Evaluating GUIs

R, as a GNU project, offers the opportunity to develop and test various different GUIs (see the Projects section). This is perhaps a fantastic opportunity to study and experiment the impact of such or such GUI feature on the productivity in data analysis for research, teaching, etc. However, evaluation of GUIs benefits in data analysis software is mostly intuitive and subjective. It depends upon the users. It is also much driven by personal feelings. There is a lack of appropriate metrics and of studies to produce experimental evidences of the efficiency of different GUIs, or of GUI versus CLI. This is a difficult field of investigation because usability is a fuzzy and complex concept that does not just summarize in one figure. We would like to encourage teachers and scientists to experiment with the various R GUIs in development, or to design new ones. We will eventually publish their results here.

A few useful readings in GUI design and evaluation:

 

Last update: 25 May 2004, by Philippe Grosjean.