Anthropometric accommodation percentage comparison for univariate and multivariate representative Mexican hand models

Area of Science:

• Ergonomics and Human Factors
• Anthropometry
• Biomechanical Engineering

Background:

• Awkward work postures negatively impact productivity and increase musculoskeletal disorder risk.
• Incorporating anthropometric data into product and workplace design mitigates these risks.
• Traditional univariate-percentile approaches in anthropometric analysis have limitations compared to multivariate methods.

Purpose of the Study:

• To develop univariate and multivariate hand models using four anthropometric dimensions.
• To compare the accommodation percentages of univariate and multivariate models.
• To assess the effectiveness of different anthropometric modeling techniques for ergonomic design.

Main Methods:

• Developed univariate percentile models for male and female populations (5th and 95th percentiles).
• Applied multivariate approaches (2D Principal Component Analysis and Archetypal Analysis) to the central 90% of populations.
• Calculated accommodation percentages and performed statistical analyses (goodness-of-fit, McNemar's tests).

Main Results:

• Principal Component Analysis yielded eight human hand models.
• Archetypal Analysis produced varying numbers of models (2-8 for males, 2-8 for females) based on archetype selection.
• Accommodation levels varied, with PCA models showing higher accommodation in males, and multivariate Archetypal Analysis (k=8) in females.

Conclusions:

• Multivariate models, particularly Principal Component Analysis, demonstrated higher accommodation than univariate models for the male population.
• For the female population, multivariate Archetypal Analysis models (k=8) achieved the highest accommodation percentage.
• These findings suggest multivariate approaches offer superior anthropometric accommodation for ergonomic design.