SUBJECT NAME: ROAD SAFETY ENGINEERING (BCV755A)

MODULE:2- Traffic Engineering Studies:

Syllabus Traffic Engineering Studies: Statistical Methods In Traffic Safety

Analysis – Regression Methods, Poisson Distribution, Chi- Squared Distribution,
Statistical Comparisons- Traffic Management Measures And Their Influence On
Accident Prevention.

Prepared by, Prof. Gowtham B, Department of Civil Engineering,

Sai Vidya Institute of Technology, Bengaluru

Statistical Methods In Traffic Safety Analysis :

Statistics plays a crucial role in traffic-safety analysis: by analyzing crash data, it helps quantify risk
factors (like road type, speed, driver behavior), identify high-risk locations, and predict the likelihood
and severity of accidents through models such as regression or machine-learning. It enables data-
driven decisions for designing interventions, evaluating the effectiveness of safety measures, and
optimizing resource allocation. Statistical analysis also supports real-time risk prediction and allows
policymaking to be backed by solid empirical evidence. Moreover, by revealing trends and patterns, it
helps in targeting enforcement, infrastructure changes, and public awareness campaigns, ultimately
contributing to reduced fatalities and injuries.

Major statistical methods used in traffic safety analysis are,

1. Regression Methods
2. Poisson Distribution
3. Chi- Squared Distribution
4. Statistical Comparisons

1. Regression Methods/analysis:
The basic principle behind this method is that the expected number of accidents, on a
certain road system during a given time period, is dependent in a linear way on factors which
are supposed to be of significance for the determination of accident frequency
The number of accidents occurring on a certain day is itself assumed to be normally
distributed, with a mean value being a linear function of regression variables, and a variance
being constant a same for all days of a certain time period studied.
The numbers of accidents on different days are assumed to be stochastically
independent.

Some of the regression variables (independent variables) that could be considered are:
 Two wheeled vehicles involved in personal injury accidents as a proportion of all
vehicles involved.
 Cost of safety improvements.
 Number of pedestrians
 Pavement width
 Number of junctions per km length of road
 Speed
Use of Regression Analysis in Traffic Safety Studies: In traffic safety, regression models
are commonly applied to develop Safety Performance Functions (SPFs), which estimate the
expected number of crashes at a location based on its characteristics. These functions are
used to identify high-risk areas, evaluate countermeasures, and perform cost-benefit analyses.
A typical regression model in this context might relate accident frequency (dependent
variable) to one or more independent variables such as:
 Traffic volume(Concentration) (e.g., Average Daily Traffic - ADT)
 Roadway design features (e.g., number of lanes, shoulder width)
 Intersection type (e.g., signalized, unsignalized)
 Environmental factors (e.g., weather, lighting conditions)
Multiple Linear Regression
 A statistical technique which will be most frequently encountered by a Traffic
Engineer and Transport Planner is the multiple linear regression analysis. The
problem concerns with the establishment of relationship between a variable which is
known to respond to changes in two or more other variables. The variable which is
known to respond, Y variable, is commonly called the dependent variable, and the
other variables influencing it are called the independent variables, i.e., X variables.
The function will be of the following form:
 𝑌 = 𝑎0 + 𝑎1 𝑥1 + 𝑎2 𝑥2 + ⋯ + 𝑎𝑛 𝑥𝑛
Where
Y = True estimate of the dependent variable, y
𝑎1, 𝑎2 ... 𝑎n = Regression coefficients of the representative m independent variables
x1, x2 … x3 = m independent variables
𝑎0 = Regression constant

P.T.O
Assumptions in the multiple linear regression analysis: The following are some of the
conditions which must be satisfied if a multiple linear regression analysis is to be used:
1. All the independent variables must be independent of each other and there should be
no correlation between them.
2. All the variables are normally distributed.
3. All the variables are continuous.
4. A linear relationship exists between the dependent variable and the independent
variables.
5. The influence of the independent variable is additive, i.e., the inclusion of each
variable in the equation contributes a distinct portion in the estimation of the
dependent variable and all the variables together contribute additively in the
estimation.
Applications of Regression in Accident Analysis

