Impact of Vehicular Countdown Signals on Driving Psychologies and Behaviors : Taking China as an Example

Countdown signal control is a relatively new control mode that can inform a driver in advance about the remaining time to pass through intersections or the time needed to wait for other drivers and pedestrians. At present, few countries apply vehicular countdown signals. However, in China, some cities have applied vehicular countdown signals for years, though it is unclear how and how much such signals influence driving psychologies and behaviors compared with non-countdown signal controls. The present work aims to clarify the impact of vehicular countdown signals on driving psychologies and behaviors on the cognitive level. A questionnaire survey with 32 questions about driving psychologies and behaviors was designed, and an online survey was conducted. A total of 1051 valid questionnaires were received. The survey data were analyzed, and the main results indicate that most of the surveyed drivers prefer countdown signal controls and think that such controls can improve not only traffic safety but also traffic operational efficiency. The surveyed drivers also think that countdown signal controls have an impact on driving psychologies and behaviors and the survey results have demonstrated that the driving behaviors of female drivers surveyed are not conservative under the clear conditions of green countdown signal control. Further studies and methods concerning the effects of countdown signals on driving psychologies and behaviors are discussed.


Introduction
While intersections are part of the road system, they are far more complex than the segments connecting them [1,2].Driving psychologies and behaviors at road intersections are considerably different compared to those at road sections.According to the control mode, there are two kinds of road intersections: unsignalized and signalized.In recent years, countdown signalized intersections have appeared in some countries or areas.Compared to traditional non-countdown signals, countdown signals can inform a driver in advance about the remaining time to pass through intersections or the time needed to wait for other drivers and pedestrians.Countdown signals can be divided into vehicular countdown signals and pedestrian countdown signals.Vehicular countdown signals further include green, red, and yellow countdown signals.
Out of more than 200 countries, only a few use countdown signal controls at road intersections, such as China, Thailand, India, Singapore, Malaysia, the United States, and the United Kingdom.Of these countries, only a small number allow pedestrian countdown signals to be used at road intersections, such as the United States and Britain.In other words, few countries currently use vehicular countdown signals.
The object of this study is to focus on vehicular countdown signals.
China, as one of the pioneer countries, is playing a leading role in the use of vehicular countdown signals.However, no specific manuals or standards are available for guiding the usage of vehicular countdown signals in China [3][4][5].Therefore, Chinese local governments encounter difficulties in making a clear decision on extending the application of vehicular countdown signals.Traffic engineers and researchers have different opinions on the application of vehicular countdown signals.Some support the use of countdown signal controls and consider such controls as capable of improving traffic operational efficiency by taking full advantage of the signal time.Others are opposed to the application of countdown signal controls and are concerned about the possibility of such controls increasing traffic crashes and having a negative impact on traffic safety [6].
In theory, vehicular countdown signals are different compared to non-countdown signals and should have an impact on driving psychologies and behaviors.However, to date, it is unclear how and how much vehicular countdown signals influence driving psychologies and behaviors compared with non-countdown signal controls.To answer these questions, the present study explored the influence of vehicular countdown signals from two aspects of driving psychologies and behaviors based on a questionnaire survey.The survey consisted of 32 questions related to vehicular countdown signals.

