Forecasting the Fatality Rate of Traffic Accidents in Jordan: Applications of Time-Series, Curve Estimation, and Multiple Linear Regression Models

Edries, Belal; Alomari, Ahmad H. · 2022 · OpenAlex-citations

DOI: 10.25103/jestr.156.09

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the critical public health and economic challenge of traffic fatalities in Jordan, aiming to forecast future fatality rates to inform safety policies. Motivated by the high cost of accidents—estimated at 2–3% of Jordan’s GDP—and the lack of comprehensive long-term trend analysis, the researchers analyzed 39 years of data (1981–2020). The objective was to evaluate the performance of three statistical modeling approaches: time-series analysis, curve estimation, and multiple linear regression, to predict the annual fatality rate per 100,000 population for the subsequent decade. The methodology utilized data from the Jordan Public Security Directorate and the World Bank, covering traffic accidents, population growth, gross domestic product (GDP), registered vehicles, and road network length. Three distinct modeling strategies were employed. First, curve estimation tested ten distribution models (including linear, logarithmic, and cubic) against time as the independent variable. Second, multiple linear regression analyzed the impact of specific socioeconomic and infrastructure variables, such as road length, injury rates, accident rates, and vehicle registration density, on the fatality rate. Third, time-series analysis applied double exponential smoothing to capture trends without seasonal patterns. Model accuracy was evaluated using the mean absolute percentage error (MAPE). The results revealed distinct historical trends, including significant population spikes in 1990–1991 and 2015 due to regional conflicts, which negatively impacted GDP per capita. Among the curve estimation models, the cubic model performed best, capturing 79.4% of the variance. Multiple linear regression models explained over 75% of the variance and identified that increasing the length of the road network significantly reduces the fatality rate, whereas higher rates of registered vehicles and injuries increase it. When comparing predictive accuracy, the time-series model (double exponential smoothing) achieved the lowest MAPE of 11.045%, outperforming the multiple linear regression models (MAPEs of 12.621% and 12.770%) and the cubic curve estimation model (MAPE of 13.264%). Consequently, the time-series model was selected for forecasting, predicting a continued damping trend in fatality rates through 2030. The significance of this research lies in providing a validated forecasting tool for Jordanian decision-makers and engineers. By identifying the time-series approach as the most accurate predictor and highlighting the protective effect of expanded road networks, the study offers evidence-based insights for infrastructure planning. The findings underscore the importance of monitoring socioeconomic variables alongside traffic data to effectively reduce the human and financial burden of road accidents in developing nations.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• fatality injury trends
• comparative international
• incidence prevalence

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: crash risk outcomes