Studying the impact of contextual factors on language barriers in maritime safety: a structural equation modeling approach

Communication failures, particularly language barriers, are critical contributors to maritime accidents, yet their impact remains underexplored. This study quantitatively assesses the effect of language and communication barriers on maritime accidents, investigating the interactions between human, operational, and organizational factors. Data from 515 maritime accident reports (2012–2022) was processed using Natural Language Processing (NLP) techniques, extracting observed variables including language proficiency, accent issues, cognitive load, and navigational complexity. Factor analysis and Structural Equation Modeling (SEM) were used to identify latent variables influencing accident risk. SEM was employed to analyze the relationships between latent variables, human factors, operational context, and organizational conditions, and the observed. Results show that human factors, especially stress and fatigue, significantly impact accident causation by compromising communication and navigation safety. Language proficiency (Low) (λ = 0.91, SE = 0.03, z = 30.33) was revealed as the most critical factor; however, operational aspects like navigational complexity (λ = 0.93, SE = 0.04, z = 23.25), time pressure, and organizational constraints exacerbate communication breakdowns and increase accident likelihood. The SEM (CB-SEM (Covariance-Based-SEM), WLSMV (Weighted Least Squares Mean and Variance Adjusted)) demonstrated strong explanatory power (R2 = 0.845) and excellent fit (RMSEA = 0.024, SRMR = 0.012, CFI = 0.962, TLI = 0.971). The study highlights the need for improved training, communication protocols, and language proficiency to reduce maritime accident risks.