Technical note: Using Johnson distributions to model trunk kinematics

Abstract

As we seek to develop high fidelity human simulation models for ergonomic applications, the characterisation of the variability in human performance is needed. This technical note describes a method for generating probability density functions (PDFs) for one performance characteristic: trunk kinematics. A PDF from the Johnson family of distributions is defined by four parameters (γ, ξ, δ and λ) and can represent a variety of distributions. In this study, previously published trunk kinematic data were fit to Johnson distributions and regression equations for each of the four parameters were created as a function of starting lift height. Using regression coefficients and Monte Carlo simulation, PDFs for novel lifting conditions were generated. These predicted PDFs were compared with histograms of empirical data collected from a new group of ten lifters performing lifts in these novel conditions. A Kolmogorov–Smirnov goodness of fit test was performed to assess the quality of the fit. Seven of the predicted distributions of these kinematic variables were found to be a good fit with the novel empirical data.

Keywords:

• Johnson distribution
• probability density function
• repetitive lifting tasks
• Kolmogorov–Smirnov
• predictive distribution

Disclosure statement

The authors have no conflicts of interest to disclose.