The Effects of Dual-Task Interference and Response Strategy on Stop or Go Decisions to Yellow Light Changes

Similar documents
Driver Response to Phase Termination at Signalized Intersections at Signalized Intersections: Are Driving Simulator Results Valid?

Every time a driver is distracted,

Covert Texting During Simulated Driving Maneuvers Effects of Head-Up versus Head-Down Posture

This is the author s version of a work that was submitted/accepted for publication in the following source:

MRI-2: Integrated Simulation and Safety

STATIC AND DYNAMIC EVALUATION OF THE DRIVER SPEED PERCEPTION AND SELECTION PROCESS

The automobile is a complex system in which humans play an important role. Driving is

OBSERVATION OF GAP ACCEPTANCE DURING INTERSECTION APPROACH

Effect of Different Distractions on Driving Performance for Drivers Using a Touch Screen

Detecting Transfer of Training Through Simulator Scenario Design: A Novice Driver Training Study

Car Following by Optical Parameters

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and

Effects on Auditory Attention and Walking While Texting with a Smartphone and Walking on Stairs

Engaging with Highly Automated Driving: To be or Not to be in the Loop?

Can Collision Warning Systems Mitigate Distraction Due to In-Vehicle Devices?

Department of Transportation

LECTUR 10 CHARACTERISTICS OF THE DRIVER, THE PEDESTRIAN, THE VEHICLE AND THE ROAD. One problem that faces traffic and transportation engineers is:

EVALUATING THE IMPACT OF TWO ALLOWABLE PERMISSIVE LEFT-TURN INDICATIONS

The Quantification of Bicyclist Conspicuity Using Eye-Tracking Technology

TOO CLOSE FOR COMFORT: EVALUATING A REWARD-BASED APPROACH TO INCREASE DRIVERS HEADWAY

Shoe-shaped Interface for Inducing a Walking Cycle

Introduction to Transportation Engineering. Discussion of Stopping and Passing Distances

Access Location, Spacing, Turn Lanes, and Medians

Crosswalk event overview (Saint Paul) How to organize a crosswalk event. 1. Choose a date, time and location

Advanced driver assistance systems status and future potential

DOT HS September Crash Factors in Intersection-Related Crashes: An On-Scene Perspective

Real Time Bicycle Simulation Study of Bicyclists Behaviors and their Implication on Safety

Investigation of driving behaviour reflecting drivers risk anticipation for pedestrian collision risk assessment of right-turns at intersections

Formative Evaluation of Engineering Designs Using Driver Performance in a Immersive Driving Simulator

c 2015 Sarah E. Banducci

Driver compliance with, and understanding of, level crossing controls

Introduction to Transportation Engineering. Discussion of Stopping and Passing Distances

An Application of Signal Detection Theory for Understanding Driver Behavior at Highway-Rail Grade Crossings

MRI-2: Integrated Simulation and Safety

Toshihisa Sato, 1 Motoyuki Akamatsu, 1 Toru Shibata, 2 Shingo Matsumoto, 2 Naoki Hatakeyama, 2 and Kazunori Hayama 2. 1.

Geometric Categories as Intersection Safety Evaluation Tools

People killed and injured per million hours spent travelling, Motorcyclist Cyclist Driver Car / van passenger

Safety Assessment of Installing Traffic Signals at High-Speed Expressway Intersections

Covert Texting During Simulated Driving Maneuvers: Effects of Head-Up versus Head-Down Posture

Why Zone In on Speed Reduction?

