This article was downloaded by: [Joshua McKeever] On: 29 January 2013, At: 14:36 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Traffic Injury Prevention Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/gcpi20
Driver Performance While Texting: Even a Little is Too Much a
Joshua D. McKeever , Maria T. Schultheis a
a b
c
, Vennila Padmanaban & Allison Blasco
a
Department of Psychology, Drexel University, Philadelphia, Pennsylvania
b
School of Biomedical Engineering, Health and Science Systems, Drexel University, Philadelphia, Pennsylvania c
Department of Biological Science, Drexel University, Philadelphia, Pennsylvania Accepted author version posted online: 16 Jul 2012.
To cite this article: Joshua D. McKeever , Maria T. Schultheis , Vennila Padmanaban & Allison Blasco (2013): Driver Performance While Texting: Even a Little is Too Much, Traffic Injury Prevention, 14:2, 132-137 To link to this article: http://dx.doi.org/10.1080/15389588.2012.699695
PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.
Traffic Injury Prevention (2013) 14, 132–137 C Taylor & Francis Group, LLC Copyright ISSN: 1538-9588 print / 1538-957X online DOI: 10.1080/15389588.2012.699695
Driver Performance While Texting: Even a Little is Too Much JOSHUA D. MCKEEVER1, MARIA T. SCHULTHEIS1,2, VENNILA PADMANABAN3, and ALLISON BLASCO1 1
Department of Psychology, Drexel University, Philadelphia, Pennsylvania School of Biomedical Engineering, Health and Science Systems, Drexel University, Philadelphia, Pennsylvania 3 Department of Biological Science, Drexel University, Philadelphia, Pennsylvania 2
Downloaded by [Joshua McKeever] at 14:36 29 January 2013
Received 13 March 2012, Accepted 30 May 2012
Objective: To examine the impact of text messaging and other in-car behaviors on driving performance under simple and naturalistic road conditions in a driving simulator. Methods: Data from 28 healthy individuals (12 female) are presented. Participant age ranged from 18 to 28 (mean = 21.0). Average driving experience was 3.8 years (SD = 2.5). Participants completed a baseline loop condition in which they drove normally through a realistic virtual environment. Next, participants drove an identical loop, and at 3 specified points during this drive, participants were required to (1) complete a radio-tuning task; (2) type and send a text message containing “Drexel University”; and (3) type and send “I am driving to the store.” Driving performance and task duration was compared between conditions. Results: Across all tasks, both lane management, F(1,27) = 11.1, P = .002, and velocity, F(1,27) = 10.3, P = .003, varied significantly more while task-engaged. Average lane deviation was significantly greater during a text messaging task than during the baseline drive of the same road segment, t(27) = −2.9, P = .007. Comparison of task durations indicated that both texting tasks took significantly longer to complete than the radio task, with the “Drexel University” text (118 s) taking almost twice as long as the radio-tuning task (60 s). Unexpected and novel findings emerged in the evaluation of duration of texting tasks using the varying text-entry methods, with touch-screen modality taking significantly longer than others. Conclusions: Engaging in secondary tasks while operating a motor vehicle may have deleterious effects on driving performance and increase risk, even under the simplest of driving conditions. Text messaging may constitute a “perfect storm” of risk compared to other in-vehicle tasks such as tuning the car radio. The current investigation demonstrated detrimental effects of text messaging on driving behaviors such as lane maintenance, speed maintenance, and shifts of attention, even under relatively ideal and naturalistic driving conditions (e.g., familiar route, good weather, no traffic). Keywords: driving, motor vehicle, distraction, cell phone, text messaging
Introduction Operating a motor vehicle is a demanding task that requires the integration of cognitive, physical, sensory, and behavioral abilities. Despite widespread acceptance of the demands of driving, in-vehicle distractions are relatively common (Young and Regan 2007), and increased availability of portable electronic devices has contributed to increases in the use of such devices while driving (Strayer et al. 2005; Young and Regan 2007). With this in mind, it is concerning that a recent study by the U.S. Department of Transportation (DOT) reported that distracted drivers were associated with 16 percent of all motor vehicle fatalities (DOT 2010). Much of driver distraction research has focused on the use of cell phones and, more specifically, examining
