WOLLO UNIVERSITY

COLLEGE OF NATURAL SCIENCE

DEPARTMENT OF STATISTICS
Emprical Analysis Of Traffic Accident Involving Human Injuries In
Bale Robe Town
A Senior Research Submitted to the
Department of Statistics Wollo University
In partial fulfillment of the
Requirements for the Bachelor Degree in
Statistics

I
Abbreviation

CSA: Central Statistics Agency

MCT: Measure of central tendency
RTIs: Road Traffic Injuries
NRSCO: National Road Safety Coordination office

II
Acknowledgment

First, we are very glad to express our heartfelt thanks to Ato Zelalem Getahun
(MSc), our advisor, for his constructive and crucial advice in shaping and
correcting this research paper. His encouragements and welcoming face at any
time will never be forgotten. We also wish to express our appreciation to Bahir Dar
town police station, traffic polices for their cooperation in providing the data.

III
TABLE OF CONTENTS

Abbreviation.....................................................................................................................................I

Acknowledgment............................................................................................................................II

Abstract..........................................................................................................................................III

CHAPTER ONE INTRODUCTION...............................................................................................1

1.1 Background......................................................................................................................1
1.2 Statement of the problem.................................................................................................2
1.3 Objective of the study......................................................................................................3
1.3.1 General objective.....................................................................................................3
1.3.2 Specific objective.....................................................................................................3
1.4 Significance of the study.................................................................................................3
1.5 Limitation of the study.....................................................................................................3

CHAPTER TWO: LITERATURE REVIEW..................................................................................4

CHAPTER THREE: METHODOLOGY........................................................................................7

3.1 Description of the study...................................................................................................7
3.2 Data Source......................................................................................................................8
3.3 Data collection Method....................................................................................................8
3.4 Variable considered in the research.................................................................................8
3.5 Statistical method.............................................................................................................9
3.5.1 Descriptive statistics................................................................................................9
3.5.2 Inferential statistics: Chi-square:χ2 distribution......................................................9

IV
 3.5.3 Logistic regression analysis...................................................................................10
3.5.4 The model..............................................................................................................11
3.5.5 Fitting the logistic regression model......................................................................12

CHAPTER FOUR: RESULT AND DISSCUTION......................................................................14

4.1 Descriptive statistics......................................................................................................14
4.2 Inferential statistics........................................................................................................16
4.3 Logistic regression analysis...........................................................................................17

CHAPTER FIVE: CONCLUSION AND RECOMMENDATION..............................................18

5.1 Conclusion.....................................................................................................................18
5.2 Recommendations..........................................................................................................18

REFERENCE................................................................................................................................19

APPENDIX....................................................................................................................................20

List of Tables in the Text

Table 1 The results of categorical variables..................................................................................14

Table 2 The results of continuous variable....................................................................................15
Table 3 Chi square value...............................................................................................................16
Table 4 Logistic regression output................................................................................................17

List of Tables in the Appendix

A Table 1 Description of factors with the respecting coding........................................................20

A Table 2.......................................................................................................................................21
A Table 3.......................................................................................................................................21

List of Figures

Figure 1 Location of the study area (Bahir Dar town).....................................................................7

V
VI
Abstract

The main purpose of this research paper is to asses which factors are primarily
responsible for the road traffic accident and to investigate road traffic accident
related to the concept like types of injuries, sex of the drivers, age of the drivers,
driving experience, atmospheric condition, light condition, educational
background of the driver, vehicle type, vehicle age, time of event, road condition,
place of accident and days of accident are independent variables and type of
injuries are dependent variable. For this research we use secondary source of
data, which is collected from seven (7) different police stations in Bahir Dar town
in May, 2012. The statistical methods we use were descriptive statistics, chi-square
and logistic regression. The finding of this study showed that experience of the
driver significantly affects the road traffic accidents. Drivers experience greater or
equals to 7 years are more responsible than that of less than or equal to 6 years
experience.