1. Predicting accident frequency based on traffic volume, road geometry, and

environmental conditions.
2. Identifying significant factors that contribute to crashes (e.g., speed, lane width,
lighting).
3. Estimating crash severity by relating variables such as speed or vehicle type to
injury outcomes.
4. Developing Safety Performance Functions (SPFs) to evaluate expected crash rates
at road segments.
5. Before–after safety evaluation, measuring the impact of road improvements or
safety measures.
6. Comparing different locations by controlling for exposure variables like AADT or
road length.
7. Quantifying the effect of countermeasures, such as rumble strips, signals, or
pavement improvements.
8. Forecasting future crash trends under changing traffic and environmental
conditions.

Limitations of Regression in Accident Analysis

1. Requires large and reliable datasets, which may not always be available for
accident records.
2. Assumes linear or specific functional relationships, which may not represent real
crash behavior.
3. Sensitive to multicollinearity, where predictor variables are highly correlated.
4. Affected by outliers, such as rare but severe crash events that distort results.
5. Cannot fully capture random variations in accident occurrences without advanced
models.
6. Model assumptions (normality, independence, homoscedasticity) may be violated
in crash data.
7. Overfitting may occur when too many variables are included relative to data size.
8. Limited for count data, requiring Poisson/Negative Binomial models rather than
simple linear regression.
2. Poisson Distribution:
The Poisson distribution is a discrete probability distribution that expresses the
probability of a given number of events (such as traffic accidents) occurring in a fixed
interval of time or space, assuming these events happen independently and with a constant
average rate. In the context of traffic safety, the Poisson distribution is widely used to model
accident occurrences at specific locations, such as intersections, road segments, or highways,
over defined time periods.
The Poisson Distribution helps us understand the probability of a certain number of
events happening within a fixed period or area. It is often used to model things like the arrival
of vehicles.

Formula to be used in Poisson distribution:

P.T.O
Application of the Poisson Distribution in Traffic Accident Modelling,

The Poisson distribution is widely used in traffic engineering to model random and
independent occurrences of accidents over a fixed period of time or road length. Since
accidents are rare events influenced by numerous unpredictable factors, the Poisson model
effectively represents their probability and frequency.

Key Applications

1. Predicting accident frequency on a particular road segment or intersection over a

given time period.
2. Estimating the probability of a specific number of accidents, e.g., exactly 2
accidents in a month.
3. Crash rate analysis by relating accident counts to exposure variables such as vehicle-
kilometers traveled.
4. Risk assessment to identify high-risk locations ("black spots") based on unusually
high accident frequencies.
5. Evaluating safety performance before and after implementing safety measures (e.g.,
signal installation, widening).
6. Modeling rare events where accident occurrences follow a random, independent
process.
7. Forecasting future accidents using historical data assuming similar traffic and
environmental conditions.
8. Comparing road segments by calculating expected vs. observed crash frequencies.
9. Supporting Poisson regression models (e.g., Poisson or Negative Binomial models)
for advanced crash prediction.
10. Determining required safety interventions when the expected accident count
exceeds acceptable thresholds.