Address correspondence to Maria T. Schultheis, PhD, 3141 Chestnut Street, Stratton Building Room Suite 123, Philadelphia, PA 19104. E-mail:
[email protected]
conversing during driving. Empirical evidence for the dangers of phone use while driving is neither unexpected nor new; in an influential study now almost 15 years old, Redelmeier and Tibshirani (1997) showed that mobile phone use was associated with a 4-fold increase in the likelihood of getting into a crash. Recent years have seen a rise in text messaging as a mode of communication, including while driving (Drews et al. 2009). In fact, a 2009 study of young drivers found that 72.5 percent of the study responders reported texting and driving at least some of the time (Nelson et al. 2009). Sending and receiving text messages while operating a vehicle may be even more distracting than other uses of cell phones (i.e., talking) and thus may pose greater risks for crash involvement. Reading or typing text messages while driving constitutes a dual task that requires resources across multiple domains (e.g., cognitive load, visual, physical/motor) that are also necessary to drive safely (Collet et al. 2010a, 2010b). Additionally, texting is likely to simply take longer to engage in compared to other distracting in-car activities, and thus
Downloaded by [Joshua McKeever] at 14:36 29 January 2013
Driving and Texting cognitive resources are compromised for a longer period of time (Strayer et al. 2005). Though the number of studies examining the effects of text messaging while driving remains relatively small compared to research on phone use and driving in general, an increasing number of studies have found empirical evidence for the unique hazards posed by texting and driving. The findings from work to date indicate that while engaged in text messaging, participants demonstrate greater numbers of driving errors (Hosking et al. 2007), increased response time to environmental stimuli (such as a leading car’s brake lights; Drews et al. 2009), greater variability in speed and lane positioning (which are both indicative of poorer vehicle control and potential for crashes; Crisler et al. 2008), increased stopping distance despite overall reduced speed, potentially increased cortical activity associated with distractibility, and increased physical, visual, and mental demands (Mouloua et al. 2010; Reed and Robbins 2008). Not surprisingly, the majority of research examining texting has employed the use of driving simulation, which allows researchers to safely examine performance under demanding conditions. In fact, many studies examining texting and driving have targeted driver errors under challenging conditions (i.e., extremely curvy roads). What is less known is how these errors might present themselves under less challenging conditions (i.e., straight roads, fewer external distractions). In addition, though it is acknowledged that it is the dual-task nature of texting and driving that relates to these errors, the contribution of the duration of the dual-tasking has not been thoroughly examined. The current study proposes to extend our understanding of the effects of texting on driving by examining performance under simple naturalistic driving conditions and by comparing durations of dualtask involvement between texting and a common in-car task (radio tuning). Evidence indicates that more than ever, younger adults are using texting as a primary means of communication and that it is important for maintaining self-esteem and social networks; clearly, text messaging is pervasive and it has become “the norm” to communicate via texting in this age group (Atchley et al. 2011; Nemme and White 2011), despite the perception of risk. Because of the relevance to this age group, the current study examined texting and driving behavior in young adults. Importantly, because ownership of cell phones is so common in this group, the current study also allowed the use of the participants’ personal cell phones, in order to reduce the effects of novelty and/or device unfamiliarity. It was anticipated that despite the simple and naturalistic driving conditions (e.g., using own phone, simple routes), texting would have a negative effect on driving performance.