VII
 CHAPTER ONE INTRODUCTION

1.1 Background

Over one million people every year are killed in road crashes, and 20-50 million are injured.
Road traffic injures (RTIs) are growing as the vehicle uses of developing-countries rise. By
2020, RTIs are expected to be the third leading cause of death and disability worldwide, by some
calculations matching the toll of AIDS. Residents of developing countries are at match higher
risk of RTIs than are residents of higher income countries .They are also at greater risk of death
when crash occurs. Developing countries also have inadequate trauma systems and are often
unable to care for crash victims. Unless action is taken to improve road safety systems, poor
countries will continue to bear the heavy toll of road traffic injuries (Lauren et al., 2005).

The 2004 world health report shows that of the 1.2 million people killed in road crashes
worldwide, 85% are in developing countries. Sub-Saharan Africa alone with only 4% of the
global vehicle registered accounts for 10% of the total road fatalities, and the economic, social
and health consequences are grave. Conversely, the high income nations, with 60% of the total
global vehicle fleet contribute only 14% of the annual road deaths. Human error, road
environment and vehicle factors are reported by the traffic police as the main causes of road
crashes. Two countries, South Africa and Nigeria, accounts for most of the reported deaths in
Sub-Saharan Africa figure of over 9,000 has been consistent over time, while Nigeria with 6,185
deaths has declined from a high of over 9,200 in the early 1990s. Ethiopia, Kenya, Uganda,
Tanzania and Ghana are the other countries that experience high numbers of road deaths (Odero,
2004).

Consistent to the above facts, another study shows that Ethiopia is a country with the largest
number of traffic accidents and fatality rate. A study was conducted to assess and describe the
extent, variation and associated cause of road traffic accidents in Bahir Dar. Based on the
understanding and appreciation of the problems, the study has come up with suggestions some
remedial measures. The research was conducted based on the archive of traffic police data

1
between 1995/96 and 2001/02.the result indicates that road traffic accident is a serious problem
in the town of Bahir Dar. The numbers have been following an increasing trend throughout the
study period and the rate of increasing has been raising rapidly due to high population number
(increase in road users) and high traffic mixes (increased of wheeled vehicles of all kinds on the
town road ). The relationships derived indicate that the increase in vehicles and population will
bring an increase in road traffic accidents in Bahir Bar town. Study analysis show that the
number and density of road traffic accidents occurring on the main roads were very high with
figure ranging 80-90 percent of the total.

Mixed allocation of land use type, high volume of traffic low, poor nature of the road, lack of
road facilities, poor traffic management and low level of enforcement contribute to these
problem. In the study, pedestrians are the road users most affected by road accidents. On the
average 52% of traffic accident are pedestrians. Passengers, driver and two wheeled vehicles
riders are also frequently involved in traffic accident.

The risks are highly among males, particularly those who are economically active age group of
18-50 years. The high accident rates of these groups of population are mainly attributed to low
level of understanding of traffic roads and low level of observance. All safety measures
introduced to protect the road users are disregarded by the large proportion of drivers,
pedestrians, school children’s and cyclists ; and that the majority of them have a tendency in
complete ignorance of the dangers encountered by traffic accident.(Yayeh A., 2003)

1.2 Statement of the problem

Road traffic accident involving human injuries has been a major problem of Bahir Bar town so
cities. As we tried to mention in the background section, Bahir -Dar town society have a great
problem in road traffic accident involving human injuries. By understanding this we want to
investigate the following problems.
 What are the major factors that can cause road traffic accident involving human injuries?
 Why so many societies are lost their life because of road traffic accident?
 Who is more responsible for the occurrence of road traffic accident?

2
1.3 Objective of the study
1.3.1 General objective

The general objective of the study is the empirical analysis of road traffic accident involving
human injuries.

1.3.2 Specific objective

More specifically the objective of the study as follows:
 To identify and describe the major variables (factors) those contribute to road traffic
accident involving human injuries.
 To analyses the impact of driver and vehicle related variables on road traffic accident
involving human injuries.