Limitations of the Poisson Distribution in Traffic Accident Modeling

1. Assumes accidents occur independently, which may not hold true because one
accident can influence the likelihood of another (e.g., congestion after a crash).
2. Assumes a constant mean accident rate, but real-world accident rates vary with
weather, time of day, traffic volume, and road conditions.
3. Cannot handle over dispersion, where the variance of accident counts is higher than
the mean (common in crash data).
4. Not suitable for under dispersed data, where variance is lower than the mean.
5. Fails to incorporate heterogeneous factors like driver behaviour, geometric
characteristics, and traffic patterns.
6. Not ideal for locations with very low or zero accidents, because rare-event data
may violate model assumptions.
7. Sensitive to exposure differences, such as varying traffic volume or length of road
segments.
8. Poor at modeling clustered crashes, where accidents tend to occur in bursts rather
than randomly.
9. Limited ability to incorporate covariates, unless extended into Poisson regression
models.
10. Oversimplifies real-world conditions, making it less accurate for complex safety
analysis.
3. Chi- Squared Distribution:
One of the situations a traffic engineer has to assess frequently is whether the safety
measures adopted at a particular location or stretch of road have been really effective in
reducing the number of accidents. Before and after data on accidents can be evaluated on
statistical principles, and one of the handy tools in this direction is the Chi-squared test.
Let “b” be the number of accidents before the improvements at a particular location
and “a” the number after the improvements. Assuming that the improvements have no effect
and the accident number is expected to increase due to the changes in traffic and weather then
let b C be the number of accidents expected if no improvements had been carried out, the
factor C being called the control ratio. Then the value of the Chi-squared is:
(𝒂 − 𝒃𝑪)𝟐
𝝌𝟐 =
(𝒂 + 𝒃)𝑪
The null hypothesis Ho stipulates that there is no real change due to the improvements.
Assuming a 5% level of significance, the value of χ2 to be 3.841 with one degree of
freedom. If χ2 > 3.841, we observe that the null hypothesis is unlikely to be true and that
there is a real change. On the other hand, if χ2 < 3.841 we conclude that the null hypothesis is
true and that there is no real change due to the improvements.

The chi-squared χ2 test is a very useful statistical tool and has many applications.

Following are the important applications to the traffic engineering. They are:
 Testing of proportions with contingency tables: A convenient application of the χ2
test is in testing the comparability of observed and expected values in two-way tables,
known as contingency tables

 Goodness-of-fit test: Another useful application of the chi-squared distribution is in

the goodness-of-fit test. Under this test, the measure of the discrepancy between a set
of observed data and the values that are to be expected if the results follow a
hypothesis distribution is evaluated. For example, the analyst will have an answer to
the following types of questions:

(i) It is observed from past experience that the spot speeds follow a normal distribution. Are
the data obtained from a particular study also in line with the previous experience.

(ii) The arrival pattern of vehicles is known to be generally random and follows the Poisson
distribution. Do the data obtained from a particular study also indicate a Poissonian arrival
pattern?

Benefits of Using the Chi-squared Test

 Simplicity and Clarity: The Chi-squared test is straightforward and easy to compute, making
it suitable for practical engineering evaluations.
 No Assumption of Normality: It is a non-parametric test and doesn’t require data to follow a
normal distribution.
 Objective Evaluation: It provides an evidence-based way to assess whether safety
improvements are effective, supporting data-driven decision-making.

Limitations and Considerations:

  The test requires a sufficient sample size; small numbers can produce unreliable
results.
 It only tests association, not causation.
 It assumes that the expected frequencies are based on a valid and consistent baseline.
The Chi-squared test serves as a valuable analytical tool in traffic engineering for
evaluating the statistical significance of changes in crash rates following the implementation
of safety countermeasures. By comparing observed and expected crash frequencies, it helps
determine whether safety interventions have had a meaningful impact, thereby supporting the
continuous improvement of roadway safety strategies.

Problems on Chi Square test:

1. In an ordinary square junction of two roads there were 20 accidents in a year. After
provision of traffic signals, the number of accidents dropped down to 8 per year. In the
sector of the city where this junction is situated, the general trend observed was that
number of accidents increased at a rate of 10 per cent during the period covered by the
above two observations. Test whether the improvement in junction design has a
significant effect at 5% significance level.
Solution:
Here, a = 8; b = 20; C = 10 % increase

110
𝐶= = 1.1
100
(𝑎−𝑏𝐶)2 (8−20∗1.1)2
𝜒2 = (𝑎+𝑏)𝐶
= (8+20)∗1.1

𝜒2 = 6.36

χ2 for 5% significance level and one degree of freedom = 3.841

As χ2 (observed) > 3.841, the data supplied provides strong evidence that the improvement in
junction design was effective, and that the reduction in the number of accidents was not
merely due to chance alone.

