PROCEEDINGS of the HUMAN FACTORS AND ERGONOMICS SOCIETY 37th ANNUAL MEETING-1993 BRAKE PERCEPTION-REACTION TIMES OF OLDER AND YOUNGER DRIVERS Neil D. Lerner COMSIS Corporation Silver Spring, MD The time drivers require to react in braking situations underlies many practices in highway design and operations. There is concern whether the perception-reaction time (PRT) values used in current practice adequately meet the requirements of many older drivers. This study compared on-the-road brake PRTs for unsuspecting drivers in three age groups: 20-40, 65-69, and 70-plus years old. The method included features to enhance the ecological validity of the observed reactions: subjects drove their own vehicles in their normal manner; driving was on actual roadways; extended preliminary driving put the driver at ease and without expectationof unusual events at the time of the braking incident; the incident occurred at a location lacking features that might enhance alertness (e.g., curves, crests, driveways). Subjects drove an extended route, under the guise that they were making periodic judgments about "road quality." At one point, a large crash barrel was remotely released from behind brush on a berm and rolled toward the driver's path. Although most of the fastest observed PRTs were from the young group, there were no differences in central tendency (mean = 1.5 s) or upper percentile values (85th percentile = 1.9 s) among the age groups. Furthermore, the current highway design value of 2.5 seconds for brake PRT appears adequate to cover the full range of drivers. values are only loosely linked to realistic driving data, to the extent they are empirically based, the underlying research has not appropriatelyconsidered elderlydrivers. The general slowing of responses with age is a broad and well establishedlaboratory finding (e.g., Salthouse, 1985), including studies showing slower foot pedal reaction times for older subjects(e.g., West VirginiaUniversity, 1988). To the extent that older drivers require longer PRTs than those established for highway design pmtice, they may be at a seriousdisadvantagefor numerous driving situations. For safety reasons, it is important that design and operational practica meet the needs of all segments of the drivingpopulation. However, increasing the assumed PRT value for highway design would have extensive ramifications, and substantial costs. INTRODUCTION Driver perception-reaction time (PRT) is a fundamental concept in highway design and safety. PRT refers to the time required to perceive, interpret, decide, and initiate a response to some stimulus. Different driving situations are characterized by different assumed values of PRT. For example, the PRT associated with reacting to an obstacle in the vehicle's path (brake reaction time) is different from the PRT associated with determining that it is safe to begin moving through an intersection, or with the time required to judge whether it is safe to initiate a passing maneuver. One of the most basic concepts for highway design and operations is that of stopping sight distance (SSD). SSD is the distance traveled before coming to a stop when a driver, travelling at roadway design speed, must brake in reaction to an unexpected obstacle in his or her path. Driver PRT is a key determinant of SSD, so that the specific PRT value assumed for design purposes influences many aspects of the roadway. Currently, the assumed value for design equations is 2.5 s (AASHTO, 1990). This parameter ultimately influences such roadway features as horizontal curve radius, vertical curves (crests), approaches to intersections, sign placement, traffic signal visibility and phasing, and other common roadway aspects. Although PRT generally slows with age, there are some reasons for presuming that current sight distance criteria may be adequate for older drivers and do not require change. Lerner (1991) has discussed a number of these. Among the more interesting is the finding that the few studies that have measured on-the-road PRTs for unsuspecting drivers of different ages have not obtained meaningful differences between age groups (Olson, Cleveland, Fancher, Kostyniuk, and Schneider, 1984; Korteling, 1990; Hostetter, McGee, Crowley, Sequin, and Dauber, 1986). Furthermore, even if older drivers require somewhatlonger response times, the "cushion"built into current design parameters may be adequateto cover their behavior. Recently, there has been a good deal of concern about whether the assumed PRT values in design equations are adequate to meet the requirements of older drivers (Lerner, 1991). Although assumed The purpose of the present study, funded