133 tiation of the testing session, all participants completed Drexel University Institutional Review Board–approved consent and Health Insurance Portability and Accountability Act authorization forms and were screened for simulation sickness. Participants A total of 28 individuals (12 women and 16 men) were included in this study. To be included, participants were required to have normal or corrected-to-normal visual acuity and a valid driver’s license. Participant age ranged from 18 to 28 (mean = 21.0, SD = 2.4), and education ranged from 13 to 20 years (mean = 14.8, SD = 1.5). None of the participants reported any significant history of psychiatric illness, neurological insult, or substance abuse. Participants had an average of 3.8 years (SD = 2.5) of driving experience. All 28 participants owned a cellular phone, 75 percent reported that they text message in general, and the same 75 percent reported that they both read and send text messages while driving. A series of t-tests and chi-square tests was conducted to evaluate demographic distinctions between the texter and non-texter groups, which revealed no significant group differences on any variables of interest, including age, t(24) = 0.27, P = .79; education, t(24) = 0.19, P = .85; gender, t(24) = 0.33, P = .97; driving frequency, t(24) = 0.48, P = .63; driving experience, t(24) = 1.08, P = .29; phone type, χ 2 (3, N = 28) = 0.67, P = .88; time to complete texting tasks, t(24) = 0.92, P = .37; lane deviation, t(24) = −1.35, P = .19; and speed variability, t(24) = −1.12, P = .27. Virtual Reality Driving Simulator A custom-made, high-fidelity virtual reality driving simulator (VRDS) was used in the study. The simulator included one desktop PC and 3 high-resolution displays, providing a 180-degree field of view. The steering wheel, gas pedal, and brake pedal were manufactured by Extreme Competition Controls, Inc. (ECCI, Minneapolis, MN), and the center console (including shifter, cup holders, ashtray, and stock stereo system) was from a Ford Taurus sedan with an automatic transmission (see Figure 1). Measures of real-time driving performance, including driving speed, lane position, brake, gas, and steering wheel inputs, were sampled at 16 Hz. The simulator software was custom-engineered by Digital Mediaworks, Inc. (DMW, Ontario, Canada) and is designed
Methods This study was part of a larger study examining performance on a virtual reality driving simulator among healthy individuals of varying ages. Specifically, individuals meeting criteria (see subsequent section) were seen for one testing session lasting approximately 1.5 to 2.5 h. At the ini-
Fig. 1. The DMW-Drexel driving simulator and screenshot of route driven by participants.
134
McKeever et al.
to target the cognitive demands of driving while remaining affordable and convenient. The virtual environment was composed of daytime dry-pavement driving conditions with good visibility and data reported were collected during straight roadway segments of between 3800 and 5000 feet.
Downloaded by [Joshua McKeever] at 14:36 29 January 2013
VRDS Task Prior to administration of the experimental VRDS tasks, participants were oriented and trained on the VRDS following standardized procedures, which included 2 practice trials, during which participants were instructed to drive through simple virtual roads and practice using the brake, gas, and steering wheel on straight segments and curves. Participants were instructed to drive as if they were driving their own car on a real road, following normal driving conventions and laws as they normally would. Following training, participants completed the baseline loop condition in which they drove through a preset virtual environment without any distractions or external stimuli. This condition lasted approximately 20 min, and when they reached the end, participants were asked to pull over and stop the vehicle. After a short break, participants drove the task-engaged loop, which was identical to the baseline loop, except that at 3 specified points, the participant was instructed to complete 1 of 3 tasks: (1) a radio-tuning task, in which participants were instructed to tune to a specific FM radio station on the stereo in the vehicle’s center console (the station was preset to 92.3 and they were asked to adjust it to 104.1); (2) the “DU” text, in which participants were instructed to type and send a text message (to the experimenter) containing the text “Drexel University”; and (3) the “Store” text, in which participants were instructed to type and send a text message (to the experimenter) containing the text “I am driving to the store.” The order of these tasks was counterbalanced across participants and the tasks were administered on 3 prespecified segments of the route selected based on similar route demand and characteristics (i.e., speed limit, minimal curvature). All participants used their own cell phones to complete the text messaging tasks, and all text messages sent by participants were accurate within a 2-character error limit (except for one participant who wrote “going” instead of “driving”). This second condition lasted approximately 20 min, and when they reached the end, participants were asked to pull over and stop the vehicle.
