distraction while driving: the case of older drivers
TRANSCRIPT
Transportation Research Part F 14 (2011) 638–648
Contents lists available at SciVerse ScienceDirect
Transportation Research Part F
journal homepage: www.elsevier .com/locate / t r f
Distraction while driving: The case of older drivers
Julia Fofanova ⇑, Mark VollrathDepartment of Engineering and Traffic Psychology, Technische Universität Braunschweig, Gaußstraße 23, D-38106 Braunschweig, Germany
a r t i c l e i n f o
Article history:Received 14 July 2010Received in revised form 19 August 2011Accepted 28 August 2011
Keywords:Older driversIn-vehicle distractionSecondary taskDriving simulationLane Change Task
1369-8478/$ - see front matter � 2011 Elsevier Ltddoi:10.1016/j.trf.2011.08.005
⇑ Corresponding author. Tel.: +49 531 391 3615;E-mail addresses: [email protected] (J. Fofanov
a b s t r a c t
As the impairment of older drivers is especially found in perception and attention, onecould assume that they are especially prone to distraction effects of secondary tasks per-formed while driving. The aim of the study was to examine the effect of age on driving per-formance as well as the compensation strategies of older drivers under distraction. 10middle-aged and 10 older drivers drove in a simulator with and without a secondary task.To assess driving performance the Lane Change Task (Mattes, 2003) was used. This methodaims at estimating driver demand while a secondary task is being performed, by measuringperformance degradation on a primary driving-like task in a standardized manner. The sec-ondary task – a self-developed computer-based version of ‘‘d2 Test of Attention’’ was pre-sented both with and without time pressure. The results show that older participants’overall driving performance (mean deviation from an ideal path) was worse in all condi-tions as compared to the younger ones. With regard to lane change reaction time bothage groups were influenced by distraction in a comparable manner. However, when thelane keeping performance (standard deviation of the lateral position) was examined, theolder participants were more affected than the younger ones. This pattern could beexplained by compensation strategies of the older drivers. They focused on the most rele-vant part of the driving task, the lane change manoeuvres and were able to maintain theirperformance level in a similar way as did younger drivers. The driving performance of theolder participants was not additionally impaired when the secondary task imposed timepressure. Overall, subjective rating of driving performance, perceived workload and per-ceived distraction was found to be similar for both age groups. The observed trends andpatterns associated with distraction while driving should contribute to the further researchor practical work regarding in-vehicle technologies and older drivers.
� 2011 Elsevier Ltd. All rights reserved.
1. Introduction
The number of older drivers in the industrialized world is rising steadily. This leads to increasing concern about trafficsafety as ageing is commonly associated with psychophysiological changes which can decrease driving ability. Literature re-views indicate that three factors are most relevant for the accidents of older drivers: an impaired visual perception, problemswith attention allocation and a general slowing in decision making, planning and execution of actions (Ball, Owsley, Sloane,Roenker, & Bruni, 1993; Owsley, Stalvey, Wells, Sloane, & McGwin, 2001; Oxley, Fildes, Corben, & Langford, 2006; Rubin et al.,2007).
Many older drivers are aware of their limitations in functional capacities and adapt their driving patterns to match thesechanges by self-regulating when, where, and how to drive (Baldock, Mathias, McLean, & Berndt, 2006; Charlton et al., 2006).Several studies indicate that older drivers are able to compensate for their impairments by not driving in situations that
. All rights reserved.
fax: +49 531 391 8181.a), [email protected] (M. Vollrath).
J. Fofanova, M. Vollrath / Transportation Research Part F 14 (2011) 638–648 639
make them uneasy and by simplifying the driving task, e.g., driving slower, driving less at night, on freeways or during badweather (Bauer, Adler, Rottunda, & Kuskowski, 2003; Charlton, Oxley, Fildes, Oxley, & Newstead, 2003; Owsley, Stalvey,Wells, & Sloane, 1999). Besides, older drivers may profit from their life-long driving experience and maturity, as well asthe flexibility to drive at times and places that they perceive as being safer. The traffic insight they have acquired may givethem the ability to anticipate possible problematic situations. Hakamies-Blomqvist, Raitanen, and O’Neill (2002) found thatcompared to middle-aged drivers, older drivers had no increased crash risk per distance driven, when driving distances werecontrolled for.
Driver distraction is a significant contributor to road traffic accidents (Horberry, Anderson, Regan, Triggs, & Brown, 2006;McEvoy, Stevenson, & Woodward, 2007). Naturalistic driving studies demonstrate that drivers tend to spend a huge amountof driving time with secondary tasks (Klauer, Dingus, Neale, Sudweeks, & Ramsey, 2006). Laboratory experiments and drivingstudies show that these secondary tasks can substantially deteriorate driving safety. Up to 23% of all crashes and near-crashes are caused by the secondary task distraction (Klauer et al., 2006). The potential for a non-driving task to distractthe driver is determined by the complex interaction of a number of factors including task complexity, current driving de-mands, driver experience and skill, as well as driver’s willingness to engage in the task.
A great deal of research has been conducted on the effect of age on dual task performance, sometimes giving conflictingresults (Lindenberger, Marsiske, & Baltes, 2000; McDowd & Shaw, 2000; Riby, Perfect, & Stollery, 2004; Salthouse & Miles,2002; Verhaeghen, Steitz, Sliwinski, & Cerella, 2003). The reasons for the variable research findings in the literature couldlie in the diversity of methods and the numerous task combinations used. The meta-analysis of Riby et al. (2004) conductedon the results of 34 studies found a strong overall effect size (d = .68), which indicated a clear age-related dual-taskingimpairment. However, this effect size was not representative of all the individual studies reported. Subsequent analysis ofstudy characteristics indicated that task domain was critical in moderating age differences in dual task performance. Nota-bly, tasks with a substantial controlled processing (e.g., episodic memory) or motor component (e.g., tracking) showed great-er dual task impairment than tasks that were relatively simple or relied on automatic processing (e.g., perceptual tasks).Lindenberger et al. (2000) suggest that a decline in sensorimotor processing in ageing results in more controlled processingmechanisms and therefore when faced with competing demands older adults are particularly impaired due to a deficit inexecutive control.