Traffic circles. February 9, 2009

Exploring Older Driver Lateral Head Rotations at Intersections Using Naturalistic Driving Data

New in the MUTCD: The Flashing Yellow Arrow Presented at the 57 th Annual Traffic and Safety Conference May 17, 2006

Human Factors for Limited-Ability Autonomous Driving Systems

Lesson #13. Review: Lesson #12 Destination: Washington (Business and Residential)

DOI /HORIZONS.B P23 UDC : (497.11) PEDESTRIAN CROSSING BEHAVIOUR AT UNSIGNALIZED CROSSINGS 1

Queue analysis for the toll station of the Öresund fixed link. Pontus Matstoms *

An Analysis of Reducing Pedestrian-Walking-Speed Impacts on Intersection Traffic MOEs

AGE, SKILL, AND HAZARD PERCEPTION IN DRIVING

Effectiveness of Visual Collision Warning Alerts on Young Drivers Latent Hazard Anticipation

At each type of conflict location, the risk is affected by certain parameters:

Final Report. Research. Investigating the Effects of Traffic Calming Strategies on Driver Behavior

ITARDA INFORMATION. No.128. Special feature

Advance Yield Markings Reduce Motor Vehicle/Pedestrian. Conflicts at Multilane Crosswalks with an Uncontrolled Approach.

Saturation Flow Rate, Start-Up Lost Time, and Capacity for Bicycles at Signalized Intersections

The Effect of Bicycle Color on Visibility

A Traffic Operations Method for Assessing Automobile and Bicycle Shared Roadways

Traffic Impact Analysis

Designing a Traffic Circle By David Bosworth For MATH 714

Pedestrian and Driver Distraction: Overview & NHTSA Prevalence and Risk Study. Heidi Coleman, NHTSA Bob Scopatz, VHB

Crash Patterns in Western Australia. Kidd B., Main Roads Western Australia Willett P., Traffic Research Services

The Quality of Behavioral and Environmental Indicators Used to Infer the Intention to Change Lanes

FYG Backing for Work Zone Signs

Empirical observations of dynamic dilemma zones at signalized intersections

Public Opinion, Traffic Performance, the Environment, and Safety After Construction of Double-Lane Roundabouts

Appendix B: Forecasting and Traffic Operations Analysis Framework Document

Driver Behavior in the Presence of Pedestrians at Signalized Intersections Operating the Flashing Yellow Arrow

Deaths/injuries in motor vehicle crashes per million hours spent travelling, July 2007 June 2011 (All ages) Mode of travel

Evaluating Road Departure Crashes Using Naturalistic Driving Study Data

A Conceptual Approach for Using the UCF Driving Simulator as a Test Bed for High Risk Locations

CROSSING GUARD PLACEMENT CONSIDERATIONS AND GAP ASSESSMENT

Use of Throw Distances of Pedestrians and Bicyclists as Part of a Scientific Accident Reconstruction Method 1

Driver Training School Instructor Curriculum Requirements for Student Learning & Performance Goals

MEET IN THE STREET MOVING TO ZERO

Simulation Analysis of Intersection Treatments for Cycle Tracks

Traffic flow optimization at sags by controlling the acceleration of some vehicles

Modal Shift in the Boulder Valley 1990 to 2009

Board & Committee Agenda Item

EFFECT OF RED LIGHT CAMERAS on INJURY CRASHES in the CITY OF VIRGINIA BEACH

Chapter 5 5. INTERSECTIONS 5.1. INTRODUCTION

Fatigue Monitoring Device for Monitoring Haul Truck Operator Alertness. 2 The Challenge

On-Road Evaluation of Driver Capability: A Medical Record Review of the Adaptive Driving Program

Standing Committee on Planning, Transportation and Environment

Study on fatal accidents in Toyota city aimed at zero traffic fatality

Roundabouts along Rural Arterials in South Africa

HSIS. Association of Selected Intersection Factors With Red-Light-Running Crashes. State Databases Used SUMMARY REPORT


IMPROVED SAFETY SURFACE ACCESS AT LOW COST AIRPORTS: KUALA LUMPUR INTERNATIONAL AIRPORT, MALAYSIA CASE STUDY

The 2002 Pedestrian Safety Awareness Campaign

This is the author s version of a work that was submitted/accepted for publication in the following source:

the Ministry of Transport is attributed as the source of the material

THE DEVELOPMENT OF MALAYSIAN HIGHWAY RAIL LEVEL CROSSING SAFETY SYSTEMS: A PROPOSED RESEARCH FRAMEWORK. Siti Zaharah Ishak