1.4 Significance of the study

The result of this study is expected to:-
 To know the cause and consequence of road traffic accident involving
human injuries.
 To reduce the road traffic accident on human injuries in Bahir Dar town by
giving recommendation and suggestions to the concerned body.
 To take action (punishment) on the factors by the concerned body.

1.5 Limitation of the study

In the process of this study we had challenges faced. The following are among the major ones:

 The data recorded was incomplete and inconsistent records.

 Lack of computer lab in the right time.
 Some police station workers have not willingness to give appropriate data.
 Lack of money to perform our task.
 Lose a lot of time to collect data (since, the data was collected from different
seven police stations, in different Kebeles, Bahir Dar town.

3
 CHAPTER TWO: LITERATURE REVIEW
Different authors defined and explained about road traffic accident. So, the researcher tries to
investigate the definition and problems of road traffic accident on human injuries.

As countries develop death rates usually fail, especially for disease that affect the young and
result in substantial life-years lost. Death due to traffic accidents are notable exception: the
growth in motor vehicles that a company’s economic growth usually brings an increase in road
traffic accidents. Indeed, the world health organization’s has predicted that traffic fatalities will
be the sixth leading cause of death worldwide and the second leading cause of disability adjusted
life-years lost in developing countries by the year 2020 (World Bank, 2003). Al-masaed et al.,
(2004) investigated the relationship between city planning and street network variables and
traffic accidents at the zone level. Damascus, the capital of Syria, which consists of fourteen
urban zones, was selected as a case study. For each zone, data on traffic accidents, population
density, land-use developments, level of travel, road network, intersection density, and the
distribution of public buildings were measured through field surveys or obtained from documents
of relevant authorities.

In the study a cross-sectional approach was adopted to focus on the differences in safety between
urban zones at a specific point in time. Mathematical modeling was used to develop relationships
between zone traffic accident and the variables mentioned above. In the analysis, a correlation
matrix was established to identify variables that had an influence on traffic accident and to check
possible multi-colinearity between pair of independent variables. In the modeling process, the
first step was to identify the best transformation for variables that had an impact on traffic
accidents. In the second step, logistic regression analysis was carried out to develop traffic
accident predictive models. Based on their study they conclude that level of travel and
population density has a strong influence on urban accidents.

Reduction of the need for travel and locating major streets on the edge of an urban zone as well
as limiting population density could enhance traffic safety. Another conclusion of their study was
traffic accidents in an urban zone are exponentially proportional to intersection density and total

4
street length. In street net work planning, the reduction of both intersection density and total
street length could provide safer net works. Finally they stated that commercial frontages and the
location of public buildings have adversely influenced urban accidents.

Geedipally (2005) shows the effect of new pavement on traffic safety in Sweden. The study
investigated whether higher pavement road standards lead to a higher or lower accident rates.
The Poisson regression model was developed taking traffic accident as dependent variable and
the factors which the accident as independent variables result of his study shows that traffic
accidents increased by 12% after one year of resurfacing on all types of roads.

The rural roads showed much worst effect with an increase of 17% after resurfacing. The urban
roads showed a positive effect with a decrease in traffic accidents by 50% after resurfacing. The
national road safety coordination office (NRSCO) (2006) states that road traffic accident in
Ethiopia is a serious problem. According to the office the death rate is estimated to be 130 per
10,000 vehicles of those killed over half are pedestrians, of which 30% are children. In Ethiopia,
one out of five people injured die as a result of road accident which again is a very high figure
mainly due to poor situation of the emergency medical services.

Based on a five-year average records the office further states that, of the personal injury
accidents,81% are caused due to drivers error,5%due to vehicles defect,4%due to pedestrians
error,1% due to road defect and 9%due to other problems. These showed that the majority of
personal injuries of accidents are caused as a result of drivers' error. The study further shows that
professional drivers are involved in 88% of such accidents. On the other hand automobiles driver
have very good safety records with only 4% of the fatal accidents, which is equivalent to a rate
of 12 fatal accidents per 10,000 vehicles. Special purpose vehicles and motor bicycles cause 8%
of the fatal accidents.