Disproportional impairments in the performance of elderly people are especially obtained with increasing complexity ofthe tasks (Kliegl, Krampe, & Mayr, 2003; Li et al., 2004). In the present experiment we increased the demands in the second-ary task inducing time pressure by means of pacing. Eisdorfer (1968) suggested that pacing induces disproportional anxiety(arousal) in old persons, resulting in a performance decline. Plude and Hoyer (1986) found that the ability to discriminaterelevant from irrelevant information is most impaired in elderly people if they have to perform the task under time pressure.Logie, Della Sala, MacPherson, and Cooper (2007) examined dual task demands on encoding and retrieval processes in youn-ger and older adults and found that older people were more sensitive to time pressure in responding under dual task con-ditions. Based on the ageing and dual-task literature, we expect that as the dual-task demands increase, the drivingperformance of older drivers will deteriorate more than that of younger drivers.
Most of the research in the area of older drivers and distraction has focused on the use of mobile phones while driving.Several studies have demonstrated that the distracting effect of concurrent mobile phone use on driving performance isgreater for older drivers compared with other age groups (Cooper et al., 2003; Hancock, Lesh, & Simmons, 2003; McPhee,Scialfa, Dennis, Ho, & Caird, 2004; Reed & Green, 1999).
In contrast, in a study by Strayer and Drews (2004) no such age differences have been found. They reported that the ef-fects of hands-free phone conversation tasks on reaction time, following distance, and speed recovery after braking did notdiffer between younger and older drivers. One explanation for this inconsistent finding is that the performance of older driv-ers (aged 65–74) was compared to that of young, inexperienced drivers aged 18–25 years, rather than older, more experi-enced drivers, and these younger drivers may also be particularly susceptible to the effects of distraction.
Several studies have reported that older drivers demonstrate difficulty with the dual-task of following a route guidancesystem while driving (Dingus et al., 1997; Green, 2001). Dingus et al. reported that older drivers (65 years and older) drovemore slowly and cautiously, while making more safety-related errors (e.g., increased lane departures) compared with youn-ger drivers (16–18 years) when using an advanced traveller information system.
Despite the observed age-related decrements in dual task performance in many driver distraction studies, research hasalso shown that older drivers engage in self-regulatory behaviour while driving, in order to compensate for their greater per-formance decrements. Horberry et al. (2006), for example, examined the effects of distraction on driving performance ofyoung, mid-age and older drivers. Participants were required to perform one of two secondary tasks while driving: operatingan entertainment system and conducting a simulated hands-free mobile phone conversation in both simple and complexsimulated driving environments. The performance decrements that occurred as a result of in-vehicle distraction were ob-served in both simple and complex road environments and for drivers in different age groups. The authors reported that old-er drivers had more difficulty performing the driving task while distracted and compensated by slowing their speed incomplex highway environments. Although it appears that older drivers regulate their driving behaviour, they performedas well as younger drivers on the mobile phone task, indicating that they did not trade off mobile phone performance to en-able them to drive safely. They slowed down to give themselves an increased margin for error, possibly because they knewthey could not respond to hazards as quickly. Whether these compensatory behaviours of older drivers are sufficient to offsetthe degradation in their driving performance and reduce their crash risk, however, should be the focus of future research.
640 J. Fofanova, M. Vollrath / Transportation Research Part F 14 (2011) 638–648
As we see, the number of studies about distraction in older drivers and their ability to cope with that is limited and theirresults are contradictory. Thus, a driving simulator experiment was conducted to analyze the following questions:
� Are older drivers particularly vulnerable to distraction while driving?� How do older drivers react when the secondary task imposes time pressure?� Are older drivers able to compensate for the possible negative effects when driving under distraction?
Additionally, we were interested in age differences with regard to the subjective rating of the driving performance, theperceived workload as well as the perceived distraction.
2. Method
2.1. Participants
Although 12 older drivers were recruited for the experiment, the results of two participants (aged 66 and 78) were omit-ted from the analysis due to extremely outlying poor driving performance as well as ambiguity errors of one of these par-ticipants in the secondary task. Thus, the performance of 10 older drivers (aged 60–73, mean age = 68.4 years, SD = 4.2) and10 middle-aged drivers (aged 31–44, mean age = 38.6 years, SD = 4.0) was compared in the study. Each group included 2 fe-male and 8 male participants. All were volunteers and they participated without any reward. The participants were recruitedfrom the circle of friends and acquaintances. Based on self-reports, all participants were free from any illness interveningwith the driving task. This included normal or corrected to normal vision. Although an objective test of visual acuity wasnot possible, observations during the experiment confirmed that all drivers were well able to see all relevant stimuli. Bothgroups consisted of experienced drivers with a valid driver’s license. Older drivers had more driving experience when mea-sured in years of active driving (older drivers M = 48.4, SD = 4.9; middle-aged drivers M = 20.2, SD = 4.05). Self report esti-mate of annual mileage was as follows: an average of 13,900 km (8637 miles) for older and 15,000 km (9320 miles) foryounger drivers. The groups were matched in regard to yearly mileage.
