Elsevier

Accident Analysis & Prevention

Volume 42, Issue 6, November 2010, Pages 1661-1671
Accident Analysis & Prevention

The effects of visibility conditions, traffic density, and navigational challenge on speed compensation and driving performance in older adults

https://doi.org/10.1016/j.aap.2010.04.005Get rights and content

Abstract

Research on how older drivers react to natural challenges in the driving environment is relevant for both the research on mental workload and that on age-related compensation. Older adults (M age = 70.8 years) were tested in a driving simulator to assess the impact of three driving challenges: a visibility challenge (clear day, fog), a traffic density challenge (low density, high density) and a navigational challenge (participants followed the road to arrive at their destination, participants had to use signs and landmarks). The three challenge manipulations induced different compensatory speed adjustments. This complicated interpretation of the other measures of driving performance. As a result, speed adjustment indices were calculated for each condition and participant and composite measures of performance were created to correct for speed compensation. (These speed adjustment indices correlated with vision test scores and subscales of the Useful Field of View®.) When the composite measures of driving performance were analyzed, visibility × density × navigational challenge interactions emerged for hazard RT and SD of lane position. Effects were synergistic: the impact of the interaction of challenge variables was greater than the sum of independent effects. The directions of the effects varied depending on the performance measure in question though. For hazard RT, the combined effects of high-density traffic and navigational challenge were more deleterious in good visibility conditions than in fog. For or SD of lane position, the opposite pattern emerged: combined effects of high-density traffic and navigational challenge were more deleterious in fog than in clear weather. This suggests different aspects of driving performance tap different resources.

Introduction

Older drivers are disproportionately at risk for certain types of collision (e.g., McGwin and Brown, 1999, Preusser et al., 1998) and are more likely to die or sustain serious injury as a result (Hauer, 1988, Skyving et al., 2009). The goal of this study was to investigate the impact of challenges inherent in different types of drive by manipulating these factors and measuring their effects, singly and in combination, using a driving simulator. This represents a different approach to the study of mental workload in older drivers insofar as it does not require secondary tasks that are not a natural part of driving. Information about the interactive effects of different types of challenge are informative in light of discussions about whether different challenges draw on the same resource or if there are different resources for different challenges. As well, given that the introduction of driving challenges may induce drivers to compensate by slowing down, and given that the degree of compensatory slowing may be related to individual capacities, this work may be relevant to theories of age-related compensatory behaviors (Baltes, 1997, Brouwer et al., 1988).

Age and age-related disorders are associated with a variety of changes in sensory, motor and cognitive function (e.g., Scialfa and Kline, 2007, Spirduso et al., 2005, Kramer and Kray, 2006, Park and Payer, 2006, Zacks and Hasher, 2006). Some of these changes result in performance deficits, though these deficits are more notable in some domains than others. In particular, there are a variety of studies that suggest that though an individual's overall store of knowledge (their crystallized intelligence) may be stable or even increase with age, performance tends to decline in tasks that require fluid intelligence: tasks that make extensive demands on executive working memory and attention (see Li et al., 2009 for a review). It has been suggested that age-related neural pathology and demyelination produce deficits in both processing speed and performance stability across time (“robustness”), and the effects are especially noticeable in complex perceptual motor tasks (Li et al., 2004, MacDonald et al., 2009). One such task is driving an automobile. Driving requires rapid response and the ability to carry out several activities at once, such as monitoring for hazards while steering and controlling the speed of the vehicle. Thus, even though many older drivers have decades of driving experience, they may begin to miss important safety-related information in complex, challenging driving environments (e.g., unfamiliar roads, information rich areas such as intersections, e.g. McGwin and Brown, 1999, Preusser et al., 1998). The importance of attention is borne out by the observation that attentional measures are the best predictors of driving performance in older drivers (e.g. Ball et al., 1993, Mathias and Lucas, 2009) and this has inspired research on the impact of attentional demands on performance in older drivers.

A large number of these studies reflect the impact of Shiffrin and Schneider's (1977) seminal paper on automatic and controlled processing. Shiffrin and Schneider proposed that although there are many types of mental process (perceptual, motor, memory), all fall into one of two categories. Automatic processes occur without awareness or intent and they can be carried out concurrently with other processes without compromising performance. In contrast, controlled processes occur with awareness, and are deliberate and goal oriented. These processes are effortful and slow and it is difficult to carry out several controlled processes at once. They are said to be attention-demanding; they produce “cognitive load”. When two controlled processes are carried out at the same time, there is interference, which is to say, performance is worse on one or both tasks when the tasks are performed together than when they are performed separately because they share a common limited capacity resource.