2. The accident data pertaining to a metropolitan city for the year 1965 and 1970 are
given below:
Year 1965 1970
Accidents 300 400
Vehicle-kilometre of travel 250 million 300 million
Test whether there is any significant increase in the accident rates in two
Solution:
Here, the traffic has increased from 250 to 300 million vehicle kilometres, and thus the
control rate C
300
𝐶= = 1.2
250

Here, a = 400; b = 300

 (𝑎−𝑏𝐶)2 (400−300∗1.2)2
𝜒2 = (𝑎+𝑏)𝐶
= (400+300)∗1.2

𝜒2 = 1.90
χ2 for 5% significance level and one degree of freedom = 3.841

As χ2 (observed) < 3.841

Conclude that the difference in the accident rates in the years might easily have arisen
due to chance.

4. Statistical Comparison

Statistical comparison refers to the process of evaluating two or more groups, datasets,
conditions, or variables using statistical methods to determine whether the observed
differences are real, significant, or due to random chance. It helps understand relationships,
patterns, and performance differences between groups.

Key Methodology of Statistical Comparison

1. Compares two or more datasets to identify differences or similarities.

2. Uses numerical and graphical tools such as mean, median, variance, box plots, and
histograms.
3. Helps determine whether differences are statistically significant using tests like t-
test, Chi-square, ANOVA, or regression.
4. Evaluates performance of two conditions, e.g., before vs. after, control vs.
treatment, two road sections, etc.
5. Quantifies the strength of differences using p-values and confidence intervals.
6. Essential in traffic safety to compare accident rates, vehicle speeds, or effect of
countermeasures.
7. Helps in decision-making by identifying which factor or intervention performs better.
8. Reduces subjective judgment and provides evidence-based conclusions.

Types of Statistical Comparison in Accident study

1. Comparing accident frequency before and after installing a speed breaker (paired t-
test).
2. Comparing accident rate on two different highways (Chi-square or Z-test).
3. Evaluating severity distribution at three intersections (ANOVA / Chi-square).
4. Comparing mean speeds during peak vs off-peak hours (t-test).
5. Trend analysis of yearly accident counts (time-series comparison).
6. Assessing lane width effect on crash numbers (regression comparison).
7. Comparing pedestrian crash involvement across different crosswalk types
(proportion comparison).
8. Understanding relationship between traffic volume and accidents (correlation
analysis).
Methods Used in Statistical Comparison

1. T-test – compares means between two groups.

2. Paired t-test – compares before–after or matched samples.
3. ANOVA – compares means among more than two groups.
4. Chi-square test – compares proportions or frequencies (e.g., accident types).
5. Regression analysis – compares effects of multiple factors simultaneously.
6. Correlation analysis – measures strength of relationship between variables.
7. Z-test for proportions – compares accident rates across two different road sections.
8. Non-parametric tests (Mann-Whitney, Kruskal-Wallis) – used when data is not
normally distributed.