2.2. Driving task
The participants’ task was to drive in a simulator with and without a secondary task. Participants were asked to prioritizethe driving task – the Lane Change Task (LCT, Mattes, 2003). As in real life, the first objective is to drive safely. The secondarytask should only be conducted in a manner that safe driving was still possible. The LCT, developed for in-vehicle informationsystems evaluation in the German research project ADAM (Advanced Driver Attention Metrics), aims at estimating driverdemand while a secondary task is being performed, by measuring performance degradation on a primary driving-like taskin a standardized manner. The LCT is currently advocated as an international standard (ISO/DIS 26022) to access distractioneffects.
A standard PC equipped with a joystick steering wheel, a gas and a brake pedal was used for the simulation. The visual LCTscene was displayed on a 1700 monitor. Engine noise was simulated to increase situation realism. The participants were in-structed to drive along a straight three-lane road. Although it was possible for the participants to alter their speed, they wereinstructed to drive at a constant speed of 60 km/h. The LCT-track is about 3000 m long and takes about 180 s. Participants areinstructed by signs on the roadside to perform a lane change manoeuvre. The mean distance from sign to sign is 150 m (min.140, max. 188, exponentially distributed), so that the mean duration between two lane changes is about 9 s. 18 lane changesper track have to be performed by the participants. Each of six possible lane changes occurs three times on a track. When notperforming a lane change manoeuvre participants are required to maintain a central position within the lane. Participantshave to perform manoeuvres as quickly and efficiently as possible. 10 different tracks are used to prevent the participantsfrom learning the lane change sequence. The tracks differ only in the order of and distance between the signs. Driving per-formance was recorded with 62 Hz corresponding to a precision of 16 ms.
2.3. Secondary task
The secondary task was a self-developed computer-based version of ‘‘d2 Test of Attention’’ (Brickenkamp, 2002). The d2-Test utilizes task elements that are required in any information processing exchange, i.e., perceive, process and respond.The d2-task is easy-to-learn; it can be interrupted and resumed. Thus, carrying out the d2-task can be comparable tovisual-manual operating of in-car devices (e.g. operating the audio system, entering route destination details or respondingto guidance instructions). Kiefer, Schulz, Schulze-Kissing, and Urbas (2006) previously used another modified in-car-version,named D2-Drive (Urbas et al., 2005) in a driving simulator-study with the objective to research cognitive heuristics inmultitasking performance.
In the present study one item (either target or distractor, font size 40) was presented in the middle of a standard com-puter screen positioned on the participant’s right-hand side. The participant was required to press the key ‘‘1’’ when the tar-get letter ‘‘d’’ marked with two dashes was displayed on the monitor. For all distractors the key ‘‘2’’ had to be pressed. 15
J. Fofanova, M. Vollrath / Transportation Research Part F 14 (2011) 638–648 641
different distractors were used, which were made up of combinations of the letters ‘‘d’’ and ‘‘p’’ with none or one to fourdashes arranged either above or below the letter. The order in which the items were presented on the screen was random-ized. This secondary task was presented both with and without time pressure. In the time pressure condition, every 1500 msa new item was presented. In the condition without time pressure, the presentation was self-paced.
2.4. Subjective rating
Participants were asked to rate their driving performance, their perceived workload as well as their perceived distractionafter completion of each driving condition. The subjective rating was done on a 15-point scale in a two-step procedure.Participants first selected of five verbal categories ranging from ‘‘very badly’’ to ‘‘very well’’ for driving performance and‘‘very little’’ to ‘‘very strong’’ for perceived workload as well as for perceived distraction. For the category selected theyrefined their assessment using a 3-point scale (�/0/+) (Heller, 1982).
2.5. Procedure
Following completion of a consent form and experiment information sheet, participants were briefed on the operation ofthe LCT. All participants practiced the driving task as well as the secondary task in isolation and in combination. After thepractice session of the LCT (a block of four LCT trials lasting 12 min) participants completed two baseline tracks (just drive),a baseline run of the secondary task (3 min) and an experimental drive accompanied by the secondary task with and withouttime pressure (two tracks each). Participants were asked to prioritize the LTC in the dual-task condition. The order of thesecondary task was counterbalanced across participants in each age group. Participants were asked to rate their driving per-formance, their perceived workload as well as their perceived distraction after completion of each driving condition.
2.6. Experimental design and statistical analysis
A 2 (age group: middle-aged, older) � 3 (secondary task: none, secondary task without time pressure, secondary task withtime pressure) repeated measures MANOVA was used to analyze the LCT performance. The driving performance parameterof major interest was the mean deviation (measured in m). It refers to the deviation between a normative model (i.e. an‘‘ideal’’ vehicle path) and the actual driving course of the participant along the track. The mean deviation covers importantaspects of the driver’s performance, namely his/her perception (late perception of the sign or missing a sign), quality of themanoeuvre (slow lane change results in larger deviation) and lane keeping quality, which all result in an increased deviation.Additionally, for the lane-keeping phases, the standard deviation of the lateral position (SDLP, measured in m) was computed.For the lane change initiation, the reaction time (in s) was measured from the point where the sign became legible to the pointwhere the driver started to steer.
To analyze the performance in the d2-secondary task a 2 (age group: middle-aged, older) � 2 (conditions: baseline,test) � 2 (time pressure: with, without) repeated measures ANOVA for every dependent variable was computed. The d2-per-formance was measured using the mean response time (in ms), the percentage of correct responses and the number of items com-pleted in a minute.
A significance level of p < .05 was adopted for all statistical tests. The reported effect size g2p (partial eta squared) is de-
fined as the proportion of the effect + error variance that is attributable to the effect.