Perhaps because of the influence of this theory, although there are many techniques for measuring workload (Verwey and Veltman, 1996), the most common way to investigate age-differences in attention is to use the dual task paradigm. Often these studies look at how adding secondary tasks interferes with driving, and there are a variety of such secondary tasks, including carrying out mental arithmetic, scanning in-vehicle displays or the immediate environment for certain probe stimuli, answering questions or engaging in conversations (e.g., Cantin et al., 2009, Pohlmann and Traenkle, 1994, Shinar et al., 2005, Verwey, 2000, Zeitlan, 1995). Some of these studies also manipulate the complexity of the drive insofar as they compare rural and urban driving (e.g. Cantin et al., 2009), low and high-density traffic (Cnossen et al., 2004, Verwey, 2000), familiar versus unfamiliar settings (e.g. Verwey, 2000), etc. Nonetheless, the focus is always on the dual task manipulation: the amount of interference produced by carrying out two tasks once. Effect magnitudes vary, but most studies show more dual task interference in older than younger drivers and these results have been interpreted as evidence that older adults have fewer resources for controlled processing (see also Riby et al., 2004). At present, most of the debate centres around whether there is a single limited capacity resource or different resources for different task modalities (e.g. Resource theory: Kahneman, 1973; Multiple Resource theory: Wickens, 2002 respectively).

The dual task paradigm has been and will continue to be very useful. However, it has drawbacks that may, at times, limit the validity and generality of the findings. For one, when a novel secondary task is introduced, participants first have to learn how to carry out the secondary task before they can combine it with driving. Older adults may take longer to learn how to perform the secondary task and as a result, when their performance suffers it may be because they have not had enough time to learn the secondary task whereas the younger adults have. Furthermore, even if the secondary task per se is not novel, the combination of tasks may be. Many dual task studies require putting together tasks that typically do not go together, at least for older drivers. For example, a number of studies use mental arithmetic as the chosen secondary task. This task generates substantial amounts of interference (e.g., Makishita and Matsunaga, 2008, Shinar et al., 2005) but most people do not carry out mathematical calculations while driving. Consequently, when older drivers show more interference between tasks, it is unclear whether it is because older drivers have fewer resources for controlled processing or because older drivers have special difficulties with peculiar combinations of tasks (see McDowd and Craik, 1988 for a discussion). In addition, whenever dual task studies are carried out, there is a danger that individuals vary in the emphasis they give to one task as opposed to the other. Cnossen et al. (2004) found that when the secondary task was not intrinsic to driving, drivers put less emphasis on it. This may be especially true for older adults, who may compensate for age-related deficits by de-emphasizing the non-driving related secondary tasks in the interests of safety (see Baltes, 1997; also Li et al., 2001). This underlines the importance of measuring both driving and secondary task performance in dual task studies that test older drivers. If the focus is solely on driving performance, it is possible that the study will underestimate the true costs of multi-tasking.

In this study we tried a more naturalistic approach, exploiting the fact that the driving task involves many different types of load that may interact in interesting ways. We explored three types of driving challenge: visibility challenges, traffic density challenges, navigational challenges. These challenges were chosen to cover a spectrum ranging from challenges that were more sensory (the visibility manipulation) to those that were more attentional and memorial (the traffic density and wayfinding manipulations).

The visibility manipulation involved comparing driving in (simulated) fog with driving in (simulated) clear weather. Collision risk increases in fog and even professional drivers find driving in fog stressful (Vivoli et al., 1993). Fog reduces image contrast. This impairs distance perception, which can prompt rear end collisions (Broughton et al., 2007, Buchner et al., 2006, Cavallo et al., 2001). It also causes drivers to underestimate how rapidly other vehicles are traveling (Horswill and Plooy, 2008, Snowden et al., 1998). Furthermore, because objects have to be closer to become fully visible, fog reduces the amount of time drivers have to react to stimuli in the environment. Given that age is associated with reductions in contrast sensitivity (Scialfa and Kline, 2007) and increases in response time even in well-practiced tasks (Voelcker-Rehage and Alberts, 2007), fog should be especially problematic for older drivers. Nonetheless, there is little research on how older drivers perform when they drive in fog.

