Small Business & Entrepreneurship

A Comparison of the Cell Phone Driver and the Drunk Driver

A Comparison of the Cell Phone Driver and the Drunk Driver
of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
  INTRODUCTION Although they are often reminded to pay fullattention to driving, people regularly engage in awide variety of multitasking activities when theyare behind the wheel. Indeed, data from the 2000U.S. census indicates that drivers spend an aver-age of 25.5 min each day commuting to work,and there is a growing interest in trying to makethe time spent on the roadway more productive(Reschovsky, 2004). Unfortunately, because of the inherent limited capacity of human attention(e.g., Kahneman, 1973; Navon & Gopher, 1979),engaging in these multitasking activities oftencomes at a cost of diverting attention away fromthe primary task of driving. There are a numberof more traditional sources of driver distraction.These “old standards” include talking to passen-gers, eating, drinking, lighting a cigarette, apply-ing makeup, and listening to the radio (Stutts etal., 2003). However, over the last decade manynew electronic devices have been developed, andthey are making their way into the vehicle. Inmany cases, these new technologies are engag-ing, interactive information delivery systems. Forexample, drivers can now surf the Internet, sendand receive E-mail or faxes, communicate via acellular device, and even watch television. Thereis good reason to believe that some of these newmultitasking activities may be substantially moredistracting than the old standards because theyare more cognitively engaging and because theyare performed over longer periods of time.The current research focuses on a dual-task activity that is commonly engaged in by morethan 100 million drivers in the United States: theconcurrent use of cell phones while driving (Cel-lular Telecommunications Industry Association,2006; Goodman et al., 1999). Indeed, the NationalHighway Transportation Safety Administration A Comparison of the Cell Phone Driver and the Drunk Driver David L. Strayer, Frank A. Drews, and Dennis J. Crouch, University of Utah, Salt LakeCity, Utah Objective: The objective of this research was to determine the relative impairmentassociated with conversing on a cellular telephone while driving. Background: Epidemiological evidence suggests that the relative risk of being in a traffic accidentwhile using a cell phone is similar to the hazard associated with driving with a bloodalcohol level at the legal limit.The purpose of this research was to provide a directcomparison of the driving performance of a cell phone driver and a drunk driver ina controlled laboratory setting. Method: We used a high-fidelity driving simulatorto compare the performance of cell phone drivers with drivers who were intoxicatedfrom ethanol (i.e., blood alcohol concentration at 0.08% weight/volume). Results: When drivers were conversing on either a handheld or hands-free cell phone, theirbraking reactions were delayed and they were involved in more traffic accidents thanwhen they were not conversing on a cell phone. By contrast, when drivers were intox-icated from ethanol they exhibited a more aggressive driving style, following closerto the vehicle immediately in front of them and applying more force while braking. Conclusion: When driving conditions and time on task were controlled for, the im-pairments associated with using a cell phone while driving can be as profound asthose associated with driving while drunk. Application: This research may help toprovide guidance for regulation addressing driver distraction caused by cell phoneconversations. Address correspondence to David L. Strayer, Department of Psychology, 380 South, 1530 East, RM 502, University of Utah,Salt Lake City, UT84112-0251;  H  UMAN   F  ACTORS  , Vol. 48, No. 2, Summer 2006, pp. 381–391.Copyright ©2006, Human Factors and Ergonomics Society. All rights reserved.  382Summer 2006 – Human Factors estimated that 8% of drivers on the roadway atany given daylight moment are using their cellphone (Glassbrenner, 2005). It is now well estab-lished that cell phone use impairs the driving per-formance of younger adults (Alm & Nilsson,1995; Briem & Hedman, 1995; Brookhuis, DeVries, & De Waard, 1991; I. D. Brown, Tickner, &Simmonds,1969; Goodman et al.,1999; McKnight& McKnight, 1993; Redelmeier & Tibshirani,1997; Strayer, Drews, & Johnston, 2003; Strayer& Johnston, 2001). For example, drivers aremore likely to miss critical traffic signals (trafficlights, a vehicle braking in front of the driver,etc.), slower to respond to the signals that they dodetect, and more likely to be involved in rear-endcollisions when they are conversing on a cellphone (Strayer et al., 2003). In addition, evenwhen participants direct their gaze at objects inthe driving environment, they often fail to “see”them when they are talking on a cell phone be-cause attention has been directed away from theexternal environment and toward an internal,cognitive context associated with the phone con-versation. However, what is lacking in the litera-ture is a clear benchmark with which to evaluatethe relative risks associated with this dual-task activity (e.g., Brookhuis, 2003).In their seminal article, Redelmeier and Tib-shirani (1997) reported epidemiological evidencesuggesting that “the relative risk [of being in atraffic accident while using a cell phone] is sim-ilar to the hazard associated with driving with ablood alcohol level at the legal limit” (p. 456).These estimates were made by evaluating the cel-lular records of 699 individuals involved in motorvehicle accidents. It was found that 24% of theseindividuals were using their cell phone within the10-min period preceding the accident, and thiswas associated with a fourfold increase in thelikelihood of getting into an accident. Moreover,these authors