3. Results
3.1. Driving performance
Table 1 presents the descriptive statistics for three driving performance measures described earlier. The MANOVA indi-cated significant main effects of age (F(3,7) = 5.4, p = .031, g2
p = .698), task (F(6,34) = 5.6, p = .001, g2p = .499) and an interaction
(F(6,34) = 3.7, p = .006, g2p = .396). The effect sizes show that the main effect of age is the strongest. Thus, driver’s age signif-
icantly influences driving performance.
Table 1Descriptive statistics for driving performance measures as a function of age and three conditions.
LCT-baseline LCT + d2 no time pressure LCT + d2 time pressure
Older Mid-age Older Mid-age Older Mid-age
Mean deviation (m) 1.14 (0.25) 0.90 (0.11) 1.43 (0.25) 1.11 (0.19) 1.55 (0.32) 1.12 (0.17)SDLP (m) 0.11 (0.08) 0.04 (0.02) 0.29 (0.19) 0.05 (0.04) 0.25 (0.15) 0.06 (0.04)Reaction time (s) 0.21 (0.18) 0.08 (0.04) 0.33 (0.13) 0.19 (0.11) 0.38 (0.23) 0.20 (0.10)
Note. Standard deviations are presented in parentheses.SDLP – standard deviation of the lateral position.
Table 2ANOVA results for dependent measures related to driving performance.
Source Mean deviation (m) SDLP (m) Reaction time (s)
F df p g2p F df p g2
p F df p g2p
Between-participantsAge (A) 15.8 1,9 .003 .638 13.5 1,9 .005 .601 7.8 1,9 .021 .465
Within-participantsTask (T) 43.9 2,18 .001 .830 11.8 2,18 .001 .568 12.4 2,18 .001 .581T � A 2.7 2,18 .091 .234 11.2 2,18 .001 .556 0.5 2,18 .602 .055
642 J. Fofanova, M. Vollrath / Transportation Research Part F 14 (2011) 638–648
The univariate analyses (Table 2) indicated significant main effect of age and secondary task for all LCT performanceparameters. Post-hoc tests for the pair-wise comparisons (Bonferonni corrected) indicated that there was no significant dif-ference between the two dual-task conditions for the three driving performance parameters. However, both dual-task con-ditions differed from the baseline (p = .001).
The older drivers’ mean deviation (see Fig. 1) was larger than that of the middle-aged participants in all three conditions(F(1,9) = 15.8, p = .003, g2
p = .638). Additionally, in both age groups an increase was found due to the secondary task. There wasalso a tendency for a stronger decrement in older participants under time pressure (Task � Age: F(2,18) = 2.7, p = .091,g2
p = .234).With regard to lane change reaction time (see Fig. 2) the older drivers reacted slower in all conditions than the mid-age
participants did (F(1,9) = 7.8, p = .021, g2p = .465). Both age groups were influenced by distraction in a comparable manner.
The significant Task � Age interaction for the standard deviation of the lateral position (F(2,18) = 11.2, p = .001, g2p = .556)
shows that the older drivers were more affected in the lane-keeping in the dual-task conditions in comparison to the youn-ger participants (see also Fig. 3).
00.20.40.60.8
11.21.41.61.8
2
Baseline LCT + d2 no time
pressure
LCT + d2 time
pressure
mea
n de
viat
ion
[m]
older drivers middle-aged drivers
Fig. 1. Mean deviation from an optimal trajectory as a function of age and of a secondary task condition (whiskers represent SD).
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Baseline LCT + d2 no time
pressure
LCT + d2 time
pressure
reac
tion
time
[s]
older drivers middle-aged drivers
Fig. 2. Lane change reaction time as a function of age and of a secondary task condition (whiskers represent SD).
Baseline LCT + d2 no time
pressure
LCT + d2 time
pressureSD
LP [m
]
older drivers middle-aged drivers
Fig. 3. Standard deviation of lateral position (SDLP) as a function of age and of a secondary task condition (whiskers represent SD).
J. Fofanova, M. Vollrath / Transportation Research Part F 14 (2011) 638–648 643
3.2. Secondary task performance
Table 3 presents the descriptive statistics for three d2-performance measures described earlier. To analyze the perfor-mance in the d2-secondary task a 2 (age group: middle-aged, older) � 2 (conditions: baseline, test) � 2 (time pressure: with,without) repeated measures ANOVA for every dependent variable was computed.
Table 4 presents ANOVA results for dependent measures related to the secondary task performance. In regard with themean response time there was a significant interaction Task � Time pressure (F(1,9) = 18.0, p = .002, g2
p = .667), indicating thatboth age groups were slower in the dual-task condition without time pressure in comparison to d2-baseline. The significantTime pressure � Age interaction (F(1,9) = 5.1, p = .049, g2
p = .365) indicates that the older drivers took significantly more timeto respond to items in the conditions without time pressure than the mid-aged participants. There was almost no differencein the mean response time between the two age groups in the conditions with time pressure (see also Fig. 4).
Fig. 5 shows the percentage of correct responses as a function of age, time pressure and a secondary task condition. Undertime pressure both age groups made more mistakes in the dual-task condition than in d2-baseline (Task � Time pressure:F(1,9) = 43.1, p = .001, g2
p = .827).With regard to the number of items completed in a minute (Fig. 6) the strongest effect was that of a dual task (F(1,9) = 110.5,
p = .001, g2p = .925). Participants in both groups completed fewer items when concurrently driving. Additionally, older
Table 4ANOVA results for dependent measures related to the secondary task performance.
