Answers are grounded in this paper and cite their sources.
The adoption of new rice varieties in western Guangdong remains low, and the role of farmers’ entrepreneurial behavior in this process has received little attention. This study investigated the impact of entrepreneurial behavior on the adoption of new rice varieties using survey data from 602 rice‑planting households in Zhanjiang, Maoming, and Yangjiang. An XGBoost model was employed to predict adoption behavior, and SHAP analysis was used to interpret feature importance and interaction effects. The results showed that the adoption rate of new varieties among entrepreneurial farmers (70.0%) was significantly higher than that among non‑entrepreneurial farmers (42.6%), and the new variety area share was also much larger (48.9% vs. 29.8%). Entrepreneurial competence score (Entre Score) and entrepreneurial farmer identifier (Entrepreneur) were the top two features in the SHAP global importance ranking, surpassing traditional variables such as farm size and income. A nonlinear threshold effect was identified: adoption probability increased slowly when Entre Score was below 2.5 but rose rapidly after exceeding 3.0. Cooperative membership and mobile payment frequency exhibited a strong synergistic effect, with adoption probability exceeding 0.78 when both were present. Sub‑sample SHAP analysis revealed heterogeneous mechanisms: entrepreneurial farmers relied more on information channels, non‑farm income, and gender roles, while non‑entrepreneurial farmers were constrained by farm size, age, and entrepreneurial competence. Propensity score matching confirmed a significant causal effect of entrepreneurship, with an average treatment effect on the treated of 44.6 percentage points. Although XGBoost did not outperform traditional models in prediction accuracy, its combination with SHAP effectively revealed nonlinear relationships and interactions that logistic regression cannot capture. Bangladesh J. Bot. 55(3): 787-795, 2026 (September) Special
No takes yet. Share an insight, caveat, or question.
Hu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: