CHAPTER 3: RESEARCH METHODOLOGY
3.2 Research data and its reliability
Various types of data related to the exposure, risk and loss of road traffic accidents are
employed in this study (Table 3.1). These data comprise the historical number of road
traffic casualties according to the types of road users, demographic data, number of
registered vehicles, socioeconomic status, infrastructure and weather data. The data used
in this study are obtained from five sources which include: (1) the Royal Malaysian Police
(PDRM), (2) Department of Statistics under the Prime Minister’s Office of Malaysia, (3)
Ministry of Transport, (4) Ministry of Works and (5) the World Bank. In terms of the
accuracy, the data are good as the data are published officially by the recognized agencies.
In fact, WHO estimations on the number of road traffic fatalities are similar with the
number reported by the police.
Table 3.1: Data and its source
Data Sources
Number of total road traffic fatalities, motorcyclist fatalities, motorcar driver fatalities, vulnerable road users fatalities, accidents due to alcohol, motor vehicle accidents, motorcycle accidents, motorcycle fatalities related to head injured
Royal Malaysian Police
Total population, male, female, urban, rural population, population age 15–24 years old, number of hospital beds, unemployment
Department of Statistics
Number of registered vehicles (total), registered motorcycles, registered buses
Ministry of Transport
Roads length Ministry of Works
GDP, road sector energy consumption the World Bank
There are a couple of constraints since this study deals primarily with historical data.
The first constraint is that the data published during the early years of the period of
that the available data are often published annually and it is generally more desirable to
use monthly data in time series analysis. Even though monthly data consist of a higher
number of observations, these data will elicit information regarding the variations in the
data series. The explanatory variables that are expected to influence the number and rate
of road traffic casualties are also published annually. However, a number of researchers
have conducted road safety research based on annual data series. This is the typically the
case for low-income and medium-income countries.
In Malaysia, road traffic safety data are collected, recorded and published by PDRM.
A standardized form named ‘Pol. 27’ is used by the police to collect data related to road traffic accidents at accident scenes. The data consist of the following information: time
and type of accident, the number of vehicles involved, road environment, injuries of
vehicle occupants (if any) as well as the location of the accident. The reports published
include the data and statistics on road traffic safety. However, the data published from the
late 1960s to 1980 are only provided for Peninsular (West) Malaysia and the data for East
Malaysia (Sabah and Sarawak) are provided beginning from 1981 onwards.
The road safety data for West and East Malaysia are available separately until 1988
and have been integrated since then. In the early years, the reports only provide the
number of vehicles involved in accidents, the main causes of the accidents, the number
of casualties, the accidents and casualties according to the type of road environments,
demographic data of road users, the type of vehicles commonly involved in accidents,
analysis of the investigations, and revenue from traffic cases. Generic data such as
population, the number of registered vehicles and length of roads are also included.
Statistics on road safety such as the rate of accidents and the rate of casualties are
generated based on these data. However, the reports in recent years include details on
Attempts have been made by road safety authorities to improve the statistics on road
safety in the reports. One such example is the reports published in 1978, in which the
number of accidents and casualties for motorcyclists are more detailed. The rate of
accidents and rate of casualties are also provided in such reports since year 1981. In 1988,
charts are made available in the reports. In 1992, more comprehensive data and statistics
on motorcyclist and pedestrian accidents are included. In 1998, accident data are divided
into two main parts: Part A and Part B. Part A contains statistics on general road accidents
derived from the monthly returns sent by all district police stations (traffic branches). Part
B contains records on road accidents and statistics generated by a computer programme
known as Microcomputer Accident Analysis Package (MAAP). The data used in this
software are obtained from the revised standard road accident information form, i.e. Pol.
27 Pin. 1/91. The definitions of the terms used in the reports are provided along with notes
to users.
The demographic, socioeconomic and weather historical data are extracted from the
annual reports published by the Department of Statistics, Prime Minister’s Office of Malaysia as well as from the online database of World Bank (2014). The number of
registered vehicles are obtained from the reports published by the Road Transport
Department, Ministry of Transport whereas the length of the roads are obtained from
reports published by the Highway Planning Unit under the Ministry of Works. In this
study, it is generally preferable to use future data estimated by national and international
official agencies – however, if these data are not available, such data will be predicted
using the models developed in this study.
Furthermore, during the data analysis in this study, the detailed of road traffic
casualties data are available up to year 2012. For this reason, the descriptive statistics
models to predict the fatalities, number of fatalities from 1981 to 2012 are used to estimate
the models. Whilst, to evaluate the forecasts, number of fatalities in 2013 and 2014 are
used.