Rethinking segmentation within the psychological continuum model using Bayesian analysis

Highlights

•	We propose a novel approach to segmentation within the Psychological Continuum Model.
•	We compare conventional segmentation, k-means clustering, and Bayesian LPA approaches.
•	Bayesian LPA outperforms the conventional staging algorithm in assigning PCM stage.
•	Bayesian LPA offers more distinct segmentation boundaries and greater predictive power.
•	We encourage the use of Bayesian analysis in future sport management research.

Abstract

The Psychological Continuum Model (PCM) represents a theoretical framework in sport management to understand why and how consumer attitudes form and change. Prior researchers developed an algorithmic staging procedure using psychological involvement to operationalize the PCM framework within sport and recreational contexts. Although this staging procedure is pragmatically sound, it rests upon a procedure that, while intuitively sensible, lacks scientific rigor. The current research offers an alternative approach to PCM segmentation using Bayesian Latent Profile Analysis (Bayesian LPA). Comparing three analyses (the conventional PCM segmentation algorithm, K-means clustering, and Bayesian LPA), results demonstrated that Bayesian LPA provides a promising and alternative statistical approach that outperforms the conventional PCM staging algorithm in two ways: (a) it has the ability to classify individuals into the corresponding PCM segments with more distinct boundaries; and (b) it is equipped with stronger statistical power to predict conceptually related distal outcomes with larger effect size.

Keywords:

• Bayesian analysis
• Psychological involvement
• Segmentation
• Psychological Continuum Model (PCM)
• Staging algorithm

Notes

1 The first two authors contributed equally to this manuscript.