From Proceedings of the Human Factors and Ergonomics Society 42nd Annual Meeting, October 5-9, 1998, Chicago, Illinois.
THE SAFETY OF ELDERLY PEDESTRIANS AT FIVE URBAN INTERSECTIONS IN MIAMI
Jose H. Guerrier and Sylvan C. Jolibois, Jr.
ABSTRACT
A study of five intersections was conducted in the city of Miami Beach to identify the problems encountered by young, middle-aged, and old pedestrians. Data were collected using questionnaires and videotapes. Both the survey and the videotape data support pedestrian difficulties at the intersections studied. Highlights of these findings are as follows: Pedestrians generally found the time available for pedestrians to cross the street too short. The data show that for the 15~ percentile older pedestrian three of the four signalized intersections studied provided insufficient crossing time. A major cause of the difficulties of older pedestrians can be attributed to physical limitations. There was a substantive number of pedestrian-vehicle conflicts and in the majority of these conflicts the pedestrians were at fault. The results of this study underscore the need for countermeasures that include engineering design considerations and educational campaigns for drivers and pedestrians.
INTRODUCTION
The elderly, (i.e., 65+) represent 13% of the U.S. population; yet, they make up 23% of pedestrian casualties (Hauer, 1988). Furthermore, when involved in collisions, elderly pedestrians are more likely to be injured and to suffer serious and disabling injuries compared to younger pedestrians (Mackay, 1988). Aging has been associated with deficits in sensory modalities such as vision and hearing (Brock, Guralnik, and Brody, 1990). For instance, older persons require 3 to 10 times more illumination than younger ones to perceive a similar target (Small, 1987; Fozard, 1981). In addition, slowing in information processing ability (McDowd and Birren, 1990), and motor control (Gottsdanker, 1982; Larish & Stelmach 1982) has been reported. These may impact on various abilities that are related to street crossing behavior. The latter may include scanning for and perceiving pedestrian signals and moving vehicles and walking speed. For instance, current walking speed used for the design of intersections is 4ft/sec. Hauer (1988) reports the results of an extensive survey of elderly pedestrians in Florida by the Institute of Transportation Engineers (ITE) which found that a walking speed of 2.5 ft/sec provided adequate crossing time for about 87% of these older pedestrians. Nevertheless, the ITE recommends walking speed standards of 3.0 to 3.25 ft/sec as more appropriate for street crossing (ibid), a recommendation that is also made by Knoblauch, Pietrucha, and Nitzburg (1996). Zegeer (1985) recommends that in areas with large concentration of elderly pedestrians, walking speeds of 2.5ft/sec or less be considered in allocating adequate crossing time. It is evident that while the need for designing intersections that consider the older pedestrian is clear, guidelines as to what should be considered appropriate vary.
Knoblauch et al. (1996) have pointed out that human performance parameters used in setting walking speed criteria have often been arbitrarily set at the 15th to 85th percentile. However, they recommend that the performance characteristics of the I 5th percentile older pedestrian be adopted as the standard. Such an approach would greatly enhance the safety of elderly pedestrians.
40% of accidents involving pedestrians occur at intersections. About 50% of the fatalities and 33% of the injuries occur at an uncontrolled intersection, while the balance tends to occur at signalized intersections. Pedestrian-vehicle accidents are manifestations of pedestrian-vehicle conflicts resulting from pedestrians crossing into the path of a car or from the latter intruding upon the pedestrian's space (e.g., crosswalk).
There were 491 pedestrian fatalities and 8,019 injuries in Florida in 1992. Persons 55 and older were the most at risk group, comprising 35% of all fatalities and 17% of all injuries (State of Florida Highway Safety Plan, Fiscal Year 1995). Moreover, national studies have shown that South Florida's major cities, specifically Fort Lauderdale and Miami, are the most dangerous communities in the country for pedestrians. In a recent study completed by the Surface Transportation Policy Project (1997), Fort-Lauderdale-Pompano Beach and Miami/Hialeah ranked first and second in the country as the two communities scoring highest in the fatality index scale. In Miami-Dade County, of the 96 pedestrians killed in 1995, at least 3 l were 65 years old or older. The number of accidents involving elderly pedestrians at intersections may not only result from the design characteristics of these intersections but also from age related deficits. Therefore, any modification to facilitate pedestrian movement at intersections must take into consideration relevant characteristics of an aging population. Given the greater concentration of older persons in South Florida compared to other regions of the U.S., this concern is particularly pertinent.
