- Email: [email protected]
Mutual interferences of driving and texting performance
Computers in Human Behavior 52 (2015) 115–123
Contents lists available at ScienceDirect
Computers in Human Behavior journal homepage: www.elsevier.com/locate/comphumbeh
Mutual interferences of driving and texting performance Jibo He a,⇑, Alex Chaparro a, Xiaohui Wu b, Joseph Crandall a, Jake Ellis a a b
Department of Psychology, Wichita State University, Wichita, KS, USA Department of Psychology, Tsinghua University, Beijing, China
a r t i c l e
i n f o
Article history:
Keywords: Texting while driving Driver distraction Mobile devices Lane Change Task Cell phone
a b s t r a c t Despite legislative and social campaigns to reduce texting while driving, drivers continue to text behind the wheel. There is abundant evidence demonstrating that texting while driving impairs driving performance. While past driver distraction research has focused on how texting influences driving, the influence of driving on texting behaviors has been overlooked. This study used a Lane Change Task and a smartphone texting application to study the mutual influences of driving and texting. Results showed that concurrent texting impaired driving by increasing the lane deviation. Meanwhile, driving impaired texting by increasing texting completion time, texting errors, and key entry time intervals, and reduced key entry speed. In addition, we show that texting behavioral data collected can be used to distinguish texting while driving from texting-only condition with an accuracy of 88.5%. The mutual interferences of driving and texting inform the theory of dual-task performance and provide a scientific foundation to develop a smartphone-based technology to reduce the risky behavior of texting while driving. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Texting while driving has become a widespread risky behavior and may impair driving performance more than talking on a cell phone (Caird, Johnston, Willness, Asbridge, & Steel, 2014; Drews, Yazdani, Godfrey, Cooper, & Strayer, 2009; He et al., 2014; Hosking, Young, & Regan, 2009; Klauer et al., 2014; Wilson & Stimpson, 2010). As many as 281,000 to 786,000 crashes in 2012 may involve text messaging, according to the estimate of the United States National Safety Council (2012). Pickrell and Ye (2013) found 0.9% of drivers were visibly manipulating hand-held devices while driving in 2010, and this percentage increased to 1.3 percent in 2011. The risk and prevalence of texting while driving has attracted the attention of the general public, auto manufacturers, legislators and safety researchers (Jacobson & Gostin, 2010; Owens, McLaughlin, & Sudweeks, 2011). Concurrent texting impairs driving in various ways. For example, texting while driving increases hazard response time (Burge & Chaparro, 2012; Drews et al., 2009; He et al., 2014; Leung, Croft, Jackson, Howard, & Mckenzie, 2012), increases lane deviations (the difference between the center of the vehicle and the center of the appropriate lane) and lane excursions (leaving the lane unintentionally) (Alosco et al., 2012; Crandall & Chaparro, 2012; ⇑ Corresponding author at: Wichita State University, 1845 Fairmount St., Wichita, KS 67260, USA. Cell: +1 217 417 3830. E-mail address: [email protected] (J. He). http://dx.doi.org/10.1016/j.chb.2015.05.004 0747-5632/Ó 2015 Elsevier Ltd. All rights reserved.
Hosking et al., 2009; McKeever, Schultheis, Padmanaban, & Blasco, 2013; Rudin-Brown, Young, Patten, Lenné, & Ceci, 2013), increases mental demand (mental demands are psychological and mental stress experienced by an individual while completing one or more tasks) (Owens et al., 2011; Rudin-Brown et al., 2013), increases gaze-off-road durations (Hosking et al., 2009; Libby, Chaparro, & He, 2013; Owens et al., 2011), causes more collisions (Alosco et al., 2012; Drews et al., 2009), and raises the risks of traffic accident as many as 8–23 times (Olson, Hanowski, Hickman, & Bocanegra, 2009). People have limited ability to perform two tasks simultaneously, such as texting and driving and doing so results in deficits on one or both of the tasks being performed (Allport, Antonis, & Reynolds, 1972). According to the theories of dual-task performance, when two tasks are carried out concurrently, the performances of one or both tasks may be impaired, causing a dual-task performance decrement (Wickens, 2002). For example, when performing a secondary auditory monitoring task (pressing a button when they hear a tone), drivers had slower reaction time when responding to a vehicle braking, compared to driving-only conditions, even when instructed to give the driving task priority (Levy & Pashler, 2008). While several studies have reported the effects of texting on driving performance, how driving affects texting performance has been ignored. Better understanding of changes in both driving and texting performance can inform theories of dual-task performance and contribute to the efforts to mitigate the risks of texting while driving.
116
J. He et al. / Computers in Human Behavior 52 (2015) 115–123
Researchers and practitioners have explored a variety of approaches to mitigate the risks of texting while driving, including legislation, social campaigns, and technological solutions. Legislative efforts have sought to discourage this behavior by making it illegal (McCartt & Geary, 2004; Mccartt, Hellinga, Strouse, & Farmer, 2010) and social campaigns have sought to educate drivers about the risks of texting while driving (Atchley, Hadlock, & Lane, 2012; Nemme & White, 2010). In addition, telecommunication companies including AT&T and T-Mobile have developed smartphone applications to discourage texting while driving, such as Drive Mode and DriveSmart. Smartphone users can enable these applications to delay or block incoming calls and messages when they drive, limiting their exposure to the dangers of texting while driving. However, despite the associated risks of texting while driving and these legislative, social, and technological efforts, drivers continue to text while driving. Ninety-one percent of college students reported having sent text messages while driving, even though they agreed or believed that texting while driving was dangerous and should be illegal (Atchley, Atwood, & Boulton, 2011; Harrison, 2011). Is it possible to develop a smartphone application to monitor texting while driving, and prevent or discourage drivers from engaging in such risk behavior? If an application can monitor texting while driving, it can potentially be easier to implement, and can complement current efforts in law enforcement or social campaigns. Researchers have attempted to detect drunk driving (Dai, Teng, Bai, Shen, & Xuan, 2010), cognitively distracted driving (Liang, Lee, & Reyes, 2007), aggressive driving (Johnson & Trivedi, 2011; Zeeman & Booysen, 2013), and drowsy driving (Hammoud & Zhang, 2008; He et al., 2013). However, to our best knowledge, no application has been developed that detects texting while driving. Thus, this study also explores the possibility using texting behavioral data to identify whether a driver is texting while driving. The popularity and risks of texting while driving highlight the continuing need for research and understanding of how texting influences driving and vice versa, and how drivers coordinate performance of both tasks (Atchley et al., 2011; Harrison, 2011). Most driving studies focus on driving performance, while texting performance is mostly ignored or less emphasized. Smartphones allow the collection of detailed data on secondary texting performance. In this study, we utilize the capability of smartphones to capture texting performance and describe the mutual interferences of texting and driving. The goal of this paper is to discover the mutual influences of concurrent texting and driving, and sought the possibility to detect texting while driving and reduce its risks. It is predicted that participants will show greater lane deviation while driving and texting compared to the drive-only condition. It is also hypothesized that participants will take longer to complete the texting task and make more errors when driving and texting than when driving-only.
touchscreen smartphone and on average, reported sending 83 text messages per day (median = 70, SD = 86.06). 2.2. Apparatus and stimuli Driving performance was assessed using a driving simulator consisting of a General Motors car seat and Logitech Driving Force GT steering wheel and pedals. The Lane Change Task (LCT) version 1.2 software simulated the driving task using a 60 inch-Sharp Aquos 3D HD LCD display. A 4.300 HTC ThunderBolt touch-screen smartphone running the Android 2.3.4 operating system was used for the texting task. The buttons on the keyboard of the smartphone were arranged in a QWERTY layout. 2.3. Experimental design There were five counterbalanced experimental conditions, including driving-only condition, two dual-task conditions, in which participants either drove while texting with one hand (drive + text one hand) or two hands (drive + text two hands), and two texting-only conditions in which participants completed the texting task with either one or two hands. We employed a within-subject design. Participants finished all the task conditions. The dependent variables measured for the driving task were mean lane deviation and standard deviation of lane deviation. Lane deviation refers to the difference between the center of the vehicle and the center of the appropriate lane. For the texting task, the dependent variables consisted of task completion time, key entry per second, texting task completion time, texting errors, input time interval, standard deviation of input time interval, and device stability (He et al., 2014). 2.4. Experiment tasks 2.4.1. Lane Change Task (LCT) In the LCT, participants were required to drive down a straight section of road with three lanes and were prompted to change lanes according to directions on signs, which appeared on both sides of the roadway. An arrow on the sign, shown in Fig. 1, indicated which lane the driver supposed to maneuver into. Participants were instructed to change from their current lane in a deliberate manner, and to do so as quickly and efficiently as possible. Participants maintained a constant speed of 60 km/h and were instructed to make lane changes as quickly and accurately as possible when they saw the lane change sign. The LCT was developed by Daimler Chrysler AG Research and Technology to test driver distraction (Hofmann, Rinkenauer, & Gude, 2008; Huemer & Vollrath, 2010; Mattes & Hallén, 2008).
2. Materials and methods 2.1. Participants Twenty-eight participants (12 men, 16 women) from a university community ages 18–35 years (M = 22.14 years, SD = 4.64 years) volunteered to participate in this driving experiment. All participants were screened prior to participation to ensure normal or corrected-to-normal vision using the Snellen Visual Acuity chart (Ferris, Kassoff, Bresnick, & Bailey, 1982). All participants completed a survey to ensure they were right-handed, active drivers, with at least two years of driving experience (M = 6.29 years; SD = 4.51 years). They all owned a
Fig. 1. Screenshot of driving simulator environment.