Automobiles shore is around 43% of the total vehicle number of the country. Fatal accidents
types were 68% pedestrian strike, 13% over turn, 6% fall from vehicles, 3% animal or cart strike
and 10% for all other remaining crash types.

5
The underlying reasons for these accidents are noted to be:-
1) Improper behavior or low skill of driver (over speeding, not respecting pedestrian
priority----etc)
2) Poor vehicle technical conditions.
3) Animals and carts using the high ways.
4) Pedestrians not taking proper precautions.
5) Poor traffic law enforcement.
6) Poor emergency medical services and
7) Safety consideration not sufficiently given in roads developments

6
 CHAPTER THREE: DATA AND METHODOLOGY
3.1 Description of the study area

Bahir Dar is a town in north western Ethiopia is the capital town of the Amhara
region (Figure 1). Administratively, Bahir Dar is considered a special zone, placing
it mid way between Addis Ababa and Dire Dawa which are organized as chartered
cities(Astedader Akabibi, equivalent to Kilil), and town like Debre Markos and
Dessie, which are organized as Woredas. Bahir Dar is one of the leading tourist
destinations in Ethiopia, with a variety of attractions in the nearby Lake Tana and
Blue Nile River. The town is located approximately 578 km north-north west of
Addis Ababa, having altitude and longitude of 1136N3723E and elevation of 1840
meters above sea level. Based on the 2007 census conducted by the central
statistical agency of Ethiopia (CSA, 2007), this town has a total population of
221,991 an increase of 130.90% over the population recorded in the 1994 census of
whom 108,456 are men and 113,535 women (http//:www.google.com.free/wiki/pedia).

Bahir Dar

Figure 1 Location of the study area (Bahir Dar town)

7
3.2 Data Source

The data being considered here are obtained from Bahir Dar town traffic police
office. The data was classified by variable, which is the cause of road traffic
accident involving human injuries. Thus, this study use secondary data obtained
from seven police stations, traffic police office, Bahir Dar.

3.3 Data collection Method

Depending on the source, there are two types of data collection methods, i.e.,
primary and secondary methods. Primary method, consist of obtaining data or
information by any one of the following ways: direct personal interview, in direct
oral interview, information from correspondents, mailed questionnaire method or
questionnaires to be filled by enumerators data are collected by the investigator for
his/her own purpose. Secondary method, in most studies, the investigator finds it
impractical to collect first hand information and he makes use of data collected by
others. The data collected with the help of primary methods is usually published in
journals, newspapers, magazines, reports or books. It may also remain unpublished
by the investigator or some other office. Such data weather published, or
unpublished could be used by another investigator for similar completely different
purpose are called secondary data.

The data used in this research paper was collected from secondary source, which is
collected from Bahir -Dar town police stations, traffic office.

3.4 Variable considered in the research

The variables that we are use in our research are two kinds. Those are independent
and dependent variable. Independent variable, the cause supposed to be responsible

8
for bringing about change(s) in a phenomenon or situations. Our research
independent variables are age of the driver, sex of the driver, experience of the
drivers, types of the vehicle, road roughness, atmospheric condition, day of the
accident, light condition, place of accident, vehicle age and educational back
ground of the driver.

Dependant variables are the outcome of the change(s) brought about by changes in
an independent variable. Our research Dependant variable is types of road traffic
accident involving human injuries (death or injury).

3.5 Statistical method

3.5.1 Descriptive statistics

This part of statistics deals with only describing the data collected without going
any further; that is without attempting to infer(conclude) anything that goes beyond
the data themselves. This statistics used to measure central tendency and variation
of the data. And values graphical and diagrammatically presentations like bar
chart, pie chart without making conclusions.

3.5.2 Inferential statistics: Chi-square:χ2 distribution

Test of association (independence)

Whenever we have a cross tabulation of two (attribute) we may ask

 Are the two variables related to each other?
 Are the two variables independence to each other?
 In other word do the variables affect each other?
 The statistically model for inference required the null hypothesis to be
tested.