General Applications of Statistical methods in traffic safety analysis,

 Accident Prediction / Risk Modeling
 Statistical models (like Poisson regression, Negative Binomial regression) are used to
predict crash frequency based on explanatory variables — e.g., traffic volume, road
geometry, traffic control, weather, driver behavior.
 These models help in estimating the probability of accidents at particular locations
(black spots), which helps in prioritizing safety interventions.
 Identification of High-Risk Locations (“Hotspots”)
 By analyzing accident data statistically, authorities can identify locations
(roads/intersections) with abnormally high accident rates.
 Evaluation of Safety Interventions
 Before-and-after studies: statistical analysis helps evaluate whether a safety measure
(e.g., road redesign, signage, new traffic control) has reduced accidents. The
“Regression-to-mean” effect is also studied in safety evaluation.
 Assessment of Intelligent Traffic Systems (ITS)
 Multi-criteria statistical analysis (e.g., cluster analysis) is used to evaluate which ITS
applications (speed warning, adaptive signals, vehicle-to-infrastructure systems) are
most effective for safety, based on expert surveys or performance data.
 Descriptive Analysis & Trend Monitoring
 Time series analysis: to study how crash rates change over time, perhaps in response
to interventions, policy changes, or external factors (e.g., holidays, weather).
 Outlier Detection and Model Calibration
 In accident modeling, outlier analysis is used to identify unusually severe or frequent
crash sites and correct / recalibrate models.
 When accident prediction models developed for one region are used in another
(“model transferability”), statistical methods test their validity and recalibrate them
for local conditions.
 Surrogate Safety Measures (Microscopic Safety Analysis)
 Use of statistical methods on high-frequency trajectory data (e.g., extracted from
video, computer vision) to detect “conflicts” or near-accidents, even when no crash
occurred.
 Data-Driven Real-Time Risk Assessment / Prediction
 Use of predictive analytics (statistical + machine learning) to estimate real-time crash
risk, combining crash history, environmental data, road conditions.
 Verification of Safety Laws / Empirical Relationships
 Empirical statistical relationships like Smeed’s Law (fatalities vs number of
vehicles/population) are studied and tested with modern data to evaluate their
relevance.
 Prioritization & Resource Allocation
  Based on statistical crash-risk modeling, transport planners and governments decide
where to allocate safety resources (road improvements, enforcement, signage) cost-
effectively.

Traffic Management Measures and Their Influence on Accident Prevention

The fundamental approach in traffic management measures is to restrain as much as
possible existing pattern of streets but to alter the pattern of traffic movement on these, so
that the most efficient use is made of the system.
Some of the well-known traffic management measures are

1. Restrictions of turning movements

The problem posed by turning traffic: At a junction, the turning traffic includes left-turners
and right-turners. Left –turning traffic dose not usually obstruct traffic flows through the
junctions, but right-turning traffic can cause serious loss of capacity.
At times, right-turning traffic can lock the flow and bring the entire flow to a halt.
One way of dealing with heavy right-turning traffic is to incorporate a separate right-turning
phase in the signal scheme, or to introduce an early cut-off or late start arrangement. These
schemes have their limitations and result in a long signal cycle. Another solution is to ban the
turning movement altogether.

Prohibited right- turning movement: Prohibition of right-turning movement can be

established only if the existing street system is capable of accommodating an alternative
routing. Depending upon the existing layout of the street system, three methods are available:
 Diversion of the right-turning traffic to an alternative intersection further along the
road where there is more capacity for dealing with a right-turn. This scheme is known
as a T turn (fig a)
 Diversion of the right-turning traffic to the left before the junction. This scheme is
known as a G turn (fig b)
 Diversion of the right-turning traffic beyond the junction. This scheme is known as a
Q turn (fig c)

2. One-way streets
One-way streets are those where traffic movement is permitted in only one direction.
As a traffic management measure intended to improve traffic flow, increase the capacity and
reduce the delays, one-way streets are known to yield beneficial results.

Advantages
 A reduction in the points of conflict: Traffic movements at junctions involve a
number of points of conflict. These generate delay, congestion and accident hazards.
Any scheme where the points of conflict are reduced in number is thus conductive to
better safety and less delay.
  Increased capacity: The removal of opposing traffic and the reduction of intersection
points of conflict results in a marked increase in the capacity of a one-way street.
 Increased speed: Since the opposing traffic is eliminated, drivers can operate at
higher speeds. This is further facilitated by the more efficient operation of the traffic
signal system that is possible under one-way street operation.