J. He et al. / Computers in Human Behavior 52 (2015) 115–123
2.4.2. Texting task A texting task was used to assess texting performance with two holding postures (texting with one hand vs. texting with two hands). An incoming text message sound was played at a random time interval, which followed a uniform distribution in the range of 40–60 s. Once the incoming sound was played, participants opened the texting application to send text messages. Participants were presented with a text message and instructed to enter the text message into the phone manually. After the message was sent, the application was closed. The text message consisted of a 10-digit telephone number. The digit-entry texting task was chosen to simulate the cognitive process and key entry requirement of a cell phone dialing or texting task while also allowing the measurement of secondary texting performance. Similar digit-entry tasks are commonly used to study the effect of distraction on driving performance (He et al., 2014; Horrey & Wickens, 2004; Reed & Green, 1999; Salvucci & Macuga, 2002). The application saved timestamps, keyboard entries, and accelerometer sensor values of the smartphone in log files for later analysis. 2.5. Procedure Participants were informed of the purpose of the experiment and then signed a consent form to participate. They then completed a questionnaire regarding driving experience and cell phone use, followed by a visual screening. The session began with instructions on how to complete the tasks and a practice session to familiarize the participants with the simulator and the cell phone. Participants received instruction on the LCT and how to correctly execute the task. Participants were instructed to prioritize the driving task during dual-task conditions and text when they felt comfortable or believed it was safe. The phone was used in portrait orientation for all texting conditions. In the two-hand texting conditions, participants were instructed to keep both hands on the phone while texting. Participants practiced each of the text-only conditions for two minutes and each of the driving conditions for six minutes, for a total of 20 min of practice. After the practice, participants began the experimental trials. Each trial consisted of four tracks lasting three minutes each, for a total of 12 min. The time to complete the survey, practice time, and the experiment took approximately 105 min. Participants could withdraw from the study at any time. After the study, participants were given course credits for their participation. 2.6. Data analysis Measures of driving performance included the mean and standard deviation of lane deviation recorded by the driving simulator. Lane deviation was the difference between the driver’s actual lane position (red line) and the ‘‘ideal model’’ of lane position (green line) for the LCT, seen in Fig. 2. A large value of the standard
117
deviation of lane deviation indicates poorer lane-keeping performance with a higher risk of lane departure and collision with vehicles in the adjacent lanes. Texting performance measures included texting task completion time, key entry per second (KES), mean and standard deviation of input time interval, mean numbers of inputs, and mean number of corrections. Texting errors were measured using the Levenshtein edit distance of two strings (simply termed as edit distance) (He et al., 2014; Levenshtein, 1966). The edit distance is the number of discrete steps required to make the strings identical, including insertions, deletions, switching, and substitutions. Larger edit distance values indicate a greater difference between the two strings or more texting errors. For example, the numbers 5200251314 and 200251319 have a Levenshtein distance of two. The pair of numbers can be made identical by inserting a 5 at the beginning (insertion) and substituting the last 9 into 4 (substitution). All the driving performance measures were submitted to repeated-measures one-way analyses of variance (ANOVA) with the driving condition (drive-only, drive + texting with one hand, and drive + texting with two hands) as the only within-subject factors. For the one-way ANOVA of driving performance, when the main effect is significant, pairwise comparisons were made between the driving conditions. Bonferroni correction was used for post hoc pairwise comparisons. The Bonferroni-corrected p value is 0.05/3 = 0.017. Texting performance variables were submitted to two-way analyses of variance (ANOVA) with task load (text-only vs. drive + texting) and texting methods (one hand vs. two hands) as within-subject factors. Partial eta-squared (g2p) was used to measure effect size. A g2p value of 0.04 indicates a recommended minimum effect size for a practically significant effect, 0.25 indicates a moderate effect and 0.64 indicates a strong effect (Ferguson, 2009). 3. Results 3.1. Driving performance The standard deviation of lane deviation was significantly different across task conditions, F(2, 54) = 29.14, p < .001, g2p = .52. As shown in Fig. 3 pairwise comparisons revealed the standard deviation of lane deviation in the drive-only condition (M = 1.09 m, SD = .20 m) was significantly smaller than that in the drive + texting with one hand condition (M = 1.21 m, SD = .26 m, p = .001) and drive + texting with two hands (M = 1.27 m, SD = .28 m, p < .001). The standard deviation of lane deviation in the drive + texting one hand condition was also significantly smaller than that in the drive + texting two hands condition. The mean lane deviation also differed significantly across task conditions, F(2, 54) = 41.43, p < .001, g2p = .61. As shown in Fig. 4, the mean lane deviation in the drive-only condition (M = .81 m, SD = .15 m) was significantly smaller than that in the drive + texting with one hand condition (M = .94 m, SD = .21 m), and
Fig. 2. The Lane Change Task. The green line represents the ideal path defined by the computer and the red line represents the path taken by the driver. The arrow symbol at the top indicates the lane, which drivers should drive in (Mattes & Hallén, 2008). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
118
J. He et al. / Computers in Human Behavior 52 (2015) 115–123
Fig. 5. Texting task completion time across conditions. Fig. 3. Standard deviation of lane deviation across task conditions. Error bars in all figures indicate within-subject standard errors based on the main effect of task (Loftus & Masson, 1994).
Fig. 4. Mean lane deviation across task conditions.
to complete the texting task while driving regardless of whether they texted with either one hand or two hands. A two-way within subjects analysis of variance was conducted to evaluate the key entry per second between driving while texting with one or two hands and texting only with one or two hands. As shown in Fig. 6, the key entry per second (KES) produced a marginally significant interaction of task load and texting methods, F(1, 27) = 3.91, p = .06, g2p = .13. The main effect of task load was significant, F(1, 27) = 226.31, p < .001, g2p = .89. Simple main effect showed that the KES in the texting-only conditions (M = 1.17, SD = .19) was significantly larger than that of drive + texting conditions (M = .67, SD = .13), t(27) = 15.04, p < .001. The main effect of texting methods was not significant, F(1, 27) = .07, p = .79, g2p = .003. Participants inputted more keys per second when texting-only compared to texting while driving. A two-way within-subjects analyses of variance was conducted to evaluate the average texting errors between texting while driving with one or two hands and texting only with one or two hands. The average texting errors (as shown in Fig. 7) differed significantly across task load conditions, F(1, 27) = 215.13, p < .001, g2p = .89. When texting while driving, drivers produced more texting errors (M = .48, SD = .66) than the texting-only condition (M = .11, SD = .14). The main effect of texting methods and interaction effect
drive + texting with two hands condition (M = .98 m, SD = .24 m), t(27) = 6.86, p < .001 and t(27) = 7.66, p < .001 respectively. The mean lane deviation in the drive + texting with one hand condition was significantly different from the drive + texting with two hands condition, t(27) = 2.56, p = .02. Participants deviated their trajectory from the ‘‘ideal model’’ when texting with two hands more than when texting using one hand or driving-only.
3.2. Texting performance A two-way within-subjects analyses of variance was conducted to evaluate the task completion time between driving while texting with one or two hands and texting only with one or two hands. The texting task completion time (as shown in Fig. 5) produced a significant main effect of task load, F(1, 27) = 175.42, p < .001, g2p = .87, with longer completion time in the drive + texting conditions (texting using either one hand or two hands) (M = 16.73 s, SD = 3.64 s) than the texting-only conditions (M = 9.57 s, SD = 2.12 s), but the main effect of texting methods and the interaction were not significant. Participants took longer
Fig. 6. Key entry per second across conditions.
J. He et al. / Computers in Human Behavior 52 (2015) 115–123
Fig. 7. The texting error measured by Levenshtein edit distance (He et al., 2014; Levenshtein, 1966).
were not significant, F(1, 27) = .59, p = .45, g2p = .02 and F(1, 27) = 1.64, p = .21, g2p = .06 respectively. A two-way within-subjects analyses of variance was conducted to evaluate the mean input time interval between driving while texting with one or two hands and texting only with one or two hands. The mean input time interval (shown in Fig. 8) also differed significantly across task conditions, F(1, 27) = 184.19, p < .001, g2p = .87. The mean input time interval was significantly higher in the drive + texting condition (M = 1.46 s, SD = .32 s) than that of texting-only condition (M = .82 s, SD = .17 s). Concurrent texting and driving delayed key entry by .64 s (78%) compared to the texting-only condition. The main effect of texting methods and interaction effect were not significant, F(1, 27) = .31, p = .58, g2p = .01 and F(1, 27) = 1.61, p = .22, g2p = .06 respectively. Participants paused more between key stroke while driving and texting compared to texting-only. A two-way within-subjects analyses of variance was conducted to evaluate the standard deviation of input time interval between driving while texting with one or two hands and texting only with one or two hands. The standard deviation of input time interval (shown in Fig. 9) differed significantly across task conditions, F(1, 27) = 46.69, p < .001, g2p = .63 and texting methods, F(1, 27) = 7.36, p = .01, g2p = .21, but the interaction was not significant, F(1, 27) = .67, p = .42, g2p = .02. The standard deviation of
Fig. 8. Mean input time interval.
119
Fig. 9. Standard deviation of input time interval.
input time interval was significantly higher in the drive + texting condition (M = 2.12 s, SD = .78 s) than that of texting-only condition (M = .91 s, SD = 1.01 s). The standard deviation of input time interval was also higher in the two-hand texting condition (M = 1.71 s, SD = 1.33 s) than the one-hand texting condition (M = 1.32 s, SD = .72 s). Participant made 10.77 key entries on average with a standard deviation of .79. Analysis of the number of key entries did not produce any significant effects, all ps > .10. Participants revised their entries by pressing the ‘Back’ button on average .26 times with a standard deviation of .25. Analysis of the number of times that the ‘Back’ button was pressed did not yield any significant main effect, all ps > .10. 3.3. Device stability We also measured the stability of the Android smartphone when participants texted while driving. Device stability was measured using the standard deviation of the accelerometer values for the X, Y, and Z coordinates of the smartphone. The Android coordinate system (as shown in Fig. 10) is defined relative to the screen of the phone in the portrait orientation. The axes are not swapped when the device’s screen orientation changes. The X-axis is horizontal and points to the right, the Y-axis is vertical and points up and the Z-axis points toward the outside of the front face of the screen. In this system, coordinates behind the screen have negative Z values. The X, Y, and Z vectors of the accelerator sensor were automatically saved to a log file when participants sent text messages using the smartphone. The accelerator’s values were sampled at a rate of 100 Hz only when participants used the texting application. For the standard deviation of the accelerometer values in the X-axis, the main effects of task load and texting methods were both significant, F(1, 27) = 47.37, p < .001, g2p = .64 and F(1, 27) = 35.20, p < .001, g2p = .57 respectively. The interaction was also significant, F(1, 27) = 39.44, p < .001, g2p = .59. Simple main effect test showed that when texting using one hand, the standard deviation of the accelerator values in the X-axis was not significantly different for the drive + texting condition (M = .58, SD = .15) and the texting-only condition (M = .53, SD = .19), t(27) = 1.26, p = .22. When texting using two hands, the standard deviation of the accelerator values in the X-axis was significantly larger in the drive + texting condition (M = 1.20, SD = .47) than the texting-only condition (M = .52, SD = .27), t(27) = 7.16, p < .001. For the standard deviation of the accelerometer values in the Y-axis, a significant interaction was found, F(1, 27) = 42.74,
120
J. He et al. / Computers in Human Behavior 52 (2015) 115–123
Fig. 10. The Android coordinate system for accelerator sensors.
