Assessing the Role of Human Factors in Traffic Accident Causation

This study employed a quantitative cross-sectional design to investigate the effect of driver behavior on traffic safety in Al-Anbar Governorate. Data were collected through a structured survey distributed to 900 drivers across various cities and districts in the province. The sample was predominantly male and included a wide age range and occupational diversity, such as students, government employees, and private-sector workers.

The questionnaire focused on 20 behavioral variables (e.g., speeding, mobile phone use, sudden braking) and key demographic factors (e.g., age, sex, occupation, vehicle ownership). After data cleaning and preparation, analysis was conducted using SPSS.

The statistical analysis included:

• Multiple Linear Regression to examine the relationship between driver behaviors and the continuous outcome (number of accidents),
• Ordinal Logistic Regression, an advanced statistical technique suitable for modeling an ordinal dependent variable (categorized number of accidents into four severity levels).

These two regression methods provided complementary insights into how individual behaviors and demographics contribute to accident risk, both in frequency and severity. Tests for normality (skewness, kurtosis, and histograms) and multicollinearity checks were also conducted to ensure equation validity.

Ethical approval was obtained from relevant academic and local authorities. All participants gave informed consent, and data were collected anonymously to ensure privacy and candid responses.

4.1 Overview of key findings

The statistical analysis revealed several significant demographic and behavioral factors that influence the frequency of traffic accidents. Both multiple linear regression and ordinal logistic regression equation were employed to ensure robust validation of the findings.

The ordinal logistic regression model identified age, vehicle ownership, and years of driving experience as significant demographic predictors. Among the behavioral factors, driving outside the speed limit, using a mobile phone while driving, intentional obstruction, chasing another vehicle, driving against traffic, and passing from the right showed statistically significant associations with accident frequency.

The linear regression model similarly confirmed the significance of several factors, including age, years of driving experience, place of work or study, sex, occupation, and behavioral factors such as driving outside the speed limit, using a mobile phone while driving, sudden braking, passing from the right, and responding when overtaken.

Notably, the ordinal logistic regression model demonstrated exceptionally high explanatory power (Nagelkerke R² = 0.944), while the linear regression model yielded a similarly strong adjusted R² of 0.984. These findings underscore the consistency and reliability of the identified risk factors and reinforce the value of combining multiple analytical methods to capture different dimensions of traffic accident causation.

4.2 Interpretation of significant predictors

The regression analyses identified several variables with statistically significant associations with the frequency of traffic accidents. Among the demographic predictors, age and years of driving experience consistently showed significant effects across both the linear and ordinal logistic regression equation. Younger drivers were found to be at greater risk of accidents, aligning with established traffic safety literature that links youth with inexperience, impulsivity, and a tendency toward risk-taking. Conversely, increased driving experience was associated with higher accident frequencies in the ordinal model—likely reflecting greater cumulative exposure over time.

Additionally, vehicle ownership was significant in the ordinal model (p = 0.024), with a negative estimate, suggesting that those who own their vehicles may experience fewer severe accident outcomes—potentially due to greater personal responsibility or vehicle familiarity.

In terms of behavioral predictors, driving outside the speed limits emerged as one of the most consistent and impactful contributors to accident occurrence, achieving statistical significance in both models. This supports the well-documented role of speeding in increasing crash risk.

Intentional obstruction was another major predictor, particularly in the ordinal logistic model, where it exhibited a high positive estimate (Estimate = 315.92, p < 0.001). Such aggressive behaviors—like deliberately blocking other vehicles—disrupt traffic flow and heighten collision risks.

Using a mobile phone while driving was statistically significant in both the linear (p = 0.012) and ordinal logistic regression models (Estimate = -10.36, p = 0.023), suggesting that mobile phone use elevates accident risk and increases the likelihood of falling into more severe accident categories. This likely reflects increased cognitive distraction and delayed reaction times under complex traffic conditions.

Other significant behavioral predictors included passing from the right (Estimate = -5.59, p = 0.002), chasing a vehicle (Estimate = 3.17, p = 0.003), and driving against traffic (Estimate = -72.90, p < 0.001), all of which represent aggressive or non-compliant behaviors that destabilize driving environments and increase collision probability.

In the linear regression model specifically, sudden braking (p < 0.001) and responding when overtaken (p < 0.001) were also found to be significant contributors. These findings emphasize the role of impulsive or defensive maneuvers in triggering accidents, particularly in urban driving contexts where vehicle spacing is limited.

Overall, the convergence of significant predictors across both regression models—particularly those related to aggression, distraction, and speed—reinforces their central role in shaping accident risk. These findings validate the robustness of the modeling approach and provide actionable insights for targeted interventions in driver behavior and enforcement policies.

