UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Assessing customer sustainability behaviours in the fashion industry through data analytics Tehranchi, Armita

Abstract

Sustainability in the fashion industry is an urgent concern as the sector faces increasing scrutiny over its environmental and social impact. This study examines consumer sustainability behavior among young adults in Canada, focusing on Generation Z and Millennials. Using a survey-based approach, it investigates which demographic, behavioural, and attitudinal factors best predict sustainability awareness and purchasing preferences. A sustainability score was developed based on engagement in five circular fashion initiatives, including clothing rental, repair, and take-back programs, and used to segment participants. The score was constructed through normalization and aggregation of behavioural indicators. The analysis combines descriptive statistics, correlation analysis, Ordinary Least Squares (OLS) regression, Analysis of Variance (ANOVA), Multivariate Analysis of Variance (MANOVA), Machine Learning (ML) classification, and association rule mining with lift metrics. The results show that income alone provides limited insight into sustainability engagement. In contrast, behavioural involvement is more closely associated with willingness to pay for sustainable products. Classification models using only demographic variables achieved moderate performance, including logistic regression, decision trees, Support Vector Machines (SVM), and random forests. Among these, the random forest model performed best, achieving an F1 score above 0.8. Model robustness was further evaluated using Monte Carlo simulation under label uncertainty. Consumers with higher sustainability awareness prioritize environmental impact and product quality, while others are more influenced by price. Statistical tests confirm that these differences are significant. Brand preferences also differ, with sustainability-oriented consumers favouring firms perceived as environmentally responsible, while less engaged individuals prefer mainstream brands. Lift analysis supports these findings by revealing distinct brand groupings based on sustainability orientation. Trust in sustainability claims is most strongly influenced by third-party certifications and clear information about environmental impact, whereas general sustainability reports have weaker effects. Take-back and repair programs receive the highest support, reflecting growing interest in circular fashion. Although many participants are willing to engage in such initiatives, traditional incentives such as discounts or charitable donations may be ineffective or discouraging for some consumers. These findings suggest that encouraging sustainable behaviour requires transparency, credible actions, and strategies aligned with consumer values rather than broad income-based targeting.

Item Media

Item Citations and Data

Rights

Attribution-NonCommercial-NoDerivatives 4.0 International