Traffic density is also a factor in collision risk and increased traffic affects performance even in professional drivers (Hanowski et al., 2009). For some drivers, high-density traffic provokes anger and aggression (Parker et al., 2002), though this may be partly because it increases physiological arousal, an effect that can be beneficial for those with a tendency to doze off while driving (Tassi et al., 2008). In this study, the amount of oncoming traffic was manipulated. Although the oncoming traffic did not venture into the driver's path it did serve as a source of dynamic visual clutter (see Horberry et al., 2006). Clutter increases demands on selective attention: the ability to select relevant items from irrelevant ones in the visual scene. Consequently, it interferes with the perception of signs and hazards. There is evidence that visual clutter is especially problematic for older drivers (McPhee et al., 2004, Horberry et al., 2006).

The third challenge variable was wayfinding – the ability to navigate while driving. Compared to younger drivers, older drivers have more difficulties in finding their way to specific destinations and this can be a source of stress and embarrassment (Burns, 1998). At present, most of the wayfinding literature focuses on whether using an in-vehicle navigation system interferes with driving (Arbesman and Pellerito, 2008, Ma and Kaber, 2007). However, even without in-vehicle technology, wayfinding constitutes a secondary task that may compromise driving performance. This may be particularly true when wayfinding involves using a map (Cnossen et al., 2004, Lee and Cheng, 2008) but a recent study suggests that even talking about spatial navigation is enough to impair distance estimation (Patrick and Elias, 2009). In the present study, there were no in-vehicle navigation systems, maps, or spatial conversations. Instead, the wayfinding task required drivers to follow directions that they had committed to memory at the beginning of the trip. Even this simple wayfinding task requires multiple action monitoring (which puts demands on executive working memory) and divided attention insofar as it requires coordinating different tasks: driving, holding directions in working memory and using the directions at the appropriate time. It also requires selective attention insofar as following the directions would require drivers to search for relevant signs and landmarks in visual clutter. Consequently the prediction was that wayfinding would interfere with driving.

A sample of healthy, active older adults was tested in order to assess the effects of these challenge variables alone and in combination. In particular, the interest was in the combined effects of these variables because it is relevant to the question of whether different driving challenges all tax a common resource, or whether these challenges tax different resources. Thus, the interactive effects of these variables were assessed using common measures of driving performance: collisions, hazard RT, and standard deviation of lane position. As well, in conditions where drivers were required to use signs and landmarks to find their way, wayfinding errors (missed or extra turns) were measured. There has never been a study looking at the combined effects of visibility, traffic density, and navigational challenge, but there is a theory that might predict interactive effects: Baldwin's sensory–cognitive interaction theory (2002). According to this theory, cognitive resources are used to order to compensate for low quality sensory information. Thus, there is reason to expect that sensory challenge (such as fog) would serve to exaggerate the effects of other cognitive challenges, including the traffic density challenge (visual clutter puts demands on selective attention), and the navigational challenge, which requires multi-tasking (driving while keeping information in working memory and searching for signs and landmarks).

Although there was reason to expect changes in performance as a function of the different types of driving challenge, there was also reason to expect that older drivers would reduce their driving speed to try to alleviate the effects of the challenges. An important component of tactical behavior is the ability to modify speed to match the conditions on the road (Summala, 1996). Drivers change their speed in response to the perceived risk of the drive (Fuller et al., 2008), increasing speed when they feel safe (e.g., Assum et al., 1999, Stanton and Pinto, 2000), or reducing it when they perceive increased risk, as sometimes occurs when people drive while using cellular phones (e.g. Haigney et al., 2000). It has been suggested that in general, older drivers adopt slower speeds to compensate for age-related increases in response time (Chu, 1994). If older drivers decrease their driving speed, and this adjustment is in fact successful in reducing these risks, it is possible that simple measures of driving performance may underestimate the true impacts of the challenge variables.