Disadvantages
 Although the journey times and delays are reduced, the actual distances to be covered
by drivers increase.
 Where buses operate on the streets, the stop will have to be relocated and in many
instances the passengers will have to be relocated and in many instances the
passengers will have to walk extra distances.
 The excessive speeds that follow as a result of one-way operation may be a hazard to
residential areas. Thus, while the number of accidents may decrease, the severity will
increase with one-way operation.

3. Tidal flow operation

One of the familiar characteristics of traffic flow on any street leading to the city
center is the imbalance in directional distribution of traffic during peak [Link] of the
method of dealing with this problem is to allot more than half the lane for one direction
during peak hours. This system is known as “tidal flow operation” or reverse flow operation.

Methods
The principle of tidal flow operation can be translated into practice in two ways:
 The first is to apportion a great number of lanes in a multi-lane street to the in-bound
traffic during morning peak and similarly a great number of lanes to the out-bound
traffic during the evening peak.
 The second requires the existence of two separate streets parallel to each other and
close to each other, so that the wider of the two can be set apart for the heavier traffic
both during morning peak and evening peak. In this case, the two streets will operate
as one-way streets.

4. Closing side-streets
Method
A main street may have a number of side-streets where the traffic may be very light. In such
situations, it may be possible to close some of these side-streets without affecting adversely
the traffic, and yet read a number of benefits.

Advantages
 Since interference from the traffic from side streets is eliminated, the speed increases
and journey time reduces.
 For the same reason as above, the accident gets reduced.
 If the side streets are too many and at close intervals, it is difficult to formulate a
scheme for the progressive system of signals.

Disadvantages
  Closure of a number of cross-streets may increase the flow to and from the remaining
cross-streets. This may necessitate signal control and other measures at these
junctions.
 When a number of side-streets are closed, the immediate effect is an increase in the
parking of vehicles on the main street itself.

5. Exclusive bus lanes

Exclusive bus lanes running against heavy one-way flow are also very common. One
experience suggests that such an arrangement nearly halves the journey time. A good
measure of enforcement is needed if serious accidents have to be avoided in this system.
Bus priority measures are a cheap and easy way to provide some aid to bus services.
The journey time can be considerably reduced and bus journey time can be made more
attractive.

6. Following 3Es in Accident Prevention

The 3 Es represent the three fundamental approaches used worldwide to reduce accidents and
improve road safety:
1. Engineering
Engineering focuses on designing safer roads, vehicles, and traffic control devices.
Examples:
 Improving road geometry (curves, gradients, lane width)
 Installation of traffic signals, signs, and road markings
 Providing speed breakers, rumble strips, guardrails
 Better lighting, pedestrian crossings, footpaths, and medians
Purpose: Eliminate or reduce physical hazards.
2. Enforcement
Enforcement ensures that road users follow traffic laws and safe driving practices.
Examples:
 Speed enforcement using cameras
 Checking drunk driving
 Wearing helmets and seat belts
 Penalizing wrong-lane driving, overloading, and red-light jumping
Purpose: Control risky behavior and ensure compliance.
3. Education
Education aims to create awareness among road users about safe road behavior.
Examples:
 Road safety campaigns in schools, colleges, and media
 Driver training programs
 Awareness on pedestrian safety, helmet use, and speed control
 Community-based safety workshops
Purpose: Develop responsible and informed road users.
Question bank: Module 2

1. List and explain the application of statistical methods in traffic safety analysis.
2. Explain regression analysis with Formula and equations.
3. Problem on regression.
4. Explain application and limitation of regression in accident analysis.
5. Explain Poisson distribution method with Formula and equations
6. Problems on Poisson distribution method.
7. Explain application and limitation of Poisson distribution method in accident analysis
8. Explain Chi square distribution with Formula and equations.
9. Explain benefits, applications and limitations of Chi square test.
10. Problems on Chi square test.
11. Write a note on Key Methodology and Types of statistical comparison in accident
study.
12. Explain methods used in Statistical Comparison.
13. Explain various traffic management measures to prevent accident studies.