9
  In case or contingence tables (cross tabulation) of data, the null hypothesis
are that the two attribute (variable) are independent.

The statistics is the appropriate test statistics for test of independence and

the critical value is χ2α (v)

Where: v= (r-1) (c-1)
Oij=observed values
Eij= expected values
r=number of rows
c=number of columns
Test of hypothesis

For our case the possible null and alternative hypothesis can be
 Ho: there is no association between types road traffic accident and explanatory variables or
(they are independence).
 Ha: there is association between types of road traffic accident and the explanatory variables
or(they are dependent).
 If Ho is true, what we expect is, there should be little difference between the observed and
the expected value.
 If our probability, P > 0.05 implying that we do not reject the NULL hypothesis that no
relationship exist between type of road traffic accident and independent variable.

3.5.3 Logistic regression analysis

Logistic regression is part of a category of statistical models called generalized linear models.
This broad class of models includes ordinary regression and ANOVA, as well as multivariate
statistics such as ANCOVA and log linear regression. Logistic regression is one to predict
discrete outcome, such as group membership, from asset of variables that may be continuous,
discrete, dichotomous, a mix of any these. Generally, the dependent or response variable is
dichotomous, such as presence/absence or success/failure. Discriminate analysis is also used to

10
predict group membership with only two groups. However, discriminate analysis can only be
used with continuous independent variables .thus, in instances where the independent variables
are a categorical, or a mix of continuous and categorical, logistic regression is preferred.

3.5.4 The model

The basic aim of modeling is to derive a mathematical representation of relationship between an

observed response variable and number of explanatory variables, together with measure of the
inherent uncertainty of any such relationship. Statistical model can be constructed for response
variable are at best an approximation to the manner in which an observed variable depends on
other variable no spastically model can be claimed to represent truth and by the same token, no
one model can be termed the correct model. Some model will be more approximate than other.
But typically, for any set of data there will be a number of models.

The dependant variable in logistic regression is usually dichotomous, that is, the dependant
variable can take the value:

1 with a probability of success ϴ or

0 with the probability of failure 1- ϴ this type of variable called Bernoulli corm binomial 1
variable

The independent or predictor variable in logistic regression could be discrete, continuous or mix
of both. Logistic regression makes no assumption about the distribution of the independent
variables. They do not have to be normally distributed, linearly related or have equal variance
with each group in logistic regression relationship between the predictor and response variable
inverse logistic function of a vector xi, which include the constant and k explanatory variables,
(β +β x +β x + .... +.β x ) (β +β x +β x + ... +.β x )
i.e., ϴ(x) = ℮ 0 1 1i 2 2i k
ki /1+ ℮ 0 1 1i 2 2i ki k
Where: βo is the constant of the equation and βi’s coefficient of the predictor variables,
alternative forms of the logistic regression model

Log it [ϴ (x)] = log [ϴ (x)/1-ϴ (x)] = β0+β1x1i +β2x2i + . . . . . . +βkxki) =x’β

11
The coefficient can be interpreted as a change in the log odd associated with a one unit change in
corresponding independent variable or the odd increase multiplicatively by ℮β for every one unit
change increase in x. The ratio of success to failure ϴ[x]/ [1-ϴ(x)] is called the odd of success.

Through the x’β =log [ϴ[x]/1-ϴ(x)] = log it [ϴ(x)].the dependence of the success probability on
explanatory variables. The probability scale transformed from the range (0, 1) to (-∞, ∞) .This
transformation ensures that the fitted success probability will lies between 0 and 1. The goal of
logistic regression is to correctly predict the category of outcome for individual cases using the
most parsimonious model.

3.5.5 Fitting the logistic regression model

According to John E., the general method of estimation that leads to the least square function
under the linear regression model (when the error terms are normally distributed) is called
maximum likelihood.