Literature Review
This paper focuses on vehicular countdown signals (hereafter referred to as countdown signals in the following text unless specified otherwise).To date, countdown signal control, as a new traffic control mode of signalized intersections, is applied in China, Singapore, Thailand, Malaysia, India, and other countries.Lum and Halim conducted a before-and-after study by observing driver reactions at a signalized intersection with a green signal countdown display (GSCD) installed [7].The finding revealed that the number of red-light running violations is significantly mitigated at the initial period after installing the GSCD.However, the effectiveness of the device tends to dissipate over time, with the number of violations bouncing back to almost the same level as before GSCD.Ibrahim et al. introduced countdown timers installed at some intersections in Kuala Lumpur, Malaysia [8].The impact of countdown timers on driving behaviors and intersection approach headways was studied by comparing three intersections with countdown timers with three intersections lacking countdown timers.The result indicated that the countdown timers have a significant impact on headways but a little impact on the initial delay.Limanond et al. studied how countdown timers affect the queue discharge characteristics of through movements during the green phase at a signalized intersection in Bangkok, Thailand [9].They pointed out that the countdown timers had a significant impact on the start-up lost time at the intersection under study, but the effect on the saturation headway was negligible.Chiou and Chang investigated the impact of GSCD and red signal countdown display (RSCD) on driver behaviors in Taiwan [10].The results showed that GSCD can reduce the late-stopping ratio, but it increases the likelihood of rear-end crashes.Although RSCD can effectively reduce the start-up delay, saturated headway, and cumulative start-up delay, it cannot significantly improve intersection safety in the long term.Sharma et al. presented the usage of countdown timers at signalized intersections in India [11].The study conducted a before-and-after analysis by comparing predata with postdata collected at a selected intersection in Chennai.The results reflected that the time information provided at the beginning of the green light (end of the red light) can enhance efficiency and reduce start-up lost time but increase red-light running violations.Papaioannou and Politis found that the percentage of early start violations at the intersection with SCD was 24%, where the percentage for intersections without SCD was less than 1% [12].Devalla et al. found that GSCD is linked with fewer red light violations (RLVs) cycles, a lower mean number of RLVs per RLV cycle, higher vehicular speeds during the phase transition at different locations upstream to the stop line, a higher number of speeding cars, and higher stop line crossing speeds during amber [13].Islam et al. found a reduction in start-up lost time at signalized intersections when a red signal countdown timer is present [14].
In China, Wang and Yang conducted a preliminary analysis on the traffic signal countdown by conducting a survey of 337 drivers regarding driver attitudes and behaviors on the green signal countdown in Longyan City, Fujian province [15].The study advised that the countdown signal should be set cautiously.Wu et al. (2009) focused on the driver's decision-making process at countdown signalized intersections [16].A logistic model was adopted to build the model of behavior decision at countdown signalized intersections based on vehicle types and speed.Zhang et al. conducted a survey on the countdown signal and collected 200 questionnaires from drivers and pedestrians at four intersections in Wanzhuang, Beijing City [17].The results showed that the drivers and pedestrians sampled had a preference for the countdown signal.Qian and Han preliminarily studied the influence of green signal countdown on traffic safety through a questionnaire investigation of 390 drivers [18].The finding indicated that the green signal countdown is good for neither traffic safety nor traffic operational efficiency.Thus, the green signal countdown should be used cautiously.Ma et al. conducted a field observation to obtain critical parameters related to drivers and vehicles at two similar intersections, one with GSCDs and the other without GSCD, in Shanghai City, China [19].They found that GSCD increased the traffic capacity at the sampled intersection and significantly reduced the number of red-light running violations.Qian carried out an eight-question driver behavior survey of 390 drivers regarding the red signal countdown to analyze driver behaviors [20].Long et al. studied the impact of the countdown timer on driver behaviors after the yellow onset and found that the countdown timer influences drivers stopping or passing through the intersection [21].Additionally, a correlation exists between the countdown timer and red-light running violations.Huang et al. found that although GSCD stimulates the drivers in dilemma zones to choose to cross the intersection during amber, which produces a higher RLR risk compared with SCD and GSFD, the intersection with GSCD has the lowest RLR violations due to its strong positive effect in cutting down the range of dilemma zones [22].Li et al. comparatively analyzed drivers' perception-reaction time (PRT) with and without a countdown timer based on the RGB color model and found that the drivers' PRT was decreased from 2.12 s to 1.48 s with countdown signals [23].Pan et al. attempted to find effects of the end of a green signal countdown on drivers' behaviors when they drive vehicles through intersections based on the data of vehicle position, time, and speed at the entrance of intersections [24].Fu et al. characterized and modeled driver's brake perception-reaction time (BPRT) to yellow signal at signalized intersections with and without countdown timer and found an increase in driver's BPRT because countdown timer may induce risky driving behaviors [25].Pan et al. did an interesting study demonstrating that the value of a driver's car has influence on driving behaviors at countdown signalized intersections [26].
As introduced above, three methods are commonly used to study the effects of countdown signals on drivers, including survey, video, and observation in the field.Evaluating the existing studies through questionnaire survey method revealed that the questionnaires are not designed comprehensively enough, because these only involve a part of driving behaviors and do not investigate driving behaviors from the psychological aspect.