Dependent Measures Data from the VRDS were analyzed with the Windshield software, a program developed by DMW. Data were analyzed for the 3 segments of the baseline loop and the 3 segments of the task-engaged loop, allowing comparisons of each individual’s performance on the identical segment of the route “on” and “off” task. The following variables were computed: average velocity (mph), velocity standard deviation, average lane deviation (distance from the vehicle’s centroid to the center of the right-hand lane, in inches), and lane deviation standard deviation. Time to complete each of the behavioral tasks was recorded in seconds.
Results The results of the driving performance measures are presented in Table I. Unless otherwise noted, analyses consisted of repeated-measures analyses of variance (ANOVAs) or repeated-measures t-tests using task engagement (baseline condition vs. task-engaged condition) as a factor. Table I provides a summary of means and standard deviations of the driving performance variables. Lane Deviation Lane deviation was significantly greater during the Store text than during the baseline drive of the same road segment, t(27) = −2.9, P = .007, d = 0.44, but was not significantly different for the DU text or the radio task. Because averaged lane deviation is a somewhat unreliable measure of overall lane management due to the fact that deviations to the left (negative values) and right (positive values) may cancel each other out, a second measure of lane management, standard deviation of lane deviation, was also examined. Note that the basic lane deviation value is not a statistic but a measure captured by the VRDS each time data are sampled representing the distance from the vehicle’s center point to the center point of the travel lane. Standard deviation of lane deviation is a summary statistic that indicates the average magnitude of this distance for each road segment. A repeated-measures ANOVA revealed a significant main effect of task engagement such that lane management varied more while task-engaged,
Table 1. Driving performance variable means and standard deviations (N = 28) Task DU text Baseline Task-engaged Store text Baseline Task-engaged Radio task Baseline Task-engaged aAverage
Average velocity (mph)
Average velocity SD
Average lane deviation (in.)
Average lane deviation SDa
44.6 (5.5) 44.8 (5.9)
2.2 (1.5) 3 (1.9)
12.1 (8.6) 13.1 (11.9)
17.5 (4.7) 18.2 (4.8)
44.4 (5.4) 45.4 (7.4)
2.1 (1) 2.8 (2.2)
11.5 (10.2) 15.9 (9.6)
17.6 (4.8) 19.1 (4.9)
44.2 (5.2) 44.3 (5.2)
2.1 (1.2) 2.8 (1.6)
12.4 (10) 12.9 (10.1)
15.8 (3.9) 19.4 (7.4)
amount of lane control (i.e., “tracking”) variability across participants.
Driving and Texting
135
Downloaded by [Joshua McKeever] at 14:36 29 January 2013
Fig. 3. Mean time to complete tasks (seconds).
Fig. 2. Effects of in-car tasks on lane management. ∗ Statistically significant difference.
