Abstract
Sequence analysis, based on the mathematical Markov model, is useful for analyzing interdependent data. Sequences can be recorded as event, state or timed event data, or by coding fixed time intervals. The probabilities of transitions in the data stream are used for modeling of the change processes. The first-order Markov model assumes that every event is influenced only by the immediately preceding event. The second-order model assumes influence from the two preceding events, etc. The models are characterized by their stationarity, the degree to which the same rules apply throughout the sequence. Statistical methods are introduced for testing of these characteristics as well as the homogeneity of sequences in a target sample. Lag sequence analysis is introduced as a method for simplifying analysis of higher order sequences. Advanced models, applications and software are introduced.