Method
3.1.Design of Questionnaire.Considering the complexity of driving psychologies and behaviors and the new mode of countdown signal controls, the questionnaire was designed to fully reflect the common and individual characteristics of the driving psychologies and behaviors involving the aspects listed as follows: (1) Gender: male and female drivers may have significantly different attitudes toward countdown signal controls (2) Age: age refers to the length of time that one has existed.It is an important indicator to reflect the differences in driving psychologies and behaviors of different age groups (3) Driving experience: driving experience is used to examine the differences of driving psychologies and behaviors for drivers with different driving experiences (4) Specific questions associated with driving psychologies and behaviors on the countdown signal controls 3.2.Online Questionnaire Survey.The designed questionnaire was released and conducted by a professional survey website in China.A total of 1051 valid questionnaires were received.

Analysis of Survey Results
The survey results were classified and analyzed to indicate the general understanding of drivers and their psychological characteristics, as well as the behavioral characteristics of different types of drivers surveyed.

Basic Characteristics of Drivers Surveyed.
The characteristics of gender, age, and driving experience for the surveyed drivers are shown in Table 1.
Table 1 shows that the majority of respondents were young drivers and that more male drivers than female drivers For driving age, the drivers with no more than three years of driving experience accounted for 36.06%, which is consistent with the rapid increase of the number of Chinese drivers in the last three years.

Attitudes and Understanding of Drivers on Countdown
Signal Controls.Six questions, questions (4) to (9), were designed to evaluate the attitudes and understanding of drivers on countdown signal controls.The survey data were analyzed, and the results are summarized in Table 2.
Table 2 shows that the majority of the surveyed drivers felt easier driving vehicles on roadway sections than at intersections; they especially felt nervous at unsignalized intersections.Most drivers also supported setting up countdown signal controls, which they considered to be beneficial to driving behavior decisions.Even more drivers thought that countdown signal controls can be conducive to improve traffic safety and traffic operational efficiency.In addition, the proportion of aggressive drivers is not high from the drivers surveyed.

Attitudes and Understanding of Drivers on Green
Countdown Signal Controls.Questions ( 10)-( 15) were designed to investigate driver attitudes on green countdown signal controls and to contrast the behavior of "race against time" at the end of the green light at countdown signal and noncountdown signal control intersections.The analysis results for questions (10)-( 13) are shown in Table 3.
According to Table 3, the surveyed drivers who supported the setup of green countdown signals accounted for the majority of respondents.The proportion of drivers who regarded the green countdown signal as having an impact on driving behaviors reaches up to 92.30%.With regard to the display modes of the green countdown, 55.19% of the surveyed drivers selected the partial countdown.Questions ( 14)-( 15) and their options are described in Table 4, and the comparative analysis is shown in Figure 1.
In comparison with non-countdown signalized intersections (Figure 1), more drivers would like to accelerate passing through intersections while the green light time shifts to the yellow light time at green countdown signalized intersections.However, the difference is not significant.More drivers surveyed agreed with option B at non-countdown signalized intersections than at countdown signalized intersections.The result shows that driving behaviors are more adventurous at countdown signalized intersections than at non-countdown signalized intersections.