by the Federal Highway Administmtion,was to measure realistic, 206 Downloaded from pro.sagepub.com at PENNSYLVANIA STATE UNIV on May 10, 2016 PROCEEDINGS of the HUMAN FACTORS A N D ERGONOMICS SOCIETY 37th ANNUAL MEETING-1993 instructedthat we had permission to continueon this road and drove around the barricades, and were told the appropriate speed was 40 mph. When the vehicle reached a location near the midpoint of the roadway section, a large yellow crash barrel, hidden on a berm behind some brush, was remotely released and suddenly became visible r o h g toward the roadway. Although it appeared to be rolling directly into the road, a set of chains held the barrel to the shoulder area. The barrel emerged into view approximately 200 feet in front of the vehicle; this provided a time-to-collision of about 3.4 s at the target speed of 40 mph. on-the-road braking PRTs for unsuspecting older and younger drivers, and to determine whether the currently assumed design value of 2.5 s is adequate for drivers of all ages. The study was designed to provide ecologically meaningful data, avoiding some of the limitations of past research. Additional detail on methods and analysis can be found in a project final report to the Federal Highway Administration (Lerner, Huey, McGee, and Sullivan, in preparation). METHOD The purpose of the study was to elicit and measure realistic brake reaction times from drivers. Given this, several methodological features were deemed important. The subjects should be driving on actual roadways. They should be driving normally and at ease. They should have no expectation of an emergency braking event and should not be leery of the motives of the experimenter. They should be driving their own vehicles, rather than trying to adapt to an experimenter-provided vehicle. This was of particular concern for older drivers, who may have selected or adapted their vehicles to meet their needs, and who may not adapt to a test vehicle as quickly. The data were recorded using a video-based data collection system. For subjects with the first procedure, the emergenceof the barrel was recorded by an in-vehicle camera, and the occurrence of the brake response by activation of a pressuresensitivetape switch attached to the brake pedal. In the other procedure, a hidden roadside microcamemrecorded both the emergence of the h l and the activationof the vehicle's brake lamps. The PRT was the interval between the emergence of the barrel and the initiation of braking. All sessions were run in daytimeand clear weather. Subjects were recruited in three age groups: 20-40 years old, 65-69 years old, and 70 or older. To minimize the selection bias toward more capable elderly, older subjects were recruited in the greater Washington, DC area, working through senior centers, churches, retirement communities,and so forth. Rather than placing initiative on the subject to volunteer, as much as possible we worked with directors of the institutions to help identify and approach individualswith a wide range of capabiities, and to provide social support and incentive for taking part. Although there can be no claim that the sample was representative, and while it is likely that those at the extreme lowest limits of ability and confidencetended to exclude themselves, the older group did appear to provide a broadly suitable range, and certainly included many individuals who would have been unlikely to participate without more active recruiting strategies. In order to meet these methodological requirements, subjects drove their own vehicles over an extended route, under the guise that they were participating in a study that was recording their judgments of "road quality." Periodically (at stop signs), they made judgments about the quality of the road sections they had just travelled. Two slightly different procedures were involved. For some subjects, the initial part of the drive was incorporated with another experiment, and the subject had been driving about an hour prior to encountering the braking event. For the other subjects, the prior part of the drive was briefer, covering about three miles. (Brake reaction times for the two groups were not significantly different, t=0.08, and the datasets were combined for analysis). It was only at the completion of the route that the subjects encountered the site of the braking event; from their perspectives, they were simply continuing the ride, and the procedure, that had been in effect all along. They were driving in a relatively relaxed, normal manner, with no expectation of any unusual event (post-session debriefings confirmed this). The subjects turned onto a four lane