It is a method that provides the foundation for our approach to estimation with the logistic
regression model .In every general sense the method of maximum likelihood yields values for
the unknown parameters which maximize the probability of obtaining the observed set of data. In
order to apply this must first construct a function called maximum likelihood function. The
maximum likelihood estimator of this parameter is chosen to be those values which maximize
this function. Thus, the resulting estimator are those which agree must closely with the observed

data.ϴ [xi] = ……………… (1)

Since the observations are assumed to be independence, the likelihood function is obtained as the
product of the term given in the above expression as follow.

L =β
The principle of maximum likelihood states that we use as our estimate of β the value which
maximize in equation (1).However, it is easier mathematically to work with the log of equation
(1).This expression the log likelihood is defined as

12
L[β] =ln (L (β)) =
To find the value of β that maximize L[β] we differentiate L[β] with respect to α and β set of
resulting equation to zero. These equations are follows:

=0 …………… (1)

=0 ………….. (2)
And these equations are called likelihood equation. For logistic regression the expression in (1)
and (2) are nonlinear in α and β, and thus require special method for their solution, which are
iterative in nature and were programmed into available logistic regression packages like SPSS.

3.5.6 Model building for logistic regression

3.5.6 Assessing the fit of the model for logistic regression

Once a model had been developed through the various steps indicated in the above sections, we
now would like to know how effective the model is in describing the outcome of the variable.
So, we need to assess the goodness of fit the model (Agersti,1996).The Hosmer and Lemeshow
test is the most commonly used measure of goodness of fit for categorical data (Hosmer and
Lemeshow,1989).So, in this study we used Hosmer and Lemeshow test to assess the goodness
of fit of model.

13
 CHAPTER FOUR: RESULT AND DISSCUTION

4.1 Descriptive statistics

Table 1 The results of categorical variables

Variable Category Frequency Valid percent
Sex M 127 99.2
F 1 0.8
Atmospheric Good 97 82.9
condition Rainy 20 17.1
Day of accident Weekend 19 16.5
Week day 96 83.5
Light condition Light 107 88.4
Dark 14 11.6
Educational Illiterate 3 2.5
background Elementary school 10 8.4
Junior secondary school 12 10.1
Secondary school 74 62.2
Above 20 16.8
Vehicle type Automobile 1 0.9
Taxi, Bajjaj, Minibus 57 50.4
Cargo 3 2.7
Bus 7 6.2
Others (truck, cart, bicycle, 45 39.8
etc.)
Time of event Morning 56 43.1
Afternoon 50 38.5
Evening 16 12.3
Night 8 6.2
Road roughness Asphalt 111 88.8
Non-asphalt 14 11.2
Place of accident Office 13 11.4
Residential center 48 42.1
Religious center 5 4.4
Recreation 17 14.9
School 8 7
Commercial center 2 1.8
Others 21 18.4
Type of accident Injuries 109 85.8
Death 18 14.2

14
Table 2 The results of continuous variable
Variable N Minimum Maximum Mean Std. deviation
Age of driver 129 14 67 28.08 8.239
Driving 62 0 20 4.0984 4.13246
experience
Vehicle age 34 0.08 11 3.4788 2.95456

Statistical table is a presentation of numbers in logical arrangement, with some explanation. Our
statistical table is discussed below. As Table 1 depicted, from the two types of road traffic
accident, injury were cover 85.8% and death covers 14.2%. From seven types of place of
accident, residential center were cover 42.1% among the whole. From two types of road
roughness, asphalt were cover 88.8% whereas, the rest is non asphalt road, which the accident
occurred. From four types of time of event, morning was cover 43.1% of the whole accident.
From five types of vehicle taxi, Bajaj and minibus were cover 50.4% from all vehicle type which
damage human. From five types of level of educational background, secondary school, were
covers 62.2% from all. From two types of light condition, light, were cover 88.4% and the rest
were dark. From two types of day of accident, week day were cover 83.5% and weekend were
cover the rest (the day which road traffic accident occurred). From two types of atmospheric
condition, good were covering 82.9 and rainy were cover 17.1% (weather conditions at which
road traffic accident were occurred).

