Multidimensional approach of objective difficulty in an aiming task

Abstract

This article deals with a new method to classify informational motor tasks, through a multidimensional approach of difficulty. The following experiment is used: 345 subjects are asked to point at computer-managed mobile targets. The 3 selected target parameters are: velocity, area and spatial uncertainty. Independence and numerical scalability are the reasons for this choice. As performance does not sump up difficulty, other dependent variables have been added: number of attempts, reaction time and motor time. The use of state-of-the-art multidimensional data analysis (HAC, CFA) makes it possible either to aggregate tasks with identical profile, or to differentiate them according to each descriptor's weight on dependent variables. Above all, the study specifies the contribution of each variable in the a posteriori definition of the difficulty by the researcher.>
KEY WORDS: Difficulty, data processing, motor task, classification, data analysis.