Consumer Behavior

Identifying and Characterizing Potential Hemp Fashion Consumers: A Supervised and Unsupervised Statistical Learning Approach

---

Abstract

This study examines U.S. consumers’ purchase intentions for hemp-based fashion products by integrating supervised and unsupervised statistical learning methods. Although hemp offers substantial environmental and functional benefits, its association with marijuana continues to shape public perceptions and limit market growth. Using survey data from 538 participants, the study tests eight hypotheses through linear regression, random forest models, and generalized additive models, followed by consumer segmentation via K-Means and Hierarchical Clustering. Seven determinants, including attitude toward hemp-based fashion products, subjective knowledge, perceived social and functional value, fashion involvement, and attitudes toward marijuana legalization, significantly predict purchase intention. Three distinct consumer segments emerged: strongly promising consumers, potential consumers, and low-intent consumers. Findings highlight the importance of reducing stigma through knowledge-building and emphasizing the social and functional value of hemp fashion. The study provides actionable insights for marketers aiming to expand the hemp fashion market.

Keywords: Hemp Fashion, Consumer Segmentation, Machine Learning, Sustainable Textiles

How to Cite:

Zhang, Y., Liu, C. C., Xia, S. & Cameron, B., (2025) “Identifying and Characterizing Potential Hemp Fashion Consumers: A Supervised and Unsupervised Statistical Learning Approach”, International Textile and Apparel Association Annual Conference Proceedings 82(1). doi: https://doi.org/10.31274/itaa.21524