Lag-sequential analysis is a method of assessing of patterns (what tends to
follow what?) in sequences of codes. The codes are typically for discrete
behaviors or states. The functions in this package read a stream of codes,
or a frequency transition matrix, and produce a variety of lag sequential
statistics, including transitional frequencies, expected transitional frequencies,
transitional probabilities, z values, adjusted residuals, Yule's Q values,
likelihood ratio tests of stationarity across time and homogeneity across groups
or segments, transformed kappas for unidirectional dependence, bidirectional
dependence, parallel and nonparallel dominance, and significance levels based on
both parametric and randomization tests. The methods are described in
Bakeman & Quera (2011)