The presence of compensatory adjustments in speed complicates the interpretation of the other measures of driving performance insofar as the compensatory adjustments may effectively negate the impact of challenge variables. There are a number of approaches to this problem. One is to require drivers to maintain a constant speed regardless of challenge condition. However, this might contribute to stress in older drivers, many of whom were very concerned about their performance. Moreover, forcing drivers to go faster than they normally would (automatically) might constitute another secondary task that requires additional attentional resources. Another approach is to allow the older drivers to change speed as they see fit, and then measure the differences in speed and use it to create individualized indices of speed adjustment for each challenge. These individualized indices of adjustment could then be used to create composite measures of driving performance that factor in both speed and driving performance for each driver. This approach (the creation of composite measures) has been used to deal with speed-accuracy tradeoffs in other tasks (e.g., Akhtar and Enns, 1989).

These indices of speed adjustment were interesting in their own right. The ability to adjust speed appropriately in the face of difficult driving conditions is an important component of driver competence (de Craen et al., 2008) and there are concerns that deficits in cognitive status may reduce the amount of compensatory adjustment (Lundqvist and Alinder, 2007). In contrast, in community samples where there is no reason to expect marked deficits in cognitive status, it is possible that drivers with reduced sensory or attentional function may compensate more than other drivers if they have a reasonably good understanding of their own strengths and weaknesses. For example, drivers with deficits in contrast sensitivity may reduce their speed more for fog; drivers with difficulties in divided attention may reduce their speeds more in situations where they have to perform two tasks at once (wayfinding while driving). This pattern of results would be predicted by Baltes’ (1997) Selection, Optimization, Compensation theory of lifespan development, which suggests in the face of age-related deficits older adults optimize their performance, compensating for their losses (loss-based compensation), in this case, by adjusting their driving speed. Thus, although this study was primarily designed to investigate average differences in performance across challenge conditions, exploratory analyses were carried out to investigate whether individual differences in sensory or attentional function (as measured by common tests of acuity and attention) predict speed compensation.

Section snippets

Apparatus and driving scenarios

A DriveSafety DS-600c simulator was used for testing: a 4-door sedan surrounded by 5 viewing screens to provide a 250° wrap-around virtual driving environment (5–50° screens enclosing the front and sides of the vehicle). Each simulated drive involved traversing a two-lane road through the country, past hills, trees, farms, and large buildings of various sorts (schools, fire halls, service stations). The speed limit was 80 kph, with speed postings appearing every 200 m. Eight different drives were

Results

The analyses involved three stages. First the raw data were analyzed. Repeated measures factorial analyses of variance were performed. Driving speed, collisions, hazard RT and SD of lane position were measured as a function of visibility (clear day, fog), traffic density (low density, high), and navigational challenge (no wayfinding, wayfinding). Navigational errors were analyzed as a function of visibility, traffic density, and type of turn cue (landmark, sign). Partial Eta squared statistics (

Discussion

This study makes a number of contributions. It was the first to use a driving simulator to measure the interactive effects of visibility, traffic density, and navigational challenge on older drivers. Information about how healthy older drivers cope with various challenges alone and in combination is useful when trying to assess the impact of disorders such as Alzheimer's and Parkinson's disease. This study has practical implications for those evaluating in-vehicle navigation devices insofar as

Conclusions

This study has several implications. First, it provides support for idea that there are interactions between factors that are more sensory in nature (such as the low contrast image conditions produced by fog) and those that involve higher order factors such as selective attention (necessary to deal with clutter produced by traffic) and executive working memory (necessary for carrying directions in memory and using them at the appropriate time). Nonetheless, the results reveal complexities that

Acknowledgements

The Ontario Neurotrauma Foundation, Canadian Foundation for Innovation, Ontario Innovation Trust, and Auto21: Network Centres of Excellence funded this research. Lauren Meegan helped with testing participants and Robert Ramkhalawansingh looked over an earlier draft of this article. Some of the data from this study were presented at the 5th International Driving Symposium on Human Factors in Driver Assessment, Training, and Vehicle Design, Big Sky, Montana, June 22–25, 2009. We would also like

References (65)

  • D. Haigney et al.

    Concurrent mobile (cellular) phone use and driving performance: task demand characteristics and compensatory processes

    Transportation Research: Part F

    (2000)
  • R. Hanowski et al.

    Evaluating the 2003 revised hours-of-service regulations for truck drivers: the impact of time-on-task on critical incident risk

    Accident Analysis and Prevention

    (2009)
  • T. Horberry et al.