When we see the distribution of sex of drivers who cause the accident, 99.2% of the total is
covered by male drivers, while the rest is for female. When we come to continuous variable
(Table 2), the minimum, maximum, mean and standard deviation for drivers’ age is 14, 67, 28,
and 8 respectively. The minimum, maximum, mean and standard deviations were 0.20, 4.0984
and 4.13246 respectively for driver’s experience. Finally, the minimum, maximum, mean and
standard deviation were cover 0.08, 11, 3.4788 and 2.95416 respectively for age of the vehicle.

15
4.2 Inferential statistics

Table 3 Chi square value

Variable Category Type of accident Chi- P value
Injury Death square
value
Sex F 100(1) 0(0) 0.168 0.682
M 85.6(101) 14.4(17)
Atmospheric Good 86.8(79) 13.2(12) 0.782 0.376
Condition Rainy 78.9(15) 21.1(4)
Light Condition Light 85.0(85) 15(15) 0.971
Dark 84.6(11) 15.4(2)
Educational back Illiterate 100(3) 0(0) 2.823 0.588
ground Elementary 75(6) 25(2)
Junior Sec. school 90.9(10) 9.1(1)
Secondary school 89.4(59) 10.6(7)
Above 80.0(16) 20(4)
Vehicle type Automobile 100.0(1) 0(0) 1.628 0.804
Taxi, Bajjaj, Mini bus 85.5(47) 14.5(8)
Cargo 66.7(2) 33.3(1)
Bus 75.0(3) 25(1)
Other 88.1(37) 11.9(5)
Time of event Morning 84.3()43 15.7(8) 0.548 0.908
After noon 89.1(41) 10.9(5)
Evening 87.5(14) 12.5(2)
Night 83.3(5) 16.7(1)
Road roughness Asphalt 85.3(87) 14.7(15) 0.424 0.515
Non asphalt 78.6(11) 21.4(3)
Place of accident Office 84.6(11) 15.4(2) 5.366 0.498
Residential 86.7(39) 13.3(6)
Commercial center 100(2) 0(0)
Religious center 80(4) 20(1)
Recreation 87.5(14) 12.5(2)
School 100.0(8) 0(0)
Other 70.0(14) 30(6)
Day of accident Weekend 100.0(18) 0(0) 3.574 0.049*
Week day 83.0(73) 17.0(15)
Age of driver <=26 96.7(58) 3.3(2) 10.891 0.001*
>=27 75.9(44) 24.1(14)
Experience <=6 88.9(40) 11.1(5) 5.288 0.021*
>=7 61.5(8) 38.5(5)
Vehicle age <=1 80(4) 20(1) 0.007 0.996
2-3 80.2(13) 18.8(3)
>=4 81.8(9) 18.2(2)

16
 Test of hypothesis
H0: there is no any association between the dependent and in dependent variable.
H1: there is association between the dependent and in dependent variable.
From the above table, we conclude that independent variable such that driver experience , age of
the driver and day of accident have a great impact on road traffic accident where as the rest
independent variables have less impact on road traffic accident involving human injuries in Bahir
Dar town.

4.3 Logistic regression analysis

Table 4 Logistic regression output

Variables in the Equation
Variable Category 95.0% C.I.for EXP(B)
B S.E. Wald df Sig. Exp(B) Lower Upper
Experience <=6 -1.804 .789 5.228 1.000 .022 .165 .035 .773
>=7
Constant -.336 .586 .330 1.000 .566 .714
a. Variable(s) entered on step 1: expire.

As shown above, Table 4, independent variable experience of the driver has significance effect
on road traffic accident involving human injuries in Bahir Dar town.
The equation is written in the form of
ln (p/1-p) = β0+β1x1
Where: β0 is constant
β1 is the parameter estimator
X1 experience of the driver which is greater than 7

ln (p/1-p) = -0.336-1.804X1
As can be seen from Table 4 experience of the driver has significantly related traffic accident.
The odds of road traffic accident of human injuries by exp (β 1)=0.165.That means road traffic
accident involving human injuries being affected by drivers experience greater or equal to 7 is
0.165 times for those drivers experience less or equal to 6 years experience. (Or, the probability
of the occurrence of road traffic accident, drivers experience >=7 years greater than <=6 year
experience by 17%).