Source Mean response time % Corrected responses Number completed (min)
F df p g2p F df p g2
p F df p g2p
Between-subjectsAge (A) 6.4 1,9 .031 .419 5.1 1,9 .049 .365 8.2 1,9 .018 .480
Within-subjectsTask (T) 15.4 1,9 .003 .632 31.4 1,9 .001 .777 110.5 1,9 .001 .925Time pressure (TP) 50.4 1,9 .001 .849 389.2 1,9 .001 .977 4.5 1,9 .062 .336T � A 1.5 1,9 .239 .150 0.4 1,9 .505 .051 1.5 1,9 .248 .145T � TP 18.0 1,9 .002 .667 43.1 1,9 .001 .827 3.9 1,9 .078 .305TP � A 5.1 1,9 .049 .365 0.6 1,9 .449 .065 1.8 1,9 .209 .169T � TP � A 2.6 1,9 .138 .228 0.3 1,9 .591 .033 2.0 1,9 .188 .184
Table 3Descriptive statistics for d2 performance measures as a function of age and task conditions.
Mean response time (ms) % Correct responses Number completed (min)
Older Mid-age Older Mid-age Older Mid-age
d2-Baseline no time pressure 1173.9 (720.7) 841.4 (389.7) 88.6 (11.7) 93.5 (7.3) 41.5 (9.9) 48.3 (8.1)d2-Baseline time pressure 657.3 (57.0) 592.5 (51.2) 54.1 (18.9) 63.5 (28.0) 38.8 (2.9) 39.7 (1.3)LCT + d2 no time pressure 2594.2 (1281.7) 1544.3 (715.9) 90.3 (17.4) 96.9 (1.6) 22.6 (6.6) 31.6 (7.9)LCT + d2 time pressure 569.3 (49.2) 598.9 (37.9) 21.2 (9.1) 38.1 (7.7) 21.2 (7.7) 29.0 (6.3)
Note. Standard deviations are presented in parentheses.
mea
n re
spon
se ti
me
[ms]
older middle-aged
Fig. 4. Mean response time as a function of age, time pressure and a secondary task condition (whiskers represent SD).
0102030405060708090
100110
corre
cted
resp
onse
s [%
]
older middle-aged
Fig. 5. Percentage of correct responses as a function of age, time pressure and a secondary task condition (whiskers represent SD).
0
10
20
30
40
50
60
Num
ber c
ompl
eted
/min
.
older middle-aged
Fig. 6. Number of items completed in a minute as a function of age, time pressure and a secondary task condition (whiskers represent SD).
644 J. Fofanova, M. Vollrath / Transportation Research Part F 14 (2011) 638–648
J. Fofanova, M. Vollrath / Transportation Research Part F 14 (2011) 638–648 645
participants completed significantly fewer items per minute than younger ones in all conditions except for d2-Baseline withtime pressure (F(1,9) = 8.2, p = .018, g2
p = .480). Also, there was a trend towards a significant Task � Time pressure interaction(F(1,9) = 3.9, p = .078, g2
p = .305).
3.3. Subjective measures
Table 5 presents the means and standard deviations for subjective measures. A 2 (age group: middle-aged, older) � 3(secondary task: none, secondary task without time pressure, secondary task with time pressure) repeated measures ANOVAwas used to analyze subjective rating of driving performance and perceived workload. To analyze the perceived distraction a2 (age group: middle-aged, older) � 2 (secondary task: secondary task without time pressure, secondary task with time pres-sure) repeated measures ANOVA was used. Because perceived distraction was not rated after driving without a secondarytask (LCT-Baseline), ANOVA included only two dual-task conditions. Accordingly, post-hoc comparisons were not necessary.
Overall, subjective rating of driving performance, perceived workload and perceived distraction was found to be similarfor mid-age and older drivers. The univariate analyses indicated that both groups judged their driving performance (Fig. 7) tobe worse in the dual-task conditions (F(2,18) = 18.6, p = .001, g2
p = .675). However the post-hoc comparisons indicated no dif-ference between the two dual-task conditions (with and without time pressure) for both age-groups. The significant
Table 5Descriptive statistics for subjective ratings as a function of age and three conditions.
LCT-baseline LCT + d2 no time pressure LCT + d2 time pressure
Older Mid-age Older Mid-age Older Mid-age
Driving performance 10.8 (1.8) 9.3 (1.9) 7.9 (2.8) 8.3 (1.7) 7.2 (1.6) 7.9 (1.4)Perceived workload 5.5 (2.4) 5.4 (2.1) 7.1 (1.4) 9.1 (1.7) 7.9 (1.7) 9.0 (1.8)Perceived distraction – – 8.4 (1.7) 9.4 (1.7) 11.0 (1.6) 10.0 (1.4)
Note. Standard deviations are presented in parentheses.
Baseline LCT + d2 no time
pressure
LCT + d2 time
pressure
subj
ectiv
e dr
ivin
g pe
rform
ance
older drivers middle-aged drivers
Fig. 7. Subjective rating of the driving performance (whiskers represent SD).
1
3
5
7
9
11
13
15
no time pressure time pressure
perc
eive
d di
stra
ctio
n
older drivers middle-aged drivers
Fig. 8. Subjective rating of the perceived distraction as a function of age and a secondary task condition (whiskers represent SD).
646 J. Fofanova, M. Vollrath / Transportation Research Part F 14 (2011) 638–648
‘‘age � secondary task’’ interaction (F(2,18) = 3.9, p = .039, g2p = .303) indicates that older drivers rated their driving perfor-
mance to be better in the baseline condition compared to driving with secondary task.Both groups experienced increased workload in the dual-task condition (F(2,18) = 24.4, p = .000, g2
p = .731) in comparison tothe baseline-drive. However the post-hoc comparisons indicated no difference between the dual-task conditions for bothage-groups. There was also a trend towards a significant ‘‘age � secondary task condition’’ interaction (F(2,18) = 3.4,p = .057, g2
p = .273) indicating that middle-aged drivers rated their perceived workload to be higher in the secondary taskconditions than in the baseline condition.