    Driver distraction: the effects of concurrent in-vehicle tasks, road environment complexity and age on driving performance

    Accident Analysis and Prevention

    (2006)
  • W.C. Lee et al.

    Effects of using a portable navigation system and paper map in real driving

    Accident Analysis and Prevention

    (2008)
  • H. Makishita et al.

    Differences of drivers’ reaction times according to age and mental workload

    Accident Analysis and Prevention

    (2008)
  • D. Parker et al.

    Anger and aggression among drivers in three European countries

    Accident Analysis and Prevention

    (2002)
  • R. Patrick et al.

    Navigational conversation impairs concurrent distance judgments

    Accident Analysis and Prevention

    (2009)
  • D. Preusser et al.

    Fatal crash risk for older drivers at intersections

    Accident Analysis and Prevention

    (1998)
  • J.G. Reed Jones et al.

    Testing assumptions implicit in the use of the 15-second rule as an early predictor of whether an in-vehicle device produces unacceptable distraction

    Accident Analysis and Prevention

    (2008)
  • D. Shinar et al.

    Effects of practice, age, and task demands, on interference from a phone task while driving

    Accident Analysis and Prevention

    (2005)
  • M. Skyving et al.

    Older drivers’ involvement in fatal RTCs. Do crashes fatal to them differ from crashes involving them but fatal to others?

    Safety Science

    (2009)
  • H. Summala

    Accident risk and driver behavior

    Safety Science

    (1996)
  • P. Tassi et al.

    Are OSAS patients impaired in their driving ability on a circuit with medium traffic density?

    Accident Analysis and Prevention

    (2008)
  • M. Arbesman et al.

    Evidence-based perspective on the effect of automobile-related modifications on the driving ability, performance, and safety of older adults

    American Journal of Occupational Therapy

    (2008)
  • C.L. Baldwin

    Designing in-vehicle technologies for older drivers: applications of sensory–cognitive interaction theory

    Theoretical Issues in Ergonomic Science

    (2002)
  • K. Ball et al.

    Visual attention problems as a predictor of vehicle crashes in older drivers

    Investigative Opthalmology and Vision Science

    (1993)
  • P.B. Baltes

    On the incomplete architecture of human ontogeny: Selection, optimization and compensation as foundation of developmental theory

    American Psychologist

    (1997)
  • Brouwer, W., Rothengatter, T. van Wolffelaar, P., 1988. Compensatory potential in elderly drivers. In: Rothengatter,...
  • A. Buchner et al.

    Car backlight position and fog density bias observer-car distance estimates

    Human Factors

    (2006)
  • V. Cavallo et al.

    Distance perception of vehicle rear lights in fog

    Human Factors

    (2001)
  • Chu, X., 1994. The effect of age on driving habits of the elderly: evidence from the 1990 national personal...
  • F. Cnossen et al.

    Adaptive strategy changes as a function of task demands: a study of car drivers

    Ergonomics

    (2004)
  • Cited by (60)

    • Back-propagation neural network model to predict visibility at a road link-level

      2020, Transportation Research Interdisciplinary Perspectives
    • Tactical and strategic driving behaviour in older drivers: The importance of readiness to change

      2020, Accident Analysis and Prevention
      Citation Excerpt :

      The findings and implications of the results for the application of the TM of behaviour change (Prochaska and DiClemente, 1982) to driving self-regulation by older drivers will be discussed. The measure of tactical behaviours, drawn from prior research on older drivers’ tactical driving behaviour (Charlton et al., 2013; LeBlanc et al., 2013; Trick et al., 2010), was best represented as a five-item scale. Two items, drive slower than the speed limit and avoid in car distractions, were unrelated to other items on the scale.

    • Safety of professional drivers in an ageing society – A driving simulator study

      2019, Transportation Research Part F: Traffic Psychology and Behaviour
      Citation Excerpt :

      For non-professional driver, degradation in steering performance by age is found remarkable under the high traffic flow condition. Such findings in line with that of previous studies (Cantin et al., 2009; Trick, Toxopeus, & Wilson, 2010). In particular, older drivers often consider driving under the high traffic flow condition as a challenging task (with higher mental workload).

    View all citing articles on Scopus
    View full text