“…A common drawback of IPW, AIPW, and PSM estimators is that they can only address the endogeneity issue associated with observed factors. Rural households decide to use ICTs on the basis of both observed factors (e.g., age, education experience, and household size) and unobserved heterogeneities (e.g., innate abilities and motivations) (Cai, Qi, & Yi, 2022; Min et al, 2020; Zheng & Ma, 2021). However, we are unaware of any research in this area that accounts for the selection bias arising from observed and unobserved heterogeneities.…”