An ordination approach to the quantification of smokers' readiness to change 1This research was supported by a grant from Société de Tabacologie, Paris, and partially supported by La Ligue Nationale Contre le Cancer.

Abstract

The paper argues that the problem of estimating a “readiness-to-change” score for smokers from processes of change measures is closely related to dating problems in archaeology. It amounts to finding a joint ordination of both smokers (“sites”) and processes of change (“tools”) in a common space, so that smokers be located close to processes they use most, in some sense. Properties of Correspondence Analysis as a reciprocal averaging technique are first briefly summarized, and its capability of recovering nonlinear structures is recalled. Application to processes of change data in smoking cessation leads to a solution which first axis is interpreted as a cognitive-behavioral continuum. Processes are shown to be scaled in a way that is in good convergence with longitudinal data previously reported ( Prochaska et al., 1991 ), and subjects' locations to be significantly related to actual quitting. Finally, it is shown how new responses profiles could be projected on the obtained solution to get a practical estimation of probability to quit.