With regard to perceived distraction (Fig. 8) there was a significant difference between the dual-task conditions with andwithout time pressure: F(1,9) = 15.4, p = .003, g2
p = .632. The significant ‘‘age � secondary task’’ interaction (F(1,9) = 6.6, p = .030,g2
p = .426) indicates that older drivers stated to be more distracted in comparison to the mid-age participants in the conditionwith time pressure.
4. Discussion
The objective of the study was to examine the following questions:
� Are older drivers particularly vulnerable to distraction while driving?� How do older drivers react when the secondary task imposes time pressure?� Are older drivers able to compensate for the possible negative effects when driving under distraction?
First of all, the overall driving performance (mean deviation from the optimal trajectory) of the older drivers was worse inall conditions as compared to the younger ones. These results are in line with previous findings of Benedict, Angell, and Dip-timan (2006) although the age of older participants in their experiment was 40–65 years. Rognin, Alidra, Val, and Lescaut(2007) to the contrary, demonstrated similar LCT-performances in baseline condition for mid-aged (aged 25–54) and older(aged 60–70) participants.
The effect of distraction depended on the different parts of the driving task (lane keeping or lane changing). With regardto lane change reaction time both age groups were influenced by distraction in a comparable manner. However, when thelane keeping performance (SDLP) was examined, the older participants were more affected than the younger ones. This pat-tern could be explained by compensation strategies of the older drivers. They focused on the most relevant part of the drivingtask, the lane change manoeuvres and were able to maintain their performance level in a similar way as younger drivers did.The division of attention between two computer screens, along with movement of the right hand away from the steeringwheel, hampered the older drivers’ ability to maintain directional control. The results of our experiment are consistent withthe findings of Merat, Anttila, and Luoma (2005), since they show that the secondary task caused marked variation in laneposition for the older drivers.
The driving performance of the older drivers was not additionally impaired when the secondary task imposed time pres-sure. But there was a trend for a stronger decrement in older participants under time pressure in terms of mean deviation.
The time pressure produced substantially faster responding in both age groups in the baseline as well as in the dual-taskconditions in the secondary task. Also apparent is that the latencies of the older participants decreased more under timepressure conditions, with the consequence that age differences were reduced. The finding that older adults had fasterresponding under time pressure conditions is similar to those of previous research (Baron & Mattila, 1989; Baron & Menich,1985).
Verhaeghen et al. (2003) examined the relations between dual-task effects and ageing through a meta-analysis of 33studies using latency as the dependent measure and 30 studies focusing on accuracy. The authors found that there was asignificant age-related deficit in dual-task performance as measured by latency costs. But no specific age deficit on accuracyassociated with dual-task processing was revealed. The authors conclude that the age effect in dual-task cost is larger thanthe general ageing effect present in single-task performance.
The results of our experiment reveal that the older drivers took significantly more time to respond to items in the dual-task condition without time pressure than the mid-aged participants. One could assume that aware of their limitations, olderdrivers abandoned the secondary task and tried to concentrate on their performance in driving. Unfortunately, this attemptof the older drivers to devote all available cognitive resources to the primary driving task was not entirely successful, as driv-ing performance was still affected.
Mean accuracy of both age groups did not differ significantly between the conditions without time pressure although theytook significantly more time to respond to items in the dual-task condition than in the baseline condition. This suggests thatspeed costs don’t lead to lower accuracy. A different pattern of outcomes was found in the time pressure conditions. Althoughparticipants responded to the items as fast as in the baseline condition, they made significantly more mistakes in the dualtask condition. Thus, accuracy costs didn’t lead to slower speed. A significant decrease in accuracy from the baseline to thedual-task condition was only due to the dual-task condition and not due to the response time.
Overall, subjective rating of driving performance, perceived workload and perceived distraction was found to be similarfor mid-age and older drivers. In the dual-task conditions both groups judged their driving performance to be worse and theyexperienced increased workload. Interestingly, although driving performance of older drivers in baseline condition was sig-nificantly worse in comparison with younger drivers, older participants rated it to be better than younger ones did. Older
J. Fofanova, M. Vollrath / Transportation Research Part F 14 (2011) 638–648 647
drivers rated their perceived workload to be lower in the dual-task conditions than the younger ones did. One can supposethat older participants tend to response either in terms of social desirability or positive self-display. Another interpretationsuggested by one anonymous reviewer would be that older adults were not so aware of any feedback the situation was giv-ing them as to their performance, i.e. inaccurate self-awareness – this is often the case in demanding conditions (spare atten-tional resources are more limited to process feedback from errors).
Caution should be taken in transferring our findings directly to on-road driving performance. The first problem with theLCT ecological validity is that the steering wheel is smaller than a normal one, so the tiniest wheel movement results in moremovement on the road than a normal wheel would give. The second problem is a small screen placed in front of the steeringwheel. This will have the effect that the middle of a lane is in front of the driver, and not to the right which is the case inreality.
Consideration should also be given to the fact that if one were required to change the lane every 9 s in a real driving sit-uation he or she would probably not engage in secondary tasks during that driving period. From this point of view, the LCT isan artificial driving task. Nevertheless it features aspects of real driving, such as perception, reaction, manoeuvring and lanekeeping.