p < .001, g2p = .61. The main effects of task load and texting methods were not significant, F(1, 27) = .05, p = .82, g2p = .002 and F(1, 27) = 2.56, p = .12, g2p = .09 respectively. Simple main effect tests showed that when texting using one hand, the standard deviation of the accelerometer values in the Y-axis was significantly smaller in the drive + texting condition (M = .72, SD = .15) than the texting-only condition (M = .91, SD = .38), t(27) = 2.59, p = .02. When texting using two hands, the standard deviation of the accelerometer values in the Y-axis was significantly larger in the drive + texting condition (M = .99, SD = .34) than the texting-only condition (M = .78, SD = .31), t(27) = 3.02, p = .01. For the standard deviation of the accelerometer values in the Z-axis, a significant interaction was found, F(1, 27) = 50.46, p < .001, g2p = .65. The main effects of task load and texting methods were also significant, F(1, 27) = 9.26, p = .01, g2p = .26 and F(1, 27) = 13.52, p = .001, g2p = .33 respectively. Simple main effect tests showed that when texting using one hand, the standard deviation of the accelerometer values in the Z-axis was significantly smaller in the drive + texting condition (M = .74, SD = .14) than the texting-only condition (M = .80, SD = .26), t(27) = 1.29, p = .21. When texting using two hands, the standard deviation of the accelerometer values in the Z-axis was significantly larger in the drive + texting condition (M = 1.07, SD = .33) than the texting-only condition (M = .71, SD = .19), t(27) = 5.25, p < .001. See Fig. 11 for the device stability as measured using the accelerometer sensor values in the X, Y, and Z-axis. The interaction effects in the above analyses is because the instability of the devices (as measured by the standard deviation of accelerator values in the X, Y, and Z-axis) is more pronounced in the texting while driving using two hands condition than the texting while driving using one hand condition. More instability to hold the smartphone in the condition when drivers texting using two hands was because the inability to share manual resources (e.g. the hands). In contrast, drivers in the one hand texting condition can hold the phone in one hand and control the steering wheel using the other hand. Thus, for drivers who texted with two hands, texting while driving significantly increased the instability of the smartphone, which indicates more manual interference of texting and driving and higher chances of dropping or even breaking the smartphone when driving (see Fig. 11). 3.4. Detect texting while driving To reduce the risks of texting while driving, it is important to detect the occurrence of such driving behaviors. We used a logistic regression to detect texting while driving, using texting data collected from the smartphone. The variables in the logistic regression include the number of key entries, mean input time interval, standard deviation of input time interval, the number of key entries,
Fig. 11. Device stability as measured using the accelerometer sensor values in the X, Y, and Z-axis.
and the number of times the ‘Back’ button was pressed. The texting task completion time and texting errors were hard to define in natural free-style texting, thus the two variables were not included in the analysis. The dependent variable was the texting conditions, either drive + texting or texting-only. To predict texting while driving we need a ‘‘training data set’’ to develop the logistic regression model. The training data set can be generic, that is based on the data from a group of participants or
121
J. He et al. / Computers in Human Behavior 52 (2015) 115–123
individualized meaning that a participant’s data is used to derive a their own user specific logistic regression model. If a common training data set can be used to monitor texting while driving reliably for all subjects, the proposed application to monitor texting while driving can be developed more easily, without the necessity to consider individual texting behaviors and does not require the collection of user specific data sets for model development. A user-specific training set may achieve higher prediction performance than a model based on a generic data set. We explored the prediction performance using both generic training set and individualized training set. 3.4.1. Prediction based on generic training set The experiment produced a total of 1748 texting instances. 1224 texting instances (70% of all instances) were randomly selected as the training set. The other 524 instances (30%) were used as the prediction set. The best logistic regression model based on generic training data set is:
Table 2 Confusion matrix to predict texting while driving based on user-specific training data. Predicted
Observed
Texting-only Drive + texting
Texting-only
Drive + texting
5.45 1.52
0.69 9.66
behavioral data, such as mean and standard deviation of the input time interval. The above data suggest that we can use texting behavior data to predict texting while driving reliably with at least 88.5% accuracy. The accuracy using generic training data set (88.5%) is only slightly lower than the accuracy based on user-specific training data set (89.7%). The minor increase in the prediction accuracy of user-specific training data should be a result of the significant large increase in the mean and standard deviation of texting input intervals. This result suggests that we can just use a generic logistic regression model to predict texting while driving, which is easier to implement than a user-specific training data set.
Task load ¼ 2:31 :0030 mean input time interval :0016 standard deviation of input time interval Where for task load, 0 indicates texting-only condition and 1 indicates drive + texting condition. The classification accuracy for the generic training set is shown in Table 1. An accuracy as high as 88.5% was achieved using texting behavior data to classify texting while driving from the texting-only condition. The sensitivity of the model based on generic training set is .88, and the specificity of the model is .94. Our accuracy in detecting texting while driving is comparable to or sometimes higher than other attempts to detect cognitive distraction (Liang, Reyes, & Lee, 2007) and visual distraction (Kutila, Jokela, Markkula, & Rue, 2007). For example, Liang, Reyes et al. (2007) achieved an average accuracy of 81.1% in detecting driver cognitive distraction using Support Vector Machines. Our high detection accuracy suggests it is possible to develop a smartphone application to detect texting while driving using texting behavioral data, such as mean and standard deviation of the input time interval. 3.4.2. Prediction based on user-specific training set User-specific training set can potential improve model prediction by learning the individual difference in texting behaviors. To explore to what extent user-specific training set can improve our ability to predict texting while driving, we also trained twenty-eight logistic regression models for each subject. Each driver produced about 58.07 texting instances in total. 70% of the texting instances for each subject were used as the training data set and the other 30% texting instances were used as the testing data set. The classification accuracy for the user-specific training data set is shown in Table 2. The average accuracy as high as 89.7% was achieved using texting behavior data to classify texting while driving from the texting-only condition. The sensitivity of the model based on generic training set is .94, and the specificity of the model is .82. The high accuracy suggests it is possible to develop a smartphone application to detect texting while driving using texting
Table 1 Confusion matrix to predict texting while driving based on generic training data. Predicted
Observed
Texting-only Drive + texting
Texting-only
Drive + texting
233 16
29 210
4. Discussion Using a classic Lane Change Task and smartphone technology, we found evidence of mutual interference between driving and texting. Texting impaired driving by increasing lane deviations, which resonates with previous studies showing that texting impaired driving performance (Drews et al., 2009; He et al., 2014; Hosking et al., 2009). To our best knowledge, existing studies have often ignored the other aspect of dual-task performance, that is, how driving influences texting? In addition to the finding that texting impairs driving, our study also showed that driving impaired texting by slowing text entry and increasing texting errors. The mutual interference of texting and driving contributes to our understanding of dual-task performance. Interestingly, we can use a logistic regression model based on texting behaviors to detect texting while driving with accuracy as high as 88.5%. Texting can increase a driver’s crash risks and endanger their lives. Drivers may continue to text for a variety of reasons. Drivers may read or send texts out of habit, regardless their subjective perception of the risks of such behaviors (Bayer & Campbell, 2012) and are more likely to take a phone call if they believe the call is important (Nelson, Atchley, & Little, 2009). They may send texts to reduce unpleasant emotions (Feldman, Greeson, Renna, & Robbins-Monteith, 2011) or maintain a social relationship by replying promptly to a text message. For these drivers, a clear message that ‘‘driving and texting interfere mutually’’ should be delivered in social campaigns to further discourage texting while driving. Practically, a better understanding of texting performance may inform the development of smartphone applications that can detect and perhaps reduce texting while driving. Several researchers are beginning to explore smartphone applications that can monitor impaired driving performance (Dai et al., 2010; Hammoud & Zhang, 2008; He et al., 2013; Johnson & Trivedi, 2011; You et al., 2013; Zeeman & Booysen, 2013;), such as drunk, distracted, aggressive and drowsy driving. However, no attempts have been made to detect texting while driving yet. Our study demonstrated that a logistic regression model using texting behavioral data (such as mean and standard deviation of input time interval) can detect texting while driving with accuracy as high as 88.5%. This finding provides a theoretical support to develop a smartphone application, which can detect and prevent texting while driving. A smartphone-based application to monitor texting while driving has the advantage to be easier to implement (Eren, Makinist, Akin, & Yilmaz, 2012), while it is very costly for law enforcement or social campaigns to prevent texting while driving.
122
J. He et al. / Computers in Human Behavior 52 (2015) 115–123
More studies are needed to improve our training data set and algorithm to detect texting while driving. For example, an on-road driving study using a natural texting task is needed to collect valuable training data set to test and improve our algorithm. Many cell phone carriers and cell phone manufacturers offer ‘‘Drive Mode’’ applications, which allow users to block incoming calls and messages or send automatic responses. These applications often require users to initiate the drive mode before driving and turn it off after driving, which is a bothersome process and limits the use of these applications. Automatic detection of texting while driving can improve existing ‘‘Drive Mode’’ applications. If smartphone applications incorporate automatic detection of texting while driving and stop or delay messages from distracting drivers, these applications can be more effective in reducing texting while driving than existing applications. Our study demonstrated that we can reliably detect texting while driving using generic training set based on a group of users. Our finding provides new knowledge to improve existing ‘‘Drive Mode’’ applications. Future studies shall investigate whether other information, including driving dynamics, can be used to develop smartphone-based technology to detect texting while driving. For example, GPS information can be used to calculate travelling velocity, which could then be used as a precondition for identifying whether the user is texting while driving. The smartphone application will detect texting while driving only when users travel at higher speeds, above 20 mph, which would greatly improve the performance of the application. Researchers have focused on how to use smartphone capabilities to improve driving safety (He et al., 2012, 2013, 2014; Ren, Wang, & He, 2013; Wang, Cardone, Corradi, Torresani, & Campbell, 2012). To our best knowledge, this is one of the first papers to demonstrate the possibility of using smartphone technology to detect texting while driving. This technology is an innovative way to mitigate the risks of texting while driving, which complements the existing social and legal approaches. With a better understanding of texting and driving behaviors and the rapid advancement of smartphone technology, smartphones can be used to make driving safer, instead of more dangerous (He et al., 2013; Ren et al., 2013). As Dr. Adesman stated, ‘‘Technological solutions will likely need to be developed to significantly reduce the frequency of texting while driving’’ (American Academy of Pediatrics, 2013; Lindqvist & Hong, 2011). Acknowledgements We appreciate the valuable suggestions from Frank Schieber, Yulan Liang, Chun Wang, and Evan Palmer. We also acknowledge the help in data collection from the undergraduate research assistants Colton Turner and Kirsten Turner. The authors would acknowledge to the National Natural Science Foundation of China with Grant 71401004 and the U.S. Department of Transportation (DOT) through the University Transportation Centers program sponsored by Research and Innovative Technology Administration (RITA). References Allport, D. A., Antonis, B., & Reynolds, P. (1972). On the division of attention: A disproof of the single channel hypothesis. The Quarterly Journal of Experimental Psychology, 24(2), 225–235. Alosco, M. L., Spitznagel, M. B., Fischer, K. H., Miller, L. A., Pillai, V., Hughes, J., et al. (2012). Both texting and eating are associated with impaired simulated driving performance. Traffic Injury Prevention, 13(5), 468–475. http://dx.doi.org/ 10.1080/15389588.2012.676697. American Academy of Pediatrics (2013). Don’t txt n drive: Teens not getting msg: 43 percent of youths admit to texting while driving. Science Daily. 4 May 2013. Atchley, P., Atwood, S., & Boulton, A. (2011). The choice to text and drive in younger drivers: Behavior may shape attitude. Accident Analysis & Prevention, 43(1), 134–142. http://dx.doi.org/10.1016/j.aap.2010.08.003.
