Experimental research methodology and statistics insights
Published in Human-Robot Interaction: Evaluation Methods and Their Standardization. Springer Series on Bio- and Neurosystems, 2020
Abstract
Methodological and statistical misunderstandings are common within empirical studies performed in the field of Human Robot Interaction (HRI). The current chapter is aimed to briefly introduce basic research methods concepts required for running robust HRI experimental studies. In addition, it is oriented to provide a conceptual perspective to the discussion regarding normality assumption violation, and describes a nonparametric alternative for complex experimental designs when such assumption cannot be fulfilled. It is concluded that HRI researchers should hold internal validity of studies as a priority and foster the use of within-subjects designs. Furthermore, the described statistical procedure is an alternative to analyze experimental data in multifactorial designs when normality assumptions are not fulfilled and may be held as a suggested practice within the field of HRI.
Recommended citation: Paredes Venero, R., Davila, A. (2020). Experimental Research Methodology and Statistics Insights. In: Jost, C., et al. Human-Robot Interaction. Springer Series on Bio- and Neurosystems, vol 12. Springer, Cham.
Download Paper