Abstract
Most of web-based systems use a fashion-driven graphical user interface design which does not necessarily provide the readers with high reading performance of colour variations of text and background. Many studies addressed this problem but none of them succeeded in offering complete and conclusive results in form of reading performance table which could be used in practice. The aim of this paper is to find reasons for these incomplete results. In our research, we firstly analyzed different experiment designs described in the literature and proposals for further research. Consequently, we tried to find an improved design and carried out an experiment involving 270 students who tested 30 web-safe colour combinations. However, our experiment also did not reveal statistically significant differences in reading speed. Therefore the multidimensional scaling (MDS) method was performed to show that the speed of reading cannot be described as a one-dimensional problem.