Approach 3 Based on theories for each indicator

Nearly all attempts to examine human factors found larger behaviour influences. Most of the accidents resulted from the violation of traffic laws. In one study carried out in Indiana State in USA presented in (Oppe, 1993), the frequencies of cause factors were calculated. The results have shown that in 70.7% of accidents the causes were human, 12,4% environmental and 4.5% vehicles. Furthermore, the measure of Traffic Risk is widely accepted in international comparisons and it deals with the fatalities rates per vehicles, per vehicles-km or per person-km. 

Again the sum of all of the weights should be 1. The results of RSDI are presented in appendix 2. The RSDI ranges between 0 and 100. The higher values indicate higher levels of safety in the country. The lower values indicate the worst performance in a country in terms of road safety level. The table also indicates that the RSDI level is high for Sweden, Singapore and Brunei. The opposite picture is the case for the countries Vietnam, Cambodia and Laos as they obtained a lower level of RSDI. 

The data of indicators is first normalised to zero, mean and unit variance, which wiU ensure that all measurements have equal weights in the analysis by using the formula  

The general forms for the created principal components in the linear combination of the indicators are (Manly, 1994)  

Indicators Eigenvalues Proportion of total Variance Cumulative variance First Eigenvectors (coefficients) PCI Second Eigenvectors (coefficients) PC2 

---