Research Objectives
The objective of this research was to determine the difficulties encountered at five intersections in Miami Beach by pedestrians in general and older pedestrians in particular, and to generate recommendations that address these difficulties. These recommendations will provide some guidelines in the projected redesign of various intersections in the area.
METHOD
Data Collection Instruments
The data were collected using both a structured interview questionnaire and filming of pedestrian traffic interactions. A general description of these data collection instruments follows.
Interview.
The pedestrians were interviewed using a fifteen item structured questionnaire developed in English and Spanish. A research assistant fluent in both languages conducted the interview. The information gathered included demographic characteristics of the respondents (e.g., age, sex) as well as the respondents' report of physical problems that are relevant to crossing intersections safely (e.g., problems with walking and seeing). In addition, the pedestrians were asked about their perception of the crosswalk dimensions, the difficulties they encountered while crossing, and their perceived safety at the intersections.
Filming.
Pedestrian and traffic behaviors were also filmed at the five intersections studied. This was done to obtain objective data that could help to identify issues relevant to pedestrian safety. Each of the five intersections was videotaped for approximately two hours. Variables collected through videotape included: traffic volume, vehicle speed, vehicle turns, pedestrian-vehicle conflicts, pedestrian walking speed, duration of Walk, Flashing, and Don't Walk phases, and pedestrians' paths. Tapes were reviewed by at least two researchers using a computerized checklist to record the pedestrian and vehicle characteristics. As mentioned earlier, the pedestrians analyzed on the videotapes were classified into two main age categories: old and young. Specifically, a pedestrian was classified as young if he/she appeared to be 40 years old or younger. Pedestrians were classified as old if they appeared to be at least 65. This method has been used in similar research and has been found to be reliable (Hoxie and Rubenstein, 1994; Knoblauch et al., 1996).
Sample
One hundred and seventy-eight (178) pedestrians were interviewed at five intersections in Miami Beach. They consisted of 93 males and 75 females in the following age categories: 66 young (between 20-40), 52 middle-aged (between 41 -60), and 60 old (61 and older). Of the 178 pedestrians, 103 spoke English and 75 spoke Spanish. The participants were administered the interview in their primary language.
In addition to the above, videotape data on the behavior of 263 pedestrians was analyzed. One-hundred and forty-eight of these pedestrians were classified as young (i.e., 40 or younger), 75 could not be classified, and 40 were classified as old (i.e., 65 or older).
Variables
The independent and dependent variables whose relationships to each other will be examined in this paper are the following:
Independent Variables.
Age, problems seeing, and problems walking.
Dependent Variables.
The dependent variables were of two types: subjective and objective measures. The subjective measures were: perceived crosswalk dimensions, perceived difficulties encountered at the intersections, perceived safety, perceived adequacy of crossing time allotted by pedestrian signal. The objective measures were: observed pedestrian difficulties, pedestrian walking speed, crossing time allotted by signal, clearance internal, and pedestrian-vehicle conflicts.
RESULTS
The results of the analyses of the interviews and the videotape data are presented below.
Analyses of Interviews
One of the objectives of the survey was to identify the problems of older pedestrians at urban intersections. In this regard, analyses comparing the perceptions of respondents of different age categories were carried out.
Characteristics of Pedestrians.
Some of the demographic characteristics of the sample obtained were reported above. Other relevant characteristics of the respondents are provided in this section. Of the 178 persons interviewed, the majority (141) were familiar with Miami Beach where sixty-one percent (109) of them resided; twenty-nine percent (52) of the respondents were tourists and 8% (14) worked in the area. Most (58%) of the pedestrians interviewed walked daily to carry out various errands. The others did so less often. Thirty-one percent of the respondents reported having problems seeing; 47% (81) wore glasses or contact lenses. Nineteen percent (33) of the persons interviewed had problems walking; only 6% (11) of the pedestrians were observed to use a walking aid (e.g., cane, walker).