F(1,27) = 11.1, P = .002, η2 = 0.053, but there was no significant main effect of task or interaction (see Figure 2). Velocity There were no significant differences in average velocity across tasks or task engagement condition. For the variable of velocity standard deviation, a repeated-measures ANOVA revealed a significant main effect of task engagement such that speed varied more while task-engaged, F(1,27) = 10.3, P = .003, η2 = 0.085, but there was no significant main effect of task or interaction. Task Completion Time Task completion time was used as a dependent variable for a comparison across the 3 tasks and also for a comparison across the 4 types of cell phones that participants used. The average time to complete each of the 3 tasks was 118.4 s (SD = 75.7) for the DU text, 103.6 s (SD = 40.8) for the Store text, and 60.2 s (SD = 39.4) for the radio-tuning task. Results from 3 separate Bonferroni-corrected t-tests comparing task completion time revealed significant differences between the DU text and the radio task, t(27) = 3.3, P = .003, and between the Store text and the radio task, t(27) = 4.9, P < .001, but no significant difference between the 2 text message tasks (see Figure 3). As a supplementary analysis, the text-entry method used to text was examined in terms of the amount of time it took to complete the texting tasks; see Figure 4 for these data. The 4 types of text-entry were ABC mode (i.e., the numberpad is used such that pressing the “2” key once outputs “A,” twice outputs “B,” etc.; N = 9), T9 mode (i.e., predictive text; N = 5), full keyboard (e.g., Blackberry devices; N = 6), and touchscreen keyboard (e.g., iPhone; N = 8). One-way ANOVAs for the DU text, the Store text, and the 2 texts averaged together revealed significant effects of text-entry method for the Store
text, F(3,24) = 4.3, P = .015, η2 = 0.349, and for the averaged texts, F(3,24) = 3.9, P = .021, η2 = 0.328, but no significant differences for the DU text. Follow-up t-tests indicated that, for the Store text, full keyboard was significantly faster than touchscreen keyboard and ABC mode but no other differences reached significance, and for the averaged texts, full keyboard and T9 mode were both significantly faster than touchscreen keyboard or ABC mode, but no other differences reached significance.
Discussion Though the effect of texting on driving performance has been examined in various types of simulated driving conditions, the findings from the current study indicate that negative effects exist even under simple naturalistic conditions (i.e., using one’s own phone, route with minimal demands and/or distractions). A significant detrimental effect on lane management was identified during a texting task but not during another common in-car task (radio tuning), and attention was diverted from the road for longer periods of time during texting than during the radio task.
Fig. 4. Time to complete texting tasks across text-entry modes.
Downloaded by [Joshua McKeever] at 14:36 29 January 2013
136 Though both of the dual-task conditions examined in this study (texting and radio-tuning) appeared to affect both speed and lane management, the findings indicated that texting was more disruptive to these driving behaviors. The assertion that speed management was adversely affected by dual-tasking was based on the fact that dual-tasking was associated with more variability in the speed participants maintained while completing an in-car task versus while driving only. This was observed despite the likelihood that drivers may have actually reduced their speed to some degree (whether consciously or unconsciously) while texting or tuning the radio, presumably in order to reduce the chance of making driving errors, a finding that is consistent with previous studies (Drews et al. 2009; Hosking et al. 2007). Similarly, the observation of variable performance on lane management in the current study is in line with prior studies that have argued that the decreased ability to stay in the intended traffic lane is associated with more hazardous driving (Drews et al. 2009; Hosking et al. 2007) and that variable or erratic driving behavior is likely to increase the risk of driving errors and crashes for several reasons, including that other drivers are less able to predict one’s maneuvers (Young and Regan 2007). Though these findings corroborate common sense and previous research, what is most striking is the fact that, compared to many real-world driving and dual-task situations, the conditions examined in the current study were considerably easier and simpler. In fact, according to Collet et al.’s (2010b) inventory of the most significant factors affecting one’s ability to engage in secondary tasks while driving, the driving conditions in the current study fall on the minimal risk side of the continuum except for the manual nature of the tasks, yet significant effects were seen on driving performance. Furthermore, participants engaged in the dual-task conditions under optimal settings that minimized novelty and/or device unfamiliarity by examining performance on a route that they had already driven and using their personal cell phones. Additionally, it is noteworthy to consider that participants in the study may have behaved differently due to the fact that they were being evaluated. They may have been more careful while driving and texting in the laboratory than on the road; if true, this would increase the implication of the findings. One important translation of these results may be that conditions where it is “okay to text