Atchley, P., Hadlock, C., & Lane, S. (2012). Stuck in the 70s: The role of social norms in distracted driving. Accident Analysis & Prevention, 48, 279–284. http:// dx.doi.org/10.1016/j.aap.2012.01.026. Bayer, J. B., & Campbell, S. W. (2012). Texting while driving on automatic: Considering the frequency-independent side of habit. Computers in Human Behavior, 28(6), 2083–2090. http://dx.doi.org/10.1016/j.chb.2012. 06.012. Burge, R., & Chaparro, A. (2012). The effects of texting and driving on hazard perception. In Proceedings of the human factors and ergonomics society annual meeting (Vol. 56, No. 1, pp. 715–719) doi: http://dx.doi.org/10.1177/ 1071181312561149. Caird, J. K., Johnston, K. A., Willness, C. R., Asbridge, M., & Steel, P. (2014). A metaanalysis of the effects of texting on driving. Accident Analysis & Prevention, 71, 311–318. http://dx.doi.org/10.1016/j.aap.2014.06.005. Crandall, J. M., & Chaparro, A. (2012). Driver distraction: Effects of text entry methods on driving performance. In Proceedings of the human factors and ergonomics society annual meeting (Vol. 56, No. 1, pp. 1693–1697). doi: http:// dx.doi.org/10.1177/1071181312561339. Dai, J., Teng, J., Bai, X., Shen, Z., & Xuan, D. (2010). Mobile phone based drunk driving detection. In Pervasive computing technologies for healthcare (PervasiveHealth). doi: http://dx.doi.org/10.4108/ICST.PERVASIVEHEALTH2010.8901. Drews, F. A., Yazdani, H., Godfrey, C. N., Cooper, J. M., & Strayer, D. L. (2009). Text messaging during simulated driving. Human Factors, 51(5), 762–770. Eren, H., Makinist, S., Akin, E., & Yilmaz, A. (2012). Estimating driving behavior by a smartphone. In 2012 IEEE intelligent vehicles symposium (IV) (pp. 234–239). doi: http://dx.doi.org/10.1109/IVS.2012.6232298. Feldman, G., Greeson, J., Renna, M., & Robbins-Monteith, K. (2011). Mindfulness predicts less texting while driving among young adults: Examining attentionand emotion-regulation motives as potential mediators. Personality and Individual Differences, 51(7), 856–861. http://dx.doi.org/10.1016/j.paid.2011. 07.020. Ferguson, C. J. (2009). An effect size primer: A guide for clinicians and researchers. Professional Psychology: Research and Practice, 40(5), 532–538. Ferris, F. L., Kassoff, A., Bresnick, G. H., & Bailey, I. (1982). New visual acuity charts for clinical research. American Journal of Ophthalmology, 94(1), 91–96. Hammoud, R. I., & Zhang, H. (2008). Alertometer: Detecting and mitigating driver drowsiness and fatigue using an integrated human factors and computer vision approach. In R. I. Hammoud (Ed.), Passive Eye Monitoring (pp. 301–321). Berlin Heidelberg: Springer. Retrieved from http://link.springer.com/chapter/10.1007/ 978-3-540-75412-1_14. Harrison, M. A. (2011). College students’ prevalence and perceptions of text messaging while driving. Accident Analysis & Prevention, 43(4), 1516–1520. http://dx.doi.org/10.1016/j.aap.2011.03.003. He, J., Chaparro, A., Nguyen, B., Burge, R. J., Crandall, J., Chaparro, B., et al. (2014). Texting while driving: Is speech-based text entry less risky than handheld text entry? Accident Analysis & Prevention, 72, 287–295. http://dx.doi.org/10.1016/ j.aap.2014.07.014. He, J., Fields, B.M., Roberson, S., Cielocha, S., Peng, J., & Coltea, J. (2012). ‘‘System and method to monitor and reduce vehicle operator impairment.’’ U.S. Patent Application 13/717,506, filed December 17, 2012. He, J., Roberson, S., Fields, B., Peng, J., Cielocha, S., & Coltea, J. (2013). Fatigue detection using smartphones. Journal of Ergonomics, 03(03). http://dx.doi.org/ 10.4172/2165-7556.1000120. Hofmann, P., Rinkenauer, G., & Gude, D. (2008). Head-up-displays support response preparation in a Lane Change Task. In Proceedings of the human factors and ergonomics society annual meeting (Vol. 52, No. 18, pp. 1233–1237). doi: http:// dx.doi.org/10.1177/154193120805201817. Horrey, W. J., & Wickens, C. D. (2004). Driving and side task performance: The effects of display clutter, separation, and modality. Human Factors, 46(4), 611–624. Hosking, S. G., Young, K. L., & Regan, M. A. (2009). The effects of text messaging on young drivers. Human Factors, 51(4), 582–592. Huemer, A. K., & Vollrath, M. (2010). Alcohol-related impairment in the Lane Change Task. Accident Analysis & Prevention, 42(6), 1983–1988. http:// dx.doi.org/10.1016/j.aap.2010.06.005. Jacobson, P. D., & Gostin, L. O. (2010). Reducing distracted driving: regulation and education to avert traffic injuries and fatalities. JAMA, 303(14), 1419–1420. Johnson, D. A., & Trivedi, M. M. (2011). Driving style recognition using a smartphone as a sensor platform. In 2011 14th International IEEE Conference on Intelligent Transportation Systems (ITSC) (pp. 1609–1615). doi: http://dx.doi.org/10.1109/ ITSC.2011.6083078. Klauer, S. G., Guo, F., Simons-Morton, B. G., Ouimet, M. C., Lee, S. E., & Dingus, T. A. (2014). Distracted driving and risk of road crashes among novice and experienced drivers. New England Journal of Medicine, 370(1), 54–59. Kutila, M., Jokela, M., Markkula, G., & Rue, M. R. (2007). Driver distraction detection with a camera vision system. IEEE International Conference on Image Processing, 6, 201–204. Leung, S., Croft, R. J., Jackson, M. L., Howard, M. E., & Mckenzie, R. J. (2012). A comparison of the effect of mobile phone use and alcohol consumption on driving simulation performance. Traffic Injury Prevention, 13(6), 566–574. http://dx.doi.org/10.1080/15389588.2012.683118. Levenshtein, V. I. (1966). Binary codes capable of correcting deletions, insertions and reversals. Soviet Physics Doklady, Johnson, 10, 707. Levy, J., & Pashler, H. (2008). Task prioritization in multitasking during driving: Opportunity to abort a concurrent task does not insulate braking responses from dual-task slowing. Applied Cognitive Psychology, 22, 507–525.
J. He et al. / Computers in Human Behavior 52 (2015) 115–123 Liang, Y., Lee, J., & Reyes, M. (2007). Nonintrusive detection of driver cognitive distraction in real time using Bayesian Networks. Transportation Research Record: Journal of the Transportation Research Board, 2018, 1–8. http://dx.doi.org/ 10.3141/2018-01. Liang, Y., Reyes, M. L., & Lee, J. D. (2007). Real-time detection of driver cognitive distraction using support vector machines. IEEE Transactions on Intelligent Transportation Systems, 8(2), 340–350. Libby, D., Chaparro, A., & He, J. (2013). Distracted while driving: A comparison of the effects of texting and talking on a cell phone. In Proceedings of the human factors and ergonomics society annual meeting, (Vol. 57, No. 1, pp. 1874–1878) doi: http://dx.doi.org/10.1177/1541931213571418. Lindqvist, J., & Hong, J. (2011). Undistracted driving: A mobile phone that doesn’t distract. In Proceedings of the 12th workshop on mobile computing systems and applications (pp. 70–75). New York, NY, USA: ACM. http://dx.doi.org/10.1145/ 2184489.2184504. Loftus, G. R., & Masson, M. E. (1994). Using confidence intervals in within-subject designs. Psychonomic Bulletin & Review, 1(4), 476–490. Mattes, S., & Hallén, A. (2008). Surrogate distraction measurement techniques. In K. Young, J. D. Lee, & M. A. Regan (Eds.), Driver distraction: Theory, effects, and mitigation. CRC Press. McCartt, A. T., & Geary, L. L. (2004). Longer term effects of New York State’s law on drivers’ handheld cell phone use. Injury Prevention, 10(1), 11–15. http:// dx.doi.org/10.1136/ip.2003.003731. Mccartt, A. T., Hellinga, L. A., Strouse, L. M., & Farmer, C. M. (2010). Long-term effects of handheld cell phone laws on driver handheld cell phone use. Traffic Injury Prevention, 11(2), 133–141. McKeever, J. D., Schultheis, M. T., Padmanaban, V., & Blasco, A. (2013). Driver performance while texting: Even a little is too much. Traffic Injury Prevention, 14(2), 132–137. http://dx.doi.org/10.1080/15389588.2012.699695. National Safety Council (2012). Annual estimate of cell phone crashes 2012. Retrieved on 20.05.14 from. Nelson, E., Atchley, P., & Little, T. D. (2009). The effects of perception of risk and importance of answering and initiating a cellular phone call while driving. Accident Analysis & Prevention, 41(3), 438–444. http://dx.doi.org/10.1016/ j.aap.2009.01.006. Nemme, H. E., & White, K. M. (2010). Texting while driving: Psychosocial influences on young people’s texting intentions and behaviour. Accident Ana⁄⁄⁄lysis & Prevention, 42(4), 1257–1265. http://dx.doi.org/10.1016/j.aap.2010.01.019. Olson, R. L., Hanowski, R. J., Hickman, J. S., & Bocanegra, J. L. (2009). Driver distraction in commercial vehicle operations (No. FMCSA-RRR-09-042). Federal Motor Carrier Safety Administration, Washington, DC.