17
4.4Assessment of Model Adequacy

The formal test applied to the model good fit of the model is Hosemer and Lemeshow test, the
result of this test suggests that model is in good fit, i.e. it is not significant at 5% level of
significance. Thus, all in all we can say that our model fits the data very well. The information
presented above is summarized in table……

18
CHAPTER FIVE: CONCLUSION AND RECOMMENDATION

5.1 Conclusion

According to our research drivers with driving experience more than 7 years are significantly
exposed to road traffic accidents causing death. On the other hand all other our independent
variables have no significance association with the influence of death/injury during road traffic
accident.

5.2 Recommendations

As our research shows that except experience of the driver all other independent variable have
less contribution to the occurrence of road traffic accident involving human injuries in bahir dar
town. Based on this result, in bahir dar town, drivers who have experience greater or equal to 7
years damage more people than that of less or equal to 6 years experience. As a result, Bahir Dar
town more experienced drivers are more careless to the occurrence of road traffic accident than
that of less experienced drivers.

Finally, we recommended that, the concerned body should take the following measure:-

 Try to teach more about road traffic accident and it’s consequence on human life.
 Try to teach more about the consequence of carelessness.
 Try to attain the day to day activities of more experienced drivers.
 A special punishment should be taken for more experienced drivers, who damage people.

19
 REFERENCE
1) Agresti A. (2007). An introduction to categorical data analysis (2nd edition)
2) Al-masaed et al., (2004).
3) CSA (2007). Central Statistical Authority, Addis Ababa, Ethiopia
4) Fitsum Zewdu (2004). Determinants of traffic fatalities and injuries in Addis Ababa. Journal
of the Ethiopian Statistical Association
5) Geedipally (2005).
6) http//:www.google.com.free/wiki/pedia /Bahir Dar/ (Internet accessed in February 2004)
7) John E. (). Frend`s mathematical statistics
8) Lauren et al., (2005).
9) NRSCO (2006). National Road Safety Coordination office, Addis Ababa, Ethiopia
10) Odero (2004).
11) Tewolde M. (2007). Analysis on traffic accidents involving human injuries in the case of
Addis Ababa. (Unpublished) (M. Sc. Thesis), Department of Statistics, Addis Ababa
University
12) Yayeh A. (2003). The extent, variation and cause of road traffic accidents in Bahir Dar.
(Unpublished) (M.A thesis), Department of Geography, Addis Ababa University
13) World Bank (2003).

20
 APPENDIX

A Table 1 Description of factors with the respecting coding

Factors Categories
Human factors
1 Sex of driver 0) Female
1) Male
2 Age of driver 0) <=26
1) >=27
3 Driving experience 0) <=6
1) >=7
4 Educational background of driver 0) Illiterate
1) Elementary school
2) Junior secondary school
3) Secondary school
4) Above

Vehicle factor
5 Vehicle type 0) Automobile
1) Taxi, Bajjaj, Minibus
2) Cargo
3) Bus
4) Other (truck, cart, etc...)
6 Vehicle age 0) <=3
1) >=4
Environmental factors
7 Atmospheric condition 0) Good
1) Rainy
8 Light condition 0) Light
1) Dark
9 Day of accident 0) Weekend
1) Week day
10 Time of event 0) Morning
1) Afternoon
2) Evening
3) Night
Road factor
11 Road roughness 0) Asphalt
1) Non-asphalt
12 Place of accident 0) Office
1) Residential center
2) Commercial center
3) Religious center
4) Recreation
5) School
6) Other category
13 Type of accident 0) Injuries
1) Death

21
A Table 2
Variables not in the Equation
Score df Sig.
Step 1 Variables dayac(1) .838 1 .360
Overall Statistics 2.184 2 .336

A Table 3
Variables not in the Equation
Score df Sig.
Step 1 Variables dayac(1) .838 1 .360
Overall Statistics 2.184 2 .336

22