Evaluation and further development of car following models in microscopic traffic simulation

Montana Teen Driver Education and Training. Module 3.3. Mixing with Traffic. Montana Teen Driver Curriculum

Module 3 Developing Timing Plans for Efficient Intersection Operations During Moderate Traffic Volume Conditions

Aging Driver and Pedestrian Safety: Human Factors Studies of Signs & Signals

Evaluation of cross-walk timing and the application of a standard crossing light timing formula

DOWNTOWN TUPELO MAIN STREET: ROAD DIET STUDY

POLICY AGENDA For Elder Pedestrian Safety

The 100 Car Study: A Pilot for Large-Scale Naturalistic Driving Research

Transcription:

University of Iowa Iowa Research Online Driving Assessment Conference 2009 Driving Assessment Conference Jun 24th, 12:00 AM The Effects of Dual-Task Interference and Response Strategy on Stop or Go Decisions to Yellow Light Changes David G. Kidd George Mason University, Fairfax, Virginia Christopher A. Monk George Mason University, Fairfax, Virginia Follow this and additional works at: http://ir.uiowa.edu/drivingassessment Kidd, David G. and Monk, Christopher A.. The Effects of Dual-Task Interference and Response Strategy on Stop or Go Decisions to Yellow Light Changes. In: Proceedings of the Fifth International Driving Symposium on Human Factors in Driver Assessment, Training and Vehicle Design, June 22-25, 2009, Big Sky, Montana. Iowa City, IA: Public Policy Center, University of Iowa, 2009: 260-267. https://doi.org/10.17077/drivingassessment.1330 This Event is brought to you for free and open access by the Public Policy Center at Iowa Research Online. It has been accepted for inclusion in Driving Assessment Conference by an authorized administrator of Iowa Research Online. For more information, please contact lib-ir@uiowa.edu.

PROCEEDINGS of the Fifth International Driving Symposium on Human Factors in Driver Assessment, Training and Vehicle Design THE EFFECTS OF DUAL-TASK INTERFERENCE AND RESPONSE STRATEGY ON STOP OR GO DECISIONS TO YELLOW LIGHT CHANGES David G. Kidd and Christopher A. Monk George Mason University Fairfax, Virginia, USA Email: dkidd3@gmu.edu Summary: Distractions can interfere with driving by causing central processing bottlenecks. In addition to performance decrements, central processing delays may also impair decision-making during critical driving maneuvers such as stop or go decisions at intersections. It was hypothesized that distractions would delay the stop or go decision leading to more go responses. Participants drove 4 simulated drives and made stop or go decisions at intersections with and without a distracting task. Distractions did not result in more go responses at intersections. Additionally, dual-task interference in braking responses was found to be dependent upon participants response strategies. Theoretical implications of response strategy on processing bottlenecks were discussed. INTRODUCTION Intersection crossings are associated with 45% of crashes in the United States (Subramanian & Lombardo, 2007). Distraction is one factor known to interfere with drivers abilities to perform critical maneuvers (Alm & Nilsson, 1995; Horrey & Wickens, 2006). One theory that has been used to explain how distractions interfere with information processing is the central bottleneck model. The central bottleneck model proposes that central processes, such as decision and response selection, are limited to serial processing (see Pashler, 1998 for a review). When two tasks try to access the central processing stage at the same time, only one task gains access, and central processing of the other task is delayed. Processing delays are greatest when there is a large temporal overlap in the central processing stage between two tasks, but decreases as the tasks become separated a phenomenon known as the psychological refractory period (PRP) effect. Recently, Levy, Pashler, and Boehr (2006) and Monk and Kidd (under review) demonstrated the PRP effect with distractions during critical driving events. These findings suggest that distractions can create a central processing bottleneck that impairs drivers ability to process information efficiently. Distracted drivers have been shown to respond slower (Hancock, Lesch, & Simmons, 2003), brake harder (Harbluk, Noy, Trbovich, Eizenman, 2007), and make more errors (Liu & Lee, 2005) when responding to traffic light changes. In addition to degrading performance, distractions may also influence stop or go decisions. If a distraction coincides with a yellow light change, central processing of the yellow light could be delayed. During this delay, drivers continue to move closer to the intersection and may be less likely to stop at the intersection when response selection finally occurs. The limited empirical evidence studying the effects of distractions on stop or go decisions is mixed. Cooper et al. (2003) found that distracted drivers tended to stop more often to yellow light changes than non-distracted drivers. However, Monk and Kidd (under review) found that distracted drivers were more likely to make go responses compared to non-distracted drivers. 260