divided highway. This highway provides access to an interstate highway, and then continues on for 0.7 miles beyond the freeway entrance. This extended stub of roadway is a functional, fully delineated roadway, but is closed to normal traffic by the use of barricades. Subjects were Although over 200 subjects participated in the study, there was a very high rate of data loss due to a combination of factors, includingequipmentfailures,video problems, weather, experimental error, inappropriate subject behavior, unauthorized traffic at the site, and so forth. Valid trials and records were obtained for 116 subjects; this included 30 in the 20-40 year old group, and 43 in each of the two older groups. 207 Downloaded from pro.sagepub.com at PENNSYLVANIA STATE UNIV on May 10, 2016 PROCEEDINGS of the HUMAN FACTORS A N D ERGONOMICS SOCIETY 37th ANNUAL MEETING-1993 automatic transmissions, while only about two-thirdsof the younger group did. However, post hoc comparisons indicatedthat transmission typehad no discernibleeffect on either the nature of the reaction to the barrel (brake, steer) or the speed of braking. RESULTS Nearly all (87%) of the 116 drivers made some overt vehicle maneuver in response to the emergence of the barrel. Of these, about 43% both steered and braked, 36% steered only, and 8% braked only. Thus, just over half of the drivers (51%) reacted by braking. This is consistent with various other on-the-road studies, which have found steering to be a more reliable reaction, and with the percentage of braking drivers varying from about 30% to about 80% (e.g., Triggs and Harris, 1982). Measurable brake reaction times were obtained for 56 subjects; this included 14 of the 20-40 year olds, 18 of the 65-69 year olds, and 24 of those 70 or older. DISCUSSION This study provided uniquedata on the brake PRTs of drivers of different ages. It used real drivers, of known ages, in their own vehicles, driving on actual roadways, under conditions where they were not expecting any unusual (emergency braking) event. None of the previous research has met all of these criteria, which are important for deriving "absolute" measures of brake reaction time that are ecologically valid. The mean brake PRT for all subjects was 1.5 s, with a standard deviation of 0.4 s. An analysis of variance revealed no significant main effect of age or gender, with the interaction of age with gender of borderline significance (p =0.055). The interaction reflects the particularly short mean reaction time of the young female group (1.22 s), while other age-bygender group means ranged from 1.40 to 1.65 s. The age groups not only showed little difference in central tendency, but also in terms of upper percentile values. The 85th percentile PRT (a level often used in developing highway design values) was about 1.9 s for all age groups. Virtually all responses were captured by the 2.5 s design value; the longest observed time was 2.54 s and the next longest was 2.39 s. consistent with some earlier on-the-d studia, there was no indication of meaningfullyslower braking by older p u p s , as measured by central tendency (mean, median) or upper percentile values (e.g., 85th percentile). Young drivers responded quickly more often, but also showed a higher proportion of slow reaction times. There may be a variety of explanationsfor this. Furthermore, this study only measured the time to initiate braking, and did not measure the braking profile (decelerationrate) or degree of driver control. The age groups could differ in these regards. However, based strictly on PRT, older groups were not slower to react. The original research plan anticipated sufficient data so that reliable frequency distributions could be obtained for each age group. However, due to data loss and the fact that only about half the drivers in such experiments respond by braking, actual brake reaction times were measured for only 56 of the original participants. Therefore, a great deal of precision should not be attached to the observed values. Nonetheless, this sample is sufficient to indicate that there is no important difference in the response times of the various age groups. However, the absence of differences in central tendency or 85th percentile values does not necessarily imply that older and younger groups were responding in the same manner. Most of the fast reaction times (e.g., < 1.25 s) were provided by the young drivers. However, the distribution of reaction times for young drivers was bi-modal. Most were less than 1.45 s, but a few were quite long (over 1.9 s). In contrast, for the older groups, about half of the cases fell in the roughly half-second