A case study in the design of Laser Eye Protection (LEP) spectacles is presented and discussed toillustrate the necessity of careful analytic thought in the variable selection process. PrincipleComponent Analysis (PCA) augments but cannot replace careful and informed reasoning aboutvariable selection. With this LEP example, we demonstrated that people who understand the designcriteria were better at selecting critical variables than PCA alone. Had the variables selected throughPCA analysis been employed without further consideration, the resulting design would have failed tofit disproportionate numbers of African-American and Asian crew members, resulting in an unintendedracial bias in fit. This paper demonstrates the process for selecting 3-D cases and emphasizes theimportance of using statistics in the decision making process, rather than relying upon statistics toprovide solutions. This project also demonstrates the flexibility provided by 3-D anthropometry datato be able to define a new, non-traditional variable that would not exist in 2-D data. The main purposeof this paper is to illustrate the importance of input from end-users and designers when deciding whichvariables are most important to use to select 3-D cases.
展开▼