Difficulties of Interpretation With Markov Chain Models
Peter D. Finch · Australian Journal of Statistics · 1982
Summary Questions such as “how useful is a Markov chain model!” and “what does it tell us about a real process?” are discussed. The relevance of context is stressed and it is argued that good data fits are not enough by themselves. It is suggested that a descriptive data analytic interpretation of the modelling process might sometimes be more appropriate than the usual probabilistic interpretation.