Using cognitive architectures to analyse the usability of human-computer interfaces is an extensively investigated strategy. A particularly powerful way to perform such analysis is through syndetic modelling, where both the interface and the chosen cognitive model are described in the same specification framework; allowing the combined behaviour of the two to be analysed. This paper proposes LOTOS as a syndetic modelling language. We highlight four reasons why syndetic modelling is so difficult

more »

... and show how the LOTOS notation addresses each of these four reasons. 1