Analysing the patterning of a sequence of discrete behavioural events

Existing techniques for analysing the organization of behavioural events are discussed. Cluster analysis, mutual replaceability, ‘melody’ detection and Markov analyses are examined in order to discover the validity and usefulness of the concepts searched for by each technique, and the mathematical and statistical background of the techniques is scrutinized. Some new techniques, designed to search for and detect possible important sequential organization patterns, are described in the light of the previous criticisms. Of these techniques, the Pre-post-state Histogram (PPSH)—a multi-order Markov graphical analysis—would appear to be the most important for future work, especially when used in conjunction with the other techniques described.