The observed trends and patterns associated with distraction while driving should contribute to the further research orpractical work regarding in-vehicle technologies and older drivers. Future research should be focused on examining which ofthe three factors contributing to the accidents of older drivers (an impaired visual perception, problems with attention allo-cation and a general slowing in decision making, planning and execution of actions) plays the greatest role in explaining old-er driver’s performance with concurrent tasks while driving.
One could assume that if older adults are aware of their difficulties in sharing attention between various tasks, they willprobably be less inclined to combine driving with other not driving related activities such as operating a radio or a CD playeror having a telephone conversation. Lerner (2008) investigated drivers’ willingness and perceived risk of engaging in varioussecondary tasks (e.g. eating, drinking, performing different functions with a mobile phone or a navigation system). In gen-eral, younger drivers expressed more willingness than middle-aged or older drivers to use in-vehicle technologies. Youngerdrivers also perceived this use as less risky than middle-aged and older drivers. Lerner concluded that older drivers’ reluc-tance to engage in distracting tasks while driving may be a process of self-regulation.
The results of our study are in line with Lerner’s conclusions. As older participants in our experiment completed signif-icantly fewer items per minute than younger ones in both dual-task conditions, one could assume that despite the forceddual task performance older drivers abandoned the secondary task.
The extent to which older drivers are willing to engage in potentially distracting activities while driving is largely unex-plored. Therefore a field study with the objective to explore the extent to which older drivers engage in distracting activitieswhile on the road is being planned. This kind of knowledge enables to examine where the problems of older drivers withregard to distraction are really found.
References
Baldock, M. R. J., Mathias, J., McLean, J., & Berndt, A. (2006). Self-regulation of driving and older drivers’ functional abilities. Clinical Gerontologist, 30(1),53–70.
Benedict, D., Angell, L., & Diptiman, T. (2006). Exploration of the lane change test. Driver Metrics Workshop, Ottawa.Ball, K., Owsley, C., Sloane, M. E., Roenker, D. L., & Bruni, J. R. (1993). Visual attention problems as a predictor of vehicle crashes in older drivers. Investigative
Ophthalmology and Visual Science, 34(11), 3110–3123.Baron, A., & Mattila, W. R. (1989). Response slowing of older adults: Effects of time-limit contingencies on single- and dual-task performances. Psychology
and Aging, 4(1), 66–72.Baron, A., & Menich, S. R. (1985). Age-related effects of temporal contingencies on response speed and memory: An operant analysis. Journals of Gerontology,
40, 60–70.Bauer, M., Adler, G., Rottunda, S., & Kuskowski, M. (2003). The influence of age and gender on the driving patterns of older adults. Journal of Women & Aging,
15(4), 3–16.Brickenkamp, R. (2002). Test d2 Aufmerksamkeits-Belastungs-Test. Manual. 9., überarbeitete und neu normierte Auflage. Hogrefe, Göttingen.Charlton, J. L., Oxley, J., Scully, J., Koppel, S., Congiu, M., Muir, C., et al. (2006). Self-regulatory driving practices amongst older drivers in the ACT and NSW,
Monash University Accident Research Centre, Melbourne. <http://www.monash.edu.au/muarc/reports/muarc254.html> (online paper).Charlton, J. L., Oxley, J., Fildes, B., Oxley, P., & Newstead, S. (2003). Self-regulatory behaviours of older drivers. Annual Proceedings Association for the
Advancement of Automotive Medicine, 47, 181–194.Cooper, P. J., Zheng, Y., Richard, C., Vavric, J., Heinrichs, B., & Sigmund, G. (2003). The impact of hands-free message reception/response on driving task
performance. Accident Analysis and Prevention, 35, 23–35.Dingus, T. A., Hulse, M. C., Mollenhaur, M. A., Fleischman, R. N., McGehee, D. V., & Natarajan, M. (1997). Effects of age, systems experience and navigation
technique on driving with an advanced traveler information system. Human Factors, 39(2), 177–199.Eisdorfer, C. (1968). Arousal and performance: Experiments in verbal rote learning and a tentative theory. In G. A. Talland (Ed.), Human aging and behavior:
Recent advances in research and theory (pp. 189–216). New York: Academic Press.Green, P. (2001). Variations in task performance between younger and older drivers: UMTRI research on Telematics. Association for the Advancement of
Automotive Medicine Conference on Aging and Driving, Southfield, MI. <http://www.umich.edu/~driving/publications/AAAMC2001.pdf> (online paper).Hakamies-Blomqvist, L., Raitanen, T., & O’Neill, D. (2002). Driver ageing does not cause higher accident rates per km. Transportation Research Part F: Traffic
Psychology and Behaviour, 5, 271–274.Hancock, P. A., Lesh, M., & Simmons, L. (2003). The distraction effects of phone use during a crucial driving maneuver. Accident Analysis and Prevention, 35(4),
501–514.Heller, O. (1982). Theorie und Praxis des Verfahrens der Kategorienunterteilung (KU). In O. Heller (Hrsg.), Forschungsbericht 1981. Würzburg: Würzburger
Psychologisches Institut.Horberry, T., Anderson, J., Regan, M. A., Triggs, T. J., & Brown, J. (2006). Driver distraction: the effects of concurrent in-vehicle tasks, road environment
complexity and age on driving performance. Accident Analysis and Prevention, 38, 185–191.
648 J. Fofanova, M. Vollrath / Transportation Research Part F 14 (2011) 638–648
ISO Working Draft 26022 (2006). Road vehicles – Ergonomic aspects of transport information and control systems – Simulated lane change test to assess in-vehicle secondary task demand. ISO TC22/SC13 N WG8 N533.
Kiefer, J., Schulz, M., Schulze-Kissing, D. & Urbas, L. (2006). Multitasking-Strategien in der Mensch-Maschine-Interaktion. MMI-Interaktiv, 11. <http://www.zmms.tuberlin.de/prometei/download/publikationen/jki/mmi11_jki_msc_dsk_lur.pdf> (online paper).
