DYNAMIC TYPOLOGY CLUSTERING WITHIN THE STAGES OF CHANGE FOR SMOKING CESSATION

Dynamic typology clustering was employed to find homogeneous subgroups of smokers within each of three early stages of change (precontemplat ion, contemplation. preparation) in a representative sample of smokers . Individual change profiles were created by coding the stage of chang e for each subject at five consecutive assessment points over a 2-year period (baseline and 6, 12, 18, and 24 months). A total of 446 unique change profiles were found in the sample of 2,088 smokers who had com plete data for all five time points. The sample was initially split in to three groups determined by baseline stage of change. Within each in itial stage, subgroups that shared similar patterns of change were int erpreted by examining the shape and elevation of the change profiles. Four major types of profiles emerged: a stable profile, a progressing profile, a vacillating profile, and a regressing profile. External val idation revealed significant differences among the dynamic typology su btypes on key transtheoretical model constructs (processes of change, decisional balance, situational temptations). These results both suppo rt and complement key construct relationships within the transtheoreti cal model and can provide important predictive information to direct a nd enhance treatment interventions for smoking cessation. (C) 1998 Els evier Science Ltd.