Perception of Crosswalk Dimensions.
A one-way Analysis of Variance (ANOVA) was carried out to determine whether there was a difference in pedestrians' perceptions of the length and width of the crosswalk. There was a significant difference between the three age groups in their perception of the dimensions of the crosswalk. Specifically, older subjects tended to perceive the crosswalk as being much longer and wider than the middle-aged and the young (E( 2,173) =10.14, p<.001) and F( 2,160) =3.96, p<.02) respectively. The respondents' perception of the length of the crosswalk is particularly relevant since it indicates the perceived demand for crossing the intersection. Thirty-five percent of the older pedestrians perceived the crosswalk as being too long compared to 10% of the middle-aged, and 8% of the young.
Perceived Difficulties and Safety at Intersections.
Chi-Square analyses were conducted to determine differences in respondents' perception of their safety and the difficulties they encountered at the intersections. Pedestrians were asked about their perception of the adequacy of the time allotted by the crossing signal. Forty-seven percent (73) reported not being provided sufficient time, followed by 44% who thought they were provided "more than enough time", and 9% who had "just enough time". The majority of the respondents felt safe crossing the intersections (59%) although a substantive number did not (41 %). There were, however, no age differences in the pedestrians' perception of safety at the intersections. As for the pedestrians' perception of the difficulties they encountered, again most pedestrians (75%) reported not encountering any difficulties while crossing the intersections. However, older pedestrians reported significantly more difficulties than their younger counterparts (II2(2, H=167) =6.22, p=.05). They complained generally about traffic being excessive and drivers preventing them from crossing. Moreover, among five categories of difficulties (medical/physical, signal phase too short, excessive traffic, drivers' interference, and signal malfunction) their most frequent complaint was that drivers prevented them from crossing (II2(4, t1= 163) =10.05, p<.04). The data show that there were significantly more older pedestrians with problems walking (42%) than younger (3%) or middle-aged (12%) pedestrians (II2(2, N=167) =34.95, p<.0001)). Furthermore, it was found that pedestrians with walking problems tended to report significantly more difficulties crossing intersections than those without walking problems (II(1, N=164) =19.22, p < .001)).
Analysis of the suggestions made by the three age groups to enhance the safety of pedestrians at intersections showed a significant difference in the types of suggestions made by the different age groups (II(6, N=96) =22.93, p=.001). The majority of pedestrians (52%) made suggestions to increase the time allotted for pedestrians to cross the intersection. Older respondents made up the largest number of pedestrians in this category (50% compared to 22% who were middle-aged, and 28% in the young category).
Analyses of Videotape Data
As stated earlier, the five intersections were videotaped to obtain objective data on the pedestrian-vehicle interactions.
Pedestrian Walking Speed.
A one-way ANOVA was conducted to assess differences in the walking speeds of the pedestrians. There was a significant difference in walking speed between older and younger pedestrians (F(2,260) =15.83, p<.001). Specifically, the average speed of older pedestrians was 3.19ft/sec compared to 4.42 ft/sec for younger pedestrians, and 4.15 ft/sec for the group that could not be classified. These speeds are lower than those found by Knoblauch et al. (1996) for their 15th percentile pedestrians. The walking speeds for the 15th percentile pedestrians in our sample were as follows: 3.09ft/sec overall, 3.31 ft/sec for the younger pedestrian, 2.20 ft/sec for the older pedestrians, and 3.44ft/sec for the unclassified group. The walking speed of the 15th percentile older pedestrians in our sample is similar to that found by Hoxie and Rubenstein (1994) and Dahlstedt (1978) (i.e., 2.4ft/sec (estimated) and 2.3ft/sec, respectively).
Pedestrian Difficulties.
