Application of Response Surface Methodology to Shape Discrimination
Abstract
The effects of systematically manipulating five physical features of random, closed figures were determined. Subjects were asked to select one of four figures presented as being least like the others. The five features investigated were: number of edges, area, radial length variance, edge length variance, and orientation of the major axis. A five-factor, five-level, central-composite response surface design was employed. The resulting regression equation indicated that four features were reliable in their prediction of performance. Additionally several interaction effects were found. The applicability of these data for computer simulation is discussed.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1974
- Accession Number
- AD0785367
Entities
People
- Lawrence A. Scanlan
Organizations
- University of Illinois Urbana–Champaign