123
Owens, J. M., McLaughlin, S. B., & Sudweeks, J. (2011). Driver performance while text messaging using handheld and in-vehicle systems. Accident Analysis & Prevention, 43(3), 939–947. http://dx.doi.org/10.1016/j.aap.2010.11.019. Pickrell, T. M., & Ye, T. J. (2013). Driver electronic device use in 2011. (Report No. DOT HS 811 719). Washington, DC: National Highway Traffic Safety Administration. Reed, M. P., & Green, P. A. (1999). Comparison of driving performance on-road and in a low-cost simulator using a concurrent telephone dialling task. Ergonomics, 42(8), 1015–1037. Ren, Z., Wang, C., & He, J. (2013). Vehicle detection using Android smartphones. The 7th international driving symposium on human factors in driver assessment, training, and vehicle design. Retrieved from. Rudin-Brown, C. M., Young, K. L., Patten, C., Lenné, M. G., & Ceci, R. (2013). Driver distraction in an unusual environment: Effects of text-messaging in tunnels. Accident Analysis & Prevention, 50, 122–129. http://dx.doi.org/10.1016/ j.aap.2012.04.002. Salvucci, D. D., & Macuga, K. L. (2002). Predicting the effects of cellular-phone dialing on driver performance. Cognitive Systems Research, 3(1), 95–102. Wang, T., Cardone, G., Corradi, A., Torresani, L., & Campbell, A. T. (2012). WalkSafe: A pedestrian safety app for mobile phone users who walk and talk while crossing roads. In Proceedings of the Twelfth Workshop on Mobile Computing Systems & Applications (pp. 5:1–5:6). New York, NY, USA: ACM. http://dx.doi.org/ 10.1145/2162081.2162089. Wickens, C. D. (2002). Multiple resources and performance prediction. Theoretical Issues in Ergonomics Science, 3(2), 159–177. http://dx.doi.org/10.1080/ 14639220210123806. Wilson, F., & Stimpson, J. (2010). Trends in fatalities from distracted driving in the United States, 1999 to 2008. American Journal of Public Health, 100(11), 2213–2219. http://dx.doi.org/10.2105/AJPH.2009.187179. You, C.-W., Lane, N. D., Chen, F., Wang, R., Chen, Z., Bao, T. J., et al. (2013). CarSafe app: Alerting drowsy and distracted drivers using dual cameras on smartphones. In Proceeding of the 11th Annual International Conference on Mobile Systems, Applications, and Services (pp. 13–26). New York, NY, USA: ACM. http://dx.doi.org/10.1145/2462456.2465428. Zeeman, A. S., & Booysen, M. J. (2013). Combining speed and acceleration to detect reckless driving in the informal public transport industry. Retrieved from .
Contents lists available at ScienceDirect
Computers in Human Behavior journal homepage: www.elsevier.com/locate/comphumbeh
Mutual interferences of driving and texting performance Jibo He a,⇑, Alex Chaparro a, Xiaohui Wu b, Joseph Crandall a, Jake Ellis a a b
Department of Psychology, Wichita State University, Wichita, KS, USA Department of Psychology, Tsinghua University, Beijing, China
a r t i c l e
i n f o
Article history:
Keywords: Texting while driving Driver distraction Mobile devices Lane Change Task Cell phone
a b s t r a c t Despite legislative and social campaigns to reduce texting while driving, drivers continue to text behind the wheel. There is abundant evidence demonstrating that texting while driving impairs driving performance. While past driver distraction research has focused on how texting influences driving, the influence of driving on texting behaviors has been overlooked. This study used a Lane Change Task and a smartphone texting application to study the mutual influences of driving and texting. Results showed that concurrent texting impaired driving by increasing the lane deviation. Meanwhile, driving impaired texting by increasing texting completion time, texting errors, and key entry time intervals, and reduced key entry speed. In addition, we show that texting behavioral data collected can be used to distinguish texting while driving from texting-only condition with an accuracy of 88.5%. The mutual interferences of driving and texting inform the theory of dual-task performance and provide a scientific foundation to develop a smartphone-based technology to reduce the risky behavior of texting while driving. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Texting while driving has become a widespread risky behavior and may impair driving performance more than talking on a cell phone (Caird, Johnston, Willness, Asbridge, & Steel, 2014; Drews, Yazdani, Godfrey, Cooper, & Strayer, 2009; He et al., 2014; Hosking, Young, & Regan, 2009; Klauer et al., 2014; Wilson & Stimpson, 2010). As many as 281,000 to 786,000 crashes in 2012 may involve text messaging, according to the estimate of the United States National Safety Council (2012). Pickrell and Ye (2013) found 0.9% of drivers were visibly manipulating hand-held devices while driving in 2010, and this percentage increased to 1.3 percent in 2011. The risk and prevalence of texting while driving has attracted the attention of the general public, auto manufacturers, legislators and safety researchers (Jacobson & Gostin, 2010; Owens, McLaughlin, & Sudweeks, 2011). Concurrent texting impairs driving in various ways. For example, texting while driving increases hazard response time (Burge & Chaparro, 2012; Drews et al., 2009; He et al., 2014; Leung, Croft, Jackson, Howard, & Mckenzie, 2012), increases lane deviations (the difference between the center of the vehicle and the center of the appropriate lane) and lane excursions (leaving the lane unintentionally) (Alosco et al., 2012; Crandall & Chaparro, 2012; ⇑ Corresponding author at: Wichita State University, 1845 Fairmount St., Wichita, KS 67260, USA. Cell: +1 217 417 3830. E-mail address: [email protected] (J. He). http://dx.doi.org/10.1016/j.chb.2015.05.004 0747-5632/Ó 2015 Elsevier Ltd. All rights reserved.
Hosking et al., 2009; McKeever, Schultheis, Padmanaban, & Blasco, 2013; Rudin-Brown, Young, Patten, Lenné, & Ceci, 2013), increases mental demand (mental demands are psychological and mental stress experienced by an individual while completing one or more tasks) (Owens et al., 2011; Rudin-Brown et al., 2013), increases gaze-off-road durations (Hosking et al., 2009; Libby, Chaparro, & He, 2013; Owens et al., 2011), causes more collisions (Alosco et al., 2012; Drews et al., 2009), and raises the risks of traffic accident as many as 8–23 times (Olson, Hanowski, Hickman, & Bocanegra, 2009). People have limited ability to perform two tasks simultaneously, such as texting and driving and doing so results in deficits on one or both of the tasks being performed (Allport, Antonis, & Reynolds, 1972). According to the theories of dual-task performance, when two tasks are carried out concurrently, the performances of one or both tasks may be impaired, causing a dual-task performance decrement (Wickens, 2002). For example, when performing a secondary auditory monitoring task (pressing a button when they hear a tone), drivers had slower reaction time when responding to a vehicle braking, compared to driving-only conditions, even when instructed to give the driving task priority (Levy & Pashler, 2008). While several studies have reported the effects of texting on driving performance, how driving affects texting performance has been ignored. Better understanding of changes in both driving and texting performance can inform theories of dual-task performance and contribute to the efforts to mitigate the risks of texting while driving.
116
J. He et al. / Computers in Human Behavior 52 (2015) 115–123
Researchers and practitioners have explored a variety of approaches to mitigate the risks of texting while driving, including legislation, social campaigns, and technological solutions. Legislative efforts have sought to discourage this behavior by making it illegal (McCartt & Geary, 2004; Mccartt, Hellinga, Strouse, & Farmer, 2010) and social campaigns have sought to educate drivers about the risks of texting while driving (Atchley, Hadlock, & Lane, 2012; Nemme & White, 2010). In addition, telecommunication companies including AT&T and T-Mobile have developed smartphone applications to discourage texting while driving, such as Drive Mode and DriveSmart. Smartphone users can enable these applications to delay or block incoming calls and messages when they drive, limiting their exposure to the dangers of texting while driving. However, despite the associated risks of texting while driving and these legislative, social, and technological efforts, drivers continue to text while driving. Ninety-one percent of college students reported having sent text messages while driving, even though they agreed or believed that texting while driving was dangerous and should be illegal (Atchley, Atwood, & Boulton, 2011; Harrison, 2011). Is it possible to develop a smartphone application to monitor texting while driving, and prevent or discourage drivers from engaging in such risk behavior? If an application can monitor texting while driving, it can potentially be easier to implement, and can complement current efforts in law enforcement or social campaigns. Researchers have attempted to detect drunk driving (Dai, Teng, Bai, Shen, & Xuan, 2010), cognitively distracted driving (Liang, Lee, & Reyes, 2007), aggressive driving (Johnson & Trivedi, 2011; Zeeman & Booysen, 2013), and drowsy driving (Hammoud & Zhang, 2008; He et al., 2013). However, to our best knowledge, no application has been developed that detects texting while driving. Thus, this study also explores the possibility using texting behavioral data to identify whether a driver is texting while driving. The popularity and risks of texting while driving highlight the continuing need for research and understanding of how texting influences driving and vice versa, and how drivers coordinate performance of both tasks (Atchley et al., 2011; Harrison, 2011). Most driving studies focus on driving performance, while texting performance is mostly ignored or less emphasized. Smartphones allow the collection of detailed data on secondary texting performance. In this study, we utilize the capability of smartphones to capture texting performance and describe the mutual interferences of texting and driving. The goal of this paper is to discover the mutual influences of concurrent texting and driving, and sought the possibility to detect texting while driving and reduce its risks. It is predicted that participants will show greater lane deviation while driving and texting compared to the drive-only condition. It is also hypothesized that participants will take longer to complete the texting task and make more errors when driving and texting than when driving-only.
touchscreen smartphone and on average, reported sending 83 text messages per day (median = 70, SD = 86.06). 2.2. Apparatus and stimuli Driving performance was assessed using a driving simulator consisting of a General Motors car seat and Logitech Driving Force GT steering wheel and pedals. The Lane Change Task (LCT) version 1.2 software simulated the driving task using a 60 inch-Sharp Aquos 3D HD LCD display. A 4.300 HTC ThunderBolt touch-screen smartphone running the Android 2.3.4 operating system was used for the texting task. The buttons on the keyboard of the smartphone were arranged in a QWERTY layout. 2.3. Experimental design There were five counterbalanced experimental conditions, including driving-only condition, two dual-task conditions, in which participants either drove while texting with one hand (drive + text one hand) or two hands (drive + text two hands), and two texting-only conditions in which participants completed the texting task with either one or two hands. We employed a within-subject design. Participants finished all the task conditions. The dependent variables measured for the driving task were mean lane deviation and standard deviation of lane deviation. Lane deviation refers to the difference between the center of the vehicle and the center of the appropriate lane. For the texting task, the dependent variables consisted of task completion time, key entry per second, texting task completion time, texting errors, input time interval, standard deviation of input time interval, and device stability (He et al., 2014). 2.4. Experiment tasks 2.4.1. Lane Change Task (LCT) In the LCT, participants were required to drive down a straight section of road with three lanes and were prompted to change lanes according to directions on signs, which appeared on both sides of the roadway. An arrow on the sign, shown in Fig. 1, indicated which lane the driver supposed to maneuver into. Participants were instructed to change from their current lane in a deliberate manner, and to do so as quickly and efficiently as possible. Participants maintained a constant speed of 60 km/h and were instructed to make lane changes as quickly and accurately as possible when they saw the lane change sign. The LCT was developed by Daimler Chrysler AG Research and Technology to test driver distraction (Hofmann, Rinkenauer, & Gude, 2008; Huemer & Vollrath, 2010; Mattes & Hallén, 2008).
2. Materials and methods 2.1. Participants Twenty-eight participants (12 men, 16 women) from a university community ages 18–35 years (M = 22.14 years, SD = 4.64 years) volunteered to participate in this driving experiment. All participants were screened prior to participation to ensure normal or corrected-to-normal vision using the Snellen Visual Acuity chart (Ferris, Kassoff, Bresnick, & Bailey, 1982). All participants completed a survey to ensure they were right-handed, active drivers, with at least two years of driving experience (M = 6.29 years; SD = 4.51 years). They all owned a
Fig. 1. Screenshot of driving simulator environment.