suggested that the interferenceassociated with cell phone use was attributable toattentional factors rather than to peripheral fac-tors such as holding the phone. However, thereare several limitations to this important study.First, although the study established a strongassociation between cell phone use and motorvehicle accidents, it did not demonstrate a causal link between cell phone use and increased accidentrates. For example, there may be self-selectionfactors underlying the association: People whouse their cell phone while driving may be morelikely to engage in risky behavior, and this in-crease in risk taking may be the cause of the cor-relation. It may also be the case that being in anemotional state may increase one’s likelihood of driving erratically and may also increase the like-lihood of talking on a cell phone. Finally, limita-tions on establishing an exact time of the accidentlead to uncertainty regarding the precise rela-tionship between talking on a cell phone whiledriving and increased traffic accidents.If the relative risk estimates of Redelmeier andTibshirani (1997) can be substantiated in a con-trolled laboratory experiment and there is acausal link between cell phone use and impaireddriving, then these data would be of immenseimportance for public safety and legislative bod-ies. Here we report the result of a controlled studythat directly compared the performance of driv-ers who were conversing on either a handheld or hands-free cell phone with the performance of drivers with a blood alcohol concentration at 0.08% weight/volume (wt/vol). Alcohol hasbeen used as a benchmark for assessing perfor-mance impairments in a variety of other areas,including aviation (Billings, Demosthenes, White,& O’Hara, 1991; Klein, 1972), anesthesiology(Thapar, Zacny, Choi,& Apfelbaum,1995; Tiplady,1991) nonprescription drug use (Burns & Mos-kovitz, 1980), and fatigue (Williamson, Feyer,Friswel,& Finlay-Brown,2001). Indeed, the WorldHealth Organization recommended that the be-havioral effects of drugs be compared with thoseof alcohol under the assumption that performanceon drugs should be no worse than that at the legalblood alcohol limit (Willette & Walsh, 1983).We used a car-following paradigm (see alsoAlm & Nilsson, 1995; Lee, Vaven, Haake, &Brown, 2001; Strayer et al., 2003) in which par-ticipants drove on a multilane freeway followinga pace car that would brake at random intervals.We measured a number of performance variables(e.g., driving speed, following distance, brake re-action time, time to collision) that have beenshown to affect the likelihood and severity of rear-end collisions, the most common type of traffic accident reported to police (T. L. Brown,Lee, & McGehee, 2001; Lee et al., 2001). Threecounterbalanced conditions were studied using awithin-subjects design: single-task driving (base-line condition), driving while conversing on a  C ELL P HONE D RIVERSAND D RUNK  D RIVERS 383 cell phone (cell phone condition), and drivingwith a blood alcohol concentration of 0.08% wt/ vol (alcohol condition). The driving tasks wereperformed on a high-fidelity driving simulator. METHODParticipants Forty adults (25 men, 15 women), recruitedvia advertisements in local newspapers, partici-pated in the Institutional Review Board approvedstudy. Participants ranged in age from 22 to 34years, with an average age of 25 years. All hadnormal or corrected-to-normal vision and a validdriver’s license with an average of 8 years of driving experience. Of the 40 participants, 78%owned a cell phone, and 87% of the cell phoneowners reported that they have used a cell phonewhile driving. Afurther requirement for inclusionin the study was that participants were socialdrinkers, consuming between three and five alco-holic drinks per week. The experiment lastedapproximately 10 hr (across the three days of thestudy), and participants were remunerated at arate of $10/hr.Apreliminary comparison of male and femaledrivers found greater variability in following dis-tance for female drivers, F  (1, 38) = 10.9,  p < .01;however, this gender effect was not modulated byalcohol or cell phone use. No other effects of gender were significant in the current sample. Ad-ditional analyses comparing the driving perfor-mance of participants who owned a cell phone withthat of those who did not own a cell phone failedto find any significant differences (all  p s > .60).Similarly, there was no significant difference indriving performance between participants whoreported that they used a cell phone while driv-ing and those who did not use a cell phone whiledriving (all  p s >.70). Stimuli and Apparatus APatrolSim high-fidelity driving simulator,illustrated in Figure 1 and manufactured by GE-ISIM, was used in the study. The simulator is com-posed of five networked microprocessors andthree high-resolution displays providing a 180°field of view. The dashboard instrumentation,steering wheel, gas pedal, and brake pedal arefrom a Ford Crown Victoria ® sedan with an auto-matic transmission. The simulator incorporatesproprietary vehicle dynamics, traffic scenario,and road surface software to provide realisticscenes and traffic conditions.Afreeway road database simulated a 24-mile(38.6-km) multilane interstate with on- and off-ramps, overpasses, and two- or three-lane trafficin each direction. Daytime driving conditions withgood visibility and dry pavement were used. Apace car, programmed to travel in the right-hand Figure 1. Aparticipant talking on a cell phone while driving in the GE-ISIM driving simulator.  