“…Methodologically, common factor models (e.g., Brown, 2006) identify the way in which variables can be grouped on the basis of shared variance, but they require all variables to be interpreted as virtually homogeneous indicators of the same latent variable. As a consequence, in the context of the SCM, it has been argued that stereotype structure is best represented by a two-factor structure (e.g., Fiske et al, 2002), three-factor structure (e.g., L opez-Rodr ıguez, Cuadrado, & Navas, 2013), second order factor structure (interpersonal perception traits, Srivastava, Guglielmo, & Beer, 2010), or bi-factor structure (Sayans-Jim enez, Cuadrado, Rojas, & Barrada, 2017). Equally important, when common factor models are applied, they can produce an infinite number of nearly equivalent models based on trivially small fit variations (Raykov & Penev, 1999), but with very different theoretical interpretations (e.g., first order, higher order, or bi-factor models; see Van Bork, Epskamp, Rhemtulla, Borsboom, & van der Maas, 2017).…”