A Cluster Analytic Test of the Transtheoretical Model Applied to Exercise Behavior

Abstract

An empirical typology was created to test the application of the Transtheoretical Model (TIM) to exercise behavior. It was reasoned that if a meaningful typology of exercise could be created that depicted construct relationships posited by the TIM, then this would be supportive evidence for the application of the model to understanding behavior change related to exercise. A sample of 342 adults between the ages of 18 and 75 completed a survey through a random telephone interview. The sample was 62% female, 95% white, and an average age of 43 years. The sample was randomly split into two subsamples to perform cluster analysis on each subsample using Ward's method and squared Euclidean distance. The three variables used in the cluster analysis were the Pros of exercise, the Cons of exercise, and Exercise Confidence. The pseudo F, pseudo t2 and hierarchical dendrograms were used initially to determine the number of clusters and indicated either a five and six cluster solution. Comparison of cluster mean profile patterns for solutions 3 through 8 between the two subsamples resulted in choosing the 5 cluster solution as optimal. Four of the cluster patterns were found in both subsamples, while a fifth cluster was unique to each subsample. This resulted in 6 clusters that were labeled Disengaged, Early, At Risk, Early Action, Maintainers, and Habituated. A series of analyses were conducted to test the internal and external validity of the exercise typology. Two internal validity tests revealed that four of the cluster patterns demonstrated high stability and replicability, while the At Risk and Early Action cluster patterns were less stable. External validity of the clusters was demonstrated with significance testing using the six clusters as levels of a categorical variable to test for differences in five different variable domains: Exercise Behavior, Processes of Change, Stage of Change for Regular Exercise, Stage of Change for other health risks, and selected demographic variables. Large effect sizes were found for exercise behavior, processes of change, and stage of change. The clusters were also differentiated by stage of change for diet, gender, and perceived health. It was concluded that this series of analyses demonstrated the verisimilitude of the exercise cluster typology and that the clusters potentially represent a limited number of Stage of Change subtypes. The results also provide empirical evidence that the spectrum of behavior change can be represented as a series of discrete groups. Issues are discussed pertaining

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 1 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.