Matthew’s and Luke’s use of Mark: hidden Markov models
As already outlined in Section 4.1, we shall now assume that behind the observed binary series (xt) and (yt) for Matthew’s and Luke’s verbal agreements with Mark there lie hidden states that represent different modes of handling the text of Mark. This will lead to the fitting of hidden Markov models (HMMs), for which the technical details were given in Chapter 4. Statistically, the analytical approach taken in this chapter is an alternative to the one taken in Chapter 3, although it shares many of the concepts presented there. We shall consider for use as regressor variables the same variables that were defined in Section 3.3 and used in the logistic regression models of Sections 3.3 and 3.4.