Chi-Square analyses were conducted to examine differences in the observed difficulties of young and old pedestrians. Older pedestrians were observed to have significantly more difficulties crossing the intersections than their younger counterparts (II(2, N=263) =40.39, p<.001). These data support the greater frequency of report of difficulties by older pedestrians while crossing the intersections.
Pedestrian Crossing Time.
The time it took pedestrians to cross at a signalized intersection was collected and compared to the clearance interval for that intersection. If one uses the average walking speed of older pedestrians (i.e.,3.19ft/sec) as a measure of the adequacy of the time allotted for crossing an intersection, it was found that the clearance interval was adequate at all intersections. However, if one uses the 15th percentile walking speed of the older pedestrian in our sample (i.e., 2.2ft/sec), or that of the I 5th percentile pedestrians in the Dahlstedt (1978) and Hoxie & Rubenstein (1994) studies, only one of the intersections provided adequate crossing time (See Table 1).
Criteria for Walking Speed at Intersections Studied.
An analysis of the criteria for walking speeds used for the clearance interval at the four signal-controlled intersections on Collins Avenue showed the following: 2.15 ft/sec (11 St), 2.43ft/sec (Lincoln Rd.), 2.44ft/sec (21 St.), and 2.78ft/sec (23 St.). While the first three walking speed criteria accommodate at least 85 percent of the older pedestrians in our sample, the last (i.e., 2.78ft/sec) would only accommodate 65% of our older pedestrians.
Pedestrian-Vehicle Conflicts.
Chi-Square analyses were conducted to determine where pedestrian-vehicle conflicts occurred and the party (i.e., pedestrian or vehicle) responsible for the conflict.
There were 96 pedestrian-vehicle conflicts recorded. Of these, the vehicle was at fault only in 29% of the time. The balance of these conflicts was caused by pedestrians (71 %). An analysis of the involvement of specific age groups in conflicts where the vehicle was at fault showed pedestrians of all age groups to be as likely to be interfered with by a vehicle (II(2, N=263)=4.91, p=.08. In instances where the pedestrian was at fault, young pedestrians were more likely than other age groups to interfere with a vehicle (II2(2, ~=263)=7.82, p=.02) . The 96 pedestrian- vehicle conflicts recorded represent about 37% of the pedestrian- vehicle interactions observed on the videotape.
Most pedestrian-vehicle conflicts occurred in the crosswalk. As expected, most cases where the vehicle was at fault occurred in the crosswalk (II(1, h=263)=4.01, p=.05, whereas in conflicts caused by the pedestrian, the latter was more likely outside of the crosswalk and very likely crossing at mid-block (II2(1, N=263)=7.63, p=.006.
DISCUSSION
The safety of elderly pedestrians is of great importance in Florida, which has the highest percentage of elderly in the country. Furthermore, the area of Miami Beach where our data collection was conducted has among the highest concentrations of elderly in Miami-Dade County as well as in the country. According to the 1990 Census 32% of persons in that area of the city are 65 years old and older. Therefore, any study that addresses the safety of pedestrians would be incomplete without analyzing issues of relevance to elderly pedestrians.
It is interesting to note that objectively the clearance intervals at the intersections permitted at least 88% of all the pedestrians we interviewed to complete crossing. Yet, 47% of these pedestrians reported not being provided sufficient time to cross. This might imply that though most pedestrians made it across the intersections, they may have felt hurried. While younger pedestrians may have been able to compensate for that by walking faster, older pedestrians, who were also more likely to have problems walking, reported experiencing significantly more difficulties than younger pedestrians at intersections. The greater demand imposed upon the elderly pedestrians by the intersections might be what is reflected in the response of more than a third of the elderly when they refer to the crosswalks as "too long".
Although the clearance intervals provided at the intersections studied are even more generous than those proposed by the ITE, based on the walking speed of our I 5th percentile elderly pedestrian, 3 of the 4 signal-controlled intersections were found inadequate and potentially dangerous for elderly pedestrians.