lane, braked intermittently throughout the sce-nario. Distractor vehicles were programmed todrive between 5% and 10% faster than the pacecar in the left lane, providing the impression of asteady flow of traffic. Unique driving scenarios,counterbalanced across participants, were usedfor each condition in the study. Measures of real-time driving performance, including drivingspeed, distance from other vehicles, and brakeinputs, were sampled at 30 Hz and stored for lateranalysis. Cellular service was provided by SprintPCS. The cell phone was manufactured by LGElectronics Inc. (Model TP1100). For hands-freeconditions, a Plantronics M135 headset (withearpiece and boom microphone) was attached tothe cell phone. Blood alcohol concentration levelswere measured using an Intoxilyzer 5000, man-ufactured by CMI Inc. Procedure The experiment used a within-subjects designand was conducted in three sessions on differentdays. The first session familiarized participantswith the driving simulator using a standardizedadaptation sequence. The order of subsequentalcohol and cell phone sessions was counterbal-anced across participants. In these latter sessions,the participant’s task was to follow the intermit-tently braking pace car driving in the right-handlane of the highway. When the participant steppedon the brake pedal in response to the braking pacecar, the pace car released its brake and acceleratedto normal highway speed. If the participant failedto depress the brake, he or she would eventuallycollide with the pace car. That is, as in real high-way stop-and-go traffic, the participant wasrequired to react in a timely and appropriate man-ner to a vehicle slowing in front of them.Figure 2 presents a typical sequence of eventsin the car-following paradigm. Initially both theparticipant’s car (solid line) and the pace car (long-dashed line) were driving at about 62 miles/hr(mph) with a following distance of 40 m (dottedline). At some point in the sequence, the pacecar’s brake lights illuminated for 750 ms (short-dashed line) and the pace car began to decelerateat a steady rate. As the pace car decelerated, fol-lowing distance decreased. At a later point in time,the participant responded to the decelerating pacecar by pressing the brake pedal. The time intervalbetween the onset of the pace car’s brake lightsand the onset of the participant’s brake responsedefines the brake onset time. Once the participantdepressed the brake, the pace car began to accel-erate, at which point the participant removed hisor her foot from the brake and applied pressureto the gas pedal. Note that in this example, follow-ing distance decreased by about 50% during thebraking event.In the alcohol session, participants drank a mix-ture of orange juice and vodka (40% alcohol byvolume) calculated to achieve a blood alcohol 384Summer 2006 – Human Factors   Figure 2. An example of the sequence of events occurring in the car following paradigm.  C ELL P HONE D RIVERSAND D RUNK  D RIVERS 385 concentration of 0.08% wt/vol. Blood alcoholconcentrations were verified using infrared spec-trometry breath analysis immediately before andafter the alcohol driving condition. Participantsdrove in the 15-min car-following scenario whilelegally intoxicated. Average blood alcohol con-centration before driving was 0.081% wt/vol andafter driving was 0.078% wt/vol.In the cell phone session, three counterbal-anced conditions, each 15 min in duration, were in-cluded: single-task baseline driving, driving whileconversing on a handheld cell phone, and drivingwhile conversing on a hands-free cell phone. Inboth cell phone conditions, the participant and aresearch assistant engaged in naturalistic conver-sations on topics that were identified on the firstday as being of interest to the participant. As wouldbe expected with any naturalistic conversation,they were unique to each participant. The task of the research assistant in our study was to main-tain a dialog in which the participant listened andspoke in approximately equal proportions. How-ever, given that our cell phone conversations werecasual, they probably underestimate the impactof intense business negotiations or other emo-tional conversations conducted over the phone.To minimize interference from manual compo-nents of cell phone use, the call was initiatedbefore participants began driving. RESULTS In order to better understand the differencesbetween conditions, we created driving profiles byextracting 10-s epochs of driving performance thatwere time locked to the onset of the pace car’sbrake lights. That is, each time that the pace car’sbrake lights were illuminated, the data for the en-suing 10 s were extracted and entered into a 32 × 300 data matrix (i.e., on the  j th occasion that thepace car brake lights were illuminated,data fromthe 1st, 2nd, 3rd, …, and 300th observations fol-lowing the onset of the pace car’s brake lights wereentered into the matrix  X  [j,1] ,  X  [j,2] ,  X  [j,3] ,...  X  [j,300] ,in which  j ranges from 1 to 32 reflecting the 32occasions in which the participant reacted to thebraking pace car). Each driving profile was creat-edby averaging across  j for each of the 300 timepoints. We created profiles of the participant’sbraking response, driving speed, and followingdistance.Figure 3 presents the braking profiles. In thebaseline condition, participants began brakingwithin 1 s of pace car deceleration. Similar brak-ing profiles were obtained for both the cell phoneand alcohol conditions. However, compared withbaseline, when participants were intoxicated theytended to brake with greater force, whereas par-ticipants’reactions were slower when they wereconversing on a cell phone.Figure 4 presents the driving speed profiles. Inthe baseline condition, participants began decel-erating within 1 s of the onset of the pace car’sbrake lights, reaching minimum speed 2 s afterthe pace car began to decelerate, whereupon par-ticipants began a gradual return to prebrakingdriving speed. When participants were intoxicat-ed they drove slower, but the shape of the speed Figure 3. The braking profile. Figure 4. The speed profile. Braking Profile Speed Profile
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!