interval between these values. Given the relatively small number of observations for the young age group, the differences between these distributions must be viewed with caution; a chi square test approached, but did not reach, conventional statistical significance levels (Chi square = 11.2, 6 df; 0.05 <p<O. 10). The 2.5 s value used for PRT in highway design applicationsappears to provide adequatecoveragefor the full range of driver age. The longest response time recorded was at this value. The mean was a full second faster, with 2.5 s being more than two standard deviations slower, and the estimate of the 85th percentilebeing more than a half-second faster than this value. The findings of this research are consistent with a number of other on-road studiesthat have observed brake or steeringresponseupper percentile (e.g., 85th percentile) times of less than 2 s (e.g., Olson et al., 1984;Triggs and Harris, 1982;Sivak, Olson, and Farmer, 1982; Sivak, Post, Olson, and Donohue, 1981; Allen Corporation, 1978; Summala, 1981). There have been some studies that have observed Because subjects drove their own vehicles, it is possible that differences (or the absence of differences) between age groups could be attributableto differences in the vehicles they drive. The major difference noted was that almost all of the older driver's vehicles had 208 Downloaded from pro.sagepub.com at PENNSYLVANIA STATE UNIV on May 10, 2016 PROCEEDINGS of the HUMAN FACTORS A N D ERGONOMICS SOCIETY 37th ANNUAL MEETING-1993 longer brake response times, but these have not been for situationsappropriate to SSD. For example, longer times may be observed in response to stimuli such as highway signs or changes in signal phase, or where there are more difficult maneuver requirements related to complex roadway geometries (treated as "decision sight distance" situations in traffic engineering terms). Some reviewers have failed to discriminate these different situations. There have also been calls to increase the PRT design value beyond 2.5 s in order to compensate for the effects of factors such as age, fatigue, or impairment. Based on typical research findings to date, the 2.5 s value already provides some design cushion (of over 0.5 s); it is not clear what the empirical basis for extending the design value further is. While any lengthening of the design value might provide some additional marginal increment in protection, it would necessitate ubiquitous changes, with substantial costs involved, throughout the highway system. Thus the safety benefits of increasing the assumed PRT for SSD beyond 2.5 s should be demonstrably significant. The absence of substantially slower brake PRTs among older groups provides an illustration of how the factorsof expertiseand compensation in complex skills can maintain performance even in the face of reduced capabilities. Virtually all of the componentpsychomotor processesthat underlie PRT -- informationpromsing rate, visual search time, responseinitiation, movement time, etc. - have been shown to slow with age, in laboratory studies. Age-relatedcompensation is poorly understood for driving, as it is for various other skilled tasks. Furthermore, the mechanisms involved in compensation for one aspect of performance might be related to degradation in other aspectsof performance (e.g., braking may be more "all-ornone," providing a greater risk of rear end collisions or loss of vehiclecontrol). Onepossible explanation for the absence of a differencebetween age groups is that older drivers might be responding in a more reflexive, stereotyped manner. Younger drivers may be prone to do more evaluation before responding, or respond in a more gradual or controlled manner, using their faster information processing capabilitiesto refine the response, rather than quicken it. It was the subjective opinion of the primary research assistant who accompanied the drivers that the older drivers tended to make more evident and dramatic foot movementswhile braking, although he did not notice a subjective sense of more severe deceleration. This observation is also consistentwith the data of Olson et al. (1984), whose instrumentationallowed the total PRT to be segmented into a "perception time" (from first target visibility to release of the accelerator) and a "response time" (from release of the accelerator to stepping on the brake pedal). Although there was little differencebetween age groups in the total PRT, the older group actually had faster "responsetimes" (estimatingfrom figures, about 0.1 s faster at the 50th percentile and about 0.2 s faster at the 90th percentile). Thus based on our observationsand the Olson