while driving” do not exist—and, in fact, the negative effects of texting while driving are seen across the spectrum of driving environments. A second focus of this study was to examine duration of distraction under dual-task conditions. Not surprisingly, texting required more time to complete when compared to another common secondary in-car task (i.e., radio tuning). This was found despite the fact that both radio tuning and texting have cognitive, visual, and physical distraction elements. Furthermore, the greater duration was seen in either short (2-word) or longer (5-word) text messages and subsequently may represent the increased risk of distracted driving while engaged in texting communication. Interestingly, an unexpected and novel finding was the evaluation of differences in duration of task using the varying text-entry methods. Although limited by the sample size, the findings indicated that it took longer to input text messages using touchscreen keyboards and ABC mode,
McKeever et al. and thus full keyboard and T9 mode allowed for quicker completion of the dual task. What is most remarkable is the fact that the slowest entry mode was in fact the most “commercially preferred” input mode, as exhibited by the increasing growth in technologies offering touchscreen input (i.e., IPad, navigation systems), many of which become in-car entertainment devices. As such, this finding may be most relevant to the growing body of research that examines the level of distraction produced by in-vehicle devices in an attempt to reduce increased risk for distracted driving given the proliferation of these technologies (Reed-Jones et al. 2008). Though this study provides novel information about the hazards of texting and driving, a few limitations warrant consideration. First, though the use of virtual reality simulation offers major advantages (e.g., safety, cost, data collection) over on-road testing and has been shown to predict real-world performance (e.g., Lew et al. 2005), simulated driving does vary from real-world driving and it is unclear how well our conditions modeled real-world driver distraction situations. Second, it is important to note that the limited sample may reduce the generalizability of these findings. Specifically, the study’s power was limited by a small sample size and restricted ranges of age and driving experience. Though statistically significant differences were identified (and effect sizes are provided and are noteworthy), this is clearly a major limitation. Future studies should examine these factors with larger samples, as well as aim for (1) a broader age range, (2) a broader range of driving experience, and (3) use of more difficult road conditions to assess potential confounds. Lastly, the notion that “driving variability equals driving hazard” is probably not quite so cut and dry, although both logic and previous research imply such a link (Young and Regan 2007). Future research directly examining the relationship of variability to dangerous driving outcomes is warranted. The practical significance of these results may be questionable, because the real-world impact of a few inches or miles per hour is debatable. However, a growing body of research has shown that driving hazard is related to both the length of time distracted and the pattern of distraction (e.g., duration of glances to and from the roadway over a period of time; Green 2000; Smith et al. 2005). Thus, though our study demonstrated statistically more erratic driving habits while engaged in a secondary task, our task completion time analyses (demonstrating distraction periods of up to 2 min, on average) may be a more practically significant finding related to driving hazard. Both safe driving and text messaging impose demands on multiple cognitive (e.g., attention), perceptual (e.g., vision), and physical (e.g., fine motor) systems, and one’s ability to process perceptual input and control behavioral responses simultaneously within these systems is presumed to be finite. As previous authors (e.g., Terry et al. 2008) have asserted, bottom-up perceptual processing of environmental details signaling danger may not be salient enough to immediately supplant top-down processes such as focused attention on typing a message, and thus the longer and more sustained a driver’s attention away from the road, the less likely he or she is to respond to road hazards in time. Text messaging requires substantial time and attention, above and beyond other in-car tasks, and attempts to text quickly may simply impose greater demands on the
Driving and Texting
Downloaded by [Joshua McKeever] at 14:36 29 January 2013
attentional system and exacerbate this problem. Our study offers only limited but suggestive elucidation of these assumptions. In sum, the research examining the effects of texting on driving is nascent, and though evidence for its negative impact is building, the popularity and relevance of this communication mode continues to grow among the young driver population. Providing evidence that driving conditions where “it is safer to text” do not exist may help to reduce this false perception. Likewise, additional knowledge about what factors may contribute to extended duration of distracted driving during this dual task may serve to educate those providing these technologies (i.e., parents, industry). And though complete discontinuation of texting while driving is an ambitious goal, continued research may be helpful to shape public policy surrounding these issues and lead to a reduction in this hazardous behavior.