The purpose of this study was to examine the effects of dual-task interference on stop or go responses to a yellow light change. In the current study participants completed four drives in a driving simulator where they responded to yellow light changes at varying distances from the intersection. During two of the drives, participants responded to a secondary task at the same time as a yellow light change. It was hypothesized that central processing bottlenecks in dualtask trials would lead to greater number of go responses compared to single-task trials. Additionally, significantly slower braking responses were expected in dual-task trials compared to single-task trials when drivers stopped at the intersection. METHOD Participants Thirty-two undergraduate students (23 men, 9 women) were recruited from the George Mason University undergraduate subject pool and received course credit for participating. All participants had a valid driver s license and were at least 18 years of age. Participants ranged in age from 18 to 31 years (M = 21.1 years) and had an average of 53.9 months (SD = 33.5) of driving experience. Three additional participants were recruited, but withdrew from the study due to simulator sickness and were replaced. Apparatus and Tasks The experiment was conducted using the George Mason University driving simulator. The driving simulator was an open-cab simulator with a motion-base system that included 90-degree yaw motion to simulate turning and a single-degree of pitch motion to simulate braking and acceleration. The traffic environment was displayed across three 42-inch plasma displays with 180-degrees field-of-view. Scenarios were developed using Realtime Technologies, Inc. s (RTI) SimVista (Version 2.24) and run with RTI s SimCreator. Participants responded to secondary tasks using the turn indicator and response buttons mounted on the left and right spokes of the force-feedback steering wheel. Data from the vehicle inputs was collected at 60 Hz. Stop/Go Task. At some intersections, the traffic light changed from green to yellow requiring participants to make a stop or go decision. Participants were told that the traffic lights may or may not change to yellow in the scenario. If a yellow light change occurred, participants were instructed to come to a complete stop if it was safe to stop, but to go through the intersection if it was safer to do so. The duration of the yellow light was 4.5 seconds. Pedestrian Task. At some intersections, a pedestrian appeared in the grass at the near right hand corner of the intersection. Participants responded to the pedestrian s shirt color and orientation. Participants pressed the right response button if the pedestrian was wearing a red shirt and pressed the left response button if the pedestrian was wearing a white shirt. An auditory feedback tone was sounded when either response button was depressed, but did not provide information about response accuracy. Participants responded to the pedestrian s orientation using the turn indicator. Participants activated the left turn signal when the pedestrian was facing left and activated the right turn signal when the pedestrian was facing right. Participants were instructed 261

