A Hierarchical Negative Binomial Regression Analysis of Turbulence-Related Injury Counts in U.S. Part 121 Airline Operations

Turbulence remains one of the leading causes of in-flight injuries in commercial aviation, yet relatively little is known about the factors associated with differences in injury outcomes across turbulence encounter events. This study examined the environmental, operational, and human factors associated with turbulence-related injury counts in U.S. Part 121 operations using hierarchical negative binomial regression. A total of 184 turbulence-related injury events reported between 1995 and 2026 were analyzed, with predictor variables entered sequentially to evaluate the incremental contribution of each factor domain. Each retained predictor domain significantly improved model fit, and the final model achieved a Generalized R2 of 0.257. Wake turbulence category, phase of flight, and night duty emerged as significant predictors. Compared with large-category aircraft, heavy-category aircraft were associated with lower expected injury counts, whereas turbulence encountered during descent and during night duty was associated with higher expected injury counts. These findings suggest that turbulence-related injury outcomes are associated with environmental, operational, and human factors and provide empirical evidence to inform assessment of turbulence-related injury counts and future safety management strategies.