Driver Attitudes and Behaviors on Red Countdown Signal
Controls.Questions ( 16)-( 24) were designed to investigate the attitudes and possible driving behaviors of drivers on red countdown signal controls.The analysis results for questions ( 16)-( 19) are shown in Tables 5 and 6.Table 5 summarizes the questions on attitudes and behaviors on red countdown signal controls.Table 6 presents the questions on turning off engines while waiting for the green signal.
Table 5 reveals that most of the drivers surveyed are supportive of red countdown signal controls, and 66.51% of the surveyed drivers considered red countdown signal controls as having an impact on driving behaviors.The proportion of drivers (52.05%) who preferred overall countdowns is close to the proportion of drivers (47.95%) who selected partial countdowns.For question (19), 19.79% of the drivers would accelerate passing through the intersection, which is a very dangerous behavior.
At signalized intersections, the reckless behavior of a driver is largely constrained by other drivers or vehicles.Therefore, the behavior of the first driver in a certain lane is focused on.In Table 6, for question (20), 79.82% of the surveyed drivers would engage the engine gear in advance and accelerate to start once the green light changes.This behavior may cause traffic accidents in conflict directions with vehicles or delayed pedestrians.However, from another angle, it can improve traffic operational efficiency.Regarding question (21), 77.38% of the surveyed drivers who would turn off their engines while waiting would start the engines in advance and then accelerate to move while changing to the green light.For question (22), 58.33% of the drivers surveyed would turn off their engines while waiting when the waiting time is longer than 30 s, which reflects the driver's awareness of conserving energy and reducing exhaust emissions.Question (23) shows that the main causes for not turning off engines at red countdown signalized intersections are feelings of inconvenience and concerns about fuel consumption when restarting the engine.Question (24) indicates that the main causes for not turning off engines are feelings of inconvenience and having no idea of how soon the red light will be over.Note that some questions have a plurality of possible causes or choices; therefore, the percentage sum of multiple-choice questions may be greater than 100%.The calculation principle is the number of times an item is selected divided by the number of drivers surveyed.

Analysis of Red-Light Running Behaviors and Other Risk
Behaviors.Questions ( 25)-(29) were designed to investigate driving behaviors on red-light running, sudden acceleration, sudden braking, and so on.Question ( 25) and its options are described in Table 7, and the results are shown in Figure 2. Question (25) indicates that the proportion of drivers who would intentionally run a red light at countdown signalized intersections is 12.33% (Figure 2).Their main reason for running a red light is to rush through intersections within a very short time, even though the drivers may be well aware of the remaining green light time.In addition, 24.63% of the surveyed drivers revealed good driving behaviors by not running a red light.
Question (26) was designed to correspond to question (25).Its options are described in Table 8, and the results are shown in Figure 3.
Question (26) reflects that cases of red-light running at non-countdown intersections caused by not knowing the time the yellow light will appear and by the sudden transition of yellow light account for a large proportion (63.6%) of the  drivers surveyed (Figure 3).Some drivers even attributed their red-light running to the lack of countdown signals.Redlight running that resulted from inattention still accounts for approximately one-third of the total (29.83%).
Question (27) and its options are described in Table 9, and the statistical results are illustrated in Figure 4. Effective countermeasures to reduce red-light running are installing     automatic-capture facilities, enhancing education, and setting up countdown signal lights based on the degree of influence.
Questions (28)-(29) and their statistical results are shown in Table 10.The proportion of acceleration at the end of the green light at countdown signalized intersections is 26.07%more than that at non-countdown signalized intersections.The key factor is the existence of the countdown signal.
For sudden braking, the possibility of occurrence at noncountdown signalized intersections is higher than that at countdown signalized intersections.Regardless of whether a driver is at countdown signalized intersections or at noncountdown signalized intersections, sudden braking will increase while the yellow light starts.11.
According to Table 11, most of the surveyed drivers selected the mode of countdown display with overall lights (red, green, and yellow).Nearly half of the surveyed drivers considered the red countdown light to be beneficial in improving traffic operational efficiency.