J. He et al. / Computers in Human Behavior 52 (2015) 115–123
2.4.2. Texting task A texting task was used to assess texting performance with two holding postures (texting with one hand vs. texting with two hands). An incoming text message sound was played at a random time interval, which followed a uniform distribution in the range of 40–60 s. Once the incoming sound was played, participants opened the texting application to send text messages. Participants were presented with a text message and instructed to enter the text message into the phone manually. After the message was sent, the application was closed. The text message consisted of a 10-digit telephone number. The digit-entry texting task was chosen to simulate the cognitive process and key entry requirement of a cell phone dialing or texting task while also allowing the measurement of secondary texting performance. Similar digit-entry tasks are commonly used to study the effect of distraction on driving performance (He et al., 2014; Horrey & Wickens, 2004; Reed & Green, 1999; Salvucci & Macuga, 2002). The application saved timestamps, keyboard entries, and accelerometer sensor values of the smartphone in log files for later analysis. 2.5. Procedure Participants were informed of the purpose of the experiment and then signed a consent form to participate. They then completed a questionnaire regarding driving experience and cell phone use, followed by a visual screening. The session began with instructions on how to complete the tasks and a practice session to familiarize the participants with the simulator and the cell phone. Participants received instruction on the LCT and how to correctly execute the task. Participants were instructed to prioritize the driving task during dual-task conditions and text when they felt comfortable or believed it was safe. The phone was used in portrait orientation for all texting conditions. In the two-hand texting conditions, participants were instructed to keep both hands on the phone while texting. Participants practiced each of the text-only conditions for two minutes and each of the driving conditions for six minutes, for a total of 20 min of practice. After the practice, participants began the experimental trials. Each trial consisted of four tracks lasting three minutes each, for a total of 12 min. The time to complete the survey, practice time, and the experiment took approximately 105 min. Participants could withdraw from the study at any time. After the study, participants were given course credits for their participation. 2.6. Data analysis Measures of driving performance included the mean and standard deviation of lane deviation recorded by the driving simulator. Lane deviation was the difference between the driver’s actual lane position (red line) and the ‘‘ideal model’’ of lane position (green line) for the LCT, seen in Fig. 2. A large value of the standard
117
deviation of lane deviation indicates poorer lane-keeping performance with a higher risk of lane departure and collision with vehicles in the adjacent lanes. Texting performance measures included texting task completion time, key entry per second (KES), mean and standard deviation of input time interval, mean numbers of inputs, and mean number of corrections. Texting errors were measured using the Levenshtein edit distance of two strings (simply termed as edit distance) (He et al., 2014; Levenshtein, 1966). The edit distance is the number of discrete steps required to make the strings identical, including insertions, deletions, switching, and substitutions. Larger edit distance values indicate a greater difference between the two strings or more texting errors. For example, the numbers 5200251314 and 200251319 have a Levenshtein distance of two. The pair of numbers can be made identical by inserting a 5 at the beginning (insertion) and substituting the last 9 into 4 (substitution). All the driving performance measures were submitted to repeated-measures one-way analyses of variance (ANOVA) with the driving condition (drive-only, drive + texting with one hand, and drive + texting with two hands) as the only within-subject factors. For the one-way ANOVA of driving performance, when the main effect is significant, pairwise comparisons were made between the driving conditions. Bonferroni correction was used for post hoc pairwise comparisons. The Bonferroni-corrected p value is 0.05/3 = 0.017. Texting performance variables were submitted to two-way analyses of variance (ANOVA) with task load (text-only vs. drive + texting) and texting methods (one hand vs. two hands) as within-subject factors. Partial eta-squared (g2p) was used to measure effect size. A g2p value of 0.04 indicates a recommended minimum effect size for a practically significant effect, 0.25 indicates a moderate effect and 0.64 indicates a strong effect (Ferguson, 2009). 3. Results 3.1. Driving performance The standard deviation of lane deviation was significantly different across task conditions, F(2, 54) = 29.14, p < .001, g2p = .52. As shown in Fig. 3 pairwise comparisons revealed the standard deviation of lane deviation in the drive-only condition (M = 1.09 m, SD = .20 m) was significantly smaller than that in the drive + texting with one hand condition (M = 1.21 m, SD = .26 m, p = .001) and drive + texting with two hands (M = 1.27 m, SD = .28 m, p < .001). The standard deviation of lane deviation in the drive + texting one hand condition was also significantly smaller than that in the drive + texting two hands condition. The mean lane deviation also differed significantly across task conditions, F(2, 54) = 41.43, p < .001, g2p = .61. As shown in Fig. 4, the mean lane deviation in the drive-only condition (M = .81 m, SD = .15 m) was significantly smaller than that in the drive + texting with one hand condition (M = .94 m, SD = .21 m), and
Fig. 2. The Lane Change Task. The green line represents the ideal path defined by the computer and the red line represents the path taken by the driver. The arrow symbol at the top indicates the lane, which drivers should drive in (Mattes & Hallén, 2008). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
118
J. He et al. / Computers in Human Behavior 52 (2015) 115–123
Fig. 5. Texting task completion time across conditions. Fig. 3. Standard deviation of lane deviation across task conditions. Error bars in all figures indicate within-subject standard errors based on the main effect of task (Loftus & Masson, 1994).
Fig. 4. Mean lane deviation across task conditions.
to complete the texting task while driving regardless of whether they texted with either one hand or two hands. A two-way within subjects analysis of variance was conducted to evaluate the key entry per second between driving while texting with one or two hands and texting only with one or two hands. As shown in Fig. 6, the key entry per second (KES) produced a marginally significant interaction of task load and texting methods, F(1, 27) = 3.91, p = .06, g2p = .13. The main effect of task load was significant, F(1, 27) = 226.31, p < .001, g2p = .89. Simple main effect showed that the KES in the texting-only conditions (M = 1.17, SD = .19) was significantly larger than that of drive + texting conditions (M = .67, SD = .13), t(27) = 15.04, p < .001. The main effect of texting methods was not significant, F(1, 27) = .07, p = .79, g2p = .003. Participants inputted more keys per second when texting-only compared to texting while driving. A two-way within-subjects analyses of variance was conducted to evaluate the average texting errors between texting while driving with one or two hands and texting only with one or two hands. The average texting errors (as shown in Fig. 7) differed significantly across task load conditions, F(1, 27) = 215.13, p < .001, g2p = .89. When texting while driving, drivers produced more texting errors (M = .48, SD = .66) than the texting-only condition (M = .11, SD = .14). The main effect of texting methods and interaction effect
drive + texting with two hands condition (M = .98 m, SD = .24 m), t(27) = 6.86, p < .001 and t(27) = 7.66, p < .001 respectively. The mean lane deviation in the drive + texting with one hand condition was significantly different from the drive + texting with two hands condition, t(27) = 2.56, p = .02. Participants deviated their trajectory from the ‘‘ideal model’’ when texting with two hands more than when texting using one hand or driving-only.
3.2. Texting performance A two-way within-subjects analyses of variance was conducted to evaluate the task completion time between driving while texting with one or two hands and texting only with one or two hands. The texting task completion time (as shown in Fig. 5) produced a significant main effect of task load, F(1, 27) = 175.42, p < .001, g2p = .87, with longer completion time in the drive + texting conditions (texting using either one hand or two hands) (M = 16.73 s, SD = 3.64 s) than the texting-only conditions (M = 9.57 s, SD = 2.12 s), but the main effect of texting methods and the interaction were not significant. Participants took longer
Fig. 6. Key entry per second across conditions.
J. He et al. / Computers in Human Behavior 52 (2015) 115–123
Fig. 7. The texting error measured by Levenshtein edit distance (He et al., 2014; Levenshtein, 1966).
were not significant, F(1, 27) = .59, p = .45, g2p = .02 and F(1, 27) = 1.64, p = .21, g2p = .06 respectively. A two-way within-subjects analyses of variance was conducted to evaluate the mean input time interval between driving while texting with one or two hands and texting only with one or two hands. The mean input time interval (shown in Fig. 8) also differed significantly across task conditions, F(1, 27) = 184.19, p < .001, g2p = .87. The mean input time interval was significantly higher in the drive + texting condition (M = 1.46 s, SD = .32 s) than that of texting-only condition (M = .82 s, SD = .17 s). Concurrent texting and driving delayed key entry by .64 s (78%) compared to the texting-only condition. The main effect of texting methods and interaction effect were not significant, F(1, 27) = .31, p = .58, g2p = .01 and F(1, 27) = 1.61, p = .22, g2p = .06 respectively. Participants paused more between key stroke while driving and texting compared to texting-only. A two-way within-subjects analyses of variance was conducted to evaluate the standard deviation of input time interval between driving while texting with one or two hands and texting only with one or two hands. The standard deviation of input time interval (shown in Fig. 9) differed significantly across task conditions, F(1, 27) = 46.69, p < .001, g2p = .63 and texting methods, F(1, 27) = 7.36, p = .01, g2p = .21, but the interaction was not significant, F(1, 27) = .67, p = .42, g2p = .02. The standard deviation of
Fig. 8. Mean input time interval.
119
Fig. 9. Standard deviation of input time interval.
input time interval was significantly higher in the drive + texting condition (M = 2.12 s, SD = .78 s) than that of texting-only condition (M = .91 s, SD = 1.01 s). The standard deviation of input time interval was also higher in the two-hand texting condition (M = 1.71 s, SD = 1.33 s) than the one-hand texting condition (M = 1.32 s, SD = .72 s). Participant made 10.77 key entries on average with a standard deviation of .79. Analysis of the number of key entries did not produce any significant effects, all ps > .10. Participants revised their entries by pressing the ‘Back’ button on average .26 times with a standard deviation of .25. Analysis of the number of times that the ‘Back’ button was pressed did not yield any significant main effect, all ps > .10. 3.3. Device stability We also measured the stability of the Android smartphone when participants texted while driving. Device stability was measured using the standard deviation of the accelerometer values for the X, Y, and Z coordinates of the smartphone. The Android coordinate system (as shown in Fig. 10) is defined relative to the screen of the phone in the portrait orientation. The axes are not swapped when the device’s screen orientation changes. The X-axis is horizontal and points to the right, the Y-axis is vertical and points up and the Z-axis points toward the outside of the front face of the screen. In this system, coordinates behind the screen have negative Z values. The X, Y, and Z vectors of the accelerator sensor were automatically saved to a log file when participants sent text messages using the smartphone. The accelerator’s values were sampled at a rate of 100 Hz only when participants used the texting application. For the standard deviation of the accelerometer values in the X-axis, the main effects of task load and texting methods were both significant, F(1, 27) = 47.37, p < .001, g2p = .64 and F(1, 27) = 35.20, p < .001, g2p = .57 respectively. The interaction was also significant, F(1, 27) = 39.44, p < .001, g2p = .59. Simple main effect test showed that when texting using one hand, the standard deviation of the accelerator values in the X-axis was not significantly different for the drive + texting condition (M = .58, SD = .15) and the texting-only condition (M = .53, SD = .19), t(27) = 1.26, p = .22. When texting using two hands, the standard deviation of the accelerator values in the X-axis was significantly larger in the drive + texting condition (M = 1.20, SD = .47) than the texting-only condition (M = .52, SD = .27), t(27) = 7.16, p < .001. For the standard deviation of the accelerometer values in the Y-axis, a significant interaction was found, F(1, 27) = 42.74,
120
J. He et al. / Computers in Human Behavior 52 (2015) 115–123
Fig. 10. The Android coordinate system for accelerator sensors.