to make both a shirt color and orientation response as quickly and accurately as possible and could respond to shirt color and orientation in any order they desired. Design Two fully crossed, within-subjects factors were manipulated in this study. First, single- and dualtask intersection trials were manipulated between drives. In single-task drives, the stop/go and pedestrian tasks never occurred together and in dual-task drives the two task stimuli were presented simultaneously (i.e., stimulus onset asynchrony was 0 ms). The second factor was time to stop line (TSL). The onset of the yellow light was based upon TSL the time it would take the participant s vehicle to reach the intersection stop line at its current velocity (e.g. Caird, Chisholm, Edwards, & Creaser, 2007). Six different TSL values were used in this study, 2.1, 2.35, 2.6, 2.85, 3.1, and 3.35 seconds. Each participant completed four 15-minute experimental drives, 2 single-task drives and 2 dualtask drives. Single- and dual-task drive order was counterbalanced using a Latin Square design. In each drive, participants encountered 20 signalized intersections, consisting of 6 single-task pedestrian task trials and either 6 single-task stop/go task trials, or 6 dual-task stop/go task trials depending on the drive condition. The stop/go and pedestrian tasks were presented at each level of TSL and were randomly assigned across the 20 intersections for each drive. Procedure First, participants completed a short demographic survey and were screened for 20/20 visual acuity and normal color vision. Participants propensity for motion sickness was also assessed. Participants were strongly encouraged not to participate if they scored as prone to motion sickness but were allowed to participate if they chose. After screening, participants completed 1 practice drive and 4 experimental drives. All 5 driving scenarios consisted of a four-lane roadway, 2 lanes traveling in each direction, with a series of signalized intersections and gentle turns. Participants were instructed to stay in the right hand lane for the duration of the drive, to avoid making lane changes, and to proceed straight at all intersections. The speed limit was 40 miles per hour for each drive and no ambient traffic was present in any of the driving scenarios. The first drive that participants completed was a 15-minute practice drive. The practice drive consisted of 16 intersections, 8 with the single-task pedestrian stimulus, 4 with the single-task stop/go task, and 4 dual-task trials. Once comfortable with the driving simulator and experimental tasks, participants completed the experimental drives with up to a 15-minute break between each. Between drives, participants completed a survey where they estimated their driving performance. The survey results were intended for another study and will not be discussed further. At the end of the study participants were debriefed and dismissed. 262

RESULTS Stop/Go Decisions The purpose of the current study was to examine the effects of distractions on stop or go decisions in response to yellow light changes. Stop or go responses were recorded at each yellow light change. A stop response was recorded if the driver s speed dropped below 0.1 m/s before clearing the intersection, and a go response was recorded if the driver did not attempt to bring the vehicle to a stop at the intersection. Overall, participants stopped at 53.4 % of the intersections. The rate of stop decisions also varied as a function of TSL (see Table 1). Participants were least likely to stop when the light change occurred 2.1 seconds from the stop line and were most likely to stop when the light change occurred at 3.35 seconds from the stop line. Similar to Caird, et al. (2007), the 50/50 stop or go decision point observed in this study was slightly less than 2.6 seconds to stop line. Comparing stop responses between single- and dual-task trials, participants stopped slightly more often in dual-task trials (54.2 % stop decisions) compared to single-task trials (52.7 % stop decisions). A similar pattern was observed across each TSL condition, except for the 2.85 condition, where drivers stopped more often in dual-task trials. Table 1. Percent of stop responses for each condition by time to stop line Condition Time to stop line (sec) 2.10 2.35 2.60 2.85 3.10 3.35 Overall Single-task 14.5 25.4 50.8 65.1 73.0 87.1 52.7 Dual-task 16.4 30.2 52.5 57.1 78.8 87.5 54.2 Overall 15.4 27.8 51.6 61.1 76.0 87.3 53.4 Response Time Braking and pedestrian task response time (RT) in single- and dual-task conditions was compared to look for evidence of dual-task interference. For stop trials, brake RT was calculated as the time from yellow light onset to the first depression of the brake pedal. Pedestrian RT was calculated as the time from the appearance of the pedestrian stimulus until the initiation of the first button or blinker response. Linear mixed models were used to analyze the response time data and t-values greater than or equal to 2 were considered significant at the =.05 level. Brake RTs and pedestrian RTs less than 200 ms and pedestrian RTs greater than 4.5 seconds were not considered in the following analyses (< 0.1% of total data). Brake RTs were 48 ms slower in dual-task trials (M = 1000 ms, SD = 270) than single-task trials (M = 952 ms, SD = 226), however, this difference was not significant (b = 18.9, SE = 91.2, t =.2). Brake RTs, however, were affected by TSL. Brake RTs were significantly faster in the 2.1 TSL condition compared to the 2.85, 3.1, and 3.35 TSL conditions (b = 223.9, SE = 73.5, t = 3.0; b = 220.5, SE = 72.8, t = 3.0; and b = 262.7, SE = 72.1, t = 3.6, respectively). A dual-task effect was observed in the pedestrian task (b = 176.5, SE = 74.6, t = 2.4). Pedestrian task RTs were 263