Another manifestation of the threats to pedestrians' safety at intersections is the nature of their interactions with the vehicles with which they share the road. As shown in the interviews one of the most frequent complaints of pedestrians in general and of older pedestrians in particular was about reckless drivers interfering with their crossing the intersections. Furthermore, although pedestrians were found to be responsible for the majority of pedestrian-vehicle conflicts, their risk of injury or death is greatly increased in these interactions. These pedestrian-vehicle conflicts may point to a disregard of these road users for each other. They could also point to misunderstandings by each of the rights of the other.
Our findings underscore the need to incorporate the physical capabilities of the elderly in the design of intersections. Furthermore, they have implications for the development of educational interventions that will inform drivers and pedestrians of their rights and obligations when interacting with other road users, thereby, enhancing the safety of all pedestrians.
ACKNOWLEDGMENTS
This paper is based on FDOT Contract No. B-9856 The views expressed in this paper are solely those of the authors and do not reflect those of the Agency or its employees.
References
Brock, D.B., Guralnik, J.M., and Brody, T. A. (1990). Demography and epidemiology of aging in the United States. In E.L. Schneider and J.W. Rowe, Handbook of the Biology of Aging (3rd edition). Academic Press, Inc. pp.3-23.
Dahlstedt, S. (1978). Walking speed and walking habits of elderly people:. National Swedish Road and Traffic Research Institute.
Fozard, J.L. (1981). Person-environment relationships in adulthood: Implications for human factors engineering. Human Factors, 23, 7-27.
Gottsdanker, R. (1982). Age and simple reaction time. Journal of Gerontology. 37, 342-348.
Hauer, E. (1988). The safety of older persons at intersections. Transportation in an aging society: Improving mobility and safety for older persons, Vol 2 Technical Papers. Transportation Research Board National Research Council. Washington, D.C. pp. 194-252.
Hoxie, R. E. and Rubenstein, L. Z. (1994). Are older pedestrians allowed enough time to cross intersections safely? Journal of the American Geriatrics Society. 42, 241 -244.
Jolibois, Jr., S.C., Guerrier, J.H., Ospina, D., Baker, T., Herrnantin, L. (1997). Interventions to promote pedestrian safety at urban intersections. FL/DOT/RMC/0510760, Contract No. B-9856. Performed by the Florida International University for the Florida Department of Transportation, Research Center, MS30, Tallahassee, FL.
Knoblauch, ILL., Pietrucha, M.T., and Nitzburg, M. (1996). Field studies of pedestrian walking speed and start-up time. In: Pedestrian and bicycle research. Transportation Research Record, No. 1538. Transportation Research Board, National Research Council. National Academy Press, Washington, D.C. pp. 27-38.
Larish, D.D., and Stelmach, GE (1982). Programming, and reprogramming of aimed hand movements as a function of age. Journal of Motor Behavior l 4, 322-340.
Mackay, M. (1988). Crash protection for older persons. Transportation in an aging society: Improving mobility and safety for older persons. Vol 2 Technical Papers. Transportation Research Board National Research Council. Washington, D.C. pp. 158-193.
Manual on Uniform Traffic Control Devices (1988). U.S. Department of Transportation, Federal Highway Administration. U.S. Government Printing Office, Washington D.C.
McDowd, J.M. and Birren, J.E. (1990). Aging and attentional processes. In J.E. Birren and K Warner Schaie, Handbook of the Psychology of Aging (3rd edition). Academic Press, Inc. pp.223- 233.
Small, A.M. (1987). Design for older people. In: G. Salvendy (ed), Handbook of human factors. A Wiley-Interscience Publication; John Wiley and Sons, N.Y. pp 494-504.
Surface Transportation Policy Project (1997). Mean Streets. pedestrian safety and reform of the nation's transportation law. Environmental Working Group, Washington, D.C.
Zegeer, C.V., Opfela, K.S., and Cynecki, M.J. (1985). Pedestrian signalization alternatives. U.S. Department of Transportation, Washington, D.C.
Copyright ©1998 by Human Factors and Ergonomics Society, All rights reserved. For a complete copy (with tables and figures), contact Lois Smith at hfes@compuserve.com.