p < .001, g2p = .61. The main effects of task load and texting methods were not significant, F(1, 27) = .05, p = .82, g2p = .002 and F(1, 27) = 2.56, p = .12, g2p = .09 respectively. Simple main effect tests showed that when texting using one hand, the standard deviation of the accelerometer values in the Y-axis was significantly smaller in the drive + texting condition (M = .72, SD = .15) than the texting-only condition (M = .91, SD = .38), t(27) = 2.59, p = .02. When texting using two hands, the standard deviation of the accelerometer values in the Y-axis was significantly larger in the drive + texting condition (M = .99, SD = .34) than the texting-only condition (M = .78, SD = .31), t(27) = 3.02, p = .01. For the standard deviation of the accelerometer values in the Z-axis, a significant interaction was found, F(1, 27) = 50.46, p < .001, g2p = .65. The main effects of task load and texting methods were also significant, F(1, 27) = 9.26, p = .01, g2p = .26 and F(1, 27) = 13.52, p = .001, g2p = .33 respectively. Simple main effect tests showed that when texting using one hand, the standard deviation of the accelerometer values in the Z-axis was significantly smaller in the drive + texting condition (M = .74, SD = .14) than the texting-only condition (M = .80, SD = .26), t(27) = 1.29, p = .21. When texting using two hands, the standard deviation of the accelerometer values in the Z-axis was significantly larger in the drive + texting condition (M = 1.07, SD = .33) than the texting-only condition (M = .71, SD = .19), t(27) = 5.25, p < .001. See Fig. 11 for the device stability as measured using the accelerometer sensor values in the X, Y, and Z-axis. The interaction effects in the above analyses is because the instability of the devices (as measured by the standard deviation of accelerator values in the X, Y, and Z-axis) is more pronounced in the texting while driving using two hands condition than the texting while driving using one hand condition. More instability to hold the smartphone in the condition when drivers texting using two hands was because the inability to share manual resources (e.g. the hands). In contrast, drivers in the one hand texting condition can hold the phone in one hand and control the steering wheel using the other hand. Thus, for drivers who texted with two hands, texting while driving significantly increased the instability of the smartphone, which indicates more manual interference of texting and driving and higher chances of dropping or even breaking the smartphone when driving (see Fig. 11). 3.4. Detect texting while driving To reduce the risks of texting while driving, it is important to detect the occurrence of such driving behaviors. We used a logistic regression to detect texting while driving, using texting data collected from the smartphone. The variables in the logistic regression include the number of key entries, mean input time interval, standard deviation of input time interval, the number of key entries,
Fig. 11. Device stability as measured using the accelerometer sensor values in the X, Y, and Z-axis.
and the number of times the ‘Back’ button was pressed. The texting task completion time and texting errors were hard to define in natural free-style texting, thus the two variables were not included in the analysis. The dependent variable was the texting conditions, either drive + texting or texting-only. To predict texting while driving we need a ‘‘training data set’’ to develop the logistic regression model. The training data set can be generic, that is based on the data from a group of participants or
121
J. He et al. / Computers in Human Behavior 52 (2015) 115–123
individualized meaning that a participant’s data is used to derive a their own user specific logistic regression model. If a common training data set can be used to monitor texting while driving reliably for all subjects, the proposed application to monitor texting while driving can be developed more easily, without the necessity to consider individual texting behaviors and does not require the collection of user specific data sets for model development. A user-specific training set may achieve higher prediction performance than a model based on a generic data set. We explored the prediction performance using both generic training set and individualized training set. 3.4.1. Prediction based on generic training set The experiment produced a total of 1748 texting instances. 1224 texting instances (70% of all instances) were randomly selected as the training set. The other 524 instances (30%) were used as the prediction set. The best logistic regression model based on generic training data set is:
Table 2 Confusion matrix to predict texting while driving based on user-specific training data. Predicted
Observed
Texting-only Drive + texting
Texting-only
Drive + texting
5.45 1.52
0.69 9.66
behavioral data, such as mean and standard deviation of the input time interval. The above data suggest that we can use texting behavior data to predict texting while driving reliably with at least 88.5% accuracy. The accuracy using generic training data set (88.5%) is only slightly lower than the accuracy based on user-specific training data set (89.7%). The minor increase in the prediction accuracy of user-specific training data should be a result of the significant large increase in the mean and standard deviation of texting input intervals. This result suggests that we can just use a generic logistic regression model to predict texting while driving, which is easier to implement than a user-specific training data set.
Task load ¼ 2:31 :0030 mean input time interval :0016 standard deviation of input time interval Where for task load, 0 indicates texting-only condition and 1 indicates drive + texting condition. The classification accuracy for the generic training set is shown in Table 1. An accuracy as high as 88.5% was achieved using texting behavior data to classify texting while driving from the texting-only condition. The sensitivity of the model based on generic training set is .88, and the specificity of the model is .94. Our accuracy in detecting texting while driving is comparable to or sometimes higher than other attempts to detect cognitive distraction (Liang, Reyes, & Lee, 2007) and visual distraction (Kutila, Jokela, Markkula, & Rue, 2007). For example, Liang, Reyes et al. (2007) achieved an average accuracy of 81.1% in detecting driver cognitive distraction using Support Vector Machines. Our high detection accuracy suggests it is possible to develop a smartphone application to detect texting while driving using texting behavioral data, such as mean and standard deviation of the input time interval. 3.4.2. Prediction based on user-specific training set User-specific training set can potential improve model prediction by learning the individual difference in texting behaviors. To explore to what extent user-specific training set can improve our ability to predict texting while driving, we also trained twenty-eight logistic regression models for each subject. Each driver produced about 58.07 texting instances in total. 70% of the texting instances for each subject were used as the training data set and the other 30% texting instances were used as the testing data set. The classification accuracy for the user-specific training data set is shown in Table 2. The average accuracy as high as 89.7% was achieved using texting behavior data to classify texting while driving from the texting-only condition. The sensitivity of the model based on generic training set is .94, and the specificity of the model is .82. The high accuracy suggests it is possible to develop a smartphone application to detect texting while driving using texting
Table 1 Confusion matrix to predict texting while driving based on generic training data. Predicted
Observed
Texting-only Drive + texting
Texting-only
Drive + texting
233 16
29 210
4. Discussion Using a classic Lane Change Task and smartphone technology, we found evidence of mutual interference between driving and texting. Texting impaired driving by increasing lane deviations, which resonates with previous studies showing that texting impaired driving performance (Drews et al., 2009; He et al., 2014; Hosking et al., 2009). To our best knowledge, existing studies have often ignored the other aspect of dual-task performance, that is, how driving influences texting? In addition to the finding that texting impairs driving, our study also showed that driving impaired texting by slowing text entry and increasing texting errors. The mutual interference of texting and driving contributes to our understanding of dual-task performance. Interestingly, we can use a logistic regression model based on texting behaviors to detect texting while driving with accuracy as high as 88.5%. Texting can increase a driver’s crash risks and endanger their lives. Drivers may continue to text for a variety of reasons. Drivers may read or send texts out of habit, regardless their subjective perception of the risks of such behaviors (Bayer & Campbell, 2012) and are more likely to take a phone call if they believe the call is important (Nelson, Atchley, & Little, 2009). They may send texts to reduce unpleasant emotions (Feldman, Greeson, Renna, & Robbins-Monteith, 2011) or maintain a social relationship by replying promptly to a text message. For these drivers, a clear message that ‘‘driving and texting interfere mutually’’ should be delivered in social campaigns to further discourage texting while driving. Practically, a better understanding of texting performance may inform the development of smartphone applications that can detect and perhaps reduce texting while driving. Several researchers are beginning to explore smartphone applications that can monitor impaired driving performance (Dai et al., 2010; Hammoud & Zhang, 2008; He et al., 2013; Johnson & Trivedi, 2011; You et al., 2013; Zeeman & Booysen, 2013;), such as drunk, distracted, aggressive and drowsy driving. However, no attempts have been made to detect texting while driving yet. Our study demonstrated that a logistic regression model using texting behavioral data (such as mean and standard deviation of input time interval) can detect texting while driving with accuracy as high as 88.5%. This finding provides a theoretical support to develop a smartphone application, which can detect and prevent texting while driving. A smartphone-based application to monitor texting while driving has the advantage to be easier to implement (Eren, Makinist, Akin, & Yilmaz, 2012), while it is very costly for law enforcement or social campaigns to prevent texting while driving.
122
J. He et al. / Computers in Human Behavior 52 (2015) 115–123
More studies are needed to improve our training data set and algorithm to detect texting while driving. For example, an on-road driving study using a natural texting task is needed to collect valuable training data set to test and improve our algorithm. Many cell phone carriers and cell phone manufacturers offer ‘‘Drive Mode’’ applications, which allow users to block incoming calls and messages or send automatic responses. These applications often require users to initiate the drive mode before driving and turn it off after driving, which is a bothersome process and limits the use of these applications. Automatic detection of texting while driving can improve existing ‘‘Drive Mode’’ applications. If smartphone applications incorporate automatic detection of texting while driving and stop or delay messages from distracting drivers, these applications can be more effective in reducing texting while driving than existing applications. Our study demonstrated that we can reliably detect texting while driving using generic training set based on a group of users. Our finding provides new knowledge to improve existing ‘‘Drive Mode’’ applications. Future studies shall investigate whether other information, including driving dynamics, can be used to develop smartphone-based technology to detect texting while driving. For example, GPS information can be used to calculate travelling velocity, which could then be used as a precondition for identifying whether the user is texting while driving. The smartphone application will detect texting while driving only when users travel at higher speeds, above 20 mph, which would greatly improve the performance of the application. Researchers have focused on how to use smartphone capabilities to improve driving safety (He et al., 2012, 2013, 2014; Ren, Wang, & He, 2013; Wang, Cardone, Corradi, Torresani, & Campbell, 2012). To our best knowledge, this is one of the first papers to demonstrate the possibility of using smartphone technology to detect texting while driving. This technology is an innovative way to mitigate the risks of texting while driving, which complements the existing social and legal approaches. With a better understanding of texting and driving behaviors and the rapid advancement of smartphone technology, smartphones can be used to make driving safer, instead of more dangerous (He et al., 2013; Ren et al., 2013). As Dr. Adesman stated, ‘‘Technological solutions will likely need to be developed to significantly reduce the frequency of texting while driving’’ (American Academy of Pediatrics, 2013; Lindqvist & Hong, 2011). Acknowledgements We appreciate the valuable suggestions from Frank Schieber, Yulan Liang, Chun Wang, and Evan Palmer. We also acknowledge the help in data collection from the undergraduate research assistants Colton Turner and Kirsten Turner. The authors would acknowledge to the National Natural Science Foundation of China with Grant 71401004 and the U.S. Department of Transportation (DOT) through the University Transportation Centers program sponsored by Research and Innovative Technology Administration (RITA). References Allport, D. A., Antonis, B., & Reynolds, P. (1972). On the division of attention: A disproof of the single channel hypothesis. The Quarterly Journal of Experimental Psychology, 24(2), 225–235. Alosco, M. L., Spitznagel, M. B., Fischer, K. H., Miller, L. A., Pillai, V., Hughes, J., et al. (2012). Both texting and eating are associated with impaired simulated driving performance. Traffic Injury Prevention, 13(5), 468–475. http://dx.doi.org/ 10.1080/15389588.2012.676697. American Academy of Pediatrics (2013). Don’t txt n drive: Teens not getting msg: 43 percent of youths admit to texting while driving. Science Daily. 4 May 2013. Atchley, P., Atwood, S., & Boulton, A. (2011). The choice to text and drive in younger drivers: Behavior may shape attitude. Accident Analysis & Prevention, 43(1), 134–142. http://dx.doi.org/10.1016/j.aap.2010.08.003.