nearly 300 ms slower in dual-task trials (M = 1393 ms, SD = 648) compared to single-task trials (M = 1099 ms, SD = 372). Thus, contrary to expectations, the pedestrian task did not appear to interfere with braking responses, but braking responses appeared to interfere with pedestrian task responses. Response Strategy It was surprising that the pedestrian task did not interfere with braking responses considering that numerous studies have found distraction effects in braking responses (see Horrey & Wickens, 2006). After closer investigation of stop decision trials, it became apparent that participants engaged in two different dual-task response strategies. In 56% of dual-task stop trials participants responded to the stop/go task before responding to the pedestrian task, while the remaining 44% participants responded to the pedestrian task first. According to central bottleneck theory, response strategies would dictate the locus of dual-task interference. If response selection occurred in the pedestrian task before the stop/go task, then the stop/go task would be delayed and slower brake RTs would result. If drivers responded to the stop/go task before the pedestrian task then evidence of dual-task interference would be seen in pedestrian task RTs. Stop decision trials were grouped by response strategy, brake first and pedestrian first. Observations were collapsed across TSL since the main effect of response order was the comparison of interest and similar proportions of observations for each response strategy were observed at each level of TSL. Response strategy had a significant influence on brake and pedestrian task RTs (see Table 2). There was no dual-task interference in brake RTs for the brake first response strategy (b = -27.1, SE = 26.5, t = 1). Braking responses in the pedestrian first strategy, however, were significantly slower than single-task braking responses (b = 238.5, SE = 36.1, t = 6.6) and braking responses in the brake first strategy (b = 265.6, SE = 39.4, t = 6.7). Pedestrian task RTs were significantly slower in dual-task trials compared to single-task trials regardless of whether a pedestrian task response (b = 123.3, SE = 26.5, t = 4.6) or braking response was made first (b = 740.3, SE = 37.6, t = 19.7). However, pedestrian RTs were significantly longer in the brake first strategy than pedestrian first strategy (b = 617, SE = 42.7, t = 14.5). As predicted by the central bottleneck model, the pedestrian first strategy interfered with central processing in the stop/go task and the brake first strategy interfered with central processing in the pedestrian task. Table 2. Brake and pedestrian task response time by response strategy Response Strategy Brake RT (ms) Pedestrian task RT (ms) M SD M SD Dual-Task: Pedestrian First 1234 296 1175 518 Dual-Task: Brake First 953 190 1942 740 Brake Task Only 989 235 -- -- Pedestrian Task Only -- -- 1099 372 264

DISCUSSION The results supported a central bottleneck account of dual-task interference in driving (e.g., Levy et al., 2006; Monk & Kidd, under review), but, unexpectedly, this effect resulted in a greater rate of stop responses. Distractions were expected to encourage more go responses at intersections, because central processing delays would cause stop or go decisions to occur closer to the intersection. Similar to Cooper at al. (2003), however, participants tended to stop more often when distracted. This is surprising considering that distractions led to more go responses in a similar experimental paradigm as the current study (Monk & Kidd, under review). A possible explanation for this discrepancy is discussed below. Different response strategies may have affected stop or go decisions in dual-task trials differently depending on the order of central processing. For example, when participants responded to the yellow light before the pedestrian task, response selection in the stop/go task would not be delayed resulting in similar rates of stopping as single-task trials. When participants responded to the pedestrian task before the stop/go task, however, response selection would have been delayed. As a result, drivers would be closer to the intersection (shorter TSL) when response selection occurred and would be more likely to make a go response. Unfortunately, response strategy could only be determined for stop trials and not go trials, so the current study could not explore this prediction. Future studies should explore the effects of task prioritization and response strategy on stop or go decisions. The two different response strategies provided an opportunity to observe the effects of processing bottlenecks on both braking and pedestrian task responses. Figure 1 shows the mean response time for each component of the stop/go and pedestrian tasks. Figure 1. Average response times by response strategy Compared to single-task trials, dual-task braking responses were 245 ms slower when participants responded to the pedestrian task first. Similarly, dual-task pedestrian task responses were 843 ms slower than single-task trials when participants made a brake response first. These findings supported the existence of a central bottleneck that affected both tasks depending on response strategy. 265