Klauer, S. G., Dingus, T. A., Neale, V. L., Sudweeks, J. D., & Ramsey, D. J. (2006). The impact of driver inattention on near-crash/crash risk: An analysis using the100-car naturalistic driving study data. Report DOT HS 810 594. Washington, DC: National Highway Traffic Safety Administration. <http://nhtsa.uptracs.com/staticfiles/DOT/NHTSA/NRD/Articles/HF/810594/pages/TRD.html> (online paper).
Kliegl, R., Krampe, R. T., & Mayr, U. (2003). Formal models of age differences in task complexity effects. In U. Staudinger & U. Lindenberger (Eds.),Understanding human development: Life-span psychology in exchange with other disciplines (pp. 289–313). Dordrecht: Kluwer.
Lerner, N. D. (2008). Driver strategies for engaging in distracting tasks using in-vehicle technologies: final report, national highway traffic safetyadministration, Washington, DC. <http://www.nhtsa.dot.gov/staticfiles/DOT/NHTSA/NRD/Multimedia/PDFs/CrashAvoidance/2008/DOTHS810919_1.pdf>(online paper).
Li, S.-C., Lindenberger, U., Hommel, B., Aschersleben, G., Prinz, W., & Baltes, P. (2004). Transformations in the couplings among intellectual abilities andconstituent cognitive processes across the life span. Psychological Science, 15(3), 155–163.
Lindenberger, U., Marsiske, M., & Baltes, P. B. (2000). Memorizing while walking: Increase in dual-task costs from young adulthood to old age. Psychologyand Aging, 15(3), 417–436.
Logie, R. H., Della Sala, S., MacPherson, S. E., & Cooper, J. (2007). Dual task demands on encoding and retrieval processes: Evidence from healthy adult aging.Cortex, 43, 159–169.
Mattes, S. (2003). The lane-change-task as a tool for driver distraction evaluation. In H. Strasse, K. Kluth, H. Rausch, & H. Bubb (Eds.), Quality of work andproducts in enterprises of the future (pp. 57–60). Stuttgart: Ergonomia Verlag.
McDowd, J., & Shaw, R. (2000). Attention and aging: A functional perspective. In F. Craik & T. Salthouse (Eds.), Handbook of aging and cognition (pp. 221–292).Mahwah, NJ: Lawrence Erlbaum.
McEvoy, S. P., Stevenson, M. R., & Woodward, M. (2007). The prevalence of, and factors associated with, serious crashes involving a distracting activity.Accident Analysis and Prevention, 39, 475–482.
McPhee, L. C., Scialfa, C. T., Dennis, W. M., Ho, G., & Caird, J. K. (2004). Age differences in visual search for traffic signs during a simulated conversation.Human Factors, 46, 674–685.
Merat, N., Anttila, V., & Luoma, J. (2005). Comparing the driving performance of average and older drivers: the effect of surrogate in-vehicle informationsystems. Transportation Research Part F: Traffic Psychology and Behaviour, 8, 147–166.
Owsley, C., Stalvey, B. T., Wells, J., & Sloane, M. (1999). Older drivers and cataract: Driving habits and crash risk. Journals of Gerontology. Series A, BiologicalSciences and Medical Sciences, 54(4), 203–211.
Owsley, C., Stalvey, B. T., Wells, J., Sloane, M., & McGwin, G. (2001). Visual risk factors for crash involvement in older drivers with cataract. Archives ofOphthalmology, 119, 881–887.
Oxley, J., Fildes, B., Corben, B., & Langford, J. (2006). Intersection design for older drivers. Transportation Research Part F: Traffic Psychology and Behaviour, 5,335–346.
Plude, D. J., & Hoyer, W. J. (1986). Age and the selectivity of visual information processing. Psychology and Aging, 1(1), 4–10.Reed, M. P., & Green, P. A. (1999). Comparison of driving performance on-road and in a low cost simulator using a concurrent telephone dialing task.
Ergonomics, 42, 1015–1037.Riby, L., Perfect, T., & Stollery, B. (2004). The effects of age and task domain on dual task performance: A meta-analysis. European Journal of Cognitive
Psychology, 16(6), 863–891.Rognin, L., Alidra, S., Val, C., & Lescaut, A. (2007). Occurrence of secondary tasks and quality of lane changes. In D. Harris (Ed.), Engineering psychology and
cognitive ergonomics: 7th international conference, HCI International (pp. 397–406). Berlin, Heidelberg: Springer.Rubin, G., Ng, E., Bandeen-Roche, K., Keyl, P., Freeman, E., & West, S. (2007). A prospective, populations-based study of the role of visual impairment in motor
vehicle crashes among older drivers: The SEE study. Investigative Ophthalmology and Visual Science, 48, 1483–1491.Salthouse, T. A., & Miles, J. D. (2002). Aging and time-sharing aspects of executive control. Memory & Cognition, 30(4), 572–582.Strayer, D. L., & Drews, F. A. (2004). Profiles in driver distraction: Effects of cell phone conversations on younger and older drivers. Human Factors, 46,
640–649.Urbas, L., Schulze-Kissing, D., Leuchter, S., Dzaack, J., Kiefer, J., & Heinath, M. (2005). Programmbeschreibung D2-Drive-Aufmerksamkeitstest. Berlin: ZMMS.Verhaeghen, P., Steitz, D. W., Sliwinski, M. J., & Cerella, J. (2003). Aging and dual-task performance: A meta-analysis. Psychology and Aging, 18(3), 443–460.