Exploring the preferences of college students for horticultural therapy activities based on a new classification framework.

To reveal college students' preferences for types of horticultural therapy (HT) activities and the influence ofbiopsychosocial factors, an online questionnaire with AI-generated illustrations was administered to assess 622respondents' preferences for 21 HT activities and their biopsychosocial characteristics. Confirmatory factor analysisconfirmed a four-dimensional model consisting of Making (MA), Planting (PL), Active Perception (AP), and PassivePerception (PP) with good fit and reliability. The preference means ranked as MA (M = 4.065) > AP (M = 3.829) > PL(M = 3.797) > PP (M = 3.646). Biologically, females exhibit a significantly stronger preference for MA (t = - 4.32) and PP(t = - 2.18) compared to males, as evidenced by t-test results. Psychologically, natural connectedness was strongly andpositively correlated with preferences for all types of HT (r = 0.522-0.648), whereas psychological well-being showed onlya weak positive correlation with PL (r = 0.146) through correlation analysis. Individuals with more experience in HT showeda significantly higher preference for PL (F = 6.80), AP (F = 8.42), and PP (F = 3.81) according to ANOVA results. Socially, theANOVA results indicated that there was no significant difference in the preferences for HT types among students ofdifferent educational backgrounds. This study validates a novel framework for classifying HT activities grounded inhuman-plant interactions and, by examining the influence of biopsychosocial factors on preferences for HT, providesstronger evidence for personalized design of HT programs tailored to college students.