1.3 Research Methodology
The research methodology in this thesis is a combination of experimentation, literature analysis and technical development of hardware and software. It is very difficult to arrive at decisive conclusions with each of the three methods alone. Experimentation is by far the best accepted method in human factors research. However, meaningful and significant experimental results depend not only on the execution of the experiment but also on the design of the experimental task and paradigm and, most importantly, on the formulation of the experimental hypotheses, which in turn relies on the theoretical analysis of the issues being investigated and the analysis of the related literature. Similarly, theoretical analysis alone is not sufficient. Due to the complex nature of human behaviour, conclusions drawn from a logical reasoning process based on first principles can hardly be accepted without experimental validation. Another important method in human interface research is creatively designing "artifacts" that are embodiments of ideas and theories (see Carroll and Kellogg, 1989) , but again one can not be sure about the validity of the resulting designs without experimentation.
This thesis takes a pragmatic approach by combining all three
methods. Each chapter begins with a first-principle analysis,
followed by a literature review. Interfaces that embody the issues
being investigated are then designed and implemented with available
state-of-the-art technologies. Experiments are then carried out
for carefully designed representative tasks. Conclusions are drawn
based upon rigorous statistical analysis of the experimental results.
Finally, these results are discussed in relation to the literature.