WIT Press


Sequential Methods For User Choices: Tests And Properties Applied To A Panel Database

Price

Free (open access)

Paper DOI

10.2495/UT080121

Volume

101

Pages

11

Page Range

121 - 131

Published

2008

Size

358 kb

Author(s)

G. Chilà

Abstract

Systematic observation is an important method for measuring behaviour. Sequential techniques applied to systematic observation allow temporal analysis of the user choice process. In this work, we propose a sequential analysis of a sample family related to the number of vehicles owned. An analysis of recorded data is presented in order to ascertain whether current decisions are directly influenced by the most recent previous decisions. Results obtained by the test application are presented in the paper. Keywords: sequential, sequence, significance, stationarity. 1 Introduction Sequential techniques result from sequential analysis, which arose as a science applied to the observation of social behaviour in psychology. As reported in Gottman and Roy [4], in a seminal paper, in 1952, George Miller introduced Markov processes into psychology, noting that probabilistic methods had proved themselves in sensory psychology and test construction. The classic introductory piece in psychology was Fred Attneave’s 1959 monograph, quoted in [4]. In this monograph, Attneave introduced the notions of temporal structure. However, the main mathematical principles behind sequential analysis were derived between 1957 and 1962. Particularly, maximum likelihood estimates for transitional probabilities in a Markov chain of any order were derived. Both likelihood ratio and chi-squared tests were also used for contingency tables to examine several properties of the chains [4].

Keywords

sequential, sequence, significance, stationarity.