4.3 Interpretation of non-significant predictors

While many predictors demonstrated strong statistical significance, several behavioral and demographic variables did not show a significant association with traffic accident frequency in either regression model. Specifically, using the signal when changing lanes, speaking loudly when provoked, and abrupt lane shifting from the far right to the far left failed to reach statistical significance. Although these behaviors are widely regarded as unsafe, their infrequent occurrence or limited variability in responses may have reduced their statistical impact. For instance, if most participants reported rarely or never engaging in these behaviors, the dataset would lack the distributional diversity necessary to detect meaningful correlations.

Additionally, slow driving on the left side did not show statistically significant effects in either model, suggesting that its impact may be context-dependent or mediated by other driving conditions not captured in the analysis.

This finding contrasts with previous research. For example, studies in Iran and India [13, 14] have identified distraction and verbal aggression as contributing factors to accidents, while studies in Japan and the UAE [5, 9] emphasized the role of situational behaviors—such as improper signaling—in crash risk. The lack of significance in this study may be due to contextual differences, such as stronger enforcement of specific behaviors, cultural norms around driving courtesy, or underreporting due to social desirability bias.

From a demographic standpoint, occupation did not demonstrate statistical significance in the ordinal logistic regression model. This suggests that job type alone may not directly predict accident involvement after controlling for other risk factors. Nevertheless, such variables may still play an indirect role through exposure frequency or type of driving (e.g., professional vs. personal), which were not modeled in detail.

In summary, the absence of statistical significance does not imply practical irrelevance. It highlights the complex and context-dependent nature of traffic accident causation, where some behaviors may only become statistically relevant under specific sample conditions or in interaction with other risk factors. This underscores the importance of interpreting quantitative findings within the broader behavioral and cultural context of the study population.

4.4 Policy and practical implications

The findings of this study have several important implications for traffic safety policies and intervention strategies, particularly in regions with similar driving environments to Al-Anbar Governorate. The identification of key behavioral predictors—such as speeding, intentional obstruction, using a mobile phone while driving, passing from the right, sudden braking (as identified in the linear regression model), and chasing other vehicles—underscores the urgent need for targeted awareness campaigns aimed at reducing these high-risk behaviors. Educational initiatives should highlight the dangers of speeding, sudden braking, and distracted driving, especially among younger drivers, who demonstrated elevated accident risk.

Enforcement also plays a crucial role. The statistical significance of aggressive and impulsive behaviors—such as intentional obstruction, overtaking from the wrong side, and driving against traffic—suggests that stricter application of traffic laws, including fines and license penalties, could help reduce accident frequency. Moreover, integrating behavioral assessments into driver licensing systems and renewal programs may help identify high-risk individuals and allow for targeted corrective training.

Urban and traffic planners should also integrate behavioral evidence into infrastructure design. For instance, the significance of passing from the right and sudden braking (in the linear model) highlights the need for improved lane discipline enforcement, clearer road markings, and adequate vehicle spacing policies to reduce rear-end collisions in accident-prone zones.

In summary, the study supports a multifaceted approach that combines education, enforcement, and engineering to address behavioral causes of accidents. The findings offer actionable guidance for policymakers and transportation authorities seeking evidence-based strategies to enhance road safety and reduce accident-related injuries and fatalities.

4.5 Comparison with previous studies

The findings of this study are largely consistent with existing international literature on driver behavior and accident risk. The significant effect of speeding aligns with multiple prior studies that identified excessive speed as a major contributor to traffic accidents, including research conducted in the United States [1], Saudi Arabia [10], and Iran [13]. Similarly, the role of aggressive behaviors such as intentional obstruction echoes findings from France [3] and China [6], where traits like hostility and competitive driving were linked to increased accident severity.

The observation that younger drivers are more accident-prone is supported by studies in Florida and Kuwait [1, 8], both of which highlighted the vulnerability of younger age groups due to inexperience, impulsivity, and peer influence. Conversely, the positive association between driving experience and accident frequency in the ordinal regression equation may reflect cumulative exposure over time—a pattern also observed in studies from North Cyprus and Ghana [11, 18].

Mobile phone use was statistically significant in both regression equation in this study, reflecting its growing impact on driver distraction. This finding aligns with results from India and Iran [13, 14], where phone use and other distraction-related behaviors were strongly associated with accident involvement, particularly in models sensitive to accident severity or driver subgroups.

Some behaviors, such as using turn signals, slow driving on the left, and abrupt lane shifting, were not statistically significant. While these results differ from expectations, similar patterns were documented in studies from the UAE and Japan [5, 9], where the effect of safety-related behaviors varied based on infrastructure design, driving culture, and enforcement levels. The lack of significance in this study may also be due to limited behavioral variation within the sample or underreporting due to social desirability bias.

Overall, the alignment with previous literature reinforces the external validity of the study, while divergences emphasize the importance of context-sensitive interpretations in traffic safety research, especially in culturally distinct and administratively diverse regions like Al-Anbar Governorate.