Atchley, P., Hadlock, C., & Lane, S. (2012). Stuck in the 70s: The role of social norms in distracted driving. Accident Analysis & Prevention, 48, 279–284. http:// dx.doi.org/10.1016/j.aap.2012.01.026. Bayer, J. B., & Campbell, S. W. (2012). Texting while driving on automatic: Considering the frequency-independent side of habit. Computers in Human Behavior, 28(6), 2083–2090. http://dx.doi.org/10.1016/j.chb.2012. 06.012. Burge, R., & Chaparro, A. (2012). The effects of texting and driving on hazard perception. In Proceedings of the human factors and ergonomics society annual meeting (Vol. 56, No. 1, pp. 715–719) doi: http://dx.doi.org/10.1177/ 1071181312561149. Caird, J. K., Johnston, K. A., Willness, C. R., Asbridge, M., & Steel, P. (2014). A metaanalysis of the effects of texting on driving. Accident Analysis & Prevention, 71, 311–318. http://dx.doi.org/10.1016/j.aap.2014.06.005. Crandall, J. M., & Chaparro, A. (2012). Driver distraction: Effects of text entry methods on driving performance. In Proceedings of the human factors and ergonomics society annual meeting (Vol. 56, No. 1, pp. 1693–1697). doi: http:// dx.doi.org/10.1177/1071181312561339. Dai, J., Teng, J., Bai, X., Shen, Z., & Xuan, D. (2010). Mobile phone based drunk driving detection. In Pervasive computing technologies for healthcare (PervasiveHealth). doi: http://dx.doi.org/10.4108/ICST.PERVASIVEHEALTH2010.8901. Drews, F. A., Yazdani, H., Godfrey, C. N., Cooper, J. M., & Strayer, D. L. (2009). Text messaging during simulated driving. Human Factors, 51(5), 762–770. Eren, H., Makinist, S., Akin, E., & Yilmaz, A. (2012). Estimating driving behavior by a smartphone. In 2012 IEEE intelligent vehicles symposium (IV) (pp. 234–239). doi: http://dx.doi.org/10.1109/IVS.2012.6232298. Feldman, G., Greeson, J., Renna, M., & Robbins-Monteith, K. (2011). Mindfulness predicts less texting while driving among young adults: Examining attentionand emotion-regulation motives as potential mediators. Personality and Individual Differences, 51(7), 856–861. http://dx.doi.org/10.1016/j.paid.2011. 07.020. Ferguson, C. J. (2009). An effect size primer: A guide for clinicians and researchers. Professional Psychology: Research and Practice, 40(5), 532–538. Ferris, F. L., Kassoff, A., Bresnick, G. H., & Bailey, I. (1982). New visual acuity charts for clinical research. American Journal of Ophthalmology, 94(1), 91–96. Hammoud, R. I., & Zhang, H. (2008). Alertometer: Detecting and mitigating driver drowsiness and fatigue using an integrated human factors and computer vision approach. In R. I. Hammoud (Ed.), Passive Eye Monitoring (pp. 301–321). Berlin Heidelberg: Springer. Retrieved from http://link.springer.com/chapter/10.1007/ 978-3-540-75412-1_14. Harrison, M. A. (2011). College students’ prevalence and perceptions of text messaging while driving. Accident Analysis & Prevention, 43(4), 1516–1520. http://dx.doi.org/10.1016/j.aap.2011.03.003. He, J., Chaparro, A., Nguyen, B., Burge, R. J., Crandall, J., Chaparro, B., et al. (2014). Texting while driving: Is speech-based text entry less risky than handheld text entry? Accident Analysis & Prevention, 72, 287–295. http://dx.doi.org/10.1016/ j.aap.2014.07.014. He, J., Fields, B.M., Roberson, S., Cielocha, S., Peng, J., & Coltea, J. (2012). ‘‘System and method to monitor and reduce vehicle operator impairment.’’ U.S. Patent Application 13/717,506, filed December 17, 2012. He, J., Roberson, S., Fields, B., Peng, J., Cielocha, S., & Coltea, J. (2013). Fatigue detection using smartphones. Journal of Ergonomics, 03(03). http://dx.doi.org/ 10.4172/2165-7556.1000120. Hofmann, P., Rinkenauer, G., & Gude, D. (2008). Head-up-displays support response preparation in a Lane Change Task. In Proceedings of the human factors and ergonomics society annual meeting (Vol. 52, No. 18, pp. 1233–1237). doi: http:// dx.doi.org/10.1177/154193120805201817. Horrey, W. J., & Wickens, C. D. (2004). Driving and side task performance: The effects of display clutter, separation, and modality. Human Factors, 46(4), 611–624. Hosking, S. G., Young, K. L., & Regan, M. A. (2009). The effects of text messaging on young drivers. Human Factors, 51(4), 582–592. Huemer, A. K., & Vollrath, M. (2010). Alcohol-related impairment in the Lane Change Task. Accident Analysis & Prevention, 42(6), 1983–1988. http:// dx.doi.org/10.1016/j.aap.2010.06.005. Jacobson, P. D., & Gostin, L. O. (2010). Reducing distracted driving: regulation and education to avert traffic injuries and fatalities. JAMA, 303(14), 1419–1420. Johnson, D. A., & Trivedi, M. M. (2011). Driving style recognition using a smartphone as a sensor platform. In 2011 14th International IEEE Conference on Intelligent Transportation Systems (ITSC) (pp. 1609–1615). doi: http://dx.doi.org/10.1109/ ITSC.2011.6083078. Klauer, S. G., Guo, F., Simons-Morton, B. G., Ouimet, M. C., Lee, S. E., & Dingus, T. A. (2014). Distracted driving and risk of road crashes among novice and experienced drivers. New England Journal of Medicine, 370(1), 54–59. Kutila, M., Jokela, M., Markkula, G., & Rue, M. R. (2007). Driver distraction detection with a camera vision system. IEEE International Conference on Image Processing, 6, 201–204. Leung, S., Croft, R. J., Jackson, M. L., Howard, M. E., & Mckenzie, R. J. (2012). A comparison of the effect of mobile phone use and alcohol consumption on driving simulation performance. Traffic Injury Prevention, 13(6), 566–574. http://dx.doi.org/10.1080/15389588.2012.683118. Levenshtein, V. I. (1966). Binary codes capable of correcting deletions, insertions and reversals. Soviet Physics Doklady, Johnson, 10, 707. Levy, J., & Pashler, H. (2008). Task prioritization in multitasking during driving: Opportunity to abort a concurrent task does not insulate braking responses from dual-task slowing. Applied Cognitive Psychology, 22, 507–525.
J. He et al. / Computers in Human Behavior 52 (2015) 115–123 Liang, Y., Lee, J., & Reyes, M. (2007). Nonintrusive detection of driver cognitive distraction in real time using Bayesian Networks. Transportation Research Record: Journal of the Transportation Research Board, 2018, 1–8. http://dx.doi.org/ 10.3141/2018-01. Liang, Y., Reyes, M. L., & Lee, J. D. (2007). Real-time detection of driver cognitive distraction using support vector machines. IEEE Transactions on Intelligent Transportation Systems, 8(2), 340–350. Libby, D., Chaparro, A., & He, J. (2013). Distracted while driving: A comparison of the effects of texting and talking on a cell phone. In Proceedings of the human factors and ergonomics society annual meeting, (Vol. 57, No. 1, pp. 1874–1878) doi: http://dx.doi.org/10.1177/1541931213571418. Lindqvist, J., & Hong, J. (2011). Undistracted driving: A mobile phone that doesn’t distract. In Proceedings of the 12th workshop on mobile computing systems and applications (pp. 70–75). New York, NY, USA: ACM. http://dx.doi.org/10.1145/ 2184489.2184504. Loftus, G. R., & Masson, M. E. (1994). Using confidence intervals in within-subject designs. Psychonomic Bulletin & Review, 1(4), 476–490. Mattes, S., & Hallén, A. (2008). Surrogate distraction measurement techniques. In K. Young, J. D. Lee, & M. A. Regan (Eds.), Driver distraction: Theory, effects, and mitigation. CRC Press. McCartt, A. T., & Geary, L. L. (2004). Longer term effects of New York State’s law on drivers’ handheld cell phone use. Injury Prevention, 10(1), 11–15. http:// dx.doi.org/10.1136/ip.2003.003731. Mccartt, A. T., Hellinga, L. A., Strouse, L. M., & Farmer, C. M. (2010). Long-term effects of handheld cell phone laws on driver handheld cell phone use. Traffic Injury Prevention, 11(2), 133–141. McKeever, J. D., Schultheis, M. T., Padmanaban, V., & Blasco, A. (2013). Driver performance while texting: Even a little is too much. Traffic Injury Prevention, 14(2), 132–137. http://dx.doi.org/10.1080/15389588.2012.699695. National Safety Council (2012). Annual estimate of cell phone crashes 2012. Retrieved on 20.05.14 from
123
Owens, J. M., McLaughlin, S. B., & Sudweeks, J. (2011). Driver performance while text messaging using handheld and in-vehicle systems. Accident Analysis & Prevention, 43(3), 939–947. http://dx.doi.org/10.1016/j.aap.2010.11.019. Pickrell, T. M., & Ye, T. J. (2013). Driver electronic device use in 2011. (Report No. DOT HS 811 719). Washington, DC: National Highway Traffic Safety Administration. Reed, M. P., & Green, P. A. (1999). Comparison of driving performance on-road and in a low-cost simulator using a concurrent telephone dialling task. Ergonomics, 42(8), 1015–1037. Ren, Z., Wang, C., & He, J. (2013). Vehicle detection using Android smartphones. The 7th international driving symposium on human factors in driver assessment, training, and vehicle design. Retrieved from