“…Latent class models capture preference heterogeneity across segments (classes) of the population and assume uniform parameter estimates within the same class (Greene & Hensher, 2003). The probabilities of class membership are estimated for each individual based on socioeconomic covariates, such as age (Geussens et al., 2019; Kassahun & Jacobsen, 2015; Sardaro et al., 2016), education (Geussens et al., 2019; Van den Broeck et al., 2017), experience (Canessa et al., 2023; Houessionon et al., 2017; Ortega et al., 2016; Rakotonarivo et al., 2017), gender (Geussens et al., 2019), income (Broch & Vedel, 2012; Geussens et al., 2019), risk perception (Tyllianakis et al., 2023), farm characteristics such as farm size (Houessionon et al., 2017), land characteristics (Jaeck & Lifran, 2013), ownership (Broch & Vedel, 2012), soil and water quality (Chang et al., 2017; Raes et al., 2017; Zandersen et al., 2016), or organic farming status (Lapierre et al., 2023; Rocchi et al., 2017).…”