y-J, Kockelman KM. Driver attitudes and choices: speed limits, seat belt use: and drinking-and-driving. J. Transp. Res. Forum. 2006;45:39–56. [Google Scholar]
  22. LaVoie N, Lauper U, Lee Y-C, Cianciolo A, Foltz P, Legree P, Psotka J. Driving wisdom: improving risk management of common driving hazards. Paper Presented at I/ITSEC; Orlando,FL. December 1–4, 2008.2008. [Google Scholar]
  23. Lam LT. Factors associated with young drivers’ car crash injury: comparisons among learner, provisional and full licensees. Accid. Anal. Prev. 2003;35:913–920. doi: 10.1016/s0001-4575(02)00099-4. [DOI] [PubMed] [Google Scholar]
  24. Legree PJ, Heffner TS, Psotka J, Medsker GJ, Martin DE. Traffic crash involvement: experiential driving knowledge and stressful contextual antecedents. J. Appl. Psychol. 2003;88(1):15–26. doi: 10.1037/0021-9010.88.1.15. [DOI] [PubMed] [Google Scholar]
  25. McKnight AJ, McKnight AS. Young novice drivers: careless or clueless? Accid. Anal. Prev. 2003;3(35):921–925. doi: 10.1016/s0001-4575(02)00100-8. [DOI] [PubMed] [Google Scholar]
  26. Moller M, Haustein S. Peer influence on speeding behavior among male drivers aged 18 and 28. Accid. Anal. Prev. 2014;64:92–99. doi: 10.1016/j.aap.2013.11.009. [DOI] [PubMed] [Google Scholar]
  27. Mueller AS, Trick LA. Driving in fog: the effects of driving experience and visibility on speed compensation and hazard avoidance. Accid. Anal. Prev. 2012;48:472–479. doi: 10.1016/j.aap.2012.03.003. [DOI] [PubMed] [Google Scholar]
  28. National Highway Traffic Safety Administration. Traffic Safety Facts Young Drivers. 2012 (Retrieved from http://www-nrd.nhtsa.dot.gov/Pubs/811622.pdf)
  29. Okun JA, Zwerman S, editors. The VES Handbook of Visual Effects: Industry Standard VFX Practices and Procedures. Focal Press; Burlington, MA: 2010. [Google Scholar]
  30. Paxion J, Galy E, Berthelon C. Mental workload and driving. Front. Psychol. 2014;5:1–11. doi: 10.3389/fpsyg.2014.01344. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Reinhardt-Rutland AH. Poor-visibility road accidents: theories entailing target risk level and relative visual motion. J. Psychol. 1992;126(1):63–71. [Google Scholar]
  32. Roidl E, Frehse B, Hoeger R. Emotional states of drivers and the impact on speed, acceleration: and traffic violations – a simulator study. Accid. Anal. Prev. 2014;70:282–292. doi: 10.1016/j.aap.2014.04.010. [DOI] [PubMed] [Google Scholar]
  33. Useche SA, Ortiz VG, Cendales BE. Stress-related psychosocial factors at work, fatigue: and risky driving behavior in bus rapid transit (BRT) drivers. Accid. Anal. Prev. 2017;104:106–114. doi: 10.1016/j.aap.2017.04.023. [DOI] [PubMed] [Google Scholar]

RESOURCES