et al. findings, it may be that in a situation where there is a surprise need for possiblebraking, older drivers are more consistent in making a rapid move to the brake pedal, once the hazard has been recognized. The response may be more stereotyped, and subject to less evaluation and modulation. Whatever the reason for the absence of observed differences in overall brake PRT between age groups for SSD situations, it is none the less clear that most older drivers can continue to react with appropriate swiftness, even to an unanticipated braking event. It should be noted, however, that where the stimulusevents and required driving maneuvers are more complex and ambiguous than emergency braking, there might bemore deleterious effects of age. There are some reasons why the present procedure may arguably have led to relatively conservative estimates of PRT. An analysis of the brake reaction times of drivers who also showed substantial steering in reaction to the barrel, compared to those that braked only, showed that those who steered had somewhat slower brake PRTs (about a quarter-second slower). Thus by providing an opportunity for avoidance steering, as well as braking, the estimate of brake reaction time may be somewhat longer than for a high-emergency situation in which braking is the only clear alternative. Furthermore, the "hazard" emerged from a hidden location off the side of the road, rather than being in more central view. Previous studies may also have used situations in which there was greater expectancy of the potential need for braking, either because of a geometric feature (coming over the crest of a hill, where sight distance was obscured) or because of the need to monitor leading traffic (where the stimulus to brake was the illumination of the brake lamps of a vehicle immediately ahead), In the present experiment, subjects were simply driving on a straight section of roadway, so that the occurrence of any conflict was quite unexpected. Whatever the reason, this study observed mean times that were somewhat slower (by about 0.15 to 0.35 s) than those reported in other onroad studies, such as those cited above. Despite this, the 2.5 s PRT design value covered the range of observed brake times. 209 Downloaded from pro.sagepub.com at PENNSYLVANIA STATE UNIV on May 10, 2016 PROCEEDINGS of the HUMAN FACTORS AND ERGONOMICS SOCIETY 37th ANNUAL MEETING-1993 REFERENCES AASHTO (1990). A Policy on Geometric Design of Highways and Streets. Washington, DC: American Association of State Highway and Transportation Officials. Olson, P., Cleveland, D., Fancher, P., Kostyniuk, L., and Schneider,L. (1984). Parameters A$ectig Stopping Sight Distance. NCHW Report 270. Washington, DC: Transportation Research Board. Allen Corporation (1978). Field Validation of Tail Lights -- Repon on Phase I. Report under contract DOT-HS-7-01756. Washington, DC: National Highway Traffic Safety Administration. Salthouse, T. (1985). Speed of behavior and its implication for cognition. In J. Birren and K. Schaie (Ed.s), Handbook of the Psychohgy ofAging. New York: Van Nostrand Reinhold Co. Hostetter, R., McGee, H., Crowley, K., Sequin, E., and Dauber, G. (1986). Improved Perception-Reaction Time Znfomtation for Intersection S i g h Distance. FHWA/RD-87/015. Washington, DC: Federal Highway Administration. Sivak, M., Olson, P., and Farmer, K. (1982). Radarmeasured reaction times of unalerted drivers to brake signals. Perceptual and Motor Skills, 55, 594. Sivak, M., Post, D., Olson, P., and Donohue, J. (1981). Driver responses to high-mounted brake lights in actual traffic. Human Factors, 23, 231-236. Korteling, J. (1990). Perception-response speed and driving capabilitiesof brain-damagedand older drivers. Human Factors, 32, 95-108. Summala, H. (1981). Driverhehicle steering response latencies. Human Factors, 23, 683-692. Lemer, N. (1991). Older driver perception-reaction time and sight distance design criteria. Z T E I991 Compendium of Technical Papers. Washington, DC: Institute of Transportation Engineers. Triggs, T. and Harris, W. (1982). Reaction Times of Drivers to Road Stimuli. Melbourne, Australia: Monash University, Department of Psychology. West Virginia University (1988). Physical Fimess and the Aging Driver, Phase I. Washington, DC:AAA Foundation for Traffic Safety. Lerner, N., Huey, R., McGee, H., and Sullivan, A. (in preparation). Older Driver Perception-Reaction Time For Intersection Sight Distance and Object Detection. Final report on Contract DTFH61-90-C00038. Washington, DC: Federal Highway Administration. 210 Downloaded from pro.sagepub.com at PENNSYLVANIA STATE UNIV on May 10, 2016
© Copyright 2026 Paperzz