Decoding the human factor through a multi-dimensional analysis of individual factors contributing to non-fatal crashes

Road traffic accidents, particularly non-fatal injuries, represent a significant global health burden, with human error identified as the primary contributing factor. This study aimed to investigate the association between individual factors and human errors among drivers. This cross-sectional study was in 2025–2026 on 150 taxi drivers selected through simple random sampling in southern Iran. Data were collected using the Driver Behavior Questionnaire (DBQ), assessing slips, deliberate violations, lapses, and unintentional violations, along with demographic and driving history variables. Data were analyzed using appropriate statistical methods (descriptive statistics, Spearman correlations, Mann–Whitney U tests, Kruskal–Wallis tests, and multiple linear regression). Participants’ mean age was 34.78 ± 9.27 years (92.7% male). Deliberate violations showed the highest mean score (0.87 ± 0.76), followed by slips (0.82 ± 0.70) and unintentional violations (0.82 ± 1.02), while lapses were the lowest (0.71 ± 0.63). Regression models explained substantial variance across all subscales (all P < 0.001). Education level was the strongest predictor; compared to drivers with a master’s degree, those with lower education reported significantly fewer errors (High School: β = − 18.22 to − 3.79, all P < 0.001; Diploma: β = − 18.46 to − 3.19, all P < 0.001; Bachelor: β =− 9.82 to − 2.59, all P < 0.001), indicating that higher education is associated with greater self‑reported aberrant behaviors. Traffic location showed that driving only in‑town ( β = − 25.15 to − 4.03, P < 0.001 for all subscales except lapses where β = − 4.85, P < 0.001) or only suburban ( β = − 18.89 to − 0.94, significant for slips and deliberate violations, P < 0.01) was associated with lower error scores, meaning that drivers who navigate mixed environments report the highest error frequencies. Female drivers reported significantly more errors than males on all four subscales (male coefficient negative: β = − 3.47 to − 1.22, P = 0.005 to < 0.001). Age of licensure was positively associated with all subscales ( β = 0.26–0.76, all P < 0.001), indicating that later licensure predicts more frequent errors. Accident frequency was negatively associated with slips ( β = − 0.37, P < 0.001), deliberate violations ( β = − 0.02, P < 0.001), and lapses ( β = − 0.13, P = 0.001), but not with unintentional violations ( β = − 0.03, P 0.110 ). The high prevalence of deliberate violations compared to lapses underscores the need for targeted interventions addressing both intentional rule‑breaking and automatic errors to reduce human errors and subsequent accidents.