The effect of response strategy on braking responses has important implications for driver safety. When participants engaged in the brake first response strategy the stop/go task was protected from the performance degradation observed in the pedestrian first strategy. Furthermore, participants seemed to place a higher priority on the stop/go task since the brake first strategy occurred more often than the pedestrian first strategy. Participants may have preferred the brake first strategy to protect driving performance from the pedestrian task, which was not in direct service of the driving task (e.g., Cnossen, Meijman, & Rothengatter, 2004). High priority distractions that directly serve the driving task (e.g., destination entry) may be more likely to cause central processing bottlenecks and interfere with driving compared to distractions of lower priority. Future research should explore how task priority influences central processing bottlenecks. In conclusion, the findings in the present study provided additional support for a central bottleneck explanation of dual-task interference in driving. Furthermore, response strategy and task prioritization seemed to play a critical role in the locus of dual-task interference and warrants attention in future research. REFERENCES Alm, H. & Nilsson, L. (1995). The effects of a mobile telephone task on driver behaviour in a car following situation. Accident Analysis and Prevention, 27, 707-715. Caird, J. K., Chisholm, S. L., Edwards, C. J., & Creaser, J. I. (2007). The effect of yellow light onset time on older and younger drivers' perception response time (prt) and intersection behavior. Transportation Research Part F, 10(5), 383-396. Cooper, P. J., Zheng, Y., Richard, C., Vavrik, J., Heinrichs, B., & Siegmund, G. (2003). The impact of hands-free message reception/response on driving task performance. Accident Analysis and Prevention, 35, 23-35. Cnossen, F., Meijman, T., & Rothengatter, T. (2004). Adaptive strategy changes as a function of task demands: a study of car drivers. Ergonomics, 47(2), 218-236. Hancock, P. A., Lesch, M., & Simmons, L. (2003). The distraction effects of phone use during a crucial driving maneuver. Accident Analysis and Prevention, 35, 501-514. Harbluk, J. L., Noy, Y. I., Trbovich, P. L., & Eizenman, M. (2007). An on-road assessment of cognitive distraction: Impacts on drivers visual behavior and braking performance. Accident Analysis and Prevention, 39, 372-379. Horrey, W. J. & Wickens, C. D. (2006). Examining the impact of cell phone conversations on driving using meta-analytic techniques. Human Factors, 48(1), 196-205. Liu, B. & Lee, Y. (2005). Effects of car-phone use and aggressive disposition during critical maneuvers. Transportation Research Part F, 8, 369-382. Levy, J., Pashler, H., & Boer, E. (2006). Central interference in driving - is there any stopping the psychological refractory period? Psychological Science, 17 (3), 228-235. Monk, C. A. & Kidd, D. G. (under review). Should I stop or should I go now? Central processing delays in yellow light stop/go decisions. 266

Pashler, H. (1998). The psychology of attention. Cambridge, MA: The MIT Press. Subramanian, R. & Lombardo, L. (2007). Analysis of fatal motor vehicle traffic crashes and fatalities at intersections, 1997-2004 (No. DOT HS 810 682). Washington, DC: US Department of Transportation. 267