Cross-Analysis of Typical
Questions.The psychological and behavioral characteristics of drivers are closely related to the driver's gender, age, and driving experience.The crossanalysis between gender and attitudes toward countdown signals is carried out based on questions (1) and ( 7) of the survey data.The analysis results shown in Table 12 indicate that the majority of male and female drivers supported the setting up of countdown signal controls, and male drivers are more inclined to support countdown signal controls.
The cross-analysis between gender and behaviors before the end of the green countdown is conducted based on questions (1) and ( 14), as listed in Table 10.Table 13 shows that the proportions of male and female drivers are very close in decelerating to stop by the end of the green countdown.Compared with the male drivers, the surveyed female drivers were also more likely to accelerate passing through the stop line by the end of the green countdown or by the end of conservative in the condition of uncertain remaining time than male drivers.

Conclusion and Discussion
According to the analysis results of the survey, several conclusions can be drawn.(1) Most of the surveyed drivers preferred countdown signalized intersections and tended to select the mode of countdown display of overall lights (red, green, and yellow).( 2) Most of the drivers considered countdown signal controls as capable of improving not only traffic safety but also traffic operational efficiency, which is consistent with the findings from earlier studies [17,19] but is contrary to the studies in [18,20].(3) Regardless of whether green countdown or red countdown controls are set up, most of the drivers considered countdown signal controls as having an impact on driving psychologies and behaviors.However, the impact may not be conducive to improve traffic safety.
(4) The proportion of drivers intentionally running red lights is relatively small at countdown signalized intersections or non-countdown signalized intersections.However, the time by the end of the green signal and at the onset of the yellow signal is the key time of red-light running at both types of intersections.According to the survey results, the installation of an automatic-capture system to catch traffic violations is conducive to reduce the occurrence of red-light running.
(5) Female drivers are traditionally viewed as having more conservative driving behaviors compared with male drivers.However, the analysis results indicate that the driving behaviors of female drivers surveyed are not conservative under clear green countdown conditions.Nevertheless, female drivers are very conservative under non-countdown conditions, which confirms the general psychological characteristics indicating that males are more adventurous than females under unknown conditions.Driving psychologies and behaviors are complex phenomena.To further study the effects of countdown signals on driving psychologies and behaviors, several ways may be recommended: (1) using professional equipment to collect indicator parameters of driving psychologies and behaviors at countdown and non-countdown signalized intersections with actual traffic conditions and then analyzing the data; (2) at countdown signalized or non-countdown signalized intersections, observing or photographing driving behaviors and then analyzing the behaviors; and (3) using a more scientific comparison and analysis of data obtained by different methods to draw reasonable conclusions.

(Figure 1 :
Figure 1: Comparison of driving behaviors before the end of the green light at two types of intersections.

Figure 2 :
Figure 2: Distribution of reasons for running a red light at countdown signalized intersections.

(
A) Inattention, not intentionally running a red light.(B) In a hurry, intentionally running a red light.(C) Intending to pass through the intersection before the end of the green countdown or the end of the yellow light, but runs a red light.(D) Others.(E) Never runs a red light.

(Figure 3 :
Figure 3: Distribution of reasons for running a red light at noncountdown signalized intersections.

Figure 4 :
Figure 4: Distribution of measures to reduce the chances of running a red light.

4. 6 .
Attitudes and Understanding of Drivers on Display Modes of Countdown Signals.Questions (30)-(32) were designed to investigate the attitudes of drivers on the display modes of countdown signals.The statistical results are shown in Table

Table 1 :
Characteristics of drivers surveyed.
responded to the survey.The drivers surveyed below the age of 40 accounted for 76.69%, indicating that young drivers use the Internet more widely in China.

Table 2 :
Attitudes and understanding of surveyed drivers on countdown signal controls.

Table 3 :
Attitudes of surveyed drivers on green countdown signal controls.

Table 4 :
Behaviors of surveyed drivers on the green signal coming to an end in the two cases.

Table 5 :
Attitudes and behaviors of surveyed drivers on red countdown signal controls.

Table 6 :
Behaviors on whether engines should be turned off while waiting for the green light.

Table 7 :
Reasons for running a red light at countdown signalized intersections.

Table 8 :
Reasons for running a red light at non-countdown signalized intersections.

Table 9 :
Measures for reducing the chances of running a red light.

Table 14 :
Cross-analysis between gender and behavior before the end of the green light at non-countdown signalized intersections.