References Atchley P, Atwood S, Boulton A. The choice to text and drive in young drivers: behaviors may shape attitude. Accid Anal Prev. 2011;43:134–142. Collet C, Guillot A, Petit C. Phoning while driving I: a review of epidemiological, psychological, behavioural and physiological studies. Ergonomics. 2010a;53:589–601. Collet C, Guillot A, Petit C. Phoning while driving II: a review of driving conditions influence. Ergonomics. 2010b;53:602–616. Crisler M, Brooks J, Ogle J, Guirl C, Alluri P, Dixon K. Effects of wireless communication and entertainment devices on simulated driving performance. Transp Res Rec. 2008;2069:48–54. Department of Transportation. National Highway Traffic Safety Administration Traffic Safety Facts Research Note: Distracted Driving 2009. 2010. Available at: http://www-nrd.nhtsa.dot.gov/Pubs/ 811379.pdf. Accessed July 11, 2011. Drews FA, Yazdani H, Godfrey CN, Cooper JM, Strayer DL. Text messaging during simulated driving. Hum Factors. 2009;51:762–770. Green M. “How long does it take to stop?” Methodological analysis of driver perception–brake times. Transp Hum Factors. 2000;2:195–216.
137 Hosking S, Young K, Regan M. The effects of text messaging on young novice driver performance. In: Faulks IJ, Regan M, Stevenson M, Brown J, Porter A, Irwin JD, eds. Distracted Driving. Sydney, Australia: Australasian College of Road Safety; 2007:155–187. Lew HL, Poole JH, Lee EH, Jaffe DL, Huang HC, Brodd E. Predictive validity of driving-simulator assessments following traumatic brain injury: a preliminary study. Brain Inj. 2005;19:177–188. Mouloua M, Ahern A, Rinalducci E, Alberti P, Brill JC, Quevedo A. The effects of text messaging on driver distraction: a bio-behavioral analysis. Poster presented at: 54th Annual Human Factors and Ergonomics Society Annual Meeting, September 27–October 1, 2010; San Francisco, Calif. Nelson E, Atchley P, Little T. The effects of perception of risk and importance of answering and initiating a cellular phone call while driving. Accid Anal Prev. 2009;41:438–444. Nemme H, White KM. Texting while driving: psychosocial influences on young people’s texting intentions and behaviors. Accid Anal Prev. 2011;42:1257–1265. Redelmeier DA, Tibshirani RJ. Association between cellular-telephone calls and motor vehicle collisions. N Engl J Med. 1997;336:453–458. Reed N, Robbins R. The Effect of Text Messaging on Driver Behavior: A Simulator Study. Transport Research Laboratory (London, UK); 2008. Published Project Report PPR 367. Reed-Jones J, Trick LM, Matthews M. Testing assumptions implicit in the use of the 15-second rule as an early predictor of whether an invehicle device produces unacceptable levels of distraction. Accid Anal Prev. 2008;40:628–634. Smith DL, Chang J, Cohen D, Foley J, Glassco R. A simulation approach for evaluating the relative safety impact of driver distraction during secondary tasks. Paper presented at: 12th World Congress on Intelligent Transport Systems; November 6–10, 2005; San Francisco, CA. Strayer DL, Drews FA, Crouch DJ, Johnston WA. Why do cell phone conversations interfere with driving? In: Walker WR, Herrmann D, eds. Cognitive Technology: Essays on the Transformation of Thought and Society. Jefferson, NC: McFarland & Co; 2005:51–68. Terry HR, Charlton SG, Perrone JA. The role of looming and attention capture in drivers’ braking responses. Accid Anal Prev. 2008;40:1375–1382. Young K, Regan M. Driver distraction: A review of the literature. In: Faulks IJ, Regan M, Stevenson M, Brown J, Porter A, Irwin JD, eds. Distracted Driving. Sydney, Australia: Australasian College of Road Safety; 2007:379–405.
