In this research, the road safety data from 1991-2022 were used, to get the longitudinal data set without any breaks in series to explore the _change_ of road safety across countries. Data was processed using R-software. Data was transformed and scaled, irrelevant data was modified and missing data was imputed for the better predictive analysis. The datasets of Eurostat, Estonian Road Authority and Statistics Estonia were used (Eurostat, 2022; Statistics Estonia, 2023). Data of some countries was received from European Transport Safety Council (ETSC) (ETSC, 2022). The investigated variables were the number of fatal accidents per day, number of fatalities, size of the country and the year.

Road safety can be evaluated by the number of (1) road fatalities and (2) fatal road accidents, both per million of inhabitants. Road fatalities means any person killed immediately or dying within 30 days as a result of a road accident. Fatal accident is an accident resulting in a person killed (UNECE Statistics of Road Traffic Accidents)

Regarding the relationship between the road fatalities and fatal road accidents, there is some evidence that the ratio is around 1.07 (_Reported Road Casualties Great Britain, Annual Report_, 2020.; Statistics Estonia, 2022). As the rate takes into account the size of the country, thus fatality rate is rather small when the country's population (in the denominator) is big. Furthermore, road safety change should also be statistically significant (T_esting-for-Statistically-Significant-Changes_; UK Public Data, 2021).

This study presents an empirical analysis of road traffic safety (RTS) in a small country—Estonia. The objective is to predict the occurrence of accidents involving fatalities using the Poisson regression model, with particular attention to how these accidents vary by weekday and month. This approach allows for an assessment of seasonal and weekly patterns in road traffic safety.

The dataset comprises 3,287 daily observations of road traffic accidents in Estonia between 2010 and 2018 (; ; ). Since accident data are count-based (e.g., ranging from 0 to 5 accidents with fatalities per day), Poisson regression is employed to estimate the daily number of accidents with fatalities. Given that the data conform to a Poisson distribution—where the mean equals the variance—the SPSS Generalized Linear Model (GLM) function is used. A chi-squared test is conducted to compare observed and expected frequencies.

The analysis draws on data from Eurostat, the Estonian Road Authority, Statistics Estonia, and the European Transport Safety Council (; ). The average ratio of road fatalities to road accidents with fatalities is approximately 1.07 (; ; ).

In this paper, we apply two complementary statistical methods: Factor Analysis and Path analysis.

The modelling strategy for the prediction of is to use the path analysis. The path analysis is a technique to analyze causal relationships that occur in multiple regression (Retherford 1993). The model is depicted in the form of circle and arrow images where a single arrow shows as a cause. Regression is imposed on each variable in a model as dependent variable while the other as the cause. Regression weighting is predicted in a model compared to the observed correlation matrix for all variables and also the calculation of statistical goodness of test. The path model is a diagram that links the network of relationships of several variables placed in sequence to be studied in the research. The conventional term is the association between independent, intervening and dependent variables. The pattern of relationships in path analysis is indicated by using arrows. The single arrows indicate a causal relationship between the independent variables which in the path analysis is then referred to as exogenous variables variables with one or more dependent variables which is in the path analysis referred to as the endogenous variable. However, just noting a change isn't enough and therefore be sure that the observed change is likely due to a actual effect or just the result of random variation.

**Path Analysis:**

We use path analysis to assess whether fatalities, accidents with fatalities, and total accident count can represent a latent RTS variable.

Recent studies highlight the usefulness of factor analysis in RTS research, as it allows for the modelling of relationships between observed variables and latent constructs (; ; ; ). Structural equation modelling (SEM) has been proposed to further explain the underlying mechanisms of RTS (; ).

In our factor analysis model, RTS is treated as a latent variable. The observed variables included are:

1. road accidents with fatalities,    
2. number of fatalities,    
3. road accidents involving both fatalities and injuries.    

**SEVERITY**
![[Pasted image 20260228123122.png]]
_Factor analysis model for evaluating RTS_
Source: Transportation Data; using SPSS Amos

_Stadardized regression weights for Figure [1](#fig:image1)_

| **Variable**                   | **Estimate** | **P-value** |
| ------------------------------ | ------------ | ----------- |
| Road deaths per day            | 0.943        | < 0.001     |
| Road accidents with deaths/day | 1.009        | < 0.001     |
| Road accidents per day         | 0.251        | < 0.001     |
According to the model, an increase of one standard deviation in RTS leads to a 0.943 standard deviation increase in road deaths per day, indicating a strong positive influence of RTS on fatality rates. Likewise, an increase of one standard deviation in RTS results in a 0.251 standard deviation increase in the number of road accidents per day, suggesting a weak positive relationship. This implies that RTS has a stronger association with accident severity than with accident frequency.

Overall, the results emphasize that RTS is significantly impacted by severe (fatal) accidents, reinforcing the notion that accident